Eliminativism, Objects, and Persons: The Virtues of Non-Existence 2018031811, 9780367000219, 9780429444944

Table of contents :
Cover
Title
Copyright
Contents
Acknowledgements
Introduction: Eliminativism: A Method
Part I Eliminativism
1 Eliminativism and Ordinary Objects
2 Eliminativism, Common Sense, and Intuitions
3 Eliminativism, Reductionism, and Composition
4 The Fundamental Ontology of Eliminativism
Part II The No-Self View
5 Exceptionalism
6 The Self: A Substance or a Bundle?
7 The No-Self View
8 Eliminativism, Life, and Death
Part III Aesthetic (and Other) Objects
9 ‘Upper Level’ Objects, Musical Works
10 Photographs
Concluding Remarks on Eliminativism and Monism
References
Index

Citation preview

Eliminativism, Objects, and Persons

In this book, Jiri Benovsky defends the view that he doesn’t exist. In this book, he also defends the view that this book itself doesn’t exist. But this did not prevent him to write the book, and although in Benovsky’s view you don’t exist either, this does not prevent you to read it. Benovsky defends a brand of non-exceptionalist eliminativism. Some eliminativists, typically focusing on ordinary material objects such as chairs and hammers, make exceptions, for instance for blue whales (that is, living beings) or for persons (that is, conscious organisms). Benovsky takes one by one all types of allegedly existing objects like chairs, whales, and persons and shows that from the metaphysical point of view they are more trouble than they are worth – we are much better off without them. He thus defends an eliminativist view about ordinary objects as well as the ‘no-Self’ view, where he explores connections between metaphysics, phenomenology, and Buddhist thought. He then also considers the case of aesthetic objects, focusing on musical works and photographs, and shows that the claim of their non-existence solves the many problems that arise when one tries to find an appropriate ontological category for them, and that such an eliminativist view is more natural than what we might have thought. The arguments provided here are always topic-specific: each type of entity is given its own type of treatment, thus proving a varied and solid foundation for a generalized, non-exceptionalist, full-blown eliminativist worldview. Having been struck by Descartes’s evil demon thought-experiment, Jiri Benovsky began to study metaphysics to try to find a proof that the world really exists. He did not find that proof, but at least he found a job at the University of Fribourg, in Switzerland, where he can not only go in the mountains but also spend his days thinking about things like existence, reality, personal identity, possible worlds, material objects, and time. He is the author of several books, mainly in the fields of metaphysics and aesthetics, and he published a recent book Meta-metaphysics (2016) where he argues that at the end of the day the evaluation of metaphysical theories is (and should be) based on their aesthetic properties – we choose a metaphysical theory as being the best because we find it beautiful. More information on Benovsky’s work can be found online at www.jiribenovsky.org.

Routledge Studies in Metaphysics

Properties, Powers and Structures Issues in the Metaphysics of Realism Edited by Alexander Bird, Brian Ellis, and Howard Sankey The Puzzle of Existence Why Is There Something Rather than Nothing? Edited by Tyron Goldschmidt Neo-Davidsonian Metaphysics From the True to the Good Samuel C. Wheeler III Neo-Aristotelian Perspectives in Metaphysics Edited by Daniel D. Novotný and Lukáš Novák Nominalism about Properties New Essays Edited by Ghislain Guigon and Gonzalo Rodriguez-Pereyra Substance and the Fundamentality of the Familiar A Neo-Aristotelian Mereology Ross D. Inman Philosophy of Time and Perceptual Experience Sean Enda Power Fundamental Causation Physics, Metaphysics, and the Deep Structure of the World Christopher Gregory Weaver Eliminativism, Objects, and Persons The Virtues of Non-Existence Jiri Benovsky For a full list of titles in this series, please visit www.routledge.com

Eliminativism, Objects, and Persons The Virtues of Non-Existence Jiri Benovsky

First published 2019 by Routledge 711 Third Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2019 Taylor & Francis The right of Jiri Benovsky to be identified as author of this work has been asserted by him in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data Names: Benovsky, Jiri, author. Title: Eliminativism, objects, and persons : the virtues of non-existence / by Jiri Benovsky. Description: 1 [edition]. | New York : Taylor & Francis, 2018. | Series: Routledge studies in metaphysics ; 13 | Includes bibliographical references and index. Identifiers: LCCN 2018031811 | ISBN 9780367000219 (hardback) Subjects: LCSH: Ontology. | Philosophy of mind. | Object (Philosophy) | Nonexistent objects (Philosophy) | Materialism. | Self-knowledge, Theory of. Classification: LCC BD331 .B463 2018 | DDC 111—dc23 LC record available at https://lccn.loc.gov/2018031811 ISBN: 978-0-367-00021-9 (hbk) ISBN: 978-0-429-44494-4 (ebk) Typeset in Sabon by Apex CoVantage, LLC

Contents



Acknowledgements

vi

Introduction: Eliminativism: A Method

1

PART I

Eliminativism

7

  1 Eliminativism and Ordinary Objects

9

  2 Eliminativism, Common Sense, and Intuitions

25

  3 Eliminativism, Reductionism, and Composition

49

  4 The Fundamental Ontology of Eliminativism

58

PART II

The No-Self View

63

 5 Exceptionalism

65

  6 The Self: A Substance or a Bundle?

78

  7 The No-Self View

91

  8 Eliminativism, Life, and Death

114

PART III

Aesthetic (and Other) Objects

139

  9 ‘Upper Level’ Objects, Musical Works

141

10 Photographs

156



Concluding Remarks on Eliminativism and Monism

167

References Index

173 183

Acknowledgements

Part II of this book defends the no-Self view. Therefore, I have nobody to thank, since persons do not exist (and neither do I, for that matter). So let me just mention some names, for no reason at all. Marie Benovska, Vlastimil Benovsky, Céline Chevalley, Annabel Colas, Damiano Costa, Barry Dainton, Coralie Dorsaz, Fabian Dorsch, Michael Esfeld, Akiko Frischut, Richard Glauser, Ghislain Guigon, Jean-Baptiste Guillon, Mark Heller, Rob Hopkins, Robert Howell, Thomas Jacobi, Dan Korman, Kathrin Koslicki, Uriah Kriegel, Baptiste Le Bihan, Robin Le Poidevin, Dan López de Sa, Jonathan Lowe, Olivier Massin, Trenton Merricks, Anne Meylan, Kevin Mulligan, Martine Nida-Ruemelin, Donnchadh O’Conaill, Nathan Oaklander, Laurie Paul, Roger Pouivet, Gonzalo RodriguezPereyra, Sven Rosenkranz, Martine Nida-Ruemelin, Thomas Sattig, Jonathan Schaffer, Daniel Schulthess, Ted Sider, Mark Siderits, Gianfranco Soldati, Tuomas Tahko, Cain Todd, Achille Varzi, and Adrien Zanarelli. I have been working on eliminativism for a long time. Some of the ideas present in this book have already appeared as parts of journal articles: ‘Are we causally redundant? Eliminativism and the no-Self view’, 2017, Croatian Journal of Philosophy; ‘Nothing is alive (we only say so)’, 2017, Think; ‘Endurance, dualism, temporal passage, and intuitions’, 2016, Review of Philosophy and Psychology; ‘Eliminativism and gunk’, 2016, Teorema; ‘From experience to metaphysics: on experience-based intuitions and their role in metaphysics’, 2015, Noûs; ‘ “Nothing over and above” or “nothing”? On eliminativism, reductionism, and composition’, 2015, Polish Journal of Philosophy; ‘Alethic modalities, temporal modalities, and representation’, 2015, Kriterion; ‘Vague objects with sharp boundaries’, 2014, Ratio; ‘Tropes or universals: how (not) to make one’s choice’, 2014, Metaphilosophy; ‘I am a lot of things: a pluralistic account of the Self’, 2014, Metaphysica; ‘What photographs are (and what they are not)’, 2011, Disputatio; ‘The Self: a Humean bundle and/or a Cartesian substance?’, 2009, European Journal of Analytic Philosophy.

Introduction Eliminativism: A Method

Desert landscapes have often attracted a number of philosophers. An eliminativist is someone who takes a stroll in the desert, wearing a nerd t-shirt that says “less is more” (or perhaps, more precisely, “less is enough”1). There are many things one can be an eliminativist about. This book is concerned with this variety of things in two different ways. First, I explore in detail some of the places where it is a good idea to be an eliminativist. These include artefacts (tables, chairs), natural ordinary macroscopic objects (trees, mountains, human bodies), the Self (the person, the subject), and aesthetic objects (musical works, photographs). Second, I explore eliminativism as a method. Indeed, as the first-level discussions of different eliminativist claims will show, there is a common eliminativist strategy that can be applied to various cases. This strategy, this method of dealing with problems, is what eliminativism is all about.2 It would be nice to have a clear and clean definition of eliminativism from the beginning, before I start discussing it. But in order to see what eliminativism is we need to see it at work, as applied to the various types of cases, and only then can we see the overall methodological picture. A hike in the desert can be long, and requires endurance and patience. To put it shortly, in order to have some idea of what we’re talking about, the general strategy eliminativists use is often said to rely on the principle of parsimony: we should not postulate the existence of something if we do not really need it. For instance, eliminativism about ordinary macroscopic material objects is the view that we don’t need to postulate the existence of many things like trees, tables, planets, and so on. There are many variants of this view,3 but the common idea to all such eliminativist claims is that without having to postulate the entities at hand, we already have all we need to account for all phenomena that need to be accounted for: we do not need tables because, as eliminativists put it, we have ‘simples arranged tablewise’ (this is the most standard version; I’ll discuss alternative versions of eliminativism that do not appeal to the existence of simples in Part I, Chapter 4). As I am now typing this sentence on my computer, it seems to me that I see a table in my visual field. But, as Merricks (2001, pp. 8–9) rightly claims, my visual experience would be exactly the same even if there were no table and if there

2 Introduction were ‘only’ simples arranged tablewise. The idea here is that our sensory experiences can be accounted for in terms of some fundamental (and existing) entities, and so there is no need to postulate a further entity (a table). Thus, eliminativism does not make us abandon our ordinary experience of the world and it is fully compatible with it (more on this in Part I, Chapters 1 and 2). It does not make us abandon ordinary language either. Indeed, the eliminativist can say that the English sentence “There is a table” can be understood in terms of “There are simples arranged tablewise” (more on this in Part I, Chapter 1 as well). Some eliminativists argue that we need to make exceptions – for us. Indeed, eliminativists such as Peter Van Inwagen or Trenton Merricks famously argued (for different reasons and in different ways) that while eliminativism is the best theory around when it comes to tables, planets, or statues, it is not to be endorsed in the case of humans – an exception is to be made. Van Inwagen focuses on all living entities, and Merricks focuses on conscious organisms. In my earlier work (see Benovsky [2014]), I also followed this line of reasoning, and I argued in favour of an exception concerning the Self. But as we will see in Part II such exceptionalist views are unnecessary and I will argue that no exception is needed. As we have just seen, this rough idea of what eliminativism is, or can be, appears to heavily rely on a principle of parsimony. Many metaphysicians – and not only eliminativists – appeal to and use the principle of simplicity and parsimony in their work, but few of them explicitly say why. “If you can do it with less, don’t do it with more”, the slogan goes (often only assumed in silence). But why? And what does “less” and “more” mean? Several criteria can be distinguished here: (i) simplicity of the structure of the theory (ii) number and complexity of primitives in the theory (iii) qualitative parsimony (that concerns the number of kinds of entities) (iv) quantitative parsimony (that concerns the number of entities). These four different criteria are, of course, related. For instance, with greater complexity with respect to (i) it is likely that a theory will be able to be more parsimonious with respect to (ii), and vice versa. The same may apply for (iii) and (iv) as well: more kinds of entities perhaps allow for being more parsimonious with respect to the number of entities. For instance, a theory of objects and properties that appeals to particulars and universals (i.e. two kinds of entities) will perhaps contain less entities than trope theory that includes only particulars – tropes – where for each instantiation of a universal (one entity) there has to be a different trope (i.e. there are as many entities as there are instantiations of the one universal). Thus, being more parsimonious with respect to some of the criteria above is very likely to force one to be less parsimonious with respect to others. So, the question is,

Introduction  3 are some of these criteria more important than others? Many think so as, for instance, does David Lewis: I subscribe to the general view that qualitative parsimony is good in a philosophical or empirical hypothesis; but I recognize no presumption whatever in favour of quantitative parsimony. (Lewis [1973, p. 87]) When it comes to theories of properties, contra friends of universals, nominalists would probably generally agree (see also for instance RodriguezPereyra [2002, p. 202]). Here is a naïve but genuine question: are nominalists nominalists because they hold that qualitative parsimony is more important than quantitative parsimony, or do they hold this view because they are nominalists? It should be (and probably is) the former, otherwise their overall strategy could be regarded as question-begging. But then, if this meta-theoretical criterion is not supported by their ontology, by what is it supported? Note that the price to pay for this criterion is not only that their ontologies are presumably less quantitatively parsimonious, but also that they usually have a more complex structure (compare, for instance, the structure of Rodriguez-Pereyra’s resemblance nominalism, which includes a hierarchy of pairs of resembling particulars [Rodriguez-Pereyra (2002, chapter 9)] in order to account for one particular’s being F, to the structurally simpler view that the particular instantiates a universal). The support for this claim comes, as Rodriguez-Pereyra (2002, p. 202) puts it, from the view that “the generality of metaphysical theories makes the existence of a certain kind of entities matter more in assessing their truth or falsity than the number of entities of that kind”. Not all nominalists agree. For instance, Melia (2005) prefers to take the route of parsimony with respect to the structure of his nominalist theory, but completely sacrifices parsimony with respect to the kinds of entities included in the theory – indeed, according to his view, there are as many kinds as there are individuals. The very general idea behind this sort of meta-theoretical attitude is that qualitative parsimony is a virtue of a theory and we should be as parsimonious as possible, but no more – that is, if introducing a new kind of entity is justified by its usefulness in generating a structurally simpler view, then it should be introduced. Thus, ironically enough, Melia’s nominalism joins the side of the friends of universals – for it is precisely this kind of considerations that defenders of universals use in order to support their view: universals are not superfluous entities that can be dispensed with since they do useful work in allowing for a powerful and structurally simple theory (here (i) and (iv) are privileged). Tim Maudlin puts the general methodological point nicely as follows: It is not clear that [Ockham’s Razor] can withstand much critical scrutiny. If by ‘necessitas’ [in “Entia non sunt multiplicanda praeter

4 Introduction necessitatem”] one means logical necessity, then the Razor will land us in solipsism. But if one means something milder – entities ought not to be multiplied without good reason – then the principle becomes a harmless bromide: nor should one’s ontology be reduced without good reason. (Maudlin [2007, p. 3]) Tallant (2013) defends quantitative parsimony in an interesting way:4 in his view, it is a virtue of a theory to be quantitatively parsimonious because quantitative parsimony often carries with it other theoretical virtues – quantitative parsimony is thus virtuous here in an indirect way, not for itself but for what interesting side effects it can have. Tallant takes the case of the debate between presentism and eternalism, and claims that the fact that presentism is more quantitatively parsimonious than eternalism (since eternalism postulates the existence of, say, dinosaurs while presentism does not) means that presentism will have at least one less worry to face: in short, it will not need to explain why we do not accept sentences such as “There are dinosaurs” to be true (see Tallant [2013, pp. 694–697]). He concludes: when engaging in the kinds of weights and balances deliberations that we typically do in metaphysics, more quantitatively parsimonious theories are preferable to their less quantitatively parsimonious counterparts in at least some ways and in some situations; namely, when they carry with them certain other virtues – virtues of the sort outlined above. These considerations give us some reason to prefer presentism to eternalism, even if that reason is ultimately outweighed. (Tallant [2013, p. 704]) As the various considerations above illustrate, even if there is disagreement as to what sort of simplicity and parsimony should be privileged, everybody seems to agree that it is a general virtue of a theory to exhibit some sort of it.5 Why is simplicity and parsimony interesting? Perhaps, for its side effects, as Tallant claims. Also, there clearly is a pragmatic reason: a simpler theory will be easier to work with, to manipulate, to understand, to explain to others, and so on (Tallant also makes this point). While this certainly is an interesting feature of a metaphysical theory, it can hardly be a reason to choose one as being better than the others – such a pragmatic advantage is also more of a welcome by-product of simplicity and parsimony than the main reason for desiring it. So, perhaps the criterion of simplicity and parsimony is desirable because it is truth-conducive?6 Could it be that a simpler theory is a better guide to metaphysical truth than a more complex one? But why? What is it that makes a simpler and more parsimonious theory more likely to be true than a more complex one? Perhaps behind this claim lies the general idea that the world is as simple as possible – as Newton put it: We are to admit no more causes of natural things than such as are both true and sufficient to explain their appearances. To this purpose the

Introduction  5 philosophers say that Nature does nothing in vain, and more is in vain when less will serve; for Nature is pleased with simplicity and affects not the pomp of superfluous causes. (Newton [1687, rule I]) When defending his own version of eliminativism, Sider (2013, p. 240) takes this Newtonian line. His preferred kind of parsimony is (ii) – this is what he calls a theory’s ideology: “its ideology consists of the undefined notions it employs”. Eliminativism, Sider remarks, does not only eliminate objects – that’s its ontological parsimony advantage – but it also eliminates the notion of part from our ideology. He then concludes that eliminativism7 is “an ideologically simpler theory, and so is more likely to be true. . . . [S]implicity is a guide to truth, together with the thought that eliminating primitive notions makes a theory ‘structurally’ simpler”. The trouble with such a view is that it is under-motivated. Why would ‘Nature prefer simplicity to complexity’? What guarantees that metaphysical truths are simple rather than complex? Isn’t it more plausible that to assume such a claim does no more than to express one’s own (and not Nature’s) preference for simplicity and parsimony? I think it is, and for two good reasons: the pragmatic one mentioned above, and the fact that many of us just like it. Let us sum up where we stand here: first, the criterion of simplicity and parsimony is taken as important by virtually all metaphysicians but, second, it seems to be very hard to say what kind of simplicity and parsimony is the one to be preferred, and third, the claim that this criterion is truth-conducive is under-motivated.8 What I want to suggest, then, is that it is, in a good sense, ‘wishful thinking’: the requirement for parsimony and simplicity comes from us rather than from metaphysical reality. This does not make parsimony and simplicity useless, on the contrary. After all, we are the ones who build our theories, and if we build them to our liking it can only be a good thing. I take this very seriously in Metametaphysics (2016, Springer), where I argue that, at the end of the (long) day, metaphysical theories are chosen in virtue of aesthetic properties they possess.9 Concerning simplicity and parsimony, Michael Huemer (2009) takes this to be a “discouraging result”. He recognizes that “many philosophers’ ‘taste for desert landscapes’ is indeed an aesthetic rather than an epistemic preference” (p. 235) but he claims that “in typical philosophical contexts, ontological simplicity has no evidential value” (p. 216). I don’t agree that this result is in any way discouraging (see especially Chapter 7 in Benovsky [2016b]), but this is not the place to argue (again) for this point. The question here is: where do the considerations above leave us when it comes to eliminativism? Eliminativism seems to be the view that relies the most heavily on parsimony when it comes to the number of entities. But if it only relied on that, it would be, as we have seen, a shaky ground to stand on. To my mind, the best reason to embrace any brand of eliminativism is thus not its parsimony or simplicity (although these are relevantly interesting features of eliminativism), but rather its explanatory power – that is, its

6 Introduction power to avoid or solve problem cases such as vagueness, composition, and others that we will come across in this book. Eliminativism, especially when taken as an overall strategy that can be applied to many different cases, is a very powerful tool indeed. Eliminativism is an incredibly useful tool for solving metaphysical problems. In short, eliminativism works, it nicely solves or avoids many problems that alternative views have – and that’s the best reason to endorse it. It does its job in a parsimonious and elegantly simple way, and that’s a further reason as well, but the eliminativist who takes a stroll in the desert wearing a nerd t-shirt that says “less is more” should rather be wearing a t-shirt that says “powerrrrr!”.10 This is why this book is written in the way it is: the plan is to show that in many important places eliminativism works. Part I is about ordinary objects (artefacts and natural objects), Part II is about the no-Self view, and Part III is concerned with social and aesthetic objects. At the end of the day, we should all be impressed by the power of eliminativism and that’s the reason to select it as being the best theory around.

Notes 1 Or perhaps “Ockham rules”. 2 Compare to Van Fraassen’s (2002) idea that empiricism (as well as other philosophical views) is a stance, and to Alyssa Ney’s (2008) discussion of physicalism as being an attitude. 3 Various versions of such an eliminativist view can be found, for instance, in Unger (1979), Van Inwagen (1990), Heller (1990, 2008), Merricks (2001), and Sider (2013). I will discuss the exceptions that Van Inwagen and Merricks make for living beings or for organisms in Part II. 4 See also Jansson and Tallant (2017). 5 In conversation, John Bigelow expressed the view that the best theory would be one that embraces both tropes and universals. While this can sound very unparsimonious and unappealing at first sight, it is no more than following a little further the idea that explanatory power and simplicity of structure of this ‘mixed’ view is more important than qualitative parsimony. 6 For a discussion, see Paul (2012), Kriegel (2013, §3), Brenner (2017). 7 Sider uses the term “nihilism”. 8 Sober (1981) and Rodriguez-Pereyra (2002, p. 207) discuss the idea that a (qualitatively) more parsimonious ontology is more likely to be true than a less parsimonious one because a conjunction (of claims) always has a lower probability than any of its conjuncts (if these are mutually independent). But this only works, following what Rodriguez-Pereyra rightly says, if it is possible to attribute comparable probability values to the axioms of a theory – but it seems to me that asking, for instance, whether the existence of tropes is more or less probable than the existence of universals is wrongheaded, for how could any such claim ever be justified? 9 See also Willard (2014) for a discussion. 10 This should probably be shout loudly in the manner of Jeremy Clarkson.

Part I

Eliminativism

1 Eliminativism and Ordinary Objects

§1 It appears to us that we live in a world populated by a great number and great variety of ‘ordinary objects’. These are the macroscopic objects that we interact with on a daily basis, including our own bodies and the chairs that our bodies often sit on, but also more distant objects like planets or stars, and even objects that we have never actually seen or interacted with in any way but that we have good reasons to believe there to be (for instance, I have never seen a blue whale, but I am told that there still are some). Nobody in her right mind would dream of denying that these objects exist. This is what ‘folk ontology’ is all about. Eliminativism then owes its name to the idea that we should eliminate these objects from our ontological worldview. But why? What’s wrong with such objects? Why should anybody want to get rid of blue whales, chairs, or even her own body? Let us have a look at some reasons why.

§2 Ordinary objects are troublemakers, that’s the thing. A first problem stems from the fact that they are composed. The idea is that there are some fundamental components, perhaps fundamental particles or properties, that compose ordinary objects. One can then ask: when do some fundamental components compose an ordinary object, and when do they not?1 If the answer is that sometimes they do and sometimes they don’t, then such a restriction on composition will have to be arbitrary. A first kind of arbitrariness is due to the vagueness of ordinary objects. Lewis (1993) famously takes the example of a cloud – it seems to be nicely delimited in the middle of a clear blue sky, but when you look closer its boundaries are in fact very vague, in the sense that it is indeterminate which molecule is part of the cloud and which is part of the air surrounding the cloud. The same is true for chairs and whales. Look closely enough, and the borders of such macroscopic material objects will be vague, in the same sense. Think of the borders of planets or stars. The trouble is, it will then

10 Eliminativism be a vague matter whether these components (including these molecules, droplets, atoms . . .) compose my chair, or whether these other components compose my chair. They’re almost the same. But only almost. There actually are very many possibilities here – put in this way, this is the well-known ‘problem of the many’. Some unpalatable consequences follow. If we say that all of the bunches of components compose an object, we then have very many chairs instead of the one I am sitting on. If we say that only one of the bunches of components compose an object, it will then appear that either composition is vague and that material macroscopic objects are metaphysically vague (this is Lewis’s own objection, since he rejects metaphysical vagueness), or that it is entirely arbitrary to say that it is this bunch of components rather than this other.2 I want to focus on this latter point, and examine in more detail what this consequence amounts to. As we will see in the next section, none of the currently available strategies for dealing with the problem of vagueness avoids an undesirable amount of arbitrariness.

§3 We have seen that ordinary objects are vague. This puts then pressure on theories of vagueness to account for their vague nature, but as we shall see, all theories of vagueness have to accept the existence of sharp thresholds – and, sharp thresholds create arbitrariness. This is obvious in the case of epistemicism, and it is a well-known defect of supervaluationism, but as we will see friends of metaphysical vagueness do have to endorse the existence of sharp thresholds in their theory as well. To start, let us make the claim that ordinary objects are vague a bit more precise and focused. Many intuitive cases of possession of vague properties spring to mind, such as somebody’s baldness, the reddish-but-still-abit-orange colour of a tomato, and many others. The cases we are mostly interested in here concern a type of vagueness one might wish to call “existential vagueness”, that is, vagueness concerning the question of whether an entity is (counts as) an ordinary object or not. Ordinary objects are those that populate the domain of ordinary quantification (chairs, whales, planets, mountains, and so on). Extra-ordinary objects are those that typically come out of a process of unrestricted mereological composition, as for instance the aggregate of the tail of a blue whale and the legs of my chair. Extraordinary objects might be vague (for instance, there might be vagueness due to their colour, baldness, and so on), but their spatio-temporal boundaries and their existence are typically well-delimited, or at least they usually easily can be. Ordinary objects, on the other hand, seem to suffer from irremediable existential vagueness. Let us then examine how various types of theories of vagueness on the market try to provide an account of this kind of vagueness. The epistemicist is the only one who explicitly endorses and defends the view that the prima facie claim about vagueness of ordinary objects is, strictly speaking, false. Indeed, at the very heart of epistemicism lies the

Eliminativism and Ordinary Objects  11 central claim that, contrary to appearances, an ordinary object like a mountain, say Mont-Blanc, does have precise spatio-temporal boundaries – it’s just that we don’t know them. This ignorance is usually said to come from considerations about language and about our use of words like “MontBlanc” or “the chair”. Supervaluationism is a type of view that claims that vagueness is a linguistic phenomenon. One can say here that Mont-Blanc is spatially vague simply because nobody ever defined precisely enough the term “MontBlanc” in a way that would delimitate with full precision its spatial boundaries (and similarly for all vague predicates, like “bald”, “red”, “orange”, “big”, and so on). There are, however, possible admissible precisifications of vague terms like “Mont-Blanc”, whose referents are, in short, candidates for being the Mont-Blanc. Truth values of sentences containing such vague terms are then given by the rule of supervaluations, as follows. A sentence like “Mont-Blanc is a high mountain”, that contains vague terms, is such that it is true under all precisifications of these vague terms – thus it is “super-true”. Similarly, a sentence like “Mont-Blanc is the smallest mountain in the Alps” is “super-false” since it is false under all precisifications. A sentence like “Mont-Blanc is a difficult mountain to climb” is true under some precisifications of “Mont-Blanc” and of “difficult” but false under others – in such cases, according to supervaluationism, the sentence is then neither true nor false (its truth-value is Indeterminate). Supervaluationism thus endorses a three-valued logic, rejects bivalence, but preserves the law of excluded middle. (Indeed, “P˅ ¬ P” is super-true. Classic verifunctionality is thus not preserved since even if P has the truth-value Indeterminate “P˅ ¬ P” is super-true.) The point of endorsing such a non-classical three-valued logic is to account for genuine cases of vagueness, but this account fails, for two reasons. The first is well-known and concerns the fact that in the supervaluationist’s own terms it is super-true that ∃n(Fan Λ¬Fan+1 ) The illustrative example often used to paraphrase this expression is that “There is an n such that a person with n hairs is bald and a person with n + 1 hairs is not bald”, which in our case of spatial vagueness of MontBlanc becomes something like “There is a spatial location such that MontBlanc is not located at it, and there is an immediately adjacent location such that Mont-Blanc is located at it”. Such sentences come out as super-true – i.e. true under all admissible precisifications – from the supervaluationist’s machinery. Of course, supervaluationism does not imply and does not say that there is a particular sharp threshold between someone who is bald and someone who is not (say, 147 hairs) or that there is a particular point where Mont-Blanc begins and where it ends. But it does imply that there is some such threshold, which in addition to going against the idea of there being

12 Eliminativism genuine cases of vagueness yields a situation where one has to accept that there are true existential statements which have no true instances (see Hyde [2011]). The second reason why this account fails and why it forces us, contrary to the desideratum supervaluationists had at start, to reject the existence of sharp thresholds, is one that it shares with the theory of metaphysical vagueness which also appeals to a three-valued logic, to which I shall turn my attention now. (Note that some versions of the theory of metaphysical vagueness actually turn out to be quite close to supervaluationism, such as the modal view of Akiba [2004] and Barnes [2010]. The main point I want to make concerning these views – namely that they have to endorse the existence of sharp thresholds – applies to all of them.) According to what we may call metaphysicism, vagueness is a metaphysical phenomenon. In short, it is not our language or our ignorance which are responsible for there being genuine cases of vagueness, rather the world itself is vague – Mont-Blanc, for instance, is thus a metaphysically vague object. This view, then, takes the phenomenon of vagueness at face value. The world is vague. Objects are vague. Mont-Blanc is a metaphysically vague object that has vague boundaries, and the reason why it is hard to say where/when it begins is that it does not have a precise spatial/temporal beginning. The notion of metaphysical vagueness is often taken to be a primitive one (see, for instance, Williams [2008]), or it is defined, probably with some amount of circularity, in terms of the operator “it is indeterminate that” (see, for instance, Tye [1990]). However one defines – or not – this notion, it is a very useful one in solving problems concerning the sorites: in a sorites, at least one (but usually more) premise of the argument is neither true nor false, and consequently it is not true, and thus the argument is not valid. Classical bivalent logic is thus abandoned in favour of a many-valued one, and typically, both bivalence and the law of excluded middle are thus discarded. Perhaps the simplest and most obvious logic to be used here is a three-valued one, such as Kleene’s. Truth-tables for disjunction, conjunction, and negation look then like this:

˅

T

I

F

˄

T

I

F

¬

T I F

T T T

T I I

T I F

T I F

T I F

I I F

F F F

T I F

F I T

where “I” stands for “neither true nor false” (as opposed to “true and false”). To focus on disjunction, it is true when at least one of the disjuncts is true, and false when both disjuncts are false, exactly as in classical bivalent logics. Otherwise, it is Indeterminate. The idea behind this is that if, say, one

Eliminativism and Ordinary Objects  13 disjunct is F and the other is I, we then don’t have enough information to calculate the classical truth-value of the disjunction, and thus it is I. Similarly for other cases. This is how in such a three-valued logic P˅¬P is not a logical truth (that is, the law of excluded middle is abandoned), and P˄¬P is not a logical falsehood – a contradiction which is not false (when P has the truth-value I)! Classical tautologies are thus not valid. Still focusing on disjunction, in classical logic, one can use disjunction elimination in proofs like this one: ¬P ∨ ¬Q  ¬ ( P ∧ Q ) 1 2 3 4 5 6 7 8 9 10 11 12

¬P ˅ ¬Q ¬P

P˄Q ¬P P ¬(P ˄ Q) ¬Q P˄Q ¬Q Q ¬(P ˄ Q) ¬(P ˄ Q)

2 ˄E, 3 ¬I, 3-5 7 ˄E, 8 ¬I, 8-10 ˅E, 2-6, 7-11

In Kleene-like three-valued logics, however, such a classical disjunction elimination, which is crucial for this type of proofs, is not available. Indeed, there are more than the two cases where ¬P is true (step 2) and where ¬Q is true (step 7): there are also the cases where one of the disjuncts has the truth-value I. But from these latter cases we cannot derive the conclusion, since if, for instance, ¬P is I and P is I we don’t have a contradiction of the type P˄¬P and thus we cannot proceed to step 6. These considerations show us the technical costs of rejecting classical bivalent logics. Such costs, of course, might be worth paying if they can help us in solving problems concerning existential vagueness, the sorites, and related worries. However, this strategy fails – and, it fails for conceptual reasons which are nicely exhibited by the use of formal tools such as Kleenelike three-valued logics, but which can also be formulated independently of these formal tools, as we shall now see. So, let us now get back to the metaphysical theory of vagueness. Even if one were ready to pay the aforementioned costs of a three-valued logic, or if one used a strategy similar to supervaluationism, it would not do the trick. Indeed, at the heart of metaphysicism lies the idea that the bivalent passage from truth to falsehood is too abrupt, and that it needs to be replaced by a smoother passage through the truth-value Indeterminate. Trouble is, a transition between T and I, and between F and I is just as abrupt and just as counter-intuitive as the abrupt transition between T and F (accepted from the start by epistemicists): such sharp thresholds, even if there is now two of

14 Eliminativism them instead of one, are just not what we have in the case of a sorites, where philosophical problems arise precisely because of the lack of such sharp transitions. In cases of existential vagueness and in the case of a sorites, we need to account for a smooth and continuous change in truth values between the different steps of the argument, and as a formal tool a threevalued logic just does not give us what we need. Perhaps then, a fuzzy logic à la Lukasiewicz and Tarski can do the job. Such a logic takes truth values to be like real numbers ranging between 0 and 1 – {0 ≤ x ≤ 1} – where, say, 0.5 means “half-true”, 0.1297763 means something like “a little bit true”, and so on. But even without going into the details of such a logic, it is not only clear that it bears the same amount of ‘technical costs’ as a three-valued logic, but it appears that it cannot avoid the existence of sharp thresholds either. To see this, let us consider a game Mark Heller played with God in Heller (1996) and let us see how it is relevant for our discussion. God stands at the top of Mont-Blanc and says “I am on Mont-Blanc”. She then starts descending in the direction of Chamonix and every time she takes a step she says “I am on Mont-Blanc”. The thing is, God is omniscient, she is capable of taking very, very, tiny steps, and she is always cooperative and telling the truth. Thus, there is some point where she will take a step and where she will stop saying “I am on Mont-Blanc”. If epistemicism is true, she will say “I am not on Mont-Blanc”. If supervaluationists have it right or if God is a friend of the metaphysical theory and of Kleene-like three-valued logics, then she will probably say “It is now indeterminate whether I am on Mont-Blanc or not”. What exactly she will say under metaphysicism and a fuzzy logic framework is not entirely clear – perhaps it will be something like “It is now almost true but a little less than true that I am on Mont-Blanc” – but what is clear is that there will be one precise point where she will stop saying “I am on Mont-Blanc”. In terms of a fuzzy logic framework this situation amounts to the passage from 1 (truth) to something else than 1 (something else than truth). Numerically speaking, so to say, this makes a tiny difference – the real number closest to 1 will be very close to 1 indeed – but conceptually speaking what we have here is a sharp and precise threshold between truth and something else than truth, which is close to it but which just is not truth. The conceptual and philosophical situation we find ourselves in here is one where we see that, by trying to avoid an allegedly inacceptable commitment of the epistemicist to the existence of sharp thresholds, the competing views and their logics actually only ‘postpone’ the problem, but do not avoid it. Epistemicism endorses the existence of sharp thresholds from the start, the others just stumble against it later on in their theory. Neither conceptually and philosophically, nor from the logical point of view, the competition thus does not seem to do a significantly better job than epistemicism on this point (that is, when it comes to trying to avoid sharp thresholds).3 Thus, ordinary objects are vague, and their vagueness then leads to the existence of sharp thresholds, which are bound to be arbitrary.

Eliminativism and Ordinary Objects  15

§4 I have focused in §2 above on the vagueness and the arbitrariness involved in the boundaries of ordinary objects, but there also is another, related, kind of arbitrariness involved in the distinction between ordinary and extraordinary objects. The friend of ordinary objects divides the world into chairs and whales, but what about the object that is composed of all of my chair except one leg, or the object that is composed of a whale except the tip of its tail? Such objects have a shape and a mass, they can be referred to, they can be seen, measured, weighted, touched, and so on, so why would they not count as objects? What about an object that is composed of all of my chair except one leg and my table except one leg? Granted, such objects are perceptually less salient than chairs and tables, they are less familiar, and not often referred to or talked about by us, but why would that have anything to do with their existence?4 One can say: they are here, whether we think of them or interact with them in any special way or not. I am not saying that there is an argument in favour of the existence of extra-ordinary objects. Such an argument could look like this: I have a perceptual experience of a chair and a table. I see a chair and I see a table. But there is also this object: “top half of the chair + left half of the table”. Let us call this object Peter. I see Peter as well as I see the chair and the table, it’s right there, in front of me. Thus, I have a reason – an “argument from experience” – in favour of the Theory of Unrestricted Mereological Composition. But this is a bad argument. What the defender of unrestricted mereological composition (who would embrace this type of argument) could want to say is that we see Peter, it’s there, just in front of us. We can see it, touch it, and so on. So, there is no reason for saying that only the chair and the table exist, while Peter does not. We experience all three objects equally well, so we should treat them accordingly as ontologically equal. Perhaps this is a simplified picture of what a friend of this line of argument could say, and of course not everybody would agree on the datum – namely, that we see Peter equally well as the chair and the table. Indeed, this line of argument can be reversed and can be taken to purportedly show the exactly opposite claim – that since we do not perceive Peter in the same way we do perceive the other two objects, we should not put it ontologically on a par with them. But, as Merricks (2001, p. 9) rightly remarks, these considerations about our perceptions of Peter, if taken as leading to the doctrine of unrestricted (or restricted) mereological composition (depending on what you think of the perceptual datum), is entirely misguided: our visual, olfactory, and tactile evidence would be exactly the same whether Peter existed or not. That is, whether the top half of the chair and the left half of the table compose an object is an ontological question, and a mereological claim, which is entirely

16 Eliminativism neutral with respect to what we see. Our visual phenomenal experience when contemplating this scene would be exactly the same whether Peter existed (as a sui generis object) or not, simply because our visual phenomenal experience is caused by light reflected by the objects in front of us, and it would be reflected in exactly the same way whether Peter were an object or not. The questions surrounding Peter’s existence should then be settled by other than perceptual means; for instance, by invoking philosophical arguments such as the ones we have seen above concerning vagueness and arbitrariness. There are many other cases of this kind populating metaphysical debates. For instance, about the statue and the lump case (more on this case below), Maclaurin and Dyke (2012, §5) say: We can talk about whether the statue and the lump of clay are really two objects or actually just one, but the singularity or duality of the statue and the lump is not something that can impinge on human experience. Again, I guess their idea is the same as above: whether there are two objects or only one, our visual phenomenal experience would be exactly the same. To bring the point closer to home, let us consider the case of eliminativism. Eliminativists claim that there are no chairs and no tables, only fundamental components arranged chairwise or tablewise. But, an objector could say, we see the table and the apple, they’re right there, just in front of us. An unfriendly objector could then point to eliminativism as to a doctrine that is almost absurd and entirely in conflict with any good common sense. A less unfriendly objector could want to stress that there is a strong tension between the eliminativist’s claim about the world and our experience of the world; the eliminativist, then, owes us at the very least a good explanation of why there is such a tension, and why we should reject what our eyes see rather than what the eliminativist wants to say. One way to go for the eliminativist could be to show that eliminativism is such a theoretically virtuous doctrine that we should indeed reject our intuitions and common sense beliefs about the world based on our experience of chairs and tables. But at this dialectical point, the more adequate strategy is to realize that in fact there is no tension at all. What we see and what the world is like from the ontological point of view are just two entirely orthogonal issues, simply because our visual phenomenal experience would be the same whether there were chairs and tables in front of us or whether there were ‘only’ fundamental components arranged chairwise and tablewise (see Merricks [2001, p. 8] and Sider [2013, §5]). Chairs and fundamental components arranged chairwise both reflect light in the very same way, and consequently the way the world ‘hits’ our perceptual system would be exactly the same in both cases. Thus, our phenomenal experience of the world, which is an experience as of chair and as of tables, is entirely neutral with respect to the existence of chairs, tables, or arrangements of fundamental components. As Merricks (2001, p. 8) puts it, fundamental

Eliminativism and Ordinary Objects  17 components arranged chairwise can do just about anything chairs can do: they can be seen, purchased, eaten, and so on. Thus, there is no tension at all between the eliminativist’s picture of the world and our phenomenal experience. Thus, as Sider (2001, p. 259) remarks, to anyone who understands the challenge [of eliminativism] and takes it seriously, any prior perceptual justification in favour of tables vanishes. Arguing against [eliminativism] on the basis of perception is no better than arguing that the atomic theory of matter must be false because tables look solid. But make no mistake: this is not an argument in favour of the eliminativist claim that chairs do not exist. Rather, the point here is that our phenomenal experience is neutral with respect to ontology – it is compatible with both a realist view about ordinary objects and an eliminativist view.5 (I’ll come back to this at length in the next chapter.)

§5 Let us sum up where we are. We have seen in §2 that ordinary objects are vague in the sense that they have vague boundaries and that this creates a problem of arbitrariness, regardless of what kind of theory of vagueness one chooses to embrace. In §3 we have then seen that there is also arbitrariness involved in the distinction between ordinary and extra-ordinary objects, and we have seen that our perceptual experiences are entirely neutral with respect to the existence or non-existence of ordinary objects (as well as with respect to the existence or non-existence of extra-ordinary objects). In short, if one wants to restrict composition to only the familiar ordinary objects, one will have to use a criterion to decide what counts as an ordinary object and what does not, and this criterion, as we have seen, will generate existential vagueness that will lead us to arbitrariness, and the distinction between the familiar ordinary objects and the other less familiar objects will be a brute arbitrary matter from the start. The unavoidable amount of arbitrariness involved in the claim that there are ordinary objects is then one good reason to reject their existence in the first place – a good reason to endorse eliminativism. But one could think that eliminativism also bears an amount of arbitrariness. One reason to say this lies in the fact that some eliminativists make exceptions – they eliminate everything except us (living entities, organisms, persons. . .). I’ll discuss various versions of this alleged need for exceptions in Part II in detail, and I’ll claim that no exception is needed. But there is another reason why one might think that some amount of arbitrariness cannot be avoided even by the eliminativist who does not endorse exceptionalism. Indeed, in order to make eliminativism compatible with ordinary thought and ordinary language, many eliminativists claim that there still is a sense

18 Eliminativism in which we can talk about ordinary objects. Heller (1990) contains an excellent defence of such a view, where the idea is to reconcile full-blown eliminativism with our language and our common sense view of the world (I’ll discuss in detail the role of intuitions in this debate in Chapter 2). Indeed, eliminativists like Heller claim that ordinary objects are conventional objects. They do not enjoy a mind-independent existence, and so, ontologically speaking, they do not exist at all, but they are conventional objects in the sense that we have conventions to the effect that we can say that such-and-such a bunch of fundamental components arranged in suchand-such a way are a chair. This idea comes together with a paraphrase strategy for ordinary language, since even eliminativists want to be able to talk about ordinary objects, and say things like “Mont-Blanc is a beautiful mountain”. The eliminativist claim is that for any practical purposes fundamental components arranged mountainwise can play the same role mountains would play if they existed, and thus we do not need to populate our ontology with mountains. An eliminativist à la Heller can then understand this claim in a conventionalist way: mountains, as ordinarily quantified over in ordinary talk, are conventional. In this view, strictly speaking, there are no mountains, but it is useful for us to call such-andsuch an arrangement of fundamental components a “mountain” and thus we have a linguistic convention governing the word “mountain” which is useful to us for many practical purposes. Thus, eliminativism does not make us abandon ordinary language. Indeed, the conventionalist eliminativist can say that the word “mountain” is about plurality of entities (say, a plurality of simples arranged mountainwise6) in the same way the term “the USS Enterprise crew” is about a plurality of entities, rather than a single entity composed of the crew’s members. The sentence “There are no mountains” is then true when it is a sentence uttered in Ontologese (the fundamental language of the metaphysician), but this is compatible with the sentence “There is a mountain in front of me” being true since it is a sentence in ordinary English. In Ontologese, it is true that there are simples arranged mountainwise in front of me, and this is all that is required for the ordinary English sentence “There is a mountain in front of me” to be true. This can simply be seen as a convention of ordinary English that whenever there are simples arranged mountainwise, it is true in ordinary English (but not in Ontologese) that there is a mountain.7 It is worth noting that this brand of eliminativism closely resembles the view put forward by some of the most prominent thinkers in the Buddhist tradition.8 Many Buddhist thinkers, like Nagarjuna (150–250 ad), interpret Buddha’s teachings by appealing to a distinction between two kinds of truth: conventional truth (samvrtisatya) and ultimate truth (paramārtha satya). Conventional truths are about conventional objects, and ultimate truths are about ultimately real objects. This brand of Buddhist thought explicitly embraces an eliminativist strategy: conventional objects (i.e. the

Eliminativism and Ordinary Objects  19 familiar ordinary objects of common sense) are eliminated. Thus, strictly speaking, conventional objects do not exist at all, but because it is convenient for us to talk about them we conventionally introduce names such as “mountain”, “chair”, and so on. The way this is often stated in the Buddhist tradition is to say that the conventional ordinary objects are mere fictions. Ultimate truths are those that describe how the real ultimately existing simple entities are arranged. Such ultimate truths then explain why we can say, at the level or ordinary conventional truths, that there are mountains, since there are ultimate real entities arranged in such a way that it is useful for us to say that there are mountains, even if these are mere fictions. Now, let us come back to the worry about arbitrariness. We have seen that everybody who thinks that there are ordinary objects has to deal with an unpalatable amount of arbitrariness. If an eliminativist wants to make exceptions, for instance for persons, like Merricks or Van Inwagen, the arbitrariness worry will arise here as well. But what about eliminativists such as Heller, Sider, or Nagarjuna, who do not make any exceptions? As we have seen, even this kind of eliminativism does not deny all possibility of talking about ordinary objects. But only such a hard-core approach – one that would deny even such a possibility – would entirely eliminate all worries concerning vagueness and arbitrariness (indeed, any such worries would simply not arise). Conventionalist eliminativists who do accept that there is some sense, even if only a conventional one, in which ‘there are ordinary objects’, have to accept that in this sense they are vague and/or arbitrary. One could then argue that the eliminativist simply has the same problems as her realist opponent. But of course, the arbitrariness that arises here is perfectly harmless. Indeed, the conventionalist eliminativist accepts from the very beginning that mountains and chairs are conventional objects, and any convention will be, by its very nature, arbitrary.9 As such conventional objects do not exist (that is, “conventional objects” must not be reified!), the ‘arbitrariness worry’ is no worry at all, since it merely states that conventions are arbitrary, which is correct – and entirely innocent. This is then where we are: everybody – except a very hard-core eliminativist who would reject even the possibility of talking about ordinary objects – has to deal with some kind of arbitrariness. In the case of nonexceptionalist eliminativism, the arbitrariness is perfectly harmless, but not so in the case of realist views about ordinary objects or in the case of exceptionalist eliminativism.

§6 In addition to the arbitrariness worry, the claim that ordinary objects exist also makes room for other kinds of trouble, as we shall see shortly. Unlike in the case of the arbitrariness issue, there are solutions available for these

20 Eliminativism other problems. The point to be appreciated here is then to see that not only eliminativism provides an elegant way to deal with such problems, but that it deals with them in a very simple and systematic way, while other solutions are much heavier pieces of metaphysical machinery. Consider the well-known case of the statue and the lump. Let us say that at t1 there is a lump of clay that at t2 an artist shapes into a statue. Thus at t2 a statue is created. Let us say that it persists until some later time t3 and is then destroyed. At some time after its destruction, at t4, the statue of course does not exist anymore but the lump of clay still does. The lump of clay persists from t1 to t4: it existed at t1 in some (let’s say, cubic) shape, then it was shaped into the form of a statue and, after the destruction, it was shaped again into some other form. A puzzle arises here because it seems that in the interval of time from t2 to t3, the lump of clay and the statue are one and the same object: they have the same shape, the same location, they are made up of the same components. But, if they were the same object, they should, following the principle of Indiscernibility of Identicals, share all their properties. But this is not the case: the lump of clay has, for instance, the property of being cubical at t1 that the statue has not. So, after all, the statue and the lump of clay are different objects. But then, how can they, from t2 to t3, coincide? How can they be made of the same components, how can they have the same spatial location, etc.? This is the well-known problem of coincident entities and there are wellknown solutions to it (or attempts at solutions, in any case). To my mind, the best of the non-eliminativist solutions available is a four-dimensionalist treatment of such cases. Let us have a quick look at what it amounts to. In short, according to the four-dimensionalist, the t2-part and the t3-part of the statue are numerically identical, respectively, to the t2-part and the t3-part of the lump of clay. For instance, the t2-part of the statue and the t2-part of the lump of clay do share all of their properties – they have no different historical properties such as being cubic at t1 because none of them existed at t1. But, of course, this does not entail that the statue and the lump of clay are identical: the lump of clay, for instance, has (temporal) parts at t1 while the statue does not. So they are not identical but they share identical temporal parts; one could say that they overlap. Thus, as Sider (2001, p. 6 and p. 152) points out, if we accept temporal parts, the puzzle with the ‘coincident’ statue and lump of clay is no more remarkable than the spatial case of two overlapping roads, one of them being a sub-segment of the other. There is then a modal analogue of the problem with coincident entities; an argument that was developed as an objection to four-dimensionalism by Peter Van Inwagen (1981, p. 92).10 Let us consider my pet lizard Jean-Luc. Some lizards can release a part or all of their tail when they are grabbed by a predator. Once the tail is broken off, the lizard quickly runs for shelter and is safe for the moment, leaving a squirming tail to confuse or distract the

Eliminativism and Ordinary Objects  21 predator. Let us imagine that such an adventure happens to Jean-Luc. The argument runs then as follows: Jean-Luc Jean-Luc-8

t0

t8

t10

(i) Jean-Luc actually lived for 10 years. (ii) There was a temporal part of Jean-Luc that existed from t0 to t8 – it lived for 8 years only. Let us call it ‘Jean-Luc-8’. (iii) Jean-Luc could have lived only 8 years. (There is a possible world W where Jean-Luc was killed at the age of 8.) (iv) So, Jean-Luc could have had the same temporal extent as Jean-Luc-8 has. (In the possible world W, Jean-Luc has the same temporal extent as Jean-Luc-8 has in the actual world.) (v) Jean-Luc-8 could exist in W with the same temporal extent it actually has (even if Jean-Luc does not exist for more than 8 years in W). (vi) Jean-Luc and Jean-Luc-8 would occupy exactly the same space-time region in W. (In W, Jean-Luc and Jean-Luc-8 would coincide.) (vii) But coincident entities are impossible. (viii) So, there is no such thing as Jean-Luc-8 (a temporal part of Jean-Luc). In Van Inwagen’s hands, the objection was thus raised against undetached temporal parts – that is, against the four-dimensionalist ‘worm view’. In reply, the four-dimensionalist can then say that four-dimensionalism is a view about temporal persistence, and so it is a solution to the standard nonmodal case of coincidence, but that when it comes to a possible worlds scenario, a theory about modal persistence is needed – and, there is a good one available, namely, modal counterpart theory.11 Thus, a four-dimensionalist can answer the objection by endorsing a counterpart-theoretical analysis of de re modal statements.12 Counterpart relations are similarity relations. Different objects can be similar in different ways, and so, there are different counterpart relations at hand. There can be, in our case, a lizard counterpart (L-counterpart) relation, and another temporal-part counterpart (T-counterpart) relation. Surely, those are different relations. The L-counterpart relation involves

22 Eliminativism different criteria for similarity than the T-counterpart relation. For instance, it matters for T-counterparts that they must have the same temporal extent – if not, they would not be similar enough to be T-counterparts. And it could be claimed that two objects are L-counterparts if and only if they are sufficiently similar maximal aggregates of temporal parts of lizards. So, in W, there is just one thing (and not a coincident entity) that has two counterparts in the actual world – under different counterpart relations. It has Jean-Luc as an L-counterpart and Jean-Luc-8 as a T-counterpart. As a lizard, Jean-Luc can have a smaller temporal extent, but not as a temporal part. This is not the place to discuss in detail, and even less a place to defend, a four-dimensionalist-cum-counterpart-theoretical way of dealing with the threat of coincident entities. What I want to do here is simply to point out that there are solutions available to this worry, and to note that they carry quite heavy ontological commitments, which need an independent defence. Eliminativists do not have to appeal to any additional metaphysical machinery – they can deal with such situations just by appealing to eliminativism itself. In short: no objects, no worries. Since statues, lumps of clay, or lizards, strictly speaking, do not exist, there is no threat of coincidence at all. If there are conventions to the effect that we can talk about statues, lumps of clay, or lizards this does not create any problem either, since there is no problem with the idea that conventions could somehow ‘coincide’ (whatever that could mean). This is then one of the places where we see how serviceable eliminativism can be. McGrath (2005) discusses such situations involving puzzle cases where eliminativists claim that their view can avoid them, and he asks whether such puzzles really disappear or whether they survive even under eliminativism13 – if so, eliminativism would lose the support of such arguments from puzzle cases. McGrath cites the Ship of Theseus famous puzzle case and argues that it will be just as difficult for the eliminativist to solve it as it is for the realist (see McGrath [2005, pp. 478–479]). Indeed, in his view, the paraphrases that the eliminativist will provide (such as “there are simples arranged shipwise”) specify ‘factual content’ (see McGrath [2005, p. 466]). Fair enough. Eliminativism alone will not be able to tell which of the two ships (that is, the ship that was reassembled using the original planks, or the ship that was continuously repaired by replacing planks) is the original ship – under the convention that there are ships. Eliminativism will avoid any metaphysical worries arising in this situation, and it will move the discussion to a much more harmless level of conventions. At this level, we can still discuss which of the two ships is the original ship, and in this sense the puzzle is yet to be solved – we will need to discuss identity criteria for ships persisting through time, and so on. But, at the ontological level, since there are no ships, there is no threat of coincidence of anything, and this is where eliminativism plays a role as a problem-solving tool. To my mind, (i) troubles with arbitrariness, (ii) the way eliminativism avoids problems with coincidence and persistence in a very efficient and

Eliminativism and Ordinary Objects  23 simple way, and (iii) the fact that we do not need ordinary objects (since there are simples arranged objectwise) are at the heart of the motivation for eliminativism. (The latter point is particularly well-developed in Merricks [2001] who defends in detail the claim that ordinary objects, if they existed, would be causally redundant. Merricks [2001, pp. 7–8]: “If things like statues and baseballs existed, everything they allegedly cause would be caused by their parts; if statues and baseballs existed, they would be – at best – wholly causally redundant. This, I argue, leads to their elimination”.)14 In the next chapter, I am now going to focus on the claim that eliminativism is unacceptable since it violates too strongly too many of our intuitions.

Notes 1 The philosophical literature on this issue is as abundant as it is varied. The main book to refer to remains to this day Van Inwagen (1990). 2 In reply to Lewis, Carmichael (2011) offers an intriguing answer. In his view, instead of a vague existence of a cloud, we should accept the existence of a proto-cloud. A proto-cloud is something about which it is vague whether it is the sum of the molecules. The proto-cloud exists in a non-vague, determinate, way. But, in Carmichael’s view, it is vague whether it is composed of the molecules – whether it has them as parts. In this sense, composition is vague, but existence is not. As he puts it: “the boundary between non-being and being is not the boundary between non-composition and composition” (p. 4). 3 Wright (2009) offers a possible strategy to deal with higher-order vagueness. He claims that higher-order vagueness is an illusion. 4 Heller (1990, p. 44): “[We] divide up the world into objects one way rather than another because doing so will serve our purposes better. . . . [T]here is little chance that the resulting ontology will be the true ontology”. Sider (2001, pp. 156–157): On [conservative views] the entities that exist correspond exactly with the categories for continuants in our conceptual scheme: trees, aggregates, statues, lumps, persons, bodies, and so on. How convenient! It would be nothing short of a miracle if reality just happened to match our conceptual scheme in this way. Or is it rather that the world contains the objects it does because of the activities of humans? This is an equally unappealing hypothesis. 5 “Indeed, it would seem that ordinary evidence is neutral over whether composite objects or merely appropriately arranged particles exist. Which hypothesis is correct is thus an open philosophical question, like the question of whether there exist holes and smirks” Sider (2013, p. 237). 6 In Part I, Chapter 4, we will see that eliminativism is also compatible with ontologies that do not postulate the existence of simples. The paraphrase strategy can then be modified accordingly. 7 This strategy is widely used by many eliminativists. See, inter alia, Heller (1990), Merricks (2001, p. 10), Sider (2013). 8 See, for instance, Siderits (2003, 2007) and Garfield, J. and Edelgass, W. (2009). 9 Heller (1990, p. 43): “What is characteristic of conventionality is arbitrariness. This arbitrariness, as I am understanding conventionality, is based on the fact that a convention is the creation of humans”. 10 Nice discussions of it can be found in Heller (1990, pp. 27–28) and Sider (2001, p. 218).

24 Eliminativism 11 Lewis (1968, 1986) provides the standard version under modal realism, and Heller (1998a, 1998b) provides a nice version under ersatzism. 12 Since Van Inwagen takes modal counterpart theory to be unacceptable, this amounts then, in his view, to be a reductio against four-dimensionalism. Sider (2001, pp. 218–222) mentions other possible solutions that do not require modal counterpart theory. 13 See also Bennett (2009) and Rettler (2018). For a discussion, see Tallant (2014), Brenner (2015b), and Thunder (2017). 14 See Le Bihan (2015) who extends Merrick’s overdetermination argument to the case of eliminativism about physical particles.

2 Eliminativism, Common Sense, and Intuitions

§1 Here is what Ted Sider says about the relationship between eliminativism and common sense: Mooreanism is utterly implausible. Why should the inherited prejudices of our forebears count for anything? We should, of course, trust common sense in some particular domain if there is independent reason to think that it is reliable about that domain. But there is no independent reason to think that common sense is reliable about whether there exist tables and chairs as opposed to there merely existing suitably arranged particles. (Sider [2013, p. 247]) Sider insists here on common sense. His view has two components: first, as we have just seen, common sense counts for little or nothing when it comes to fundamental metaphysical questions; second, once properly understood, eliminativism is not as absurd as it may seem at a first glance: [Eliminativism] may seem absurd. For the world of common sense and science consists primarily of composite entities: persons, animals, plants, planets, stars, galaxies, molecules, viruses, rocks, mountains, rivers, tables, chairs, telephones, skyscrapers, cities. . . . According to nihilism, none of these entities exist. But it is not absurd to reject such entities if one accepts their noncomposite subatomic particles. . . . Since I accept the existence of the particles, my denial of an object composed of them isn’t absurd. Denying that T exists in addition to a, b, and c is no more absurd than denying that holes exist in addition to perforated things, or denying that smirks exist in addition to smirking faces. Similarly, denying the existence of persons, animals, plants, and the rest is not absurd if one accepts subatomic particles that are “arranged person-wise” (to use Van Inwagen’s phrase), animalwise, plant-wise, and so on. (Sider [2013, p. 237])

26 Eliminativism This view is to be combined with the paraphrase strategy that eliminativists appeal to, as we have already seen in the previous chapter. Indeed, the eliminativist claims that instead of saying “There is a chair” we can say “There are simples arranged chairwise”. The former is a sentence uttered in ordinary English – and it can be said to be true in ordinary English, because the latter is a true sentence uttered in Ontologese, the metaphysician’s fundamental language. As Sider (2013) puts it, the former sentence is correct – it is, strictly speaking, false, because Ontologese is the fundamental language while English is just a simpler way of talking, but it is useful in successful communication in the circumstances where there are simples arranged chairwise. The conventionalist’s way to put it is to say that we have a useful convention that it is correct to say “There is a chair” in the circumstances where there are simples arranged chairwise. (More on this below in the last section of this chapter.) Later on in this chapter, I will discuss what one might call ‘philosophical intuitions’, that is, intuitions of philosophers who are experts in a given field (say, metaphysics) – but let us first continue to focus here on beliefs and intuitions of ‘ordinary folk’. Here is what Dan Korman says when he objects to eliminativism: Some may think that it counts for very little that hordes of nonphilosophers (who, let us not forget, have never encountered the arguments for eliminativism) find eliminativism implausible. (Korman [2009, p. 244]) All false beliefs of otherwise intelligent people stand in need of explanation. Understood in this way, even opponents of eliminativism should be able to explain why some reasonable and intelligent philosophers believe that there are no statues. But here the explanation is straightforward: these philosophers have been moved by powerful arguments that there are no statues. Most false folk beliefs admit of equally straightforward explanation. What is special about the mistakes alleged by eliminativists and other revisionary metaphysicians is that it is exceptionally difficult to provide any satisfactory explanation of their reasonableness. (Korman [2009, p. 245]) Korman is a friend of common sense and of ordinary objects. There is something peculiar in the way he puts his point here: he says that nonphilosophers find eliminativism implausible, but in the same breath he says that non-philosophers have never encountered the arguments for eliminativism – and, we should add, have actually never encountered eliminativism itself. But since non-philosophers have really no idea of what eliminativism is, how can they then find it implausible? I submit that they can’t. They simply have no opinion one way or the other, since they are entirely unaware of the subject matter. In this view, it’s not just that we cannot trust common sense

Eliminativism, Common Sense, and Intuitions  27 in the domain of fundamental metaphysics (see what Sider says above), it’s that there is nothing really to trust or to distrust at all. In this view, the claim that non-philosophers are “otherwise intelligent people”, as Korman puts it, is simply not relevant. To be able to know what eliminativism is, one has to go through at least some of the arguments in its favour (or against it) – one has to see eliminativism at work. This is indeed what I am trying to accomplish in this book, and it does take a whole book to do so. If one is entirely unaware of these issues, and if one is only told something like “eliminativism is the view that mountains, chairs, and your body don’t exist” without any more information/motivation/arguments, then one can perhaps finds this implausible, but this does not amount to finding eliminativism implausible, since in order to get the idea of what eliminativism as a metaphysical claim and as a piece of philosophical methodology is about, much more is needed. (If it were that easy to get to know what eliminativism is, it would be much quicker to teach our students.) So, here is the claim that I want to put forward and defend in this chapter: common sense, understood as a philosophically uninformed set of beliefs and intuitions, simply has nothing to do with, and nothing to say about, philosophical theories like eliminativism (just like it has nothing to say about scientific theories, such as, for instance, Darwin’s theory of evolution by natural selection or general relativity). The stance I take here is thus different from the claim that Horgan and Potrč (2008) put forward when they face the objection that their view (a brand of monism, which I will compare to eliminativism in the concluding chapter of this book) is counter-intuitive. They defend a view that prima facie flies in the face of common sense, but they take very seriously the need to reconcile this view with ordinary beliefs and intuitions. They say that “serious metaphysical inquiry . . . ought to pay great respect to deeply held common sense beliefs. . . . Philosophical theorizing should accommodate common sense as much as possible” (p. 4). They add that intuitions play a key role here: we take it that one serious obligation of a philosophical theory is to give a suitably respectful treatment of intuitively plausible judgements or statements that the theory rejects – all the more so if the judgments are as intuitively plausible as are judgments like ‘Of course there really are tables, chairs, and people’. (p. 121) To repeat it, my claim is different: I do not try to reconcile common sense with eliminativism. Rather, I maintain that common sense and eliminativism are entirely orthogonal. Korman seems to go in a similar direction in the sense that even if, in his view, there is a conflict between eliminativism and folk beliefs, the fact that there is such a conflict doesn’t count for much since one party is woefully uninformed. He says: “You can probably just ignore the objection that

28 Eliminativism people who know nothing about philosophy would disagree with your view. But you cannot ignore the objection that your view is open to counterexamples” (Korman [2015, p. 28]). Thus, in his view, even if we set aside the way ordinary folk intuitively react to eliminativism, we still can object to eliminativism by pointing to the allegedly obvious fact that there are tables and by producing an argument such as this one (from Korman [2015, p. 27]):1 (i) If eliminativism is true, then there are no tables. (ii) There are tables. (iii) So, eliminativism is false. Korman provides himself the reason why this argument is not satisfactory and why it cannot convince any eliminativist – it’s just that he does not take this reason seriously. The reason is, of course, that this argument is question-begging: Some will say that my arguments beg the question against revisionists. . . . To beg the question against someone in the dialectical sense is to present an argument against their view that takes as a premise something that they are already on record as denying. My arguments certainly do that. (Korman [2015, pp. 28–29]) The situation is similar here to the way Amie Thomasson (2015) defends her “easy ontology” strategy. In her view, existence questions are not deep questions of metaphysics requiring some special philosophical insight or formulation of a total ‘best theory’ to answer. . . . [E]xistence questions . . . are straightforwardly answerable by making use of our conceptual competence and (often) conducting straightforward empirical enquiries. (p. 20) As she puts it, “the result of the easy approach . . . is typically a simple realism about philosophically disputed entities, giving us the answers that yes, there are such disputed entities as numbers, tables and chairs, properties and propositions” (p. 22). As she develops her view, existence questions are answered by appealing to application conditions for ordinary language terms. A competent speaker who masters such terms correctly can thus provide the right answers. A competent speaker, for example, who has mastered the use of the noun ‘table’, is in a position to know that the term may be successfully applied in restaurants all over the country, and so to conclude that there are tables. (p. 113)

Eliminativism, Common Sense, and Intuitions  29 This “easy ontology” is a type of easy naïve realism, similar to Korman’s, that yields a rich ontology. Thomasson uses as the basis for her ontology linguistic and conceptual competence of ordinary thinkers. She clearly says that the entities that correspond to our linguistic and conceptual practices typically exist, independently of these practices (p. 217). She thus focuses on the use of the word “object” in ordinary English (p. 218). In Korman’s version of this type of realism, the main point lies in the key role intuitions play.2 This kind of intuitions in favour of ordinary objects plays a very traditional role in the construction of realist ontologies. More or less NeoAristotelian theories of substance, such as Jonathan Lowe’s (2006) or Hoffman and Rosenkrantz (1994, 1997),3 also rely on such intuitions. One of the core notions that lie at the heart of such theories of substance is the notion of a kind.4 What substances are is to be understood in terms of kinds being instantiated. Such ontologies are very hierarchical. The point is to provide a system of categories hierarchized in a relevant way. Let us consider the case of Hoffman and Rosenkrantz (1994, 1997) as an illustrative example. They provide such a system, which has four levels, and where the most fundamental distinction is between the abstract and the concrete:

Entity

A

B

C

D

Concrete

Abstract

Property Relation Proposition Event Time Substance Limit Collection Privation Trope

Material Object

Spirit

Where does this classification come from? In Hoffman and Rosenkrantz’s view, when doing metaphysics, we should start with an “intuitive notion of a hierarchy of levels of generality among ontological categories” (1997, p. 48). That is, they start with what they call “folk ontology” (1997, p. 1) and they add that “If entities of a certain kind belong to folk ontology, then there is a prima facie presumption in favour of their reality” (1997, p. 7). In their view, the ontological categories correspond then to the way common sense sees the world. Thus, substances – objects – are enduring (rather than perduring) things, they are continuants such as human persons, rocks, flowers, and houses. There also are events, places, times, properties, surfaces, edges, shadows, and holes.5

30 Eliminativism All of these views – Korman’s realism, Thomasson’s easy ontology, Hoffman and Rosenkrantz’s theory of substance – share the common idea that a rich ontology6 is a good thing, and one of the main reasons to believe this is that such an ontology stems from our intuitive common sense picture of the world. But, as we have already seen above, some friends of desert landscapes, such as Horgan and Potrč (who reject the aforementioned “levelinfested”7 ontologies), also insist on intuitions as being crucially important and they set themselves the task of accommodating intuitions even in their austere ontology. Many different sides agree, then, that intuitions play a key role in the construction of ontological theories. But what role is this, exactly? I claimed above that it is not the case that eliminativists disagree with common sense; rather, I want to say that common sense beliefs and intuitions are entirely orthogonal to the issue. We have already seen in the preceding chapter that our perceptual beliefs about the alleged existence of ordinary objects are entirely independent on the truth or falsity of eliminativism. The objector says that we see a chair in front of us and claims that eliminativism is absurd and entirely at odds with common sense. But as we have seen, and as I will argue in detail in this chapter below, what we see and what the world is like from the ontological point of view are just two entirely orthogonal issues. Our visual phenomenal experience would be the same whether there were chairs in front of us or whether there were ‘only’ fundamental components arranged chairwise (see Merricks [2001, p. 8] and Sider [2013, §5]). Chairs and fundamental components arranged chairwise both reflect light in the very same way, and consequently the way the world ‘hits’ our perceptual system would be exactly the same in both cases. Thus, our phenomenal experience of the world, which is an experience as of chairs and as of tables, is entirely neutral with respect to the existence of chairs, tables, or arrangements of fundamental components. I’ll defend this claim in more detail below. As with perception, I will defend in this chapter an independence/neutrality/irrelevance claim concerning ordinary intuitions. Furthermore, this independence claim is true of most metaphysical theories, like relationism/ substantivalism about time, bundle theory, substratum theory, perdurantism, endurantism, restricted/unrestricted composition, A-theory, B-theory, modal realism, ersatz possible worlds, and so on. Some of these views have been said to gain support (or the contrary) from common sense beliefs and/or intuitions. In what follows, I propose to examine some of these views and the role intuitions play here in more detail. In this way, I want to establish the claim that intuitions simply have nothing relevant to say here, and more specifically that they cannot establish statements like (2) in Korman’s argument.

§2 Endurantism is a good case study. As opposed to perdurantism, it is often said to be the intuitive view. But the ‘endurantist intuition’ – roughly, that

Eliminativism, Common Sense, and Intuitions  31 objects persist through time by being numerically identical and wholly located at all times at which they exist – is behind more than just endurantism. Indeed, it plays an important role in the motivation of some theories about the passage of time (namely, a family of A-theories), and some theories about the nature of the subject (namely, some forms of dualism, and some conceptions of the Self). This endurantist intuition is often taken in these cases to be a reason, or at least a motivation, to endorse one or the other of these metaphysical theories. Sometimes, it is even said to constitute a basis for an argument in their favour. I don’t deny that there is an endurantist intuition of this kind, but I do deny that it can play any philosophically significant role as a reason – and even less as a basis for an argument – in defending metaphysical theories such as endurantism, dualism, or the A-theory: it simply has nothing to do with the truth or falsity of these views, exactly as intelligent non-philosophers’ beliefs or intuitions have nothing to do with the truth or falsity of eliminativism. One often hears – even from perdurantists! – that endurantism is ‘the intuitive view’. This is often expressed in a negative form where the opposite view – one version or another of perdurantism – is said to be “obviously false” (Thomson [1983, p. 213]) or even unintelligible (Van Inwagen [1981, p. 90]). Thus, Mellor (1998, p. 86) objects to the four-dimensionalist perdurantist view as follows, clearly labelling the endurantist view as being the common sense intuitive one: No one else [than the four-dimensionalist] would say that only [temporal] parts of Sir Edmund Hillary and Tenzing Norgay climbed only a part of Everest in 1953. The rest of us think those two whole men climbed that one whole mountain, and that all three parties were wholly present throughout every temporal part of that historic event. In the same vein, Simons (2000, p. 62) protests that “outside philosophical seminars a four-dimensionalist never says ‘a two-hour phase of me last night was a waking phase’; he says, with the rest of us, ‘I was awake for two hours last night’ ”. These endurantist intuitions are often, even if not always of course, expressed using examples involving conscious subjects (e.g. Hillary, Norgay, philosophers). This is not entirely innocent and not very surprising, since our conception of the persistence through time of ourselves and the conception of our Selves are often strongly influenced by the same underlying thought: one entity, numerically one and the same, existing wholly at different times. (Substance) dualists, I take it, often have such a conception in mind and take it to be needed to explain the unity of the Self through time, the continuity of experience, the possibility of libertarian free will, moral responsibility, the alleged fact that qualia and thoughts need a bearer, or even the possibility of ‘life beyond death’. Few dualists make this link between the endurantist intuition and their dualist theory explicit in print.

32 Eliminativism Martine Nida-Rümelin (2009) makes a salient exception when she includes the endurantist idea in the very definition of what dualism is: according to subject body dualism the subject is an individual that exists wholly at any given moment of its existence and persists across time while changing its properties. According to subject body dualism subjects endure and they do not perdure. She then adds that one of the main motivations for subject body dualism “is based on intuitions concerning the identity across time” (Nida-Rümelin [2009, p. 191 and 194, my italics]). What Nida-Rümelin has here particularly in mind are considerations of thought experiments of fission, halfbrain transplants, and duplication, well-known from the literature about personal identity and persistence through time (I shall come back to the role of such exotic scenarios later), and in another article she explores the role that counterfactual thoughts (‘In such-and-such a situation, would this still be me?’) play in these thought experiments. She thus “describes the conceptual origin of an intuition and in doing so uncovers the intuition itself” (Nida-Rümelin [2011, p. 39]). I’ll have more to say on such ‘expert intuitions’ in I.2.§6 below. Velleman (2006, pp. 12–13, my italics) connects the endurantist intuition with a strong conception of the Self, even if not necessarily dualist, as well as with the further issue which I will also be interested in, namely, the passage of time: Whatever the future draws nearer to, or the past recedes from, must be something that can exist at different positions in time with its identity intact. And we have already found such a thing – or the illusion of one, at least – in the form of the enduring self. . . . I exist in my entirety at successive moments in time, thereby moving in my entirety with respect to events. As I move through time, future events draw nearer to me and past events recede. Time truly passes, in the sense that it passes me. Endurantism seems here to be the natural and intuitive way of understanding the passage of time – time’s passing appears here to require an endurantist conception at least of the Self. We experience ourselves as being one and the same entity at different times – this is the endurantist intuition as applied to ourselves, based on our experience of ourselves – and this numerical stability gives us then the sense of temporal passage since we experience this enduring Self as existing at different successive moments in time. Thus, the endurantist intuition about ourselves plays a role in the way we can understand the passage of time. Metaphysically speaking, the theory that claims to best account for there being genuine passage of time is the A-theory:8 it claims that time passes, and that time is not like space, contrary to what the B-theory9 says, namely that time is very much space-like – a further dimension in which things can be located by bearing relations of ‘being

Eliminativism, Common Sense, and Intuitions  33 earlier than’, ‘being later than’, and ‘being simultaneous with’ to each other. Thus, the A-theory has as one of its main motivations the idea that it alone can account for the fact that time passes, a fact that is taken to be apparent in our experience of time as of passing, and in Velleman’s terms, in our experience of ourselves as ‘being passed by time, while enduring’. Skow (2009, §4) makes a similar claim, although not related to endurantism: “I cannot survey all the motivations philosophers have had for the moving spotlight theory. But the motivation that I like best appeals to the nature of our conscious experience”. Here, we have an (alleged) argument in favour of the A-theory: we have an experience of time as passing (this is given to us by our phenomenology), but only the A-theory (as opposed to the B-theory) can account for temporal passage, and thus we conclude that the A-theory is a better theory (we draw a metaphysical conclusion from considerations about our phenomenal experience). When it comes to the passage of time, such statements about the role of our phenomenal experience are relatively common. Davies (1995, p. 275) could not be more explicit: Does our impression of the flow of time . . . tell us nothing at all about how time is as opposed to how it merely appears to us? . . . I find it impossible to relinquish the sensation of a flowing time and a moving present moment. It is something so basic to my experience of the world that I am repelled by the claim that it is only an illusion or misperception. It seems to me that there is an aspect of time of great significance that we have so far overlooked in our description of the physical universe. The claim involved here is strong: our experience is one of temporal passage, and so one concludes that there has to be temporal passage in the world. Perhaps even more strongly, Craig (2000, p. 138) claims that “[experience is] a defeater-defeater that overwhelms any B-theoretic arguments against the reality of tense”. Thus it appears that the fact that our experience of time is an experience of time as passing is taken to be a strong motivation and an argument in favour of the A-theory. Endurantism is not explicitly involved in these statements, but as Prosser (2012, §1 and §6) rightly diagnoses, the illusion of passage derives at least in part from the fact that experience represents objects as enduring, rather than perduring . . . in order for there to be a representation of change (‘alteration’) there must also be a representation of something that retains its strict identity through the change – the very same thing is first F, then not F. Our phenomenology is then an endurantist one: a key factor in time seeming to pass is that change is experienced as dynamic, and change is experienced as dynamic because the experience involves the representation of something enduring through the change.

34 Eliminativism It is this notion of a single entity passing ‘through’ a change that captures at least a very important element of the experience of temporal passage. Prosser himself does not think that there is an argument from these considerations about our phenomenology to the metaphysical claim that the A-theory is true. But, as we have seen, the claim is there, and the role of a kind of endurantist intuition or a kind of endurantist experience is pervasively relevant in metaphysical debates about endurantism, the Self, dualism, and temporal passage.

§3 We have seen in the preceding section that the passage of time, the A-theory, endurantism, as well as some conceptions of the Self and some versions of dualism all appear to be motivated by or (at least partly) rely on some kind of an endurantist intuition, apparently arising from our phenomenal experience in some cases. The idea here is that, on the one hand, we have an endurantist phenomenology as well as endurantist intuitions, and on the other hand, we have metaphysical views supported by this phenomenology and these intuitions. I will not deny that we have such a phenomenology or such intuitions. What I will deny is that these have any real philosophical significance when it comes to the metaphysical theories they allegedly motivate or support, exactly as in the case of eliminativism. Thus, I wish to defend here a general point about the relationship between such intuitions and metaphysical theories. These intuitions are not deep philosophical insights, and thinking otherwise can be misguiding and can lead one to embrace metaphysical theories for the wrong reasons. The origin of the endurantist (and connecting) intuition is simply to be found in our experience – a point that is already apparent in some of the statements from the preceding section. But our experience of ourselves as enduring, our experience of persisting objects, and our experience of temporal passage are merely products of the contingent way our perceptual system is built – similarly to what we have already seen when it comes to eliminativism and perceptual beliefs in I.1.§4 above. The endurantist intuition is supposed to lead us towards a metaphysical truth, but since this and other neighbour intuitions are contingently due to the way we humans are built, it cannot play this role. To start off, let’s talk about cinema. You watch Clint Eastwood’s Trouble with the Curve, where a player during a baseball game catches the ball and starts running. Thus, you have an experience of movement – you can see the player running. But of course, as you well know, there is no genuine movement at all, only a succession of static images projected at (usually) 24 or 25 images per second. This is a phenomenon standardly exploited in the

Eliminativism, Common Sense, and Intuitions  35 film industry where apparent motion in a movie is generated by a series of static images succeeding each other at an appropriate speed. Thus, we have an illusion of movement, and interestingly, this illusion has nothing to do with our knowledge or ignorance of whether there is genuine movement or not, since the appearance of movement persists even when we know that it is illusory. Now, what about real movement, and real change, outside the cinema? Suppose you’re at a real baseball game, and you see a player catching a ball and starting to run. As she moves along, she constantly and continuously changes her position and her shape. Not only is she running, you can see her running. Thus, at least prima facie, you can have a visual experience of change and movement. One striking feature of the experience we have of the player’s movements is the continuous way in which she changes her position and shape. We are in some sense aware of the fact that at a certain time she is in such-and-such a state and that at a subsequent moment she is in a different state, and so on – but this is not how things appear to us. We do not have a succession of experientially isolated instantaneous experiences of different states of the player succeeding each other; instead, we have a continuous experience of her running. This feature of our experience has not only been insisted upon famously by James (1890, p. 629) who says that A succession of feelings, in and of itself, is not a feeling of succession. And since, to our successive feelings, a feeling of their own succession is added, that must be treated as an additional fact requiring its own special elucidation but also, for instance, by Husserl (1964, p. 31) who claims that “The duration of sensation and the sensation of duration are different. . . . The succession of sensations and the sensations of succession are not the same”. These considerations lead proponents of some of the most influential accounts of temporal experience to embrace the notion of a ‘specious present’ as an explanatory tool in order to account for James’s and Husserl’s observations and, more generally, for the continuous and unfolding aspect of our experiences such as our seeing a person running.10 Our experience of movement and change can be better understood by considering three types of situations, i.e. three different types of experiences of moving objects: very slow movement, very fast movement, and neithertoo-slow-nor-too-fast movement. The hour hand case is a standard example of the first kind. The hour hand on a mechanical watch moves so slowly that we do not see it moving. We can see that it has moved by having a succession of experiences of the hour hand and by remembering its position from one experience to another. But we do not have an experience of movement or change, because the hand’s continuous movement is too slow for us to be able to perceive it. Our

36 Eliminativism capacities to notice change and movement have a lower limit and anything that moves too slowly will not be registered by our perceptual system as moving. On the other side of the spectrum of cases of experience of movement, there is the ‘moving dot’ example. Take a dot on a computer screen that moves so quickly along a circular path (more precisely, in this situation, some pixels go on and off on the circular path) that we don’t see it moving: we see a complete circle and we don’t see any dot at all. Thus, there is a dot that moves very quickly, but this is not how things appear to us, even if we know about it. We don’t see a succession of dots, and we don’t see a moving dot either – the only thing we see is a circle. In this case, then, we don’t have an experience of succession, and we don’t have a succession of experiences either. Indeed, our capacities for detecting and perceiving movement have an upper limit: any movement that is too fast will simply not be registered by our perceptual system as a movement. Thus, in the hour hand case, due to the lower limit of our perceptual faculties, we have a succession of experiences but no experience of a succession, and due to the upper limit of our perceptual faculties, in the moving dot case, we have neither a succession of experiences nor an experience of a succession. We only have an experience of movement and change in cases like the baseball player case. Indeed, the player moves just fast enough, but not too fast, for us to see her moving. The speed of this moving object is between the lower and upper limits of our perceptual system’s recognitional capacities, and thus it is interpreted as moving. The perceptual system uses here the inputs it gets from the world in order to provide an experience as of movement. The cinema is a good example of the same kind, where this feature of our perceptual system is used to create an illusion of movement by static images following each other at an appropriate speed. Thus, only a succession of experiences that are appropriately linked gives rise to an experience of succession, that is, of movement and of change. “Appropriately” means here that the successive experiences must occur at an appropriate speed. Robin Le Poidevin seems to hold a view close to this: The perception of E is accompanied by the memory of C. But because of the proximity of the perceptions the experience of succession is not consciously inferential. The conjunction of the very recent memory of C with the perception of E gives rise to an experience of ‘pure succession’. (Le Poidevin [2007, p. 91, my italics]) When Le Poidevin says that the phenomenon here is not consciously inferential, he comes close to the account provided by Paul (2010) who defends “an account of how temporal experience could arise from the way the brains of conscious beings experience and interpret cognitive inputs from series of static events” (Paul [2010, p. 7]). She provides some well-tested

Eliminativism, Common Sense, and Intuitions  37 and well-documented experiments from cognitive science and experimental psychology that fit well the claim concerning the lower and upper limits of our perceptual system’s recognitional faculties, one of which is the following different variant of a moving dot experiment: I will first describe an interesting and empirically well-documented fact about our experience – namely, the illusion we have when, first, one small dot is shown on the left-hand side of a computer screen and then, very quickly, that dot disappears and a small dot is shown on the right-hand side of a computer screen. Then, the right-hand dot disappears, and the left-hand dot appears, again and again, in rapid succession. Even when we are told that what the computer is actually doing is merely blinking different dots on alternating sides of the screen, as long as the succession is rapid enough and spatio-temporally close enough, the effect is that we have the illusion of the dot moving back and forth across the screen. This is what cognitive scientists usually describe as ‘apparent motion’. (Paul [2010, pp. 15–16]) Here we have a good experiential basis for the claim discussed above. Paul highlights here a crucial point: it is prior to any conscious experience that our perceptual system, on a psycho-neurological but not on a phenomenal level, interprets the successive stimuli (say, of the running baseball player), which constitute the inputs that the perceptual system receives ‘from the world’, in such a way as to produce an experience of movement – that is, an experience of succession, generated by the perceptual system from a succession of inputs.11 This allows us to understand why we don’t have an experience of a dot in the ‘circle case’, or why we don’t have an experience of singular static images in the case of cinema – our perceptual system ‘computes’ for us an experience as of movement precognitively before we become aware/conscious of it (before, so to speak, it reaches the phenomenal level of our experience). As we have already seen, this is entirely independent of whether we know that there is a genuinely moving object that we are perceiving (like a baseball player) or whether we know that we are facing an illusion of movement (like in the case of the moving dot on the computer screen or in the case of cinema). What we have here is an understanding of our temporal experience in two distinct stages, one of them being precognitive, unconscious, and non-phenomenal (where our perceptual system gets ‘brute’ inputs from the world and interprets them), and the other being phenomenally accessible to us (where we have – or not, depending on in which case we find ourselves – an experience of movement, change, and succession). Thus, we see that the way we perceive movement and change is largely due to the way our perceptual system (especially, our brain) is built. This is a crucial point. Indeed, what we realize here is that our phenomenology – the way

38 Eliminativism we perceive the world – is one thing, and metaphysics – the way the world is – is another. Think of the A-theory and the B-theory, in the light of what we learned about our experience of temporal passage. Our phenomenal experience of time as passing was taken to be a reason in favour of the A-theory. But, as we have seen, our experience can very well be illusory – indeed, we have an experience of movement even when there is none (like in the cinema) and we lack an experience of movement even when there is one (like in the case of the hour hand). We have also seen the reason for this: our brain does the job before an experience of the world becomes available in our consciousness. Our perceptual system, precognitively and preconsciously, interprets the stimuli it gets from the world to produce an experience as of movement and as of change, independently on whether there is genuine movement or genuine change in the world. Suppose, now, that we find ourselves in an endurantist A-world with genuine flow of time and objects that endure. In such a world, our phenomenology is indeed rightly endurantist and A-like. Suppose, then, that we find ourselves in a perdurantist B-world which consists in a series of numerically different stages or temporal parts related by relations of anteriority, simultaneity, and posteriority. Would our phenomenology be any different? It would not. Exactly as we have an illusory experience of movement in the cinema, and even when we know that there only is a series of static images projected one after another, we would (or at least could) experience a B-world as a world with movement, change, continuity, and temporal passage – this just is how we are made. This is not something we choose to do: independently of what we know, or of what we want, we do have experiences of movement, continuity, and passage of time, even in cases where there are no such genuine phenomena, for example, in a B-world. This is not an argument in favour of the B-theory. But it is an objection to those who think that there is a link from our endurantist A-like phenomenology to metaphysical theories such as the A-theory or endurantism (more on the latter in the next section). Such a way of argument is simply wrong, since again, our phenomenal experience is what it is in a way that is neutral with respect to the two competing metaphysical theories.

§4 Until now, I focused on temporal experience, and the experience of movement and of change. This was directly relevant to the issue concerning the A-theory (which lost one of its main motivations on the way) but it already was, even if less directly, also relevant to the claim that our phenomenal experience of the world is endurantist-like. Indeed, what we saw above is that due to the way our perceptual system is built, we tend to experience the world as containing genuine movement and change of one and the same object rather than as containing successive numerically and qualitatively distinct objects. This has been well-documented and well-tested by Scholl

Eliminativism, Common Sense, and Intuitions  39 (2007) who discusses a series of ‘multiple object tracking’ experiments, where participants are presented with moving objects, some of them being labelled as ‘targets’, and where they are asked to track these objects while they move around. This is tested both in cases where the objects (usually, dots or other figures on a computer screen) move along continuous paths, and in cases where there are occluders that temporarily mask them (they ‘continue’ their paths while being hidden behind these occluders):12 visually speaking, the moving objects’ trajectories are discontinuous and the objects temporarily disappear when they are hidden behind occluders. What such experiments show is that we always have the tendency to perceptually privilege identity through time and spatio-temporal continuity, even for objects that ‘disappear’ on one side of an occluder and ‘reappear’ on the other side. To see this, Scholl discusses not only cases of objects that have discontinuous paths, but even cases of objects that, at times, disappear behind an occluder and emerge from the other side completely changed (in their shape, size, and colour): even in such cases, when what emerges from the other side is a very different thing (qualitatively) than what entered behind an occluder, it is still interpreted as being one and the same object that changed. Spatio-temporal continuity and numerical identity simply trump qualitative difference. The only case which participants tend to interpret as a case of two different objects is the one where a delay is introduced – the second object emerges from behind the occluder later than what one would have expected given the speed at which the first object entered behind the occluder. Thus, what these experiments show is what Scholl (2007, p. 573) calls a “temporal same-object advantage” – that is, an endurantist rather than perdurantist way our perceptual system presents us with the world. As already discussed before in the case of temporal passage and our experience of movement and change, it is here also relevantly true that we cannot help but having this kind of phenomenology – it is, as before, independently of whether participants know that there is one object or several objects that they have the tendency to privilege continuity and identity over qualitative difference. As Scholl (2007, p. 569) puts it, “the principle of spatio-temporal continuity . . . seems to be wired into our minds in a deep way, controlling how we experience the world”. (Besides, in a general way, participants are well aware that they are facing a computer screen where, as in the case of cinema, there is no genuine movement or genuine identity at all – rather there are simply dots or other shapes on a screen illuminated one after another, to give the illusion of, say, a dot moving.) These experiments and findings reinforce then the conclusions drawn at the end of the preceding section. Our perceptual system (especially, our brain) is built in such a way that it tends to interpret the various stimuli it gets from the world in an A-like endurantist fashion, even in cases where it is clear that there is no genuine identity between the objects perceived. This just is how we are built, and one can speculate about what type of evolutionary advantage this gives us – perhaps it just is easier for us to have

40 Eliminativism a representation of one single object enduring through time than a representation of a series of different objects (there might be some economy that our brain achieves in this way and helps us then to react more quickly to our environment – see Prosser [2012]). But this is beside my point: I am not interested in why we have the phenomenology and experience of the world we have, but simply in the fact that it is so, and that it is so in a very possibly and plausibly systematically illusory way, that is, even if we live in a perdurantist B-world. Of course, this does not mean that we indeed do live in a perdurantist B-world, but it does not mean that we live in an endurantist A-world either. A series of successive different temporal parts in a B-world and one single enduring object in an A-world would indeed produce in us, given the way our perceptual system is built, the same phenomenal experience. The overall point then is: the two issues – phenomenology and metaphysics – are simply orthogonal.

§5 In the preceding sections, I vaguely mixed talk about our experience (our phenomenology) and talk about our intuitions. It is now time to get clearer about the relationship between experience and intuitions, and this amounts to a further claim I want to make: a relevant part of the so-called endurantist intuition is the result of the phenomenal experiences we have of the world. If we think that an endurantist intuition is a deep philosophical and conceptual insight, we are wrong. It is no more than an expression of how we perceive the world, and the way we perceive the world is contingently due to the way our perceptual system is built, so there is no “conceptual” or “philosophically significant” insight, but merely a sophisticated description of our sensory experience (our phenomenology). As a consequence, such an intuition cannot play any argumentative work in favour (or against) any metaphysical theory, as we have already seen above. Thus, metaphysicians should pay attention to experimental psychology and cognitive science findings, such as Scholl’s (2007), in order to help to identify which of our so-called intuitions – that we so much work with when doing metaphysics – are the possibly misleading ones. Similar claims have been raised in other philosophical domains as well, for different reasons. For instance, Machery, Mallon, Nichols, and Stich (2004) offer persuasive evidence that semantic intuitions are culture-relative, thus making it difficult to rely on such intuitions when developing a philosophical theory of reference. Bartel (2017) offers a detailed critical discussion of the role intuitions can(not) play in the philosophy of music. When it comes to thought experiments and intuitions involved in these thought experiments, Swain, Alexander, and Weinberg (2008) show that depending on the order in which participants are presented with the thought experiments, their responses vary – thus, intuitions are shown to be easily manipulated (the intuitions

Eliminativism, Common Sense, and Intuitions  41 involved in this case were epistemic intuitions, arising from Lehrer’s [1990, pp. 163–164] Truetemp Case). As they put it (§1): Intuitions track more than just the philosophically relevant content of the thought experiments; they track factors that are irrelevant to the issues thought experiments attempt to address. . . . Such sensitivity to irrelevant factors [socio-economic status, cultural background, or the order in which one considers various thought experiments] undermines intuitions’ status as evidence. Evidence so unstable risks being discounted as not being truly evidence at all. Furthermore, given that intuitions vary in these ways, there is unlikely to be a fixed set of intuitions about a particular thought-experiment to which we can appeal. The point here is that the case of ‘the endurantist intuition’ that I focused on above is far from being an isolated one. The so-called endurantist intuition is misleading because it might present itself as having a conceptual source and as being a philosophical insight, while it actually merely describes our phenomenal (or introspective, in the case of the Self) experience. Other ‘metaphysical intuitions’ suffer from the same defect, as we have also seen in the case of eliminativism and in the case of restricted/unrestricted composition, and as the studies I just mentioned show, some semantic and epistemic intuitions are unreliable for reasons that are different (that is, not related to our phenomenal experience) but similar on a general level: they arise from the ways we humans are variously biased (perceptually, culturally, or otherwise). Eli Hirsch raises a similar point concerning our sense of unity and similarity, involved in our understanding of persistence through time. He says: we think of the world in terms of our criteria of bodily unity because we are innately disposed to think in this way. . . . [A] sense of bodily unity is part of our inborn constitution, and this is what determines us to interpret our experience in the way we do. . . . [A]s a matter of empirical fact, our sensory fields are ‘naturally’ and ‘spontaneously’ organized in terms of distinctive kinds of units. (Hirsch [2010, pp. 6–7, see also p. 20]) According to Hirsch, we have an innate sense of similarity (and unity) and this is what makes us divide up the world in objects – we have an innate disposition to classify things in the way we do rather than in another way. This echoes the claims and arguments we have seen above – remember for instance Paul’s claim that our perceptual system, precognitively and preconsciously, interprets the stimuli it gets from the world to produce an experience as of movement (independently on whether there is genuine movement or genuine change in the world). In this way, when it comes to ordinary material objects, the way we intuitively conceive of them is based

42 Eliminativism in the – contingent – way our brains and minds are made. Similarly to what we have seen above in the case of Scholl’s experiments, Hirsch insists that “the primitive power of an object, especially of a moving object, to fix our attention is unmistakable” – which is a contingent fact about the way we are built (Hirsch [2010, p. 23]). This contingent fact is central to the way we experience the world. Hirsch uses this example to illustrate the point:

On the first of the two figures above, we immediately notice that there is a portion that is shaped like the letter “H”. On the second figure, there also is an H-shaped portion, but we don’t notice it, at least not immediately, and not in the same salient way. We have the tendency to focus on the first H-shaped portion but not on the second. In Hirsch’s view, this is analogous to the way we perceive the world around us: we have a natural innate – and contingent – tendency to focus on some space-time portions rather than on others. The intuitive manner in which we divide up the world, and in which we come up with an ordinary-objects-friendly ontology is then rooted in our natural tendency to exclude the portions of the world that we do not perceive in the particular way in which we do not perceive the H-shaped portion on the second figure. (Note that Hirsch himself is a friend of ordinary objects, but mainly because he believes that the disputes involving ordinary objects and their persistence through time are merely verbal. I discuss this meta-ontological matter in Meta-metaphysics [Springer, 2016].) To repeat, the main point that arises as a cumulative effect of all of the considerations we have seen in this chapter is that our so-called ‘endurantist intuitions’ are misleading in the sense that they might present themselves as

Eliminativism, Common Sense, and Intuitions  43 having a conceptual source and as being deep philosophical insights, while they in fact merely arise from our phenomenal experience, which is itself rooted in the contingent way we are built. Our intuitions and our ways of dividing up the world into ordinary objects are then, I submit, entirely orthogonal to claims concerning fundamental ontology, like eliminativism.

§6 A possible reaction a defender of the use of intuitions in philosophy could have, would be to insist that philosophers, since they are experts, have better intuitions than any simply pre-theoretical intuitions of people who are, say, presented for the first time with a thought-experiment or with a ‘multiple objects tracking’ task. Nida-Rümelin, in the articles I cited in §2, typically bases her arguments and observations on detailed considerations of various thought experiments of (Parfit-like) scenarios of fission, and what she wants to do is to make us see an intuitive and conceptual point, that we can get to by thinking hard about such cases. The idea here is not to count as an intuition the first thing that springs into our minds when we are first presented with such cases, but rather to think about them and discover a kind of ‘conceptual intuition’ – a philosopher’s, that is an expert’s, intuition. But it is far from being clear that being experts gives us, qua philosophers, any advantage. One way to raise this critical point is mentioned by Scholl (2007, p. 585) who focuses on the use of thought experiments involving imaginary examples such as, precisely, cases of fission of objects or people. He says: This analysis might seem to lead us to question the use of such outlandish examples in cases involving object persistence. . . . The outlandishness per se may not be problematic, but these scenarios may violate assumptions about the world that are made in a reflexive way by our perceptual and cognitive processes. This is consistent with previous discussions of the danger of such thought experiments. . . . Such perceptual illusions may, in the service of metaphysical theorizing, become cognitive illusions that can lead us astray. In short, the outlandishness of such scenarios may introduce even more possible bias than more mundane experiments. Weinberg, Gonnerman, Buckner, and Alexander (2010) object to the claim that philosophers, as experts, have intuitions any better than the layman’s for a similar reason: expertise, in general, is not a cure for cultural or other biases, and there is no reason to think that philosophers are an exception. A higher level of expertise, as they argue, can actually be a source of more ways to go wrong, when appealing to one’s expert intuitions. Suppose, for example, that a philosopher has at her disposal better conceptual schemata than a non-expert, thanks to her training. While such schemata might indeed help to avoid some of the misleading

44 Eliminativism folk intuitions, working with such schemata and concepts is working with a heavy cognitive machinery which might bring new possibilities of biases on its own – one may be introducing a new source of error without realizing it.13 Besides, there is no reason to think – and even less is there a proof – that philosophers are immune to the defects of other scientists’ or non-experts’ thinking, like overgeneralization, overconfidence, belief bias, or belief perseverance. While believing that we, as philosophers, have finely tuned deep intuitions, better than the layman’s, we might actually only be reformulating early intuitive impressions.14

§7 To sum up, we have seen that there is no perceptual evidence that would count pro or contra eliminativism, and we have seen that there is no tension between common sense beliefs about chairs and eliminativism, because common sense simply has no idea at all of what eliminativism is in the first place. We may have perceptual experiences as of chairs (when we face simples arranged chairwise) and we may have perceptual experiences as of temporal passage, and this may give rise to an ordinary belief that there are chairs or that time passes, but all of this has simply nothing to do with and nothing to say about the truth or falsity of the metaphysical claim that there are chairs or that time genuinely passes. This latter claim needs to be established via other means, namely by doing philosophy and/or science. We have seen, in the case of endurantism/perdurantism and the case of A-theory/Btheory, that the ‘endurantist intuition’ is no more than a description of the contingent way we perceive the world, given the perceptual apparatus and the brain we have. In short, bringing these points together, we learn here that we are contingently made in such a way that we divide up the world into enduring chairs, mountains, trees, and so on, but this does not constitute any argument in favour or against metaphysical theories such as eliminativism or endurantism/perdurantism. Intuitions and common sense worldviews cannot play any argumentative work here. They are simply entirely orthogonal to what metaphysicians do. In this sense, there is no argument against eliminativism on the basis of intuitions or folk beliefs. This brings us back to the discussion concerning ordinary language and eliminativism. In Chapter 1 as well as earlier in this chapter, I used the distinction between ordinary English and Ontologese – the fundamental language of the metaphysician – to say that the sentence “There are no chairs” is true when it is uttered in Ontologese but that this is compatible with the truth of the ordinary sentence “I am sitting on a chair” since this is a sentence in ordinary English. (See, for instance, Sider [2013].) One way of understanding the situation here is to take the idea of different languages seriously and to claim that ordinary English and Ontologese are genuinely different languages and that ordinary folk and metaphysicians are simply talking past each other when they say contradictory things about the existence of chairs.

Eliminativism, Common Sense, and Intuitions  45 There is no contradiction because the two contradictory claims are uttered in entirely different languages. Dan Korman objects to this strategy on the grounds that the separation between the two allegedly different languages is too strong to be plausibly useful. Indeed, he claims that when we consider the arguments from vagueness, arbitrariness, or material constitution we realize that there is a tension in our beliefs (see Korman [2015, p. 50]). As he remarks, if we are convinced by such arguments, we are then changing our minds and not just starting to speak a different language. Thus, when we discuss the truth or falsity of eliminativism, we are (and we should be) discussing it in plain ordinary English – again, this is why the aforementioned arguments (can) have a strong grip. If the metaphysician insists that she speaks an entirely different language than the folk, then when she denies the existence of chairs she does not say that chairs don’t exist, but she only says, as Korman (2015, p. 67) puts it, that chairs don’t existO – where “existO” is a word in Ontologese that refers to existenceO but not to existence. But, Korman adds, we want to know whether chairs exist, not whether they existO – indeed, it might be entirely trivial that chairs don’t existO, but the interesting point is to establish whether they exist or not. However, we do not have to embrace such a strong view concerning the ‘two languages claim’, even if we keep in mind the results concerning perception and intuitions from this chapter – i.e. the point that ordinary intuitions based in perception are entirely orthogonal to the fundamental metaphysical issues. Here is a way to put it, that perhaps everybody, including Korman, can accept.15 Let’s accept that we all – metaphysicians and folk included – speak plain English. But – very importantly – when we say something in the ontology room (that is, when we speak Ontologese, in this sense), we carefully avoid the biases that come from the contingent way our perceptual apparatus is built and from the resulting ordinary intuitions we have (see above) and we bear in mind the arguments from vagueness, arbitrariness, material constitution, and so on. In short, we are speaking seriously, and in an informed way – that’s what speaking Ontologese is about. Speaking Ontologese is not about changing the subject or changing the language. It’s about changing the standards of the discussion. When I tell my 5-year-old son to stop jumping on the kitchen chair, I am not having a discussion involving high philosophical standards, exactly as I am not having a discussion involving a scientific standard according to which the chair is made of – say – vibrating quantum strings and a lot of empty space. Similarly, when I have a discussion with Dan Korman and I say to him that there are no chairs, I do use a high philosophical standard for our discussion. In both situations, I speak plain English (well, I speak French in the case of my son). But in the first type of situations, I systematically – and harmlessly – adopt the useful folk fiction/convention that there are chairs, etc. Once I become aware of all of the arguments from the preceding chapter and this chapter, I become aware of the fact that fundamentally speaking there are no chairs, and that my perception as of chairs has nothing to do

46 Eliminativism with this, and so on. But this does not prevent me from acknowledging in everyday life the usefulness of the fiction/convention that there are chairs – exactly as, similarly, the fact that I have the everyday intuition that there are chairs does not prevent me from acknowledging the fact that there are no chairs, strictly speaking. When it comes to truth, it is only when we speak Ontologese – that is, when we speak in a serious and philosophically informed way – that we get to it. So, strictly speaking, it is simply true (in any language) that there are no chairs, and that’s it. However, nothing forces us to ‘speak fundamentally seriously’ all the time, and it is often useful to say something that is false while being, as Sider (2013) puts it, correct – that is, useful in successful communication in the circumstances where there are simples arranged chairwise. The Heller-style conventionalist’s way to put it is to say that we have a useful convention according to which it is correct to say “There is a chair” in such circumstances. Thus, ordinary folk speak usefully and correctly, even if what they say would be false if they were speaking with Ontologese standards in mind (and recognized as such by them once they stop being ordinary folk and they become aware of all the arguments in favour of eliminativism). In the conventional sense – and only in the conventional sense – it is correct to say that there are chairs, and that’s all common sense really needs.16

Notes 1 Korman says here “nihilism” and not “eliminativism”, but this does not matter for our present purposes. 2 Korman (2015, p. 29): “What reasons do we have for accepting [(2)]? As I see it, intuition plays a key role in explaining both why it is reasonable to accept [this premise] and also why we do accept [it]”. 3 Thomas Sattig’s ‘perspectival hylomorphism’ is another interesting example of a friend of common sense ordinary objects. In his view, this is even the first desideratum for a philosophical theory of objects to satisfy: “A philosophical account of ordinary objects should aim to preserve our common-sense conception of the latter” (Sattig [2015, p. vii]). His view is a mixture of classical mereology and Aristotelian substance theory: The thesis, in a nutshell, is that ordinary objects lead double lives: they are compounds of matter and form; and since their matter and form have different qualitative profiles, ordinary objects can be described differently from different perspectives. Perspectival hylomorphism carves a middle way between the two accounts that have dominated traditional metaphysics of material objects, namely, classical mereology and Aristotelian hylomorphism. It is a fundamentally classical-mereological framework with an Aristotelian twist. . . . A philosophical account of ordinary objects should aim to preserve our common-sense conception of the latter. The task of saving the appearances, however, has proven difficult. For our familiar worldview faces a range of hard problems: it is riddled with paradox and clashes with plausible principles from metaphysics. The orthodox position in contemporary discussions is that these problems show our familiar worldview to be defective and in need of substantial revision. What recommends perspectival hylomorphism is that it does a better job than its rivals in preserving our folk conception of the

Eliminativism, Common Sense, and Intuitions  47 world in the face of a range of such problems. The unified type of response in the proposed framework is compatibilist: seemingly inconsistent judgements about ordinary objects are really consistent because they manifest different perspectives on the same double-layered objects. (Sattig [2015, p. vii]) 4 The idea is that substances fall under kinds. For instance, you fall under the kind ‘human being’, your dog falls under the kind ‘dog’, and so on. The notion of ‘falling under’ in not to be understood in extensional terms, which means that kinds are not like sets. Sets are individuated by their members, they are what they are because of the members they have. Kinds, as it is often put, are prior to their members. If you ask ‘What is this object?’, the best answer is the kind of the object at hand. To be a human being, or a dog, is to instantiate the kind ‘human being’ or the kind ‘dog’. When a kind is instantiated, this is enough for there to be an object of that particular kind. This is how objects get their identity conditions and their numerical diversity – for instance, if the kind ‘dog’ is instantiated four times, then there are four dogs. The instantiation of a kind is all it takes for objects such as dogs to exist. In this sense, objects – substances – are not built from more fundamental elements; rather, they are understood as basic entities that simply instantiate their kind. In this view, the notion of a kind is thus central to the understanding of the notion of substance. 5 Among the central intuitions concerning substances, in Hoffman and Rosenkrantz’s view (as well as in Lowe’s view), there is the intuition that substances have features. Such features are thus features of substances, unlike under the bundle theory for instance, where they are parts of substances. The features of substances can be accidental or essential, and substances can undergo intrinsic change. Also, their very existence is contingent – substances can be created as well as destroyed. The system of ontological categories on the figure above is organized in four levels of generality. There is a hierarchy of levels. Everything is an entity (level A). The first distinction is between the abstract and the concrete, which are mutually exclusive (level B). At level C lie various species of concreta and abstracta. Properties, relations, and propositions are abstract, as well as sets and numbers. Substances are to be found on the same level as events, time, places, as well as tropes or mereological sums (collections). Material objects and spirits are types of substances (level D). The notion of substance is thus here understood as being a part of an interrelated and hierarchical network of notions, all related to each other as genus and species. 6 To mention another example, Jonathan Lowe’s (2006) ontology involves four fundamental categories and three types of relations of instantiation. 7 Horgan and Potrč (2008, p. 92). 8 Various versions of the A-theory can be found in Bigelow (1996), Merricks (1999), Markosian (2004), Craig (2000), Zimmerman (1996, 1997), Prior (1970, 2003b), and Chisholm (1990a, 1990b). 9 Various versions of the B-theory can be found in Quine (1960), Lewis (1976b), Mellor (1998), Sider (2001), Le Poidevin (1991), and Oaklander (1991). 10 The notion of a ‘specious present’ comes in many varieties (and the specious present comes in different lengths, from a couple of milliseconds up to a full minute) and it has been defended on various grounds by many participants in the debate on temporal experience, such as James (1890), Husserl (1964), and Broad (1923), and more recently Foster (1982, 1992), Dainton (2000, 2003), Hoerl (2009), and Phillips (2011)). Common to all accounts is the general idea that in order to understand the continuous and unfolding aspect of our experience of movement, change, and time we need the notion of a non-zero temporal interval which our mind embraces ‘at once’.

48 Eliminativism 11 This claim is no stranger to specious present theorists, as for instance Dainton (2008a, §5, my italics) puts it: “[our brains] try to work out a single, coherent version of events on the basis of the fragmentary and (at times) conflicting data available to them. Only this ‘final draft’, as it were, reaches consciousness”. 12 See these experiments at www.yale.edu/perception/Brian/demos/MOT-Occlusion. html. 13 Compare to Walton (2008, p. 110), commenting on the alleged counterintuitive ness of his ‘transparency claim’ about photographic representation: So shall we argue about whether [the transparency claim] really is counterintuitive? It would be better simply to recognize that intuitions are largely reflections of one’s currently internalized theoretical commitments (there being no such thing as entirely pre-theoretical intuitions), and that whatever authority one accords them amounts to resistance to theory change simply because it is theory change. 14 Weinberg, Gonnerman, Buckner, and Alexander (2010) offer a lot more detail (and more criticism) on these and related issues. Williamson (2011) provides a critical discussion. 15 See Korman (2015, chapter 7), as well as Merricks (2001, chapter 7), Heller (1990, chapter 2), and Sider (2013). 16 I am indebted to Dan Korman for a very useful discussion that helped me to write this chapter.

3 Eliminativism, Reductionism, and Composition

§1 In this chapter, I want to focus on an issue concerning eliminativism, discussed by Merricks (2001, chapter 1),1 that can be – at least for the sake of clarity of the discussion – formulated as the objection that the eliminativist postulates the existence of something (the simples2 arranged tablewise) but not of something that is identical to it (the table). Indeed, one way of understanding the relationship between the simples that compose a table and the table is to claim that they are identical – the table is the simples, since it is composed of them, and it is nothing over and above them. If this is true, the eliminativist is guilty of endorsing a contradiction, since on the one hand she claims that tables do not exist, but on the other hand she has to say that they do, since simples arranged tablewise exist, and tables are nothing else than simples arranged tablewise. For the purposes of the discussion, let us call this problem “the objection”. As we will see, the objection backfires – not only there is no threat to eliminativism, but considerations surrounding the objection actually provide reasons in favour of eliminativism. In order to see this, I will compare eliminativism to various versions of reductionism and to strong realism about tables as well, to claim that once we understand what lies behind this objection properly, it threatens everybody except eliminativists. On the way, I hope to offer some relevant thoughts about the difference between eliminativism and reductionism, about composition as identity, and about the notion of an ontological free lunch (Armstrong [1997]).

§2 Merricks (2001, chapter 1) answers the objection by rejecting composition as identity (CAI). He rejects CAI because he claims that identity is never one-many, but always one-one (and so the table cannot be identical to the simples), and also because he claims that if the table were identical to simples S1, S2, . . . Sn, it would then always and necessarily have to be identical to these simples, which would amount to embracing a strong (and

50 Eliminativism unpalatable) version of mereological essentialism. To escape this unpleasant commitment, Merricks shows that one would have to embrace perdurantism and modal counterpart theory, which are commitments that he finds undesirable, for independent reasons. Merricks’s argument is complex and contains a lot of interesting moves and insights, and I will not do justice to it here. I will not, because I find perdurantism and modal counterpart theory to be perfectly respectable theories, and an argument based on their rejection for independent reasons would not mean that much to me, and perhaps to many others as well. What about, then, the idea that identity is never one-many? At first sight, it seems that we can make sense of one-many identity. For instance, the crew of the USS Enterprise is composed by the members of the crew, and we could quite naturally want to understand this situation as one where the crew is identical to the members – we could say that the crew is the members, where the copula can very well be understood as identity. If you take away the members, you take away the crew. Perhaps, one could want to say something similar about the table and the simples arranged tablewise. To see better if/how this can be understood and articulated, let us try to be more precise about what eliminativism is, as opposed to reductionism and to realism about tables: Eliminavism

There is no table. Simples arranged tablewise Table

Plurality Reduconism

The table is ‘the simples’. It’s a plurality. Simples arranged tablewise = or ≠NOAA Table

Single-enty Reduconism

The table is the ‘collecon’ of the simples.

Simples arranged tablewise ≠NOAA Table

Eliminativism, Reductionism, and Composition  51 Strong Realism

The table is something different (‘more’) than the simples. Simples arranged tablewise ≠ Table

What does it mean to reduce something to something else, rather than to eliminate it? Plural Reductionism and Single-Entity Reductionism, as I label them above, are attempts to answer this question. Plural Reductionism builds on the idea that the table is a plurality – it is the simples. Such a view does not deny that the table exists. It does exist, it’s just that it is the simples arranged tablewise that compose it. Now, the copula can be understood in different ways. It can be understood as identity, for why could it not? Under Plural Reductionism, it makes perfect sense to say that the table is a plural entity, that it is the simples, and that it is them in the sense that it is identical to them. So, do we have here a case of identity as being one-many? We don’t, since ‘the table’ is a plurality rather than a single entity, so what we have here is a situation where identity is many-many, rather than one-many. But, this does not prevent here CAI to hold, even if it is a modified CAI. In this view, the table, qua plural entity, is identical to the simples that are it. As a general principle, it is true that the only proper case of identity is self-identity, where one thing is identical to itself. But nothing forbids that this ‘thing’ can be a plurality, identical to itself. (The case of the crew of the USS Enterprise, I would suggest, also falls within this category of many-many identity.) A different version of Plural Reductionism might want to say that the table is not identical to the simples, but that it is ‘nothing over and above’ them (see “≠NOAA” on the figures above). I will discuss this option, and the notion of ‘nothing over and above’, in I.3.§5 below. Single-Entity Reductionism is perhaps best understood as a kind of a bundle theory, where the bundling relation is ontologically as little loaded as possible. According to this view, the table is a kind of ‘collection’ (to use a neutral, theoretically unloaded, term), and as such it is a single entity. Thus, it is different from the simples – a collection of simples, qua collection, is not the same entity as the simples. In other words, Merricks was right to remark that identity is not one-many. But Single-Entity Reductionism does not require the relation between the table and the simples to be identity. The reductionist can very well here acknowledge the difference between the table, qua collection, and the simples, but still she can want to say that the bundling relation is ‘ontologically innocent’ and that the table is ‘nothing over and above’ the simples – a popular option that is also available to Plurality Reductionists, as mentioned above. (Her reasons for preferring

52 Eliminativism Single-Entity Reductionism to Plurality Reductionism may be her own; perhaps, she wants to secure some kind of unity for the table, which she might feel is missing in the Plural Reductionist account.) Full-blown Realism about tables is the last option on the list. It simply claims that the table is something over and above the simples, and that it is something more, and something different. This view could also be articulated in terms of a bundle theory, where the bundling relation would be ontologically loaded, but it does not need to be – other possibilities are available, in the form of the substratum theory, or an Aristotelian substance theory, or something else.3

§3 With these distinctions and clarifications in mind, let us now come back to the objection. Indeed, one way to put the objection is to ask: what is the difference between Eliminativism and Plural Reductionism? The objection wants to force the eliminativist to accept that there are tables after all – so, it wants to force the eliminativist to be a reductionist, more precisely, to be a Plural Reductionist, since Plural Reductionism is just like eliminativism, except for the claim that the table exists. The idea of the objector could then be formulated as the claim that eliminativism is impossible: it turns out, automatically, to be a Plural Reductionism. When the objector says that the eliminativist cannot – on pain of a contradiction – say that the simples exist but that the table does not, that’s what she has in mind.4 This way of understanding the objection will now help us to see how eliminativism can avoid it, and also to see what the difference between eliminativism and reductionism is. The Plural Reductionist claims that the table exists. But then, she has to explain what the relationship between the simples and the table is. The question here is not the one we have already encountered, namely, whether the relation is identity or something else like ‘nothing over and above’. Let us grant that it is one of those. The further question that arises then, and that arises for everybody who wants to say that both the simples and the table exist (so, everybody except the eliminativist), concerns the relata and the way they depend on each other. Suppose I say that my dining room table is identical to some simples. Now, one can ask: which simples? At first, we might think that there ‘only’ is the familiar problem of vagueness. Indeed, since it is vague which simples compose the table, we might be forced to buy a claim (that Peter Van Inwagen did endorse, in the end) of metaphysical vagueness. Or to buy the claim that identity (the identity that holds between the simples and the table) is vague. Or that there are many (many!) tables, instead of one (this is ‘the problem of the many’). But the problem is not only one about vagueness (although, as we have already seen, this is a problem, for everyone who wants to say that there are tables in addition to the simples that compose them). There is a more general problem and a more general question to ask: what is the – metaphysical and

Eliminativism, Reductionism, and Composition  53 necessary – law which says that whenever there are these and these simples, there is a table? What is the metaphysical and necessary connection between the simples and the table? How can we be guaranteed that whenever there are the (suitably arranged) simples, there is a table? Heller (2008, p. 86) raised this question in a different context, where he discussed the ‘donkey problem’ that arises against the modal linguistic ersatzist – to paraphrase what Heller says, the question is: “what guarantees that that particular pattern of simples can never be instantiated without a table being present in the same location? What guarantees that simples in pattern T always compose a table?” These are questions for everyone who thinks that there are tables. The menace of vagueness yields a menace of arbitrariness, as we have seen in Chapter 1, and here we have an additional menace of bruteness, if, to answer the question, we say that it is a brute fact that the simples compose a table.5 What else can the friend of tables – both the reductionist and the realist – offer? She can say that the metaphysical and necessary connection is a brute fact, she can say that it is a primitive unexplained fact, but there isn’t much more she can say. To claim that the relation between the simples and the table is identity, or the relation of ‘nothing over and above’, does not answer the question that is being raised here. Saying that the table is, say, identical to the simples is one thing, but here we have the further problem about what guarantees that this is the case – what, so to say, provides the table (and its identity with the simples at hand). Thus everybody owes us an answer to a hard question, except the eliminativist, since according to eliminativism there are no tables, and so no ‘metaphysical connection problem’ to deal with. The eliminativist is the only one who does not have this problem. Thus the objection that wanted to force the eliminativist to ‘automatically’ be a reductionist actually makes us realize that reductionism and eliminativism are crucially different on this one important issue, and that eliminativism is in a better position with respect to this problem than both the reductionist and the realist.

§4 Let me indulge in a short digression, to point to a kind of generality there might be in what I discussed above. Think of the mind-body problem and consider the reductionist and the eliminativist strategy about mental states. Quine is a good example of someone who raised a worry that is similar to the objection. He was a reductionist about mental states, but he thought that an eliminativist view would not be far from a reductionist one – eliminating mental states or reducing them to physical states, he suggests, might not make a big difference: Is physicalism a repudiation of mental objects after all, or a theory of them? Does it repudiate the mental state of pain or anger in favour of its physical concomitant, or does it identify the mental state with a state

54 Eliminativism of the physical organism (and so a state of the physical organism with the mental state)? . . . Some may therefore find comfort in reflecting that the distinction between an eliminative and an explicative physicalism is unreal. (Quine [1960, p. 265]) Without going to the details here (this is just a short digression), it is easy to see that we can have the same kind of options as in the case of eliminativism/reductionism(s)/realism about ordinary objects we have seen above. One can say that mental states don’t exist, or that they do exist but that they are (identical to, or nothing over and above) physical states of the brain – to mention only the eliminativist and reductionist general options. But then, exactly as above, it turns out that Quine’s worry, exactly as the objection, does not conflate eliminativism and reductionism. To put it briefly and bluntly, if one thinks that there are mental states and that they are reducible to brain states, saying – for instance – that the relation between the brain states and the mental states is identity does not answer all the questions, and the same question about what the metaphysical and necessary connection between the brain states and the mental states requires an answer (“What guarantees that that particular bunch of brain states can never be instantiated without such-and-such mental states to occur? What guarantees that brain states B always give rise to such-and-such mental states?”). The one who claims that mental states do not exist does not face this problem, and does not face the threat of bruteness and/or arbitrariness. End of the short digression.

§5 Let us focus now on the notion of ‘nothing over and above’. The Plural Reductionist or the Single-Entity Reductionist can both want to say that the table exists but that it is nothing over and above the simples. The objection could also very well be formulated in this framework, even if nobody says here that the table is identical to the simples. Indeed, if the table is said to be nothing over and above the simples, the objector would then say that the eliminativist is forced to accept its existence, since it would be a contradiction to say that the simples exist but that something that is nothing over and above them does not. Perhaps, the objection is here weaker than in the case of identity, but it could be raised in this way. The link between identity and the notion of ‘nothing over and above’ is strong. When appealing to this notion, Armstrong (1997) claims that it means – in our case – that the table is “no addition to being” and that “mereological wholes are identical with all their parts taken together” (Armstrong [1997, p. 12, my italics]),6 thus making identity play the central role. David Lewis also appeals to identity in the cases where an object is said to be nothing over and above its parts – this is his version of CAI

Eliminativism, Reductionism, and Composition  55 (Lewis [1991, p. 80]). Sider also explores this notion (2015, p. 204) where he examines various versions of composition as identity: “The doctrine of composition as identity that I will consider does not say that for each part, the whole is identical to that part. It says, rather, that the whole is identical to the parts taken together”. The notion of ‘nothing over and above’ is also found in the current debate about grounding: as Bricker (2006), DeRosset (2010), Schaffer (2009, 2016) claim, if a is grounded in b, a is nothing over and above b. a, in other words, is an “ontological free lunch” in Armstrong’s (1997) sense; the “ontological price”, to use Schaffer’s (manuscript) term, you pay for a and b is just whatever you would pay for b alone. If ‘nothing over and above’ centrally involves identity, it seems that at the end of the day it just is identity (as Lewis would say, I think), so concerning the objection raised against eliminativism, we find ourselves in the same situation as the one discussed in I.3.§3–4 above. Indeed, in this case, saying that the table is nothing over and above the simples does not seem to add much to the claim that it is identical to them. But perhaps, one could want to say that there is more to ‘nothing over and above’ than just identity. Perhaps, one could want to say that it is something else, close, and ‘ontologically innocent’, but not exactly the same thing as identity. Trouble is, it is hard to see what it could be. Indeed, this notion seems to be systematically taken to be a primitive one and a rather under-explained one.7 But even if we accept that the relation between the table and the simples is not just identity and that the table is nothing over and above the simples in some relevant sense we understand, there would be no threat to the eliminativist (that is, the threat of forcing her to be a reductionist). On the contrary, we would have here even more reason than before to be eliminativists: if the table is nothing over and above the simples, if it does not have any more causal powers than the simples (see Merricks [2001]), if it does not do any more work than the simples in perception (see I.1.§4 above) or in ordinary language (see I.1.§5 above), it then really is nothing over and above the simples or indeed just nothing. It does not do anything, it is nothing over and above the simples that do all the work, so claiming that it is nothing – that it does not exist – seems like the most natural thing to say. If one thought, like the Realist does, that the table is something different and more than the simples, then one would have a reason to think that it exists, but if one insists so much on saying that it is nothing over and above them, I then submit that the eliminativist’s claim that it is nothing rather than nothing over and above is a more natural one and one that we understand much better than the under-explained claim of being nothing over and above. Perhaps these dialectical and general concerns will not convince you. So, to come back to the objection and to the considerations from I.3.§3–4 above, if one does not follow the eliminativist here, one will have the very same problem to explain the necessary and metaphysical connection

56 Eliminativism between the simples and the table that we have seen in the case of identity – on this issue, replacing identity with a different notion like ‘nothing over and above’ simply does not make any significant difference. The problem is there for everybody who thinks that the table exists. To wrap things up. The eliminativist is the only one who has an (easy) answer to this problem – the problem does not arise, since there are no tables – and everybody else from the reductionist to the realist owes us an answer to a hard question, unless she is ready to embrace the bruteness claim.8 The initial objection to eliminativism backfires. The eliminativist’s answer also avoids any under-explained primitive elements like the notion of ‘nothing over and above’. The claim that the table does not exist is free of any mysteries. I tried to suggest above that if one thinks that the table is nothing over and above the simples, it would be natural to think that that it is simply nothing. Somewhat dialectically, I admit, what I am doing when it comes to this particular point of the discussion, is to reverse the direction of the objection raised against the eliminativist: I am saying that if one is a reductionist, one is pushed in the direction of eliminativism (but, I am not saying that one is forced to, or that reductionism is ‘automatically’ an eliminativism).

Notes See also Korman (2015, p. 21, §2.2). 1 2 In Chapter 4, we’ll see that eliminativism is also compatible with other fundamental ontologies, which do not involve simples. For now, it’s simpler to work with simples. 3 Le Bihan (2016) discusses the relationship between eliminativism about ordinary material objects and various forms of reductionist and non-reductionist realism about space-time (relationism and substantivalism). 4 Alternatively, a weaker objection could be formulated as the claim that given the similarity between the two views, and given that Plural Reductionism is a coherent view, then one might as well be a Plural Reductionist rather than an Eliminativist. 5 For a discussion of the way eliminativism allows us to take less phenomena than competing views as brute, see Brenner (2015a). 6 For an interesting discussion, see Smid (2015). 7 “But what does ‘nothing over and above’ mean? This slippery phrase has had a lot of employment in philosophy, but what it means is never explained by its employers” (Van Inwagen (1994, p. 210). 8 Heller (1990, 2008) is a special case. As we have seen in I.1.§5, Heller has a conventionalist view about ordinary objects. The table, according to him, is a conventional object in the sense that ontologically speaking there only are the simples but we humans have a convention that such-and-such an arrangement of simples ‘counts’, for us, as a table. There is both a possible eliminativist and a reductionist reading of Heller’s claim – depending on how much ontological weight one puts on the conventions. Heller himself puts no ontological weight on them at all, so I would say that he is best interpreted as being an eliminativist (also, in Ontologese – the metaphysician’s fundamental language – it is false, in his view, to say that there are tables [or, it even cannot be expressed, if Ontologese only contains fundamental

Eliminativism, Reductionism, and Composition  57 terms, and not terms like “table”]). But, it is worth noting that in both cases he can avoid the problem with the metaphysical and necessary connection between the simples and the table since, given the merely conventionalist understanding of what tables are, and given the fact that these conventions are in his view purely linguistic matters, there is nothing hard or mysterious to explain, except some linguistic facts about the way we use words like “table”.

4 The Fundamental Ontology of Eliminativism

§1 Here is a worry that can arise for the eliminativist: if she says that instead of there being tables there are ‘simples arranged tablewise’, then eliminativism seems to fall prey to the objection that the world could be ‘gunky’ – no simples, no arrangements of simples. This objection has been developed and explored by Sider (1993) whose target was Peter Van Inwagen’s brand of eliminativism (see, inter alia, Van Inwagen [1990, p. 109]). Van Inwagen subscribes to the existence of simples and living things. Tables do not exist, but simples arranged tablewise do, and that’s enough for any linguistic, practical, or philosophical purposes. As we have seen in Chapter 1, the way eliminativists usually articulate their claim is that “There are simples arranged tablewise” is a paraphrase for “There is a table”. Sider’s line of argument can accept this paraphrase strategy, but it questions the existence of the building blocks on which it relies – the existence of simples. In the history of physics, every time we thought we have discovered a fundamental particle (chemical atoms, protons, neutrons, quarks, . . .), it appeared sooner or later that, contrary to what we thought, it was not a simple, and that it was made of smaller elements. Experience then seems to teach us caution, and it seems to be, at least prima facie, a metaphysical possibility that there are no simples – perhaps matter is infinitely divisible, and the world is ‘gunky’: any part of any object has proper parts. This is a problem for the eliminativist (like Van Inwagen), since if there are no simples, eliminativism, as it stands, cannot be true in a gunky world. Van Inwagen himself claims his ‘atomism’ to be a metaphysical assumption he does not argue for – after all, every theory has its load of primitives, and this is his. But, Sider (1993, p. 288) objects, “Van Inwagen needs more than the truth of Atomism: he needs its necessary truth. And this is what I find implausible”. Sider then adds: Van Inwagen may reply that accepting the necessary truth of Atomism is a cost of his theory, but an acceptable one, given his unappealing rivals. I disagree: I find the possibility of gunk so compelling that I am willing to reject any theory that rules it out.

The Fundamental Ontology of Eliminativism  59

§2 There are various ways the eliminativist can react to this worry, some of them available to Van Inwagen himself, and others only available to other versions of eliminativism, as we shall see. The one that I will ultimately find to be the best will then show us something about eliminativism itself, namely that it is an ontology that provides a generic method that can be used in various ways. A first type of reaction could criticize Sider’s argument by claiming that it is rather weak (see Korman ([2014, pp. 9–10]). Indeed, Sider seems to rely on a kind of a metaphysician’s intuition that gunk is possible, and finds it so compelling that he is ‘willing to reject any theory that rules it out’ (see above). But then, there is the intuition that tables are possible which would rule out eliminativism even more directly. Perhaps, qua intuition, the latter is even stronger than the gunk intuition, and so if one decides to follow one’s own intuitions of this kind one will probably not even consider eliminativism to be a live option at all, and no ‘argument from gunk’ is needed. But, as we have seen in Chapter 2, pro or against eliminativism, intuitions are not a good guide for the metaphysician to follow in such a situation. The issue is not about whether gunk, tables, or eliminativism are intuitive. Sider explores a possibility, or at least the possibility of a possibility, and the eliminativist should have an answer to the objection, rather than claiming that his opponent’s intuition is not intuitive enough. The eliminativist should take Sider’s possibility seriously, even if only to play as fair as possible. Taking the possibility seriously, the eliminativist could claim to be a contingentist eliminativist, that is, she could say that her view is meant to apply to the actual world only, and leave it open whether in other possible worlds her view is applicable, depending on whether these worlds are gunky or contain suitable simples. Contingentism could seem to be a frustrating retreat for any metaphysician, but it is a live option (see Rosen [2006], Miller [2009, 2010]). Stricto sensu, it would avoid at least a part of Sider’s objection. Indeed, it is the possibility of gunk that Sider finds so convincingly plausible, not its actual existence in our world. Thus, the contingentist’s retreat answers the objection by accepting it: yes, gunk is possible, and there are worlds where composition takes place and eliminativism is false. The remaining piece of trouble for the eliminativist is to defend the claim that the actual world is ultimately made of simples – which is an empirical claim. Perhaps our world is made of some kind of quanta, or indivisible vibrating strings, or some other ‘atoms’. But perhaps not. The truth of contingent eliminativism relies then on future empirical discoveries that may, or may not, be conclusive. Let us see if eliminativism can do better. A more metaphysically minded option for the eliminativist is to meet the objection head-on and argue against the possibility of gunk. Williams (2006) takes this line of defence, and claims that Sider’s argument is a tricky one since it argues for the possibility of gunk based on its conceivability – and the route from intuitive conceivability to possibility can be a treacherous

60 Eliminativism one. It’s hard to say who has the burden of proof here. But suppose, for the sake of the point I want to make here, that the Williams-like eliminativist succeeds in undermining the possibility of gunk. Well, if this is the path he wants to follow, he’ll have to go even further. Indeed, a mirror objection to Sider’s could be based on the claim that monism is possible: there is (can be) only one big entity, the universe. Here again, at least prima facie, the standard eliminativist who appeals to the paraphrase of ‘simples arranged tablewise’ will not work. Thus, in the same spirit as Williams above, he could then argue that monism is not possible, and he may, or may not, succeed. In both cases, and perhaps in other cases as well, this type of eliminativism is vulnerable to arguments supporting the possibility of gunk, monism, or other ontologies which are incompatible with the claim that there are simples arranged tablewise. Success is possible, but difficult.

§3 Let us come back to the very idea of the paraphrase strategy. Eliminativists claim that instead of saying “There is a table” we can say “There are simples arranged tablewise in region R”. The former is a sentence uttered in ordinary English, and can be said to be true in ordinary English, because the latter is a true sentence uttered in Ontologese, the metaphysician’s fundamental language. Now, nobody forces the eliminativist to endorse this paraphrase. Instead, she could say “There is gunk shaped tablewise in region R” (or something like this), or she could have a suitable monistic paraphrase, perhaps “The universe is table-like in region R” or in a Spinozian spirit “The universe has a table-like aspect in region R”. If gunk is fundamental, the paraphrase will appeal to gunk, if the whole universe is fundamental, the paraphrase will appeal to the universe, and if there are simples, it will mention simples – it can be as simple as that. Simple, yes, but each of these strategies is open to the same objection. If one has a paraphrase in terms of simples, one has to face the possibility of gunk and the possibility of monism. If one has a paraphrase in terms of gunk, one has to face a similar objection about the possibility of simples and the possibility of monism. If one has a monistic paraphrase, one has to face the possibility of gunk and the possibility of simples. In any case, the same type of objection will arise. It seems that whatever route the eliminativist takes, she ends up in trouble, or will have to retreat to contingentism, or to provide strong arguments to reject the possibilities of gunk/simples/monism, and perhaps more. To my mind, this pressure put on the eliminativist is misguided and it is a misconception of what eliminativism is. Eliminativism is not a theory about the nature of the fundamental constituents of reality. It does not tell us, and it does not need to tell us, whether quarks, strings, gunk, instantiated properties, or something else is the fundamental stuff of which the world (or the worlds) is made (provided that it does not postulate composite entities, since this would deprive eliminativism from its strongest motivation and its

The Fundamental Ontology of Eliminativism  61 problem-solving powers). These are all additional claims. What it does tell us is that there is nothing more than this fundamental stuff – that’s what eliminativism is all about. Eliminativism is not, in this sense, a complete ontology. Rather, it is an incomplete ontology that provides a strategy, a method that allows to eliminate unnecessary entities from one’s ontology, whatever that ontology is. In Part III, we will see a kind of eliminativism where this is clearly apparent: eliminativism about musical works or photographs says literally nothing about the ontology of the fundamental, it only eliminates musical works or photographs, but not, say, tables, and it does not appeal at all to any paraphrases using ‘gunk’, or ‘atoms’, or ‘quarks’, etc. Things are similar with respect to eliminativism about ordinary objects. Its purpose is to say something about ordinary objects, not about the fundamental ontology. Granted, there needs to be a fundamental ontology and the eliminativist needs to appeal to this ontology in her paraphrases, but as we have seen it is easily compatible with any reasonable ontology one can have, and so it is neutral with respect to the choice of which of the fundamental ontologies is the correct one. Eliminativism, qua strategy for eliminating tables and chairs, is compatible with virtually any fundamental ontology, and it can be dissociated from the additional claim that the fundamental ontology is such-and-such. The eliminativist’s job is to eliminate, not to say what the nature of what remains is. Of course, eliminativists can have in mind their own preferred fundamental ontology, but again, what they do then is to supplement eliminativism with additional claims that they want to defend or embrace. In its simplest and most general form, the eliminativist’s paraphrase is simply this: instead of saying “There are tables” we can say “There is fundamental stuff arranged tablewise”, where the nature of ‘fundamental stuff’ and of what ‘arranged’ exactly means can be left open, without any loss of what eliminativism is there for and of its explanatory power and theoretical virtues. This does not amount to embracing contingentism.1 Eliminativism is compatible both with contingentism and with the claim that the fundamental ontology is necessary. In the picture of eliminativism I suggest here, eliminativism provides a strategy that can adapt to any (or many) situations and worlds. It is an entirely different issue to claim that all possible worlds are gunky, or that some worlds are gunky and others contain simples, or that some worlds contain only one object (monism) while others are gunky or contain simples. Eliminativism can be true in all these situations, and its claims are orthogonal to these additional issues. This is why eliminativism is best understood as being a flexible method, rather than as a complete ontology.

Note 1 Contra Le Bihan (2013, p. 5).

Part II

The No-Self View

5 Exceptionalism

§1 The main task of Part II of this book is to defend the no-Self view. But before we get there in Chapter 7, some preliminaries are in order. In Chapter 6, I’ll have something to say about two traditional metaphysical views of the Self, namely a Cartesian-like substance view and a Humean bundle view, in order to get a firm grip on the ontology of what the allegedly existing Self could be. But first, in this chapter, I’ll make some comments on the two most prominent eliminativist views that make an exception. Indeed, Peter Van Inwagen and Trenton Merricks are both influential eliminativists who embrace one form or another of exceptionalism. Van Inwagen makes an exception for living entities and Merricks makes an exception for conscious organisms. In this chapter, I will critically discuss some issues involved in these two lines of thought. These objections will not be conclusive in the sense that they would directly demonstrate the truth of the no-Self view – again, the defence of this view will take place in Chapter 7 – but they will serve as a useful introductory motivation one could have to embrace it.

§2 Let us start with Merricks. Merricks (2001) focuses on conscious organisms, and he claims that the fact that they are conscious is the reason for not being an eliminativist about them, while he is an eliminativist about everything else. Since my own interest lies in what we might call ‘the Self’, the way I would like to put the question here is: can we eliminate the Self in the same – or similar – way we can eliminate chairs and statues, without losing something important? Under ordinary objects eliminativism, we can without any loss eliminate chairs and statues because there are simples1 arranged chairwise or statuewise and those can play the same practical and theoretical roles that tables or statues could play if they existed. Thus tables and statues, that is, single objects, are eliminated because there exists a plurality of other objects that takes their place – the simples, suitably arranged. In principle, the same strategy could be applied to the case of the Self, although it needs to be articulated in a way that suits such a special case. The basic idea can be the same: an allegedly single entity, the Self, can be eliminated

66  The No-Self View because there is a plurality of other entities which satisfy all our theoretical and practical needs, namely, successive impermanent psychological states/ experiences arranged ‘Self-Wise’. We will see in Chapter 7 below a full defence of this view; here, my aim is to defend it against an objection raised by Merricks, who thinks that here is a place to make an exception. The idea of such a non-exceptionalist eliminativism about the Self is that we can eliminate us understood in an ontologically strong, reified, and ontologically committing sense of ‘the Self’, but without losing anything important – I can still say “I am happy now to be sitting on a chair and to be thinking about eliminativism” in a perfectly respectable sense, understood in terms of simples arranged my-body-wise and chairwise, and in terms of the existence of a succession of impermanent psychological states, suitably causally connected. In this way, we can hold a unified and complete eliminativist view, with no exceptions. In Merricks’s (2001) view, however, there is a disanalogy since entities like chairs or statues are causally irrelevant – whatever they can cause, can be caused entirely by the simples that compose them. So, in his line of thought, this is one of the good reasons to say that chairs do not exist. But, he adds that “we humans – in virtue of causing things by having conscious mental properties – are causally non-redundant” (Merricks [2001, p. 114, my italics]). When I decide to run, there is a cause to be understood in terms of microphysics or microbiology, but there also is a cause to be understood in terms of my decision. It is my decision, Merricks says, that causes the simples to move as they do. This is why we cannot be eliminated in the same way chairs can be, since we are causally relevant – we have causal powers over and above the causal powers of simples that compose us. Merricks’s argument to the effect that we, human organisms, are causally non-redundant in virtue of having conscious mental properties is a complex and a very long one – indeed, it stretches on almost 30 pages (see Merricks [2001, chapter 4]). In what follows, let me focus on a (rather self-standing) part of his argument in detail and see how this step can be resisted – if it can, the overall argument will then not go through. The main point of the argument is to show that we humans are not causally redundant, and that we have conscious mental properties that do “not supervene on what our parts are like” (Merricks [2001, p. 88]). We cause things in virtue of having these properties and we are therefore not causally redundant. (Thus, an argument based on the idea that we should eliminate anything that is causally redundant cannot be used to eliminate us – this is Merrick’s overall main point.) On the way of defending this claim, Merricks argues for the rejection of: Consciousness (C). Necessarily, if some atoms A1 . . . An compose a conscious object, then any atoms intrinsically like A1 . . . An, interrelated by all the same spatiotemporal and causal interrelations as A1 . . . An compose a conscious object. (Merricks [2001, p. 94])

Exceptionalism  67 Here is, in short, Merricks’s argument against (C) – it is a variant of the ‘undetached parts argument’.2 Suppose a small part of you is annihilated (say that your finger, or perhaps just an atom composing your finger, is cut away). Right after the amputation there is a conscious object – you – composed of some atoms. But these atoms existed in exactly the same way just before the amputation – indeed, they composed a (big majority) part of you. But, if (C) is true, this means that even before the amputation there was an object that was part of you – let’s call it “you-minus” – that was a conscious object. So, it seems that before the amputation there were two non-identical conscious objects, namely you and you-minus. (By the same reasoning, there actually were many you-minus-like objects before the time of the amputation.) But this is false, since there was only one conscious object before the amputation. Thus, Merricks concludes, by reductio (C) is false. In case one would be tempted to answer the objection by appealing to four-dimensionalism and by using talk about temporal parts to escape the unwelcome consequence that there were two conscious objects before the amputation, Merricks provides a temporal version of the objection as well, which can be formulated as follows (this is my formulation, close to Merricks’s own). Take a four-dimensional person named “Trenton” who lives for 100 years. Take also another four-dimensional person, inhabiting the same possible world, who lives for only 80 years and is named “Trent”. Suppose that Trent is microphysically intrinsically exactly like the temporal part of Trenton who lives for the first 80 years of Trenton’s existence, and let us call this temporal part of Trenton “Trenton-minus”, where Trenton-minus and Trent thus have atomic temporal parts exactly similar in intrinsic features and causal and spatio-temporal interrelations.

Trenton

Trenton-minus Trent

68  The No-Self View According to Merricks, according to the four-dimensionalist, assuming (C) for reductio, when it comes to Trenton between the age of 0 and 80, we then have a case where there are two coincident persons: Trenton and Trenton-minus. Trenton-minus is a conscious person in virtue of the existence of Trent and in virtue of the truth of (C). But such coincidence is unacceptable, and as a consequence, by reductio, (C) is false. But the way Merricks presents the case here can be resisted. Indeed, this is not how four-dimensionalists typically describe the situation. Here is Sider (2001, p. 6), about the Statue and Lump famous case of coincidence: At any given time it is only a temporal part of a space-time worm that is wholly present. Thus it is only temporal parts of Statue and Lump that are wholly present at the time of coincidence. How can these temporal parts both fit into a single region of space? Because ‘they’ are identical. Sider then compares this to a case of a road: Road

Middle secon Road subsegment

There is a road that has a subsegment. In the very middle of the road, should we say that there are two entities, namely, the road and the road subsegment? Of course not. There is only one entity – the middle section – common to both the road and the road segment. Similarly for Trenton and Trenton-minus. At a time where ‘both’ Trenton and Trenton-minus exist, should we say that they are two persons? Of course not. Before Trenton’s 80th birthday, there always was only one person, exactly as in the middle section of the road there is only one road. It’s just that this middle section is part of a road and of a road subsegment, and in the same way Trenton-minus is part of Trenton. This does not prevent Trenton-minus to be a person, as for instance David Lewis insists upon: “A person-stage is a physical object, just as a person is. (If persons had a ghostly part as well, so would person-stages)” Lewis (1983, Postscript B). Typical examples of the way four-dimensionalism deals with such situations also involve cases of fission.3 Let us say, for the sake of brevity, that for some reason a person undergoes fission, perhaps using a transporting

Exceptionalism  69 device such as the one commonly used on the USS Enterprise, where due to a malfunction of the device, instead of simply transporting one person from one place to another, the device also leaves the original person behind. (You can replace this example with any other case of fission, if you don’t like Star Trek stories.) Thus, after the fission, there are two persons, exactly alike. According to four-dimensionalism, we then have a situation where there are two four-dimensional persons, sharing an initial segment:

Person 1 Person 2

Under a typical endurantist reading, this situation is one where the threat of coincidence is real: Person1-after-fission is not identical to Person2-afterfission, but Person1-after-fission is identical to Person1-before-fission, and Person2-after-fission is identical to Person2-before-fission – we then seem to have a situation where Person1-before-fission is identical to Person2-beforefission, that is, where these two persons seem to coincide in an unpalatable sense. But not so under the four-dimensionalist view where Person1-afterfission is not identical to Person1-before-fission, since these are two different temporal parts, numerically distinct, and similarly for Person2. In this way, (i) four-dimensionalists do not have to face the threat of coincident entities,4 and (ii) they can say, relevantly to our present discussion, that there is only one person before the fission, in the same sense that there is only one person in the case of Trenton and Trenton-minus, and similarly in the spatial case of you and you-minus (in the finger amputation case). Metaphysically speaking, in all such cases where there seem to be two objects competing for the same space (and time), there really is only one, it’s just that it’s also part of other, spatially and/or temporally bigger, objects. It’s like a wall that’s common to two houses: if you need to repair it, you’ll only need bricks to repair one wall, not two. In the way Merricks describes the situations he uses in his argument, there seems to be something like the principle that only ‘the biggest’ object is the one that counts. Thus, only Trenton, but not Trenton-minus is a conscious object. In his argument, appealing to the existence of Trent and to the truth of (C), Merricks then wants to force his opponent to recognize, for reductio, that Trenton-minus is a conscious object as well, thus creating a situation where there apparently are two distinct coincident objects that crowd each other out. But, as we have seen,

70  The No-Self View four-dimensionalists do not, and do not have to, understand this situation (as well as the other similar situations) in this way. One issue still remains. Who is doing the thinking, in the fourdimensionalist view? Is it the whole worm, or is it only a (rather short-lived) temporal part of the worm? If it were the whole worm, then – and only then – Merrick’s objection above would go through. So, in order for the reply to work, we have to say that it is not the whole worm that has thoughts but that it has them only in virtue of having temporal parts that have them. A possible objection arises here:5 say that a temporal part that lasts for only one minute thinks the thought “I have lived for 50 years”. This is true, in a sense, because the whole worm did live for 50 years. But it is false, when thought by the temporal part, since the temporal part only lived for one minute. So, the same thought seems to be both true and false – how messy! But this only points to a specific feature of four-dimensionalism (i.e. the worm view6). In this view, worms have most of their properties in virtue of the having of those properties by their temporal parts. The temporal parts can overlap, in the sense we have just seen, or in a fission case scenario. Thus, what is being thought or said at some point by some temporal part is ambiguous. Take the case of fission we have seen above. Let us say that, before the fission, the person that is there is called “Jean-Luc Picard”. From an atemporal standpoint such a name is then ambiguous, since it refers both to Person1 and Person2, and since these two overlap at the time before fission, but not at the time after the fission. But as David Lewis points out, such an ambiguity is perfectly harmless as long as the two bearers of the name “Jean-Luc Picard” are indiscernible – that is, precisely, before the fission (see Lewis [1983, pp. 64–65]). The need to distinguish the two persons arises only after the fission, and there is no ambiguity there, since there clearly are two persons, and we will use two different names to refer to them (even perhaps in a homonymic way). Similarly, what the one-minute-long temporal part thinks or says is ambiguous, and it is true under one disambiguation and false under another. The problem then easily dissolves as a mere case of ambiguity. As a consequence, using four-dimensionalism to answer Merricks’s objection, we can say that (C) is true, and thus one cannot use the alleged falsity of (C) to argue to the effect that we humans are not causally redundant because our conscious mental properties do not supervene on what our parts are like. And one cannot then use this as a reason to make an exception for us when it comes to eliminativism.

§3 In Material Beings, Peter Van Inwagen focuses on material composition and claims that the only case where some objects compose another object is the case where their activity constitutes a life. Thus, the only non-simple objects that exist are living entities. In Van Inwagen’s ontology there are

Exceptionalism  71 then simples and living organisms, and nothing more – he is an eliminativist about everything else. Thus Van Inwagen allows for the existence of persons, and also for the existence of trees and even the smallest of living organisms. But as we shall now see “life” only leads to complications, and is an unnecessary burden for the metaphysician here. Let us examine more closely what “life” could mean, and the best way to start is to consider some test cases. Indeed, in the debate about what it means for something to ‘count’ as being alive or not, it is often asked what we should be looking for when we look for signs of life in the universe. But search for extra-terrestrial life is complicated. Not only because of the logistics of travelling long distances, but precisely because it is far from being clear what one should be looking for. Think of James T. Kirk facing for the first time the Dikironium cloud creature7 or the Beta XII-A entity.8 Or think of Lieutenant Commander Data9 arguing with Q10 – two entities talking to each other, in a situation where it is far from being clear whether either of them is a living creature. And, to encounter difficulties with borderline cases of life, one does not have to (boldly) go so far from Earth. Protocell synthetic biologists aim at producing new life forms by creating artificial cells from scratch, bioengineers work by trying to introduce new metabolic pathways in already living cells, while silico synthetic biologists work on computer simulations of synthetic organisms. In many ways humans thus set themselves the task to create new life, through rational design.11 But when exactly does something start to count as “life”? This question appears to be crucial not only when it comes to identifying something as being alive or not, but also when we try to determine the origin of life on Earth. To be alive, or not: that is the question. Or at least it seems to be. But as we will now see, there are a lot of interesting questions in the neighbourhood of this question, but not this one itself. First, we shall quickly see that attempts at defining “life” all raise problems and counter-examples; second, we will see that among the reasons for this failure of finding a definition lies the vagueness of the words “life” or “is alive” and, most importantly, that we are facing here a linguistic (as opposed to metaphysical) phenomenon; and third, we will see how much simpler an eliminativist view of life is – as opposed to Van Inwagen’s. When encountering new types of entities, or when synthesizing them in labs, we should not so much ask ourselves whether they are alive or not, rather we should focus on other interesting properties they have.

§4 The most straightforward strategy, if one wants to decide whether an entity x is alive or not, is to produce a definition of life and check whether x satisfies the definition of not. The trouble is, it is hard to produce a non-arbitrary and good definition. (Relevantly, remember our worries with arbitrariness and ordinary objects, from Part I, Chapter 1.)

72  The No-Self View Aristotle (1961, 415a22–415b2) defines life as the capacity to reproduce, but fire can perhaps be said to possess this capacity, while not counting as a case of life. Cleland and Chyba (2007, 5.2.2) critically discuss the popular Darwinian – genetic – definition of life where life is defined as a system capable of undergoing Darwinian evolution. But, as they correctly argue, such a definition (i) does not allow for the possibility of primitive cellular life where reproduction without DNA replication would take place, thus precluding Darwinian evolution, and (ii) it has the unwelcome consequence that since individual organisms do not themselves evolve they do not count as cases of life, since the Darwinian definition is a claim about systems capable of undergoing evolution. Perhaps, one could add that beings such as God or Q cannot (did not?) reproduce at all, while – controversially, but possibly – being alive, at least in the sense in which all conscious creatures are alive. Here are some often-cited examples of attempts at defining life: Short definition

Problem/objection

Darwinian definition

Life is a system capable of undergoing Darwinian evolution.

(i)  Does not allow for the possibility of primitive cellular life where reproduction without DNA replication would take place. (ii) Individual organisms do not themselves evolve, and thus do not count as cases of life.

Metabolic definition

Life is the capacity to reproduce, move, and grow by converting energy.

Possible counter-examples (such as fire).

Biochemical definition

Being alive is being based in a certain type of biomolecules.

Can in principle always face counterexamples of entities that would count as alive but that would be based in a different type of molecules.

Autopoiesis

Life is the capacity of self-production and self-maintenance.

Possible counter-examples (such as some machines).

Communication definition

Life is the capacity to interact and communicate with one’s environment.

Possible counter-examples (such as some machines, or software).

Rosen’s (M,R)-systems (see Rosen [1985]) exhibit features common to both the metabolic and the autopoiesis definitions. Every definition has its merits and brings its bit of interesting piece of information about some feature of life (at least, life as we know it on Earth), but any such feature does not seem to be a necessary one (and thus, it suffers from possible

Exceptionalism  73 counter-examples), nor a sufficient one (and thus, the definition fails). This has been argued for at length by Sagan (1970), Cleland and Chyba (2007), Cleland (2012), and Deplazes-Zemp (2012, pp. 758–759) who also addresses the case of a combination of several or all of these features into a single list-like definition. The common idea of any such definition is to produce a tool to separate all entities into two mutually excluding categories: on the one side entities that are alive, and on the other side those that are not. This is indeed something that Van Inwagen would ideally need (this would allow him to avoid his commitment to metaphysical vagueness). But in addition to all the reasons for failure of these definitions (such as the possible counter-examples and problems quickly mentioned above and discussed in Cleland [2012, §2]), there is a general very simple reason why such a strategy is bound to fail in principle: the vagueness of “life”. Indeed, there are things which are clearly alive, there are things which are clearly not alive, but there are (actual or possible) borderline cases and unclear cases, where a definition would seemingly help by categorizing them into one group or the other – but often, such a help might simply be wrong in thinking that all entities nicely fit into these two categories, ignoring the vagueness involved here. Any new borderline cases might actually always be reasons to reject the definition at hand and to revise it in the light of these new cases. I am not going to insist on the vagueness of “life” or “is alive”. But I do want to insist on the fact that what the vagueness involved here shows us is that the trouble does not come from the world (that is, a vague world, where it is hard to find a definition that would fit its vague nature), but from us (from our language). The issue is a purely linguistic one and should then be treated as such. There are three main theories addressing the problem of vagueness in general: the metaphysical approach, the epistemic approach, and the supervaluationist approach. But, the metaphysical approach and the epistemic approach work well only under a linguistic interpretation, as we have already seen in Part I, Chapter 1, and the supervaluationist approach also clearly emphasizes the linguistic nature of the problem.12 In short, vagueness is a linguistic affair. Thus, “life” and “is alive” are vague terms, and it does not seem very promising to think that we can a find a precise and non-arbitrary definition of them (Dupré and O’Malley (2009) defend a similar claim, for different reasons). Vagueness, in general as well as in the case of “life”, comes from us and not from the world. Now, this points (inconclusively but interestingly) towards the idea I wish to put forward here, namely that “is alive” is merely a predicate that we use to talk about things. Being alive is allegedly a property of those things. But, importantly, being alive is not a property like, say, having positive charge. Having positive charge is a property we can find in the world, things have positive charge independently of us and we can find out about them. Being alive is a different affair – it is not a genuine property of things in the same way. You can’t put Data under a microscope,

74  The No-Self View or measure him in any scientific way to find out whether he is alive or not. You can’t look closely at him and say: “Here it is! See there this property? He is alive!” You can do this, however, with many other of his properties: he can talk, he has such-and-such a shape, he can play the violin, he has a certain weight, and so on. (En passant, note that, relevantly, “being conscious” is a very different affair from “being alive”. If Data is a conscious being, then he does have the property of being conscious). Now, the important thing to realize is that when we list all those things we can learn about Data, we are then in a position where we know all there is to know about him – and the fact that he is alive or not is not on the list. As we have seen above, the vagueness of “is alive” is a linguistic phenomenon, a fact about our use of the predicate, which fits well with the idea I want to convey here: not only the vagueness of life comes from us rather than from the world, but the very idea that some things are alive while others are not is a mere artefact of our thought and language, it is not a metaphysical fact – i.e. nothing in the world is, independently of us, alive or not. Global eliminativism set aside just for a minute, what there is in the world is a huge variety of entities with a lot of different genuine properties like the property of being able to crawl, the property of being able to reproduce, move, and grow by converting energy, the property of being based in a certain type of biomolecules, the property of being able to interact and communicate with one’s environment, and so on. These are the genuine and the interesting properties. Categorizing the entities that possess these properties into two separate categories (alive or not alive) is then no more than expressing our comparative interests for one or another of such properties or a set of such properties. Cleland (2012) rightly argues that we are in no position to formulate a good definition of life or even a good scientific theory of life because we have at our disposal a very restricted sample of living things (those located on Earth and nearby). But would it really help to have larger, even a much larger, sample? What we could get from a larger sample of entities is a larger sample of various properties such as the property of being able to reproduce, the property of being able to crawl, etc. But none of these properties would be the property of being alive. What we see here is that asking for a definition of life is not really the interesting question to ask. Take Jean-Luc Picard who may indeed be at doubt whether his friend Data is alive or not, since he is a borderline case featuring some of the properties often associated with living things but lacking others. What is the correct attitude Picard should adopt here? In the light of what precedes, I submit that he should focus on the things he can know about Data (say, that he is a good officer and a good friend), and simply dismiss the question of whether he is alive or not as a simple matter of an arbitrary decision one could take (but that one does not have to take). When knowing that Data can talk, crawl, play the violin, and so on, Picard knows all there is to know about him, and “knowing” that he is alive would bring

Exceptionalism  75 no additional relevant knowledge at all – indeed, it is even hard to imagine what such a knowledge would amount to. Things are similar with, say, the protocell synthetic biologist who manages to create a new type of cell in the lab. The cell has then such-and-such new interesting properties, and perhaps it can behave in some ways which are similar to other cells we typically call “alive” and perhaps it is dissimilar in other respects to these typical cases. Again, the scientist knows here all there is to know about the new cell, and claiming that it is alive would be at most the statement of a comparative interest she takes in some of its features. The same is true about search for extra-terrestrial life: it is not so much life that is interesting to look for, but other kinds of things. For instance, it would be interesting to find out whether in other parts of the universe there are entities based on biomolecules similar to ours, or whether there are entities that can reproduce, or whether there are entities that are intelligent and conscious, or whether there are entities that are capable of evolution and that undergo natural selection. These are the interesting questions, and here we do know what to look for. Forgetting about life, and focusing on genuine interesting properties of things is then not only conceptually more adequate but makes programmes of search for extra-terrestrial life a lot easier, since it can set tasks in a non-vague and clear way. Some biologists and philosophers reject the whole idea of there being a need for a definition of “life”, since according to them it is an irreducible brute fact about the natural world. My suggestion is also to reject the need for a definition, but for the exactly opposite reason: there is no such genuine property in the world, and so any definition is a purely linguistic and entirely arbitrary affair (which does not prevent it from possibly having some useful practical purposes and applications, of course). Life, I submit, is better understood as not being a natural kind.

§5 The argumentative strategy used in the preceding section is now familiar to us: we should not postulate the existence of something (an entity or a property) if we do not really need it. This eliminativist view about life then parallels (but, as an isolated issue, does not entail nor require), a general brand of eliminativism such as the one we have seen in Part I, and such as the one that Van Inwagen defends as well. A generalized eliminativist view then not only claims that we do not need to postulate the existence of a genuine property of being alive that would be a natural kind, but that we even do not need to postulate the existence of many other things like trees, tables, planets, cells, and so on. The common idea – easily available to Van Inwagen himself – to both eliminativist strategies (the restricted one about life and the general one about all ordinary objects) is that without having to postulate such entities or properties, we already have all we need to account for all phenomena

76  The No-Self View that need to be accounted for: we do not need the property of being alive because we have all the interesting properties of cells, Data, and so on, and we do not need chairs because we have simples arranged chairwise. As I am now typing this sentence on my computer, it seems to me that I see a chair and a table in my visual field. But, as we have already seen, my visual experience would be exactly the same even if there were no chair and no table and if there were ‘only’ simples arranged chairwise and tablewise. Indeed, the visual experience I have is caused by light reflected by the atoms and this reflection would be exactly the same if there were a chair and a table. The idea here is that our sensory experiences can be accounted for in terms of some more basic and genuinely fundamental (and existing) entities, and so there is no need to postulate a further entity (a chair, a table). In the case of allegedly living things, the analogous (but not exactly the same) claim is that our observations of, interactions with, and experiences of living things can all be accounted for in terms of the genuine properties they have (such as, being able to reproduce, being based on such-and-such biomolecules, being able to crawl, etc.), and consequently there simply is no need to postulate the existence of a further property or a further natural kind of being alive. Remember, as we have also already seen, that general eliminativism does not make us abandon ordinary language either. Indeed, the eliminativist can say that the word “chair” denotes a plurality of entities (a plurality of simples) in the same sense the expression “the USS Enterprise crew” denotes a plurality of entities, rather than a single entity composed of the crew’s members. The sentence “There are no chairs” is then true when it is a sentence uttered in Ontologese (the fundamental language of the metaphysician), but this is compatible with the sentence “There is a chair in my office” being true since it is a sentence in ordinary English. In Ontologese, it is true that there are simples arranged tablewise in my office (that is, in ‘simples arranged officewise’), and this is all that is required for the ordinary English sentence “There is a chair in my office” to be true. This is exactly the right thing to say when it comes to ‘defining’ life. We can, and we do, in our ordinary language talk about things which are alive and things which are not, and to some extent such a talk can even be scientifically useful, for instance if one starts with an arbitrary definition of life and tries to search in the universe for things that fit it. But we cannot, and we should not, think that any such definition picks out a genuine property of being alive in the world, or that it could. To sum up, the best explanation of why there is a problem with finding a satisfactory definition of “life” is that being alive is not a genuine property. If it were in need of being summarized in a slogan, my claim would then go like this: ‘nothing is alive, we only say so’. The property of being alive simply doesn’t exist, and the best reason for thinking this is that we don’t need it. As a consequence, “life” is not a good place to make an exception, and eliminativism is a better strategy here than realism, in the same way it is a better strategy elsewhere as well.

Exceptionalism  77

§6 We have seen above some local and ad hominem reasons to reject the need for exceptionalism, focusing on a part of Merricks’s argument and on life as being an allegedly good place to motivate an exception to eliminativism. None of this is of course conclusive in the sense that there might be other reasons to make exceptions, and these would be untouched by the considerations above. Nonetheless, one can see here that exceptionalism in general, whatever its local detailed content might be, will always be a source of worries since any exception will need to be justified and motivated in a non-arbitrary way (that is, one must avoid here any kind of arbitrariness that affects ordinary objects, as we have seen in Part I, Chapter 1). Nonexceptionalist eliminativists will thus always have a simpler life – that is, if (and only if) they can show that they do not leave something important out. Indeed, the virtue of eliminativism can only be acknowledged if it can be shown that we really do not need the allegedly existing additional entities or properties. “Life” is a good example of such a situation, as we have just seen: at the end of the story, nothing of importance is lost, nothing is missing. But the story must go on much further, if we want to see that eliminativism works, in a more general way, when it comes to the Self. This will be the task of Chapter 7 below. But before we get there, let us have a look in the next chapter at what the competition looks like.

Notes 1 See Part I, Chapter 4. For the sake of brevity, I’ll stick to the use of “simples” here and in the following chapters as well. 2 See Van Inwagen (1981); for a discussion see inter alia Heller (1990) and Benovsky (2006b). 3 I discuss one such case in detail in Benovsky (2013a, pp. 162–164). 4 I simply assume here, in agreement with Merricks, that such coincident entities are not acceptable. 5 I would like to thank Trenton Merricks for raising this point in a discussion. 6 I have focused here on four-dimensionalism understood as the view that objects like chairs or people are temporally extended entities, composed of temporal parts, since this is the view that Merricks appeals to in his argument, and since it is, to my mind, the best available version of four-dimensionalism. But note that another variant of four-dimensionalism, namely the stage view (see Sider [2001] and Varzi [2003]), can accommodate the truth of (C) even more easily, since according to this view conscious objects like persons are only instantaneous entities (that persist by having temporal counterparts at different times), and there is then no worry concerning amputation or fission cases since the person after the amputation or after a fission simply is a different person. 7 http://en.memory-alpha.org/wiki/Dikironium_cloud_creature 8 http://en.memory-alpha.org/wiki/Beta_XII-A_entity 9 http://en.memory-alpha.org/wiki/Data 10 http://en.memory-alpha.org/wiki/Q 11 See Deplazes-Zemp (2012). 12 See for instance Lewis (1986).

6 The Self A Substance or a Bundle?

§1 In the next chapter, I will defend the no-Self view. But before I do so, I want to have a closer look at the metaphysics of pro-Self views – that is, realist views about the metaphysical nature of the Self. The list of the different views about the Self would be a very long one, and the point here will not be to examine this list in detail. Rather, I want to outline here the two main traditional positions – namely, a Cartesian substance view and a Humean bundle view – and see how they work and how they compare to each other. In Benovsky (2008), I have argued at length that, in general, there is much less of a difference between the bundle theory and the substratum theory than what we thought, and when it comes to these two traditionally opposed views as applied to the case of the Self we will also see that they are largely merely terminological variants. In this way, we will see that it does not make any significant difference here to go for one or the other, and so, from the point of view of the metaphysician, the competition to the no-Self view shrinks to one enemy rather than two. On our way, we will see some of the core features that a Self is supposed to have, which the no-Self view will then be required to account for as well. Although I shall talk about “the Cartesian view” and “the Humean view”, I will not be concerned here with Cartesian or Humean exegesis. I think that what I will say about these views is (at least) very close to views Descartes and Hume actually held, but the exact interpretation of their writings is a tricky endeavour that I shall not attempt. Rather, I will be content if I can contribute to the discussion about the nature of the Self that arose from the opposition between their views on this issue. Thus, I shall leave aside any specialities of Descartes’s or Hume’s view – in particular, I will be examining the two theories in a more general way comparing the view that the Self is a bundle of properties and the view that it is a substance that has properties, which is neutral with respect to what sort of things one takes to be bundled or had: perceptions, ideas, mental events, thoughts, experiences, or other. What I am interested in is to compare these two views by examining their central features that are most commonly shared by usual variants of

The Self  79 them. Furthermore, I shall limit myself to examining versions of these views where properties are said to be tropes, rather than universals; for the case of the bundle theory combined with the claim that properties are universals requires a special additional treatment because of worries that arise with the principle of Identity of Indiscernibles (in Benovsky [2008], I include a discussion of this issue and I show that the arguments that I do discuss here do apply to the bundle theory with universals as well). There are two central and commonly shared features of these two views that have played a central role in the discussion, both of which can be quickly and usefully explained if one puts them as an objection to the bundle view. First, friends of the substance view have insisted that only if one conceives of the Self as a substance is it possible to account for genuine persistence of selves through time. In his recent book, Barry Dainton (2008a, p. 341) says that “we think of ourselves as being particular things, enduring objects – in traditional parlance, substances”1 where it even seems to be part of the meaning of “substance” that it is something that endures through time, which a mere bundle is traditionally said to be unable to do. Furthermore, this substance is something that is simple which guarantees the particularity of the Self: it is the ultimate bearer of properties that it has but that it does not contain, unlike a bundle. Thus, a substance, it is argued, is needed for two things: accommodate the idea that the Self is a particular thing that has properties at a time, and that it genuinely persists through time. In what follows, I will discuss in detail these features of the two theories, and I will also discuss a special case of persistence of Selves through time – the case of a fission of a Self, and I will ask, as Shoemaker (1997) did, how such a case can be handled by the two competing theories. The second central point of traditional disagreement concerns independence: it is often said that only a substance, but not a mere bundle, is independent enough of its properties to play properly the role of a Self, and I will have something to say about this. Concerning all these points that I take to be central to the debate and that are indeed such that they are often taken to define the views at hand, my thesis will be a meta-theoretical one: contrary to appearances, both views can accommodate all the above-mentioned points (particularity at a time, persistence, fission, independence) in the same way, and I will examine two possible conclusions to be drawn from this – that is, either that the differences between the two views are no more than terminological and that the two views turn out to be equivalent, or that the differences are metaphysical but that it is epistemically under-determined which one of the views we should choose. Such a conclusion(s), especially the former one, may seem strange at first sight, since as metaphysical theories go, these two rival views could hardly seem be more dissimilar and opposite to each other. My aim here is then twofold. First, I want to make some progress with respect to the substance theory and the bundle theory of the Self, and I want to do this mainly by

80  The No-Self View examining the central points of alleged disagreement between them mentioned above. Second, I want to defend the meta-metaphysical claim that the two allegedly very different rival views are much less different than what we thought: their structure is extremely similar, their strategies are extremely similar, they can both face their theoretical challenges (particularity, persistence, fission, independence) in the same way, so that, as we will see, some central objections to one side always have a sneaky tendency to reappear for the other side as well.

§2 The substance theory of the Self claims that a Self is not just a collection of thoughts, ideas, perceptions, and so on; rather it is the thing which possesses them. The Self is an ultimate subject of these attributes, an “ultimate particular” as Gallie (1936, p. 29) put it. Thus this view can be put as a claim about what the relationship between the Self and its properties is: there is the Self and there are its properties that are had by the Self that is conceived of as being the bearer of those properties. This view of course parallels a general theory of substances where such a bearer of properties, which has its identity independently of the properties which it bears, is often called “an underlying subject”, a “substratum”, a “substance”, or a “bare particular”. Thus, analogously to what the general substance theory says about objects like tables, the substance theory of the Self sees it this way: there are properties (thoughts, perceptions, experiences, . . .) and a substance (a ‘bare Self’) that supports them and glues them together in order to make up a person (i.e. something that has particular thoughts, experiences, . . .). The Humean bundle theory claims that there are only properties and denies the existence and the need for any such substance: a Self is then taken to be a bundle (a cluster, a bunch, . . .) of its properties. The bundle theory thus claims that there is only one kind of component needed to make up Selves, instead of two as the substance theory has it, and that Selves are just bundles of properties which are held together (glued together in order to make up a person) by a special property (an n-adic relation, where n is the number of properties) often called “co-personality”. As already announced, I shall discuss and compare the substance theory and the bundle theory under the assumption that properties are tropes rather than universals, for reasons mentioned above. The bundle theory appeals to a bundling relation that is labelled in different ways like “copersonality”, “co-consciousness”, “consubjectivity”, “compresence”, “consubstantiality”, “co-instantiation”, “togetherness”, “collocation”, etc. The fact that there are so many different labels for this relation does not mean, however, that there are many different analysis of the nature of this relation. Indeed, typically, the co-personality relation is taken to be unanalysable and primitive. Thus, it is important for what follows to note that this relation is defined and individuated not by its nature or intrinsic features

The Self  81 of which we are not told much by the bundle theory, but rather by its theoretical role: it is a unifying device, a device that takes properties to make up Selves. While many friends of the bundle view only implicitly assume the primitiveness of their bundling relation, Barry Dainton, for one, claims it to be the central notion of his theory explicitly: “The key relationship is co-consciousness, a primitive but real relationship between experiences” (Dainton [2008a, p. 347]). In a parallel way, the substance theory uses properties and a bearer of properties that is labelled in different ways like “substance” (a tricky word in philosophy, but that I shall continue to use nonetheless in order to stick to tradition – but, be careful to avoid confusion with an Aristotelian substance theory, which is quite a different story2), “substratum”, “naked particular”, “bare particular”, “thin particular”, etc. As for the bundle theory, the fact that there are so many different labels for this bearer does not mean, however, that there are many different analyses of its nature. Indeed, typically, it is thus defined and individuated not by its nature or intrinsic features of which we are not told much by the substance theory, but rather by its theoretical role: it is a unifying device, a device that takes properties to make Selves. It is then I think very important to note that both the bundle theory and the substance theory contain a unifying device that is primitive and underdefined, and whose purpose is to tie together properties of a single Self. Paraphrasing Locke, who was interested in the general case of the bundle theory and the substance theory of material objects, in both cases this unifying device is a “we-know-not-what” . . . but it is a “we-know-what-it-does”, that is, we know its theoretical role. The parallel between the two unifying devices also shows in the way the substance view is often objected to. Indeed, an objection to the substance view goes: “In virtue of what is one Self, one ‘bare’ substance, distinct from another Self? No attributes or properties can distinguish between them, since they are bare!”. But I think that the very same question can be asked about co-personality: “Tu quoque: In virtue of what is one co-personality relation (involved in the bundling of a Self A) distinct from another (involved in the bundling of another Self B)? No attributes or properties can distinguish between them”. Both views, I submit, answer these questions by a primitivist claim. Let us now turn our attention to the case of persistence of Selves through time. There is allegedly an importantly different way in which Selves conceived of as substances can persist through time while Selves conceived of as ‘mere’ bundles cannot. Here is Peter Van Inwagen on this: One who took the general Humean line might of course say that the word ‘I’ referred to some collection of these qualities, but collections of ideas aren’t really suitable candidates for the referent of ‘I’ (or so it might be argued) because it is part of the meaning of the word ‘I’ that

82  The No-Self View its referent is something that persists through changes of qualities, and that is just what collections of qualities don’t do. (Van Inwagen [2002, p. 176]) This objection parallels a general traditional complaint against the bundle theory of material objects, that can be found, for instance, in Van Cleve (1985, p. 122): If a thing were a set of properties, it would be incapable of change. For a thing could change its properties only if the set identical with it could change its members, but that is impossible; no set can change its members. Taking an example of an individual that is supposed to change one of its properties over time, he adds: “what we have is replacement of one individual by another, not change in the properties of one and the same individual” (Van Cleve [1985, p. 124]). The objection certainly has some clear initial force: if a Self is a bundle of properties then even if only one of the properties making up the bundle changes, the bundle itself changes and consequently the Self that is the bundle changes as well – it is simply not the same entity, and we have a case of an entity that ceased to exist and another entity that took its place, which has the unwelcome consequence that selves cannot genuinely change and cannot persist through time because they cannot persist through intrinsic change. The substance view is supposed to be able to avoid such objections, but actually it does not. What could amount to genuine change under the substance view? The substance itself (the ‘bare’ entity) does not change since it does not have in itself any properties it cannot undergo qualitative change at all, and the ‘thick’ Self composed of the substance and the properties it has cannot genuinely change for a reason that relevantly parallels what we have seen above concerning why the Self conceived as a bundle cannot change: there is no change, rather there is replacement of one ‘thick’ particular by another. True enough, the substance view can claim that the substance remains the same over different times, and that this guarantees that the Self, while changing its properties, is the same Self, but if this were an acceptable reply, then the bundle view can give exactly the same: the co-personality relation remains the same over different times, and this guarantees that the Self, while changing its properties, is the same Self. Both views thus have a primitive unifying device that allows them to do the job: both views can answer the objection in the same way by appealing to their unifying device, and calling the device different names (“co-personality” or “substance”) does not change anything relevant since both unifying devices just play the same role in the same way – that is, in a primitive way. Again, this is something

The Self  83 that is often only passed along in silence, but Jonathan Lowe, for one, puts it as clearly as one can wish: “The simplicity of the Self is seen to imply that its diachronic identity – its persistence through time – is irreducible and ungrounded, and hence criterionless” (Lowe [1996, p. 10]); “the diachronic identity of simple substances, including the Self, is primitive or ungrounded” (Lowe [1996, p. 41]). In order to get a better and more precise grip on this claim, I shall in the next section examine the two main strategies there are to face the problem of persistence through time, namely a version of perdurantism and a version of endurantism, and see whether there is any difference between the use of a substance or a co-personality relation. (As we shall see, there isn’t.) The general idea behind my claim here is that, as we have already seen and as we shall see in more detail below, both views appeal to their unifying device which is what I will call a “problem-solver”. Simply put, a problem-solver is something that is a primitive in a theory and that solves a problem (in our case, the problem of how a Self can persist through time and undergo change in intrinsic properties). Perhaps, every primitive in every theory is a problem-solver – for why do we introduce primitives in the first place, if not for them to do an explanatory-power job? And how do they do this explanatory-power job? By having a primitive capacity to do so. The idea is simple: how can the theory account for a Self to be one and the same at different times while having different qualities? By having a primitive unifying device that is primitively the same over time (in a sense we will see in more detail below). The premise that there are such primitive unifying devices is thus a “problem-solver” in the sense that without it the theory would not be able to face the theoretical challenge of accounting for change over time for Selves and that it succeeds to do so only in virtue of the postulation that it can do so. The latter claim may sound a bit pejorative, but it is not: every theory has its primitives and every primitive is, at least to some extent, a problem-solver. As I see it, the use of problem-solvers is commonplace in all philosophy, and without it we would not get very far – it just is one among the components of the philosopher’s toolbox. The conception of primitives I have in mind here is a view about their nature that takes very seriously the functional role they play in the theory. By its very nature, a primitive being primitive, it is non-analysable and we are not really given any information concerning its nature; we are told what it does rather than what it is. So it is what it does that counts – after all, that’s what any primitive is introduced for in a theory in the first place (otherwise there would be little justification for having it). Thus, primitives are individuated by what they do, what their functional role in a theory is, and as a consequence two primitives that do the same job just turn out to be equivalent for all theoretical purposes. In what follows I intend to show in more detail that this is indeed the case for the unifying devices ‘co-personality’ and ‘substance’ with respect to the problem of change of Selves over time, and see what consequences follow from that for the two theories at hand.3

84  The No-Self View

§3 If one combines perdurantism and the bundle theory, one gets a bundlebundle theory. Let us take the example of Jean-Luc Picard, captain of the starship Enterprise, who is first happy, because the Enterprise is getting a new warp engine, and later he is sad because the engine does not work as it should. Under perdurantism, Picard is a collection of his temporal parts, each numerically and qualitatively distinct from the others. It is then by having different temporal parts at different times that Picard is said to be able to undergo qualitative intrinsic change. Schematically, the picture looks as follows, where “C” stands for “co-personality”: Picard t1-part

t2-part

t3-part

F, G, Happy

F, G, Sad

………..

C

C'

C''

Now, what happens if we combine perdurantism with the substance view? Interestingly, the picture we get is not really different, except that I had to change “C” into “S” (for “substance”): Picard t1-part

t2-part

t3-part

F, G, Happy

F, G, Sad

………..

S

S'

S''

Consequently, we now have a case for the claim that the only difference between the two views is terminological: both views contain a unifying device included in the momentary temporal parts of Picard, and saying that in one case it’s co-personality and that in the other case it’s a substance does not make any difference in the way this unifying device achieves to play its theoretical role (more on this at the end of this section below). Let us now have a look at the endurantist alternative. According to endurantism, Picard persists through time and through intrinsic change by existing wholly at different times, rather than by having temporal parts. Under this view, the Self is thus said to be genuinely one

The Self  85 and the same at different times: one and the same (numerically identical) substance exists wholly at t1 and t2 and has the two incompatible properties of being happy and being sad. In order to avoid a contradiction, endurantists will typically embrace some kind of temporal indexation strategy, as for instance Peter Van Inwagen (1985) does. According to this indexicalist version of endurantism, Picard does not have incompatible properties, for instead of having properties like “being happy” he has time-indexed properties like “being-happy-at-t1” and “being-sad-at-t2”, and these are non-contradictory. Thus, since the indexicalist will claim that all properties are always indexed, no contradiction can ever arise from intrinsic change of an object that is numerically one and the same at different times, as the endurantist claims. Combined with the bundle view, the endurantist claim thus yields a picture of Picard as being a bundle of time-indexed properties. (Such a picture parallels closely the endurantist picture Peter Van Inwagen draws [see Van Inwagen (1985, p. 195)]). If combined with the substance view, as before, the picture will not look very different: instead of having as a unifying device the relation of compresence, we will have as a unifying device a substance that has all of the time-indexed properties that are, under the bundle view, tied by co-personality. Until now, I have examined a ‘normal’ case of persistence of a Self through time involving simple change in intrinsic properties. Interestingly, Shoemaker (1997) discusses a fictional case of fission (and also of teletransportation) of persons and suggests that friends of the substance view have the natural tendency to regard these cases as not person-preserving, while friends of the bundle view typically find it natural to say that they are. If so, it could mean that there is some difference with respect to persistence through time of selves as substances and as bundles that only becomes apparent in these extravagant cases. But I think that there is no reason why one party should find it more natural than the other to claim that such procedures are person-preserving or not – rather, such a claim is independent of the choice between a substance or a bundle tied by co-personality; it is for independent reasons that one might think (or not) that the procedure is not person-preserving and then model the situation either in terms of substances or bundles, with no significant difference. To illustrate this, let us examine just one case of what might happen in a fictional case of fission of Picard. When Picard uses the transporter aboard the Enterprise, say just before his ship gets the new engine installed, let us suppose that the transporter malfunctions and as a result Picard undergoes a fission where two individuals are standing there after the ‘transport’. Suppose then that one of them is sad because the new engine is not working, but the other is happy because for some reason he just was not told about the new engine’s failure. There are then four possibilities: both of the resulting post-fission selves are Picard(s), none of the two resulting selves is Picard, or one of the two is Picard and the other is

86  The No-Self View not. All four possibilities can be equally well modelled and accounted for by the substance view and the bundle view; let us quickly see how this looks for the first option, the other three cases being easily done in a similar way. If both post-fission selves are said to be Picard(s), this is how it looks under perdurantism and the bundle view:

t2-part F Happy t1-part

C'

F G t2-part

C

F Sad C''

According to perdurantism, the case of fission is here no more than a case of sharing an overlapping temporal part. The bundle view can easily provide such an account as pictured above, and the substance view can do the same – just replace the Cs with Ss, in the same manner I have done it in the normal case of persistence through time before. There is no reason why one could not do this as simply as that. Under endurantism and the bundle view, the two Picard(s) are simply two different bundles: Sad-at-t2

Happy-at-t2 F-at-t1 F-at-t2 G-at-t1

C1

F-at-t1 F-at-t2

C2

G-at-t1

Here again, it takes no more than replacing the Cs with Ss to get the substance view picture. It is thus easy to model under both views the idea that a fission case is person-preserving and that both post-fission resulting persons are Picard(s), and it would be equally easy, using the same strategy, to model the other options. So, the question whether fission or teletransportation procedures are person-preserving must be decided on other grounds (most likely, the answer will depend on the nature of the relation that ‘glues together’ temporal parts of worms like Picard under perdurantism, and on

The Self  87 the nature of criteria for diachronic identity under endurantism), and then modelled in one or the other way, using with equal efficiency substances or bundles. The point that arises from all these cases is, as we have already seen, that the substance and the co-personality relation play the same theoretical role. Both the bundle view and the substance view use their unifying device in the same way, and so they have the same means to face the case of intrinsic change over time and the case of fission. It seems then that the difference between them is merely terminological – one has a unifying device called “C” and the other has a device called “S” but since both devices are theoretical entities (they are there to do some theoretical work) and are thus individuated by their theoretical role, and since they play their theoretical role in the same way, they just seem to be one and the same thing under different disguises. And if that’s the case, there just does not seem to be any real difference between the bundle view and the substance view. But let us not be too quick. Granted, the substance and the co-personality relation do the same theoretical work, but perhaps it is not so obvious that they are ‘just’ theoretical entities: even though they play the same theoretical role, they are metaphysically different things. If one thinks that theoretical equivalence entails metaphysical equivalence, one will embrace the strong metaphysical equivalence thesis. (I myself feel strongly inclined to embrace this strong claim, for if there is a theoretical equivalence, that is equivalence with respect to what substance and copersonality can do, what would justify the claim that there is no equivalence with respect to what they are? The very reason to postulate their existence was to get a theoretical job done!) But if one thinks that theoretical role does not exhaust the nature of substances and relations like co-personality, one may only wish to embrace the weaker conclusion that, given that both substance and co-personality do the same theoretical work, it is epistemically under-determined which one of the bundle view or the substance view we should choose. The idea behind this weaker claim is that, granted, C and S are metaphysically different entities, but they play the same theoretical role in the same way, and consequently the bundle view and the substance view have the same explanatory power (as far as we metaphysicians are concerned, they both do the job we want them to do).

§4 When asking myself why anyone would want to insist that, despite their playing the same theoretical roles, the substance and the co-personality unifying devices are metaphysically different entities (see the Weak Conclusion above), I come to think that it might be because one sees an important difference between substance and co-personality in terms of their independence. This is often put as an objection to the bundle view: the Self has to be a substance since only a substance is ontologically independent enough

88  The No-Self View to play such a role – ideas and experiences have to be had by something, they cannot “float free” without any supporting subject, and the relation of co-personality is not of any help since it is itself also no more than a property. A different way to express this worry is to say that properties, and bundles of properties, are not ‘substantial’ enough to play the role of a Self or even to exist without being had by any substance. So, this can be used either to say that the bundle view does not provide enough independence or ‘substantiality’ to serve as a theory of Selves, or worse to say that the bundle view in general fails altogether. (And of course, the same objection could be raised against the no-Self view as well, as we shall see in the next chapter.) But what exactly does “being independent” or “being substantial enough” mean here? Most often, in the mouths of the critics, it means something like “is ‘ontologically stronger’ than the other elements (properties) that are themselves not substantial enough to sustain a Self’s existence, and is itself such that it can confer the ‘substantiality’ we need” – which is something that co-personality is said not be able to do, since it is itself no more than a property. To my mind this is no more than an unfortunately very familiar prejudice against properties, relations, and bundles. What I call prejudice, Hawthorne and Cover call ‘incredulous stare’ (while speaking about the bundle theory combined with universals): Perhaps some philosophers will claim to find it just self-evident that universals are had by something. We don’t have much to say to such philosophers. We do note, however, that the polemic against the bundle theory has rarely taken the form ‘It is simply self-evident that anything quality-like is directly or indirectly predicated of something that isn’t like a quality’. If opponents of . . . the Bundle Theory wish to retreat to this form of an incredulous stare, so be it. (Hawthorne and Cover [1998, §2]) Barry Dainton in his The Phenomenal Self also makes it clear that he does not see why bundles of properties that are suitably unified by a copersonality relation would not be ‘substantial enough’ to serve as selves (his bundles are “C-systems”, the details of which I unfortunately do not have the space to discuss here): Entities such as pigs and planets may well satisfy the unity criterion and the independence criterion, so they might reasonably be thought to be substantial in a way subjects (construed as extended C-systems) are not. But by virtue of satisfying the unity criterion in so resounding a manner, subjects are undeniably entities, indeed substances, of an eminently respectable sort. (Dainton [2008a, pp. 347–348])

The Self  89 Yet a different and interesting way to address this issue can be found in Galen Strawson’s “The Self”: “But if there is a process, there must be something – an object or substance – in which it goes on. If something happens, there must be something to which it happens, something which is not just the happening itself”. This expresses our ordinary understanding of things, but physicists are increasingly content with the view that physical reality is itself a kind of pure process – even if it remains hard to know exactly what this idea amounts to. The view that there is some ultimate stuff to which things happen has increasingly ceded to the idea that the existence of anything worthy of the name ‘ultimate stuff’ consists in the existence of fields of energy – consists, in other words, in the existence of a kind of pure process which is not usefully thought of as something which is happening to a thing distinct from it. (Strawson [1997, p. 427]) I have now used arguments from authority and my gut feeling to claim that a bundle of properties unified by co-personality can equally well satisfy requirements for independence and ‘substantiality’ to play the role of a Self as well as a substance that has properties does it. Perhaps my main reason behind this unrigorous way of arguing for it is simply that to my knowledge there is no argument that exists that would show that a bundle is indeed guilty of the accusations that are made against it,4 apart from ontologically prejudiced and un-argued for complaints. This being said, let me say something more about the Strong Conclusion, that is, the claim that the bundle view and the substance view are no more than terminological variants. To illustrate it again with the case of fission, here is how I think that it works: first, for independent reasons, a philosopher more-or-less explicitly decides whether it is a desirable feature of her theory or not to allow for person-preservation in such a case (and for persistence through time in general); second, she has to decide how to accommodate it and model it and does this by incorporating into her primitive problemsolver the power to do so; and third, she has to make her problem-solver graspable by others and express it in a way that conveys well the concept she has in mind – here terminology plays an important role, since by calling her problem-solver a “substance” the theorist conveys perhaps better the idea of substantiality and independence or sameness across time, or other relevant features of an existing and persisting Self that she wants to insist on (and by calling it a “bundle” she may insist on other features, like the idea cherished by Hume that one is never aware by introspection of the unifying device). Thus, as we have seen, although both the bundle view and the substance view are no more than different ways of expressing oneself and both can equally well fulfil the same role, there is a choice between the bundle view and the substance view that is a choice between alternative ways of

90  The No-Self View formulating the same thing, where nothing really depends on the formulation, except that of course one formulation can be better than another in order to express in a more understandable way what one wants to say.5

Notes Dainton actually rejects this view. 1 2 For a discussion, see my Routledge Encyclopaedia of Philosophy entry “Substance” (forthcoming online, www.rep.routledge.com). 3 One may think that co-personality is not primitive since it can be analysed in a Humean fashion in terms of resemblance and causality. But this will not do since distinct selves can also enter in close relations of similarity and causality, while being distinct. To make them apart and count them as distinct bundles they must therefore be unified by something else than just resemblance and/or causality, so primitive bundling is required. 4 Quite recently, Donnchadh O’Conaill gives it a try in a draft manuscript “The Identity of Experiences and the Identity of the Subject”. 5 Cases of metaphysical equivalence like this one can lead to an anti-realist metametaphysical view; I defend such a view in Benovsky (2016b).

7 The No-Self View

§1 Having indulged in long preliminaries in the two preceding chapters, let us now get to one of the main points of this book: the no-Self view. To avoid a misleading start, let me first say that the strategy here is not to say that physicalism in the philosophy of mind is true, i.e. that mental states and properties can be reduced to brain states, and that brain states can be eliminated in virtue of the existence of simples arranged brain-states-wise. This could be a way to go, perhaps, but this is not how I plan to tackle the issue concerning the alleged existence of the Self. Indeed, this reductionist/ eliminativist physicalist claim can be held in addition to what I am going to argue for below (although my own version will involve dual-aspect monism rather than physicalism, as we shall see), but mostly what I want to do here is to show that we can eliminate the Self from our ontology in virtue of the existence of individual impermanent mental states – and whether these can then be further reduced to physical simples is an additional, and different, question. Thus, most of what I am going to say in this chapter concerns the phenomenology and the metaphysics of the Self, rather than a discussion of eliminativism in the field of philosophy of mind (that is, I will say nothing about eliminativism such as discussed in Churchland [1981, 1988], Churchland [1986], and Dennett [1978, 1988]). It would be nice to have something like a clear definition of ‘the Self’, before I start defending the view that it does not exist. But things are not that easy, of course. To begin, let me rather say what role we expect ‘the Self’ to play in our philosophical thinking – that is, what it is supposed to do as a piece of theory. First, we need to have an account of subjectivity, and this is where the Self is supposed to play its central role as being the subject of experience. This, I take it, is the main and strongest reason to believe in the existence of the Self. Second, the Self is supposed to account for the sense of unity and continuity of our experience. These are the two core ingredients that any theory of us must account for, and the best argument I can provide in favour of the ‘no-Self view’ is to show that it can account for these two desiderata. That’s exactly what I will do in this chapter.

92  The No-Self View To put it shortly, using the same overall strategy as eliminativism about ordinary objects, the no-Self view claims that instead of there being a Self in a metaphysically strong sense (like a substance or a bundle), there ‘only’ are experiences arranged Self-wise. In this view, nothing more than the existence of individual impermanent psychological states is needed. To defend this view, I will appeal to the notion of mineness, which I will discuss and elaborate in detail, and which I will intimately link to dual-aspect monism, to have a complete and coherent – and I hope quite natural and convincing – picture. We will see how such a view can satisfy the two desiderata above.

§2 The notion of ‘mineness’ – as well as cognate notions such as ‘for-me-ness’, ‘myness’, or ‘me-ness’1 – appears to be a key element in the understanding of subjectivity (see, for instance, Levine [2001], Kriegel [2003, 2009], Zahavi [2005]). As we will see, appealing to this notion will be of great service in order to get some welcome results concerning the metaphysics of the Self – the notion of mineness is indeed one of the important pieces of philosophical theory that allows us to build a well-behaved eliminativist view of the Self. The idea of mineness has roots in many philosophical traditions, famously tracing back to Sartre (1943) and the phenomenological tradition in general, as well as to a long tradition in Buddhist thought. Indeed, it has been argued in many places in Buddhist philosophy that eliminating the subject of experience, understood in an ontologically strong sense, does not amount to eliminating subjectivity. Dharmakīrti, a 7th-century Buddhist thinker, held a fully developed view of this kind.2 When it comes to us, in the Buddhist tradition, similarly to what Parfit (1984) as well as many other Western thinkers have in mind, there always only is a succession of impermanent psychological states, variously causally related – but, no ‘bearer’ of these states. The important point for Dharmakīrti here is that these states are always conscious states, which means, in his view, that they are selfconscious and self-reflexive. However, in Dharmakīrti’s view, they are not such in a relational way, since this would lead to a vicious regress where some kind of a subject would be needed to be conscious of the first-order states and then again another subject would be needed to be conscious of the first subject, and so on. According to Dharmakīrti, this self-reflexivity/selfconsciousness is a first-order feature of the various conscious psychological states. The phenomenology of conscious experience is thus understood here as involving in a non-relational way not only features such as, for instance, colour qualia, but also self-consciousness as a necessary constitutive feature of any conscious experience in the first place: if, per impossibile, an experience lacked this feature, it could not be a (conscious) experience at all. The ‘mineness’ or ‘first-personal’ character of individual psychological states is thus understood here as a first-order feature of these states – it’s simply part of their very nature to be that way. This is how, in Dharmakīrti’s view, there

The No-Self View  93 is room for subjectivity without a subject. We make a mistake, Dharmakīrti argues, when we reify this piece of our phenomenal experience and when we think of it as of a subject, a ‘bearer’, that is somehow different from the experience itself. To put it shortly, the view I will defend here is very close to what Dharmakīrti had in mind. As already announced, I will discuss the alleged need for a ‘bearer’ of individual psychological states, and I will discuss the issue concerning the alleged diachronic unity of the ‘Self’, only to conclude that nothing more than the existence of the individual impermanent psychological states is needed. But first, let us see in more detail what the notion of ‘mineness’ amounts to and what role it plays here. When I have a rich experience, as for instance when I admire the beauty of the Alps at sunrise, this experience contains various elements – reddish colours, shapes of mountains, the feeling of the wind on my face, the biting cold, the smell of the air, the cracking of snow under my feet, the lightheadedness coming from being located at a high altitude, etc. All of these elements, as well as the experience taken in its entirety, are distinctly mine. Where does the mineness feature of the experience come from? Here are four possibilities: (i) It comes from a sense of the Self, understood as a bearer of the various elements of the experience. (ii) It comes from a sense of unity that arises from the fact that the various elements of an experience are both synchronically and diachronically united by a ‘unifying device’, often understood as a relation of ‘compresence’ or ‘co-consciousness’, in a Humean way.3 (iii) It is one of the elements of the total – rich and complex – experience at time t (or at a short interval of time, such as the ‘specious present’4), on a par with the other elements. Perhaps, it can be understood in this way as a quale of a special kind that is one of the elements of an experience (like, say, redness is). (iv) It is an aspect of each of the individual elements of an experience. In this sense, it is not on a par with the other elements that make up the total experience. Rather, it is understood here as ‘accompanying’ or as ‘being an aspect of’ or as ‘being a part of’ each of the individual elements. Thus, each element has its own intrinsic what-it-is-likeness and it also necessarily has a mineness aspect. Option (i) is the one that I – following Dharmakīrti and others – want to reject. Option (ii) may look like it is perhaps less metaphysically heavy, and contains as a primitive unifying device a relation that does the job of primitively ‘gluing together’ individual psychological states, at a time and through time, but as we have seen in the preceding chapter, at the end of the day, (i) and (ii) really amount to the same kind of view. My aim here is to show that such pieces of metaphysical machinery are unnecessary. The

94  The No-Self View argumentative strategy is thus here to show that a view that does not appeal to pieces of theory such as a bearer (or any other ontologically reified unifying device) works and that it accommodates the desiderata we have for an understanding of what a Self is (supposed to be). In this way, I am not going to argue against (i) or (ii), rather, I want to defend the view that one can do without them, thus making them unnecessary. So, (iii) and (iv) are the interesting options here. Option (iv) seems to be phenomenologically more adequate than option (iii). In the total experience of the sunrise in the Alps, taken as a whole located at a time (or a short interval of time, i.e. the specious present), it is both conceptually and phenomenologically possible – and actually easy – to selectively concentrate on each of the elements of the total experience. I can focus on the biting cold, for instance, or on the particular colour the rising sun has on that particular day. In such cases of selective attention, it is always the case that each of the elements, considered in isolation of the others, not only has its particular qualitative character but it also has the aspect of mineness. It thus seems that the mineness feature of an experience is a feature of all of the elements of it – even the smallest ones that one can selectively focus on (conceptually and/or phenomenologically speaking). This is also what explains that one cannot selectively isolate, in one’s phenomenal experience, the mineness itself. Indeed, mineness seems to always accompany one of the qualitative elements of an experience, but is not such one element itself, and in this way it is impossible to select it and focus solely on it, independently of any other element. This is perhaps also what makes this aspect of one’s experiences so elusive and mysterious. This is even perhaps the place where we can locate the mystery that surrounds our understanding of the nature of consciousness – the mineness aspect of each of the elements seems to be a genuinely primitive and individually unanalysable aspect of experiences. We need help here. We need help because we need to get a better grip on what an aspect is in this context. In order to do this, I will now make a detour and examine a view that is rarely embraced but that is very compelling in the discussion of the mind-body problem, namely, dual-aspect monism. I shall start by briefly discussing some general motivations for dual-aspect monism, in order to understand what the view is all about, and I will then offer an understanding of what an aspect is – and what it isn’t. I will then come back to the notion of mineness and see what a dual-aspect view of experience can look like. This will then provide us with an account of subjectivity without the need to postulate the existence of a Self.

§3 When it comes to the mind-body problem, dual-aspect monism can be put forward as a cure to problems that the other – more mainstream – views have.

The No-Self View  95 To put it very shortly, physicalism builds upon the central idea of a causal closure of the physical world, and it succeeds in this respect perfectly well. But, critics object, the price is quite high: by reducing the mental to the physical, physicalists overlook what makes the mental to be mental. They seem to leave behind the phenomenal character of experience, its so-called ‘what-it-is-likeness’. Feeling pain, the objector insists, is simply not the same thing as there being such-and-such a physical activity taking place in my brain. The idea here is that mental/phenomenal properties or states are irreducible, and that if we reduce them, we lose their phenomenal character, and we thus lose them entirely. On the other side of the spectrum, dualism is built around the core idea that mental states and properties are not reducible to physical states and properties. They are of different ontological categories, and dualism thus avoids the above-mentioned problem raised by physicalism. But, here again, the critics say, the price is also very high: the interrelations and interactions between the two ontological realms are very much under-explained and rather mysterious. Mental states/properties and physical states/properties interact a lot, and they are very intimately (cor)related: a pain is always found in the presence of some kind of physical process. Thus, under dualism (both property dualism and substance dualism), the mental and the physical are genuinely ontologically separate and different, but they are somehow strongly united – and the explanation of this union is loudly lacking. How can a material, spatio-temporal body interact with a mental entity, unlocated in space? There are many answers to be found throughout the history of philosophy (Descartes’s causal interactionism, Leibniz’s pre-established harmony between the mind and the body, Malebranche’s occasionalism, or epiphenomenalism), but they all are as interesting as they are unconvincing. In short, dualism suffers from the problem of mental causation. This is not the place for a detailed discussion of physicalism or dualism, and the rough and sketchy way of introducing the main well-known issues concerning these attempts at solving the mind-body problem is meant here only as a means to introduce dual-aspect monism. When asking how many (types of) entities there are when I have a toothache, dualists say “two” while physicalists say “one”. Dual-aspect monism also says “one” but without embracing any kind of reduction, contrary to physicalism. The core idea of dual-aspect monism is this: there is no ontological or conceptual priority, both the mental and the physical are on a par. Tracing back to Spinoza’s Ethics, Russell’s neutral monism is perhaps closest to this view, which is also discussed by Chalmers (2003, §11), and of which a different variant is defended by Strawson (2003, 2006). I’ll try to put forward here a similar but different version of dual-aspect monism, as simple as possible. The idea is this: the body and the mind are two aspects of one and the same thing, and they do not reduce to each other. My toothache is one thing that has two aspects: a mental one, and a physical one. The physical aspect corresponds to an activity in my body (most relevantly, my brain), which can

96  The No-Self View be monitored in an ‘objective’ way by neuroscience tools, while the mental aspect corresponds to the qualitative way it is given to me, with its phenomenal character. The main insight of dual-aspect monism is that these two aspects are different aspects of one and the same metaphysical reality. The toothache is thus not physical, and it is not mental either, rather we could say that it is ‘phental’. One way to understand this locution is to say that it is somehow both-mental-and-physical (say it quickly, all in one breath!), by being dual-aspectual (but not by being mental and by being physical, in a separable way). Another option here would be to follow Russell’s neutral monism, based on the claim that it is neither mental nor physical, and that it is neutral. Perhaps it does not matter so much to choose between these two options (and perhaps, at the end of the day, they collapse entirely), although the first seems to me more in the spirit of dual-aspect monism, and I’ll keep using the term ‘phental’ when needed. Now, what is an aspect? It is not a property – we are not in property dualist territory here. The idea is not that the toothache has a higher-order property M (corresponding to the mental aspect) and that it also has a higher-order property P (corresponding to the physical aspect), and that somehow these two higher-order properties are intimately related. This would be missing the point of dual-aspect monism entirely. Rather, the idea is that when I have a toothache, I have one property of ‘having a toothache’ and there are two aspects to this situation. Having a toothache is a phental property (or, a neutral property, if you prefer Russell’s reading). I myself, as a person, am a phental entity, for that matter. I am a phental entity exemplifying a phental property, that’s how the situation is to be understood from the metaphysical point of view. Thus, according to dual-aspect monism, ‘having an aspect’ and ‘having a property’ are two very different things. Having two aspects is just part of the nature of my toothache. ‘Having a toothache’ is the only property around, and this property has two aspects (which, again, are not high-order properties). Thus, we can negatively characterize an aspect by saying what it is not: it is not a (higher-order) property. How can we characterize it positively? What is it? The point here is precisely that it isn’t anything at all – we should not reify aspects. If we did reify aspects, that would make them suspiciously look like properties, but the idea is not to think of aspects as properties in disguise. The ‘aspects-talk’ is just talk, a way of speaking. A way of speaking to say: X is phental. This is the only positive, ontologically committing, and correct thing to say. That’s what there is. Talking about aspects is a useful way to talk about how things are, but – again – this talk should not be reified, otherwise the whole point of it is lost. One way to better understand what dual-aspect monism amounts to, is to see what it is good for: indeed, it avoids the two main issues with physicalism and dualism. There is no reduction here, the mental is not reduced to the physical (or vice versa). The mental and the physical are both equally aspects of one and the same thing. There is no ontological priority of one

The No-Self View  97 over the other, and they are both existentially dependent on the one thing that ‘exhibits’ them (say, the property of having a toothache). Relevantly to the problems that physicalism has, dual-aspect monism does not reduce the mental aspect to anything and it thus can preserve its genuine and irreducible qualitative mental character – its what-it-is-likeness (and, as we shall see below, its subjective character as well, but let us not put the cart before the horses). Relevantly to the problems that dualism has, dual-aspect monism does not generate any worries about mental causation. There is no mystery about how the mental and the physical can ‘interact’ since metaphysically speaking there is only one – phental – entity, and not two separate ontological categories. Dual-aspect monism, being a monism, does not need to postulate any cross-categorical relations. Furthermore, strictly speaking, there are no causal relations at all between mental properties/states/events and physical properties/states/events, given that, in this view, these are aspects of one and the same thing, and consequently there is no causality between them. In short, since there is only one ontological category, namely the realm of the phental, the mental causation worry simply does not arise. Causal closure of the physical world is to be abandoned of course, since the main claim here is that the world is not physical but that it is phental. But such an abandonment is not a problematic one, since what counts when it comes to the causal closure principle is not the claim that the world has to be physical, but rather the claim that causality always has to occur between entities of the same ontological category (whatever its nature is). Dual-aspect monism is in fact the most natural explanation of how and why there is a strong correlation between the mental and the physical, since according to this view these are aspects of one and the same thing, and it is then only natural that they ‘go together’.

§4 I hope that this short discussion of dual-aspect monism accomplishes two things. Firstly, I hope that it makes the view at least prima facie plausible. Secondly, I hope that it provides at least some understanding of how such a dual-aspect view works, and of what an aspect is and what it isn’t (especially, we saw that it isn’t a property, or a high-order property). Until now, then, we have seen that when I have a toothache, there are two aspects to this, a physical one and a mental one. But the mental aspect does not merely consist in what is often called the “qualitative character”, or the “what-it-is-likeness”. It also has a distinct subjective character, in addition to its qualitative character. This is where the idea of ‘mineness’ comes from. Mineness is a phenomenal feature that grounds a pre-reflective sense of ownership that is an invariant component of phenomenal character. To repeat what I already said earlier, when I have a complex experience (say, when observing a sunrise in the Alps) all of the elements of the experience as well as the experience taken in its entirety, are distinctly

98  The No-Self View mine. There is a quality to the experience I have, and also a subjectivity – there is something it is like for me. This mineness character of experiences always ‘feels the same’: all qualitative elements of experiences change with the time passing, but the subjective character is invariant, it is a feature that is common to all my experiences. But it would be a mistake to think of mineness as being one of the elements of my overall experience, on a par with the others. Mineness is not a quale of a special kind. Rather, it ‘attaches’ to any and all of the qualitative elements of any experience. It is an aspect of any element of any experience. Thus any such element, like the perceived particular colour of the rising sun, has a qualitative aspect and a subjective aspect. I can selectively focus on each element of the total experience, and any such element of an experience, considered in isolation of the others, has its particular qualitative character, and it also has the subjective aspect – its mineness – but I cannot selectively isolate, in my experience, the mineness itself. This does not mean that mineness is phenomenologically inaccessible – on the contrary, it is a pervasive and always present aspect of any experience. When Hume (1738/1978) says that he always ‘stumbles on particular perceptions’ and that he can never ‘catch himself’ via introspection, it is not the mineness aspect of experiences that he finds to be absent; rather he has in mind something like a Self, in the sense of a reified owner of experiences, and that’s quite a different story, as we shall see. Mineness is thus best understood as a (necessary) aspect of any qualitative element of an experience, but it is not such one element itself, and this is why it is impossible to select it and focus exclusively on it. We can – and we should – make a conceptual distinction between the qualitative character of (an element of) an experience and its subjective character (its mineness), but this does not mean that they are really separable. We cannot ‘isolate’ mineness from any experience it accompanies. Again, the best way to put it is that mineness is an aspect of an experience. An experience is thus something that has two aspects: a subjective aspect and a qualitative aspect, not separable from each other. Similarly to what we have seen in the case of dual-aspect monism, these two aspects are not two higher-order properties of experiences. As before, the idea is that it just is in the nature of an experience to be dual-aspectual in this way. We could combine dual-aspect monism and the dual-aspect account of mineness in order to say: when I have an experience, say of a sunrise or of a toothache, there are three aspects to it – the physical aspect, the qualitative aspect, and the subjective aspect. A complete view of the nature of experiences thus amounts to a trial-aspect monism. Such a view would not only be nicely complete, but it would also help with a problem concerning the identification and individuation of mineness. Consider what Howell and Thompson (forthcoming) say: Another important distinction is whether phenomenal me-ness is indexical or singular. In other words, is the me-ness I experience the same

The No-Self View  99 phenomenally (and semantically) as the me-ness you experience? Or is there a distinct phenomenal character that each of us has in virtue of which we are presented to ourselves? Do Sally and Jim each have their own ‘Sallyishness’ or ‘Jimness’ or is there a generic me-ness which each of them has and which presents each of them to themselves? (my italics) As we have seen, mineness is not a qualitative element of an experience like a perceived shape or colour, it attaches to elements of a qualitative nature but it is not a quale itself. It is a phenomenal feature that grounds a sense of ownership of each of the qualitative elements. In this sense, it is invariant and it is always ‘phenomenally the same’. The question is, is it then numerically the same? That would be quite unwelcome, meaning that any experience, experienced by any individual, would exhibit the very same (numerically speaking) mineness. But how should we then distinguish between ‘my’ mineness and ‘your’ mineness? How is mineness to be identified and individuated? Here is a sceptical way to put the question: how do I know that I am me? If ‘my’ mineness and ‘your’ mineness were the very same thing, a sceptical challenge could perhaps arise. There are two answers to such worries, and one of them uses the resources of trial-aspect monism. Indeed, in a complete trial-monist view, a toothache is something that has three aspects that are conceptually separable, but not separable in reality. Metaphysically speaking, they stand or fall together – since ‘they’ are one and the same (phental) thing. It is then easy to individuate and identify this particular toothache I am having now, and to distinguish it from your toothache that you are having now, since the brains (and teeth nerves . . .) involved are not the same. The physical aspect is not the same and so even if, say, the qualitative aspect is qualitatively identical and even if the mineness is ‘phenomenally the same’ (in what Howell and Thompson call the generic way), the two toothaches are not the same and can be distinguished. Furthermore (this is the second answer), any experience is typically a part of a stream of experiences, related by causal relations and by relations of similarity as well as temporal continuity and contiguity.5 This way to distinguish between different experiences, and thus different minenesses, is less strong than the trial-aspect way since it does not allow for intrinsic individuation but only for relational individuation – it is by relation to other experiences that an experience is individuated and identified as being part of this continuum rather than another continuum (that is, by being an experience of this individual rather than of another individual). This second way of answering the question is then open to extra-ordinary thought experiments, perhaps involving two possible individuals that both exist only long enough to have one single experience and that both have qualitatively exactly the same experience – in this case, the relational way of individuating the experiences, and their respective minenesses, would fail. Thus, trial-aspect monism would seem to be a more systematic and

100  The No-Self View less objectionable way to understand what mineness is and how it is to be individuated. To the question “Can one subject have two experiences with different minenesses?”, the answer is “yes” and “no”. Phenomenally speaking, the answer is “no” since mineness is always phenomenally invariant – in this phenomenal sense there is only one mineness, and there is no difference between ‘my’ mineness and ‘your’ mineness. Numerically speaking, the answer is “yes”. The idea here, built around a perdurantist-like worldview (that is very close to the Buddhist claim about the impermanence of all things), is that when it comes to material objects as well as when it comes to experiences, these are best conceived of as being constituted by timebound momentary6 temporal stages, suitably interconnected, and nothing is ever numerically identical at different times. Thus the mineness involved in my toothache from yesterday is numerically different from the mineness involved in my toothache from today. Nothing endures here: no enduring mineness, no enduring Self (more on that soon). At the end of the day, I think that – in general – trial-aspect monism is the right way to go. But to demonstrate the truth of this, a proper and independent defence of dual-aspect monism would be needed, and this is a task for another day (I give it a try in Benovsky [2016a]). For now, then, I suggest to keep this idea in mind, but to investigate further the dual-aspect account of subjectivity for its own sake, taking lessons from dual-aspect monism, but in an independent way.

§5 As we have seen, aspects are not higher-order properties. The dual-aspect account of subjectivity is thus clearly different from any kind of higherorder theory of consciousness of which there are many variants (see, for instance, Armstrong [1968, 1984], Lycan [1996, 2001], Gennaro [1996, 2005]), and Rosenthal [1986, 2002]). In what follows, I am not going to try to add to the existing pile of objections against such HOT views; rather, I am going to focus on a family of views that are very close to the dualaspect view I have in mind, but that contain some unnecessary components to which I am going to object. I have in mind here the self-reflexive accounts of subjectivity. Brentano held the view that all conscious states are reflexive. He famously says: Every mental act . . . includes within it a consciousness of itself. Therefore, every mental act, no matter how simple, has a double object, a primary and a secondary object. The simplest act, for example, the act of hearing, has as its primary object the sound, and for its secondary object, itself, the mental phenomenon in which the sound is heard. (Brentano [1874, pp. 153–154])

The No-Self View  101 In Brentano’s view, a mental act, as he puts it, is always directed towards something – that is, it is essentially intentional – and it is also always directed towards itself, not only in the case of deliberate introspection, but in all cases: even when we do not reflect on our experiences, we are secondarily aware of them. An experience thus always has itself as an object, and this awareness of itself is also a conscious state. Brentano often refers to Aristotle (De Anima) who warns against a possible vicious regress where a perception would be needed to be conscious of the first-order experience and then again another perception would be needed to be conscious of this perception, and so on. (Note that, as we have seen in §2 above, one also finds this warning in a very similar form in the Dharmakīrti view.) The idea is simply this: since a mental act is always perceived, it is perceived either by another mental act or it is perceived by itself; but if it were perceived by another mental act this would lead to an infinite regress, so it is perceived by itself. In this sense, all mental acts are reflexive, and one can thus understand consciousness in terms of a reflexive self-representational state. In the recent debate about the nature of consciousness and subjectivity, Uriah Kriegel holds a view of this kind. He says: there is a more specific division of conceptual labor between qualitative and subjective character: a phenomenally conscious state’s qualitative character is what makes it the phenomenally conscious state it is, while its subjective character is what makes it a phenomenally conscious state at all. . . . [T]he qualitative character of a conscious state consists in its representing certain response-dependent properties of external objects; its subjective character consists in its representing itself in a suitable way. . . . It follows that what makes something a phenomenally conscious state (at all) is suitable self-representation. (Kriegel [2009, pp. 1 and 2]) The self-representation claim avoids the worry about a possible infinite regress. But so does the dual-aspect view of subjectivity. The dual-aspect view does not appeal to self-representation or reflexivity; rather, it claims that the subjective aspect is one of the aspects of any experience, on a par with the other (two) aspect(s). Any experience, in this view, just is such that it has these aspects, including the subjective aspect. From this point of view, the self-representation claim looks like an additional claim – and, an unnecessary one, since the job is already done. Furthermore, it is not clear at all how self-representation could do the job – that is, the job of ‘making a conscious mental state a conscious state at all’. During his famous travels, Gulliver arrives in the city of Lagado where there are people who want to eliminate words altogether and they propose that, in order to communicate, one should always carry all the things that are necessary to express what one wants to express – thus, the idea is not to have a word like “chair” to refer to a chair, but to have a chair that

102  The No-Self View self-represents itself (and thus self-refers to itself), in order to ‘talk’ about a chair. While such a language is certainly not very practical for everyday conversations, a Lagadonian language can nevertheless be very useful for philosophers (for instance, David Lewis [1986, p. 145] puts it to good use when it comes to some issues in the field of the metaphysics of possible worlds). The core idea of a Lagadonian language is to solve the reference issue by using the idea that any object trivially self-represents itself (and self-refers to itself). Self-representation can indeed be understood in this way and can be seen as a very trivial and widespread phenomenon. And, crucially, it has nothing to do with consciousness or subjectivity. The fact that a chair selfrepresents itself does not make the chair aware of itself, it does not make it conscious, and it does not involve any subjective character at all. Basically anything can be said to self-represent itself, trivially, and whether we want to say that this is the case or not, what is certain is that this question is entirely orthogonal to the question concerning the nature of consciousness and subjectivity. Self-representation is thus not the good tool to use in order to explain ‘what makes a conscious mental state a conscious state at all’ – it simply can’t. (Besides, note that such an understanding of subjectivity, even if it were possible, would be vulnerable to the sceptical worries we have seen above concerning the identification and individuation of mineness.) Kriegel seems to agree with worries of this kind, when he says that “it is simply false that a mental state’s representation of itself is conscious in virtue of the state representing itself to represent itself” (Kriegel [2009, p. 127]). But, then, how can a self-representational view of subjectivity work? Kriegel puts forward a self-representation theory where a first-order representation is combined with a second-order representation in order to generate one complex state in the following way: “(i) a first-order representation, (ii) a higher-order representation of that first-order representation, and (iii) some relationship of cognitive unity between the two, in virtue of which they form a complex” (Kriegel [2009, p. 233]). But even such a more sophisticated account is entirely prone to the objection we have just seen: representing something, or representing the representation of something, or self-representing itself, are phenomena that do not ‘generate consciousness’, they are commonplace phenomena, interesting, but not relevant for that purpose.

§6 It is not unreasonable to think of consciousness and subjectivity as being primitive phenomena. In philosophical theorizing, one has to stop somewhere,7 and here seems to be a very good place to stop indeed. Trying to analyse these notions, for instance in terms of higher-order perceptions/ thoughts or in terms of (self-)representation, does not seem to work. The dual-aspect account of subjectivity does not try to analyse the subjective aspect of experiences any further. It helps to understand the nature of experience and the nature of subjectivity by showing what dual structure (or trial

The No-Self View  103 structure, if you follow me all the way) experiences have. It is phenomenologically adequate, not raising any Humean worries. It takes mineness to be a primitive aspect, which has a proper place in an overall theory of the nature of conscious experience. In the preceding section, I tried to get a grip on the notion of mineness and I argued that we should appeal to a notion of mineness devoid of the burden of self-representation. My suggestion is to work with an understanding of mineness as a first-order, non-relational, primitive and unanalysable aspect of (elements of) experience. Such a notion of mineness plays, as we have seen, a crucial role when it comes to understanding subjectivity, and – this will be the topic of this section – it also plays an important role when it comes to diachronic unity. Diachronic unity is a datum about ourselves that any theory of the (no-) Self needs to address. It seems to us that we endure (that is, that we are one and the same subject of experience wholly located at different times [at least for some stretches of time]), and this needs explaining, especially if one holds a kind of eliminativist no-Self view, as Dharmakīrti did, and as I do as well. Saying that it is a datum that it seems to us that we endure does not mean that we endure ontologically speaking. Paul (forthcoming) puts forward such a view very clearly: she distinguishes between subjective ontology, that is given to us by analysis of phenomenal experience, where, as she argues, we endure at least for short periods of time (perhaps, we could say, the specious present), and objective ontology, that is given to us by metaphysical analysis, where the best theory around seems to be one or another version of four-dimensionalism according to which, in short, we perdure by having numerically distinct temporal parts at different times.8 In Paul’s view, subjective endurance is perfectly compatible with objective perdurance; this is then one place where results from phenomenological analysis can be disconnected, at least in some relevant ways, from results of metaphysical inquiry (see Benovsky [2013b] and [2015] for a discussion of various other, and related, cases of such disconnections). We have already seen in Part I (especially Chapter 2) that, when it comes to eliminativism about ordinary objects, perceptual evidence is entirely orthogonal to the metaphysical issue. The same type of considerations apply here as well, and it will be useful to go through this line of thought again, as applied to our current concerns. The view I have in mind here is that while there is appearance of diachronic unity on the phenomenal level, there is no objective diachronic unity in the world. Mineness is a central point here. Indeed, as we have seen, mineness is what remains constant across different experiences at different times (while the qualitative elements of experiences change). The mineness aspect of (elements of) experiences always ‘feels the same’ and is phenomenally invariant across time. In fact, since all qualitative elements of experiences change (sooner or later), mineness is the only thing that does not change and that is a common feature to all of my experiences. This is where, I submit (in a Dharmakīrtian way), the appearance of

104  The No-Self View diachronic unity comes from. It is a natural mistake for us to make: since there is always something that feels the same in all of our experiences (that accompanies all of them), we have a tendency to think that it is one and the same thing indeed – and, furthermore, that we are that thing, namely, a Self in the sense of options (i) or (ii) (see §2 above). But in reality, there is merely a succession of impermanent and numerically different psychological states/ experiences/elements of experiences. Any theory of the (no-)Self should accommodate the appearance of diachronic unity and continuity there seems to be to our experiences. Indeed, phenomenologically speaking, it is correct to say that our experience appears to us as being at least partly continuous and somehow ‘stream-like’. Prima facie, this may sound like a problem for a no-Self view, since – the objector could say – the view claims that there are only ‘disconnected’ experiences, in the sense that they are not united in a bundle or had by an underlying substance: there just is a disunited series of singular experiences, without continuity. The task is then to show where the unity and continuity that we experience comes from. These worries have been largely debated, especially since James’s famous discussion of the “specious present”: A succession of feelings, in and of itself, is not a feeling of succession. And since, to our successive feelings, a feeling of their own succession is added, that must be treated as an additional fact requiring its own special elucidation. . . . [T]he prototype of all conceived times is the specious present, the short duration of which we are immediately and incessantly sensible. (James [1890, p. 629]) The idea is this: think of the apple falling from a nearby tree, and think about how you experience its fall – its movement. Your experience is continuous and smooth, you experience movement, you do not have a succession of static isolated experiences of the apple being at various points of its fall. In order to provide a phenomenologically adequate understanding of such a continuous experience, James and many others9 introduced the notion of a “specious present”, that is, a non-zero temporal duration which, in our experience, our mind embraces “at once”. This notion is indeed very useful to understand the phenomenology of our experience of continuity, such as the apple’s movement, and it has given rise to the debate between (i) extensionalists10 who claim that the specious present is temporally extended and that our experience is composed of short but temporally extended overlapping experiences featuring temporally extended and temporally ordered content unified into a single experiential episode which appears to us “at once” as present, and (ii) retentionalists11 who argue that the specious present is not temporally extended but rather instantaneous, although it has temporally extended content, because at any point the experience of a falling apple has

The No-Self View  105 as components of its content states of the apple at the previous instants all of which are given to us in the one single instantaneous experience together. In one way or the other, the important thing for our discussion now is that somehow (i) our experiences such as the one of the falling apple are experiences as of movement and continuity, and (ii) they sort of smoothly flow one into the next. (i) and (ii) is what we need to account for, when we talk about the unity and continuity of our experiences. Of course, some will have in mind something stronger when they speak about diachronic continuity and unity of the Self – something like a unity over the whole life of a person – and thus it might seem that my desideratum is not strong and demanding enough. But here, I join with Hume: phenomenologically speaking, there is nothing like such a lifelong unity we experience, we only ever experience the kind of continuity and unity mentioned under (i) and (ii) above, and nothing more. This is then what a theory of the (no-)Self needs to account for. Let us remember the discussion from I.2.§3 (if your memory is good enough, you may want to read through this section rather quickly – but don’t skip it entirely, the point plays an important role here). The claim to defend here is that the alleged diachronic unity, or (alleged) diachronic identity, is merely an appearance of unity or identity that is itself a product of how our perceptual system and our brain (contingently) works. Remember the familiar case of cinema, where we have an experience as of movement, when a character moves – say, when Superman flies – but what we’re actually looking at is a series of static images projected at the speed of (usually) 24 or 25 images per second. We do not have an experience of a series of static images, and we do not have a series of static experiences either. Rather, we have a single unified experience as of continuous movement. John Locke12 provides a nice discussion of other cases, relevantly related. The first type of cases concerns situations where we have an experience of a movement that is too slow like, for instance, the movement of the hour hand on a watch. In this situation, we do not have an experience as of movement because the actual movement of the hour hand is too slow for us to be registered as a movement. On the other side of the spectrum, when we are faced with a movement that is too fast, we do not experience it as a movement either. If an object, say a pendulum, moves very fast following a circular path, we will perceive a circle, because the movement is simply too fast for us to register the single object (the pendulum) as moving along a circular trajectory. This is a fact about our perceptual abilities, already noticed by Locke: our perceptual system has lower and upper limits beyond which we are not able to perceive movement as movement. For us to perceive movement as movement, it has to happen at a speed that lies in between these two limits. Similar to the case of cinema, as we have seen, Paul (2010) discusses the phenomenon of “apparent motion”. The standard example is a dot on a computer screen, where the dot is displayed first on one side of the screen and then in quick succession on the other side. We then perceive the dot as moving from one side to the other, while in reality nothing moved there at all, since

106  The No-Self View there only were different pixels activated at different times. But, like in the case of cinema, we have an experience as of movement. This happens even when we know about the apparent motion phenomenon. It’s not an illusion in the sense that we would make an epistemic mistake, incorrectly believing that there is genuine movement. But the appearance of movement persists even when we know how things actually are. The reason for this (which is at play in the other cases we have seen above as well) is that the appearance of movement arises from the way our brain (our perceptual system) works, in the sense that it registers and interprets a succession of static stimuli, on a neurological but not phenomenal level, before we become conscious of it, to produce an experience as of movement (or as of a circle, in the pendulum case, etc.). Our brain does this work before we have any conscious experience at all, and this is why there is no way we can resist having, for instance, an experience as of movement in the cinema case. The analysis of our experience of movement in general comes then in two stages. First, a precognitive, unconscious, and non-phenomenal stage where our perceptual system registers inputs it gets and interprets them; and second, a stage where the result of this precognitive work reaches our consciousness and becomes phenomenally accessible to us. The way we experience movement is then strongly dependent on the (contingent) way our perceptual system is built. A final point here concerns the fact that our perceptual system is (contingently) built in such a way that when producing for us, precognitively, experiences such as an experience of a movement, it exhibits what Scholl (2007, p. 573) calls a “temporal same-object advantage”. Scholl (2007) analyses experiments where subjects have to follow objects like dots or other shapes on a computer screen, and where these objects are sometimes easy to follow, but sometimes they move behind occluders that temporarily mask their trajectory, in such a way that the trajectory is discontinuous and the objects temporarily disappear. In such situations, subjects invariably have a natural tendency to privilege continuity and identity (of the object they are tracking though time), even in cases where the objects undergo important qualitative change (they change shape, for instance, while hidden behind an occluder, and they reappear on the other side of an occluder as, say, a square instead of a circle). Even in such cases where the object that enters the masked zone is qualitatively quite different from the object that emerges on the other side, subjects interpret the situation as one where one object changed, rather than as a situation where one object remains hidden and another qualitatively different object emerges on the other side of the occluder. Here again, this is something that subjects ‘cannot resist’; this ‘same-object prejudice’ is simply built in the way our perceptual system works.13 The relevant thing we learn from all these considerations concerning the way we experience movement is that our brain creates unity and temporal continuity for us even in cases where there is none. I submit that something similar happens when it comes to us. There is a series of distinct and impermanent psychological states that follow each other but because (i) they all bear

The No-Self View  107 the ‘same’ mark of mineness, as we have seen, and (ii) there is appearance of continuity in many of our experiences (even when this does not correspond to any such continuity in the world), we have this strong and irresistible natural tendency to think that something continuous lies behind these experiences, which can lead to the idea of a subject of the experiences, ontologically objective and reified, that endures through time, while in reality there is a series of distinct successive experiences that all have a mineness aspect. This is the mistake Dharmakīrti, and many others, condemned. As we have seen, our phenomenal experience (the subjective ontology side) is relevantly different from how the world is (the objective ontology side). From the objective point of view, what we come to appreciate here is that what really counts in temporal continuity and diachronic ‘identity’ is not any kind of continuous unity guaranteed by some kind of enduring subject, but rather causal connectedness, similarity, and phenomenal continuity between various experiences/states, as is familiar not only from Buddhist thought but also from Western philosophy of personal identity (think, for instance, of Lewis [1983] or Parfit [1984]). This then paves the road to an eliminativist no-Self view: indeed, as we have seen, there is no need, when it comes to diachronic unity and temporal continuity, and when it comes to subjectivity, for a Self understood in an objectively ontological strong sense, that ontologically ‘glues together’ different successive experiences or psychological states. There is no such need because, to repeat, these are connected to each other in virtue of what they are, how similar they are, how causally related they are, and so on. As we have seen, the appearance of unity, continuity, and subjectivity, are to be explained in such a way that it does not require the existence of any such Self.

§7 In the two preceding sections, we have seen that the no-Self view is nicely compatible with views that emerge from careful considerations about the nature of subjectivity and temporal continuity. But, one could argue, we still need a Self to play an important role in our ontology as being the bearer of experiences. The idea that properties need a bearer in general, and that psychological properties/states need a bearer in particular has a long tradition. According to this tradition, an experience is always an experience of a subject – no subject, no experience. Here are some philosophers who subscribe to this type of view (for different reasons and articulated in different ways): I think, therefore I am. (Descartes [1637, Part IV, §1]) Individual mental states are necessarily states of persons: they are necessarily ‘owned’ – necessarily have a subject. (Lowe [1996, p. 25])

108  The No-Self View The claim that atoms arranged humanwise fail to compose a human but are united, in part, by thinkings and speakings implies something that parallel claims about statues . . . do not imply. It implies the exotic (and perhaps demonstrably impossible) claim that there are thinkings and speakings but neither thinkers nor speakers. (Merricks [2001, pp. 121–122], my italics) There may be such things as bundles of thoughts, but it is a metaphysical or logical blunder to suppose that such things are the subjects of the thoughts that compose them. . . . A particular thought is one thing; the being that has or thinks it, if anything does, is another. (Olson [2007, pp. 139–140], my italics) there is the claim that we are identical with particular episodes of experience. This runs counter to our ordinary conception of ourselves as things which have experiences. (Dainton [2012, p. 185], my italics) The opposite tradition also has many different voices; here are some examples: For my part, when I enter most intimately into what I call myself, I always stumble on some particular perception or other, of heat or cold, light or shade, love or hatred, pain or pleasure. I never can catch myself at any time without a perception, and never can observe anything but the perception. (Hume [1738–1740, Book I, Part IV, §6]) As far as the superstitions of the logicians are concerned: I will not stop emphasizing a tiny little fact that these superstitious men are loath to admit: that a thought comes when ‘it’ wants, and not when ‘I’ want. It is, therefore, a falsification of the facts to say that the subject ‘I’ is the condition of the predicate ‘think’. It thinks: but to say the ‘it’ is just that famous old ‘I’ – well that is just an assumption or opinion, to put it mildly, and by no means an ‘immediate certainty’. In fact, there is already too much packed into the ‘it thinks’: even the ‘it’ contains an interpretation of the process, and does not belong to the process itself. People are following grammatical habits here in drawing conclusions, reasoning that ‘thinking is an activity, behind every activity something is active, therefore – ’. Following the same basic scheme, the older atomism looked behind every ‘force’ that produces effects for that little lump of matter in which the force resides, and out of which the effects are produced, which is to say: the atom. More rigorous minds finally learned how to make do without that bit of ‘residual earth’, and

The No-Self View  109 perhaps one day even logicians will get used to making do without this little ‘it’ (into which the honest old I has disappeared). (Nietzsche [1886, Part I, §17]) The Cogito affirms too much. The certain content of the pseudo-‘Cogito’ is not “I have consciousness of this chair”, but “There is consciousness of this chair”. (Sartre [1936, pp. 53–54]) There is no such thing as the subject that thinks or entertains ideas. If I wrote a book called ‘The world as I found it’, I should have to include a report on my body, and should have to say which parts were subordinate to my will, and which were not, etc., this being a method of isolating the subject, or rather of showing that in an important sense there is no subject; for it alone could not be mentioned in that book. (Wittgenstein [1921, 5.631]) Georg Lichtenberg’s objection14 to the first citation by Descartes nicely sums up the disagreement: according to Lichtenberg, what Descartes should have said is that “thinking is going on” instead of saying that there is a thing that thinks (the res cogitans). Sartre, Nietzsche, and Wittgenstein were all impressed by this objection and embraced it, each in his own way. The “particular perceptions” on which Hume “stumbles” correspond to what constitutes the core of Buddhist thought: there is a series of impermanent psychological states, and nothing more. In Hume’s own view, a bundle theory of the Self was then the adequate conclusion to draw, while for most Buddhists it is the no-Self view that captures in the most suitable way the phenomenological data and that provides the most elegant and efficient metaphysical picture.15 The more recent concerns, raised by Merricks, Olson, and Dainton are all differently expressed and they are differently used for different purposes, but they all share the same common ground, namely the worry that we are ‘things which have experiences’ rather than ‘things which are experiences’. As we have already seen at the end of Chapter 2, there is a very general metaphysical problem involved in the background of this worry: a classical problem for the bundle theory about ordinary material objects. A tree, according to the bundle theory, is a bundle of properties. But, says the classical objection, a tree is rather something that has properties, and not that is a bunch of properties. According to the objector, the bundle theory has it all wrong from the very beginning. The idea here is that properties cannot ‘float free’, that is, that they cannot exist independently of something that ‘supports’ them (see Armstrong [1997, p. 99]). This worry then translates to the case of psychological properties/states as the idea that they require the existence of a subject, which can be referred to as ‘the Self’. The nature

110  The No-Self View of this bearer can take various forms, but what all the variants have in common is the broadly Cartesian idea of something – some thing – that has the properties/states. When it comes to the bundle theory and to the classical objection to it, a reply is available and well-known. The situation is here dialectically peculiar in the sense that the objector objects to the bundle theory by denying precisely what the friend of the bundle theory wants to say. Indeed, the claim that lies at the core of the bundle theory is the claim that to have a property is to ‘contain’ a property in a bundle – the having of a property is analysed in terms of the property’s being an element of a bundle. The objector is then simply denying the very claim the bundle theorist makes. In short, the bundle theorist’s reply is then that the objector is begging the question and that there is no reason here to reject the analysis of ‘having’ in terms of ‘containing’. But, to come back to the case of the no-Self view, the objection is only partly similar and it is a perhaps deeper one. What objectors like Merricks, Olson, and Dainton seem to have in mind is that it’s wrong to say that experiences can exist without being had by something. In this view, this just is what being an experience amounts to. The issue here seems to lie in the (alleged) fact that experiences are dependent entities, as Lowe claims: The deepest problem with this sort of view [i.e. the bundle theory of the Self] is that the entities out of which it attempts to construct the self – psychological states and processes – are themselves quite generally not individuable and identifiable independently of the selves that are their subjects, so that fatal circularity dooms the project. (Lowe [1996, p. 8]) Lowe objects to the bundle theory of the Self à la Hume, but of course, the same worry could be raised against the no-Self view. Armstrong is even more explicit (again, his objection is raised against the bundle theory, but it could target the no-Self view as well): There is a fundamental difficulty with all bundle theories. It is that properties and relations, whether universals or particulars, seem not suitable to be the ultimate constituents of reality. If they are the ultimate constituents, then, it appears, completely different (non-overlapping) properties and relations will be ‘distinct existences’ in Hume’s sense of the phrase: entities logically capable of independent existence. But are properties and relations really capable of independent existence? Can a certain determinate mass, for instance, whether the universal of that mass or the trope mass of this particular body, exist in logical independence of anything else? It hardly seems so. The case against the possibility of independent existence for relations seems even stronger. (Armstrong [1997, p. 99])

The No-Self View  111 The worries about properties translate in the case of the no-Self view, in agreement with the concerns raised in the citations above, as worries about the possible independence of experiences. Are experiences independent enough in order to being metaphysically capable of existing without there being a subject, understood as a bearer of these experiences? The dialectical situation is, here again, delicate. On one side of the debate, there is a claim of a need for a bearer of experiences (or properties, in the more general debate), and on the other side there is the claim that no such thing is to be found in our phenomenology and that it is very unparsimonious and unnecessary to include it in our ontology. To my mind, the latter claim surpasses the former in virtue of the fact that it is indeed more parsimonious, of the fact that it fits better with what is found in our phenomenal experience (that is, qualitative character and subjective character, but no subject), and that there does not seem to be any reason to think that experiences – or properties – are not metaphysically independent enough: there is a long tradition that says so, but once we realize (as Dharmakīrti did) that this latter claim most likely comes from mistakenly reifying a sense of subjectivity and continuity, as we have seen in §1 and §2, we can see that this is mere prejudice against experiences (or properties).16 The non-prejudiced, parsimonious, and phenomenologically adequate (when it comes to experiences) idea is then, in short, this: in order for an entity to be ‘metaphysically independent enough’, it does not have to be an object-like substance. Properties and experiences are metaphysically independent enough to exist without the need for a bearer – there just is no argument that would show that they are not.

§8 The picture that emerges from the considerations of this chapter is the noSelf view. This view is able to accommodate the explanatory need when it comes to subjectivity, unity and continuity, and the notion of a subject of experience. The no-Self view is a form of eliminativism that has a lot in common with eliminativism about ordinary objects. As we have seen in the first Chapter of Part II, one issue that divides eliminativists is precisely the question to know whether eliminativism of this kind can also be applied to us – that is, the Self, in my way of putting it – or not. We have seen that some eliminativists such as Peter Van Inwagen or Trenton Merricks embrace one brand or another of exceptionalism. In this chapter, I hope I have succeeded in showing that an exception for us is unnecessary – that is, I hope to have shown that the no-Self view works. In the no-Self view, we can eliminate the subject understood in an ontologically reified sense, because there are impermanent psychological states, and because no more than those is needed to explain what we want to explain (subjectivity, unity, . . .). Thus, in a way similar to chairs, we can eliminate the Self because there are impermanent psychological states ‘arranged Self-Wise’: an allegedly single entity, the Self,

112  The No-Self View can be eliminated because there is a plurality of other entities which satisfy all our theoretical and practical needs. The idea is that we can eliminate us, understood in an ontologically strong, reified, and ontologically committing sense of Selves, but without losing anything important – I can still say “I am happy now to be sitting on a chair and to be thinking about eliminativism” in a perfectly respectable sense, understood in terms of simples arranged my-body-wise and chairwise, and in terms of the existence of a succession of impermanent psychological states, suitably causally connected. In this way, we can hold a unified and complete eliminativist view, with no exceptions.

Notes 1 Guillot (2016) includes a detailed discussion of this somewhat chaotic terminology. 2 See MacKenzie (2008) and Krueger (2011) for a discussion. 3 See Part II, Chapter 6 above, as well as Benovsky (2008) for a more general discussion. 4 The notion of the specious present can be found in various forms in James (1890), Husserl (1964), Broad (1923), Dainton (2000, 2003), Hoerl (2009), Phillips (2011). I discuss it in Benovsky (2013b). 5 See, for instance, Lewis (1976a, 1976b) and Parfit (1984). 6 This can involve short intervals of time (in the case of experiences, the idea here is the one of the specious present; see, inter alia, James [1890], Husserl [1964], and Broad [1923], Dainton [2000, 2003], Hoerl [2009], Phillips [2011], and my discussion in Benovsky [2013b]). 7 Most, perhaps all, philosophical theories rely heavily on their primitives to do the job they do. I discuss in detail the role of primitives in philosophical theories in Benovsky (2016b, especially chapter 4). 8 Various versions of four-dimensionalism are defended, inter alia, in Heller (1990, 1992, 1993), LePoidevin (2000), Lewis (1983, 1986), Quine (1960), and Sider (1997, 2000, 2001). For a detailed discussion of all variants of perdurantism and endurantism, see Benovsky (2006a, 2009a, 2009b). 9 See inter alia Kant (1781), James (1890), Husserl (1964), Broad (1923), Foster (1982, 1992), Dainton (2000, 2003), Hoerl (2009), Phillips (2011). 10 See Dainton (2000, 2003, 2008b, 2010). 11 See Husserl (1964). 12 Locke (1975, Book II, especially chapter 14). 13 Perhaps, this is because it has an evolutionary advantage; see Prosser (2012). 14 See his notebook K, in Hollingdale (2000). 15 Besides, many Buddhist thinkers argued that the no-Self view has interesting and welcome practical consequences, since it helps us to liberate ourselves from existential suffering; in Benovsky (manuscript) I critically discuss this claim, and I defend a Humean approach with respect to this aspect of the no-Self view. 16 Remember Strawson’s claim that we have seen in Chapter 6: ‘But if there is a process, there must be something – an object or substance – in which it goes on. If something happens, there must be something to which it happens, something which is not just the happening itself’. This expresses our ordinary understanding of things, but physicists are increasingly content with the view that physical reality is itself a kind of pure process – even if it remains hard to know exactly what this idea amounts to. The view that there is some ultimate stuff to which things happen has increasingly ceded

The No-Self View  113 to the idea that the existence of anything worthy of the name ‘ultimate stuff’ consists in the existence of fields of energy – consists, in other words, in the existence of a kind of pure process which is not usefully thought of as something which is happening to a thing distinct from it. Strawson (1997, p. 427) Hawthorne and Cover (1998, §2) defend the bundle theory against what they call an “incredulous stare”: Perhaps some philosophers will claim to find it just self-evident that universals are had by something. We don’t have much to say to such philosophers. We do note, however, that the polemic against the bundle theory has rarely taken the form ‘It is simply self-evident that anything quality-like is directly or indirectly predicated of something that isn’t like a quality. . .’. If opponents of . . . the Bundle Theory wish to retreat to this form of an incredulous stare, so be it.

8 Eliminativism, Life, and Death

§1 Suppose you follow me in what I said in the preceding Chapter, and suppose that you come to believe that you, in the sense of an ontologically reified Self, don’t exist. Does this make any difference for the way you (should) live your everyday life, and for the way you (should) relate to others? In different debates, and in different philosophical traditions, metaphysical theories have been said to lead to practical (often, moral) conclusions. In some cases, it is argued that we should reject a metaphysical theory because of its unpalatable moral consequences. In other cases, on the contrary, it has been argued that a metaphysical theory has desirable moral consequences. In cases such as these, it is often not entirely clear how the dialectic goes: we can see such a situation as one where this is an argument in favour of the metaphysical theory, since it has the advantage – over its competitors – of yielding a desirable moral consequence; or one can interpret the situation as one where the metaphysical theory is an argument in favour of the practical claim. Often, there is a mixture of both. In this chapter, I will shortly discuss an example of a case of the first kind, namely Mark Heller’s argument against David Lewis’s modal realism based on the claim that it has unacceptable moral consequences, but I will mostly focus on a case of the second kind (where a metaphysical theory is supposed to have desirable moral consequences), of which I will discuss two variants coming from two different philosophical traditions. The first variant is Parfit’s reductionism about personal identity, where he has argued that this metaphysical theory has desirable practical consequences such as helping us to deal with the fear of death (and related issues). This is a case of someone who argues – or at least one can interpret him as doing so – that the fact that reductionism about personal identity has this welcome consequence is an argument in its favour. From the methodological point of view, this is then the same strategy as Heller’s, except that Heller explores negative consequences of a metaphysical theory while Parfit discusses positive ones – so we have an objection in one case, and an argument in favour of the theory

Eliminativism, Life, and Death  115 in the other case. The second variant of the second kind of case can be found in some Buddhist theories of personal identity. There are reductionist and eliminativist versions of this family of views, and in a way similar to what Parfit says, it is argued that they can help us to liberate ourselves from existential suffering, which is due, at least in part, to the inevitability of our death. This is a case that we can meaningfully interpret as one where the main focus is on the relevant practical issues, and where the main purpose of the metaphysical theory of personal identity is to help us with these practical issues. In short, the metaphysics is used as a path to and an argument in favour of these practical conclusions. This is not a fundamental difference between Parfit and Buddhist philosophy – rather, it is a difference in what one can focus on: a Parfit-like view can be seen as defending a metaphysical theory, while the Buddhist tradition is (arguably) mainly interested in the practical issues. But, the reasoning behind both strategies is very similar and goes in the same direction – from metaphysics to practical conclusions. These similarities are then much more relevant than the differences in focus. To put it crudely, in both cases, it is possible to formulate the argument as going in both directions. My aim in this chapter is twofold. First, I want to examine Heller’s, Parfit’s, and the Buddhist arguments and see whether the practical conclusions they put forward really do follow from their metaphysics. Second, I want to make a general methodological claim about the path from metaphysics to practical conclusions. I will take a rather critical stance: in line with the general Humean idea that ‘ought’ cannot be derived from ‘is’, I will argue that, at least in the case of my three case studies, the metaphysical theories at hand do not allow us to derive the practical conclusions, and that it is plausible to think that in general there is no direct path from metaphysics to such conclusions. I will examine a Strong and a Weak claim of this kind. In this way, I want to distinguish eliminativism (both eliminativism about ordinary objects and eliminativism about the Self) from the ‘practical consequences’ that it may seem to have. Metaphysical theories are one thing. Moral considerations are another.1 (To be sure, Parfit is a reductionist about personal identity, not an eliminativist about the Self. But the point I will focus on in this chapter is common to both types of views.) But first of all, as an important preliminary discussion, I will devote §2-§5 to a discussion of what is often referred to as ‘the Humphrey objection’ in order to better understand how temporal counterpart theory works. Indeed, this view about persistence through time can implement reductionism about personal identity particularly well, and it will be instructive to have a clear picture of how it does its job, and what its problems are. In doing so, as the starting point, I will examine in detail the differences there are between alethic modalities and temporal modalities. This will be a rather long preliminary detour, but a useful one I hope – please, bear with me.

116  The No-Self View

§2 In the actual world, I have a doppelganger. His name is Jerry and not only is he very much like me, but he also likes doing the same things I do. He physically resembles me a lot, he is the same size, the same build, the same strength. Like me, he likes to go running, hiking, and climbing. He is exactly as fit as me – no more, no less. He climbed the same summits I did, in the same weather conditions, at the same speed, with the same technical skill; he is exactly as good – or bad – as I am. He looks and behaves so much like me that if you’d see him running up a mountain you’d think it’s me. Last week, Jerry did something I did not: he managed to climb up the Everest. That’s a big performance, far above my (our) usual level, so when I learned about his achievement I was really happy for him. But, after a few seconds of thought, I also started to feel happy for me, because if he managed to do it, that means that I could do it as well, given how similar we are. Great news, I can go on the Everest, I can do it! The feeling I have – namely, the idea that something is possible for me because something is actual for somebody else – will play an important role in the discussion below. Indeed, I will be interested here in four versions of what is often called ‘the Humphrey objection’. This objection was initially raised by Saul Kripke (1972, p. 45)2 against David Lewis’s (1968) counterpart theory, so this is where I will start the discussion. As we will see, there is a perfectly good – well-known – answer to the objection, along the lines of the story of Jerry and me. I will then examine other places where a similar objection can be raised: it can arise in the case of temporal counterpart theory (in fact, it can arise in the case of all kinds of counterpart theories, independently on modal realism) but a very similar worry can also arise against modal realism itself or against an ersatzist theory of possible worlds itself. For similar reasons, in similar situations, a similar objection will arise. Now, what is interesting is that it is not the case that a similar response can be given in these similar cases. So, in the end, we will see what dissimilarities there are and how and why they are relevant. In particular, we will see the differences there are between alethic modalities and temporal modalities. In the case of alethic modalities (metaphysical necessity and possibility), the objection can be answered by appealing to the notion of representation, while this does not work very well in the case of temporal modalities. Let us start with David Lewis’s modal counterpart theory, under his own realist framework of possible worlds (see Lewis [1968, 1986]). This version of modal counterpart theory claims that all individuals are world-bound (that is, they inhabit only one world) and provides an analysis of de re modal statements in this way: (i) (X is possibly F) ↔ (X is F or at least one counterpart of X is F)3 (ii) (X is necessarily F) ↔ (X is F and all of X’s counterparts are F)

Eliminativism, Life, and Death  117 The picture such an analysis provides is the following (I put it here mainly because it will be useful to compare it to another picture, in §3 below). counterpart Jerry

Jerry*

W1

W2

In W1 (say, the actual world), Jerry is, as we know, an alpinist and he is 185cm tall. He does not exist in any other world, but he has counterparts in other worlds. In W2, there is an individual, Jerry*, who is exactly like Jerry except that he is only 175cm tall and instead of going in the mountains he likes to do Kung-Fu. In W2, he is the most similar person to our actual Jerry. The idea behind the counterpart-theoretic analysis of modal statements is then that Jerry could have been smaller and could do Kung-Fu, because one of his counterparts (Jerry*) is such. What is a counterpart? The counterpart relation is based on similarity. In short, something is an other-worldly counterpart of Jerry if it is an individual inhabiting a different possible world who resembles him in his important features and who resembles him more closely than any other object in that world. It is important that such individuals, his counterparts, are not (numerically) identical to him: they all exist in their own worlds and he is the only one inhabiting the actual world. There are neither causal nor spatio-temporal relations between Jerry and his counterparts – what there is, is a relation of similarity and that’s how we can say that they are what he would be, would the world be different. (More on this right below.) Among the traditional motivations one can have to endorse modal counterpart theory is the fact that it easily avoids any worries with accidental intrinsic properties. Jerry is 185cm tall in W1 but he can be said to have the modal property of being possibly 175cm tall in virtue of having at W2 an other-worldly counterpart that is 175cm tall. But since the two incompatible properties (being 185cm tall, and being 175cm tall) are had by different things – different modal counterparts – no contradiction threatens to arise (see Lewis [1986, p. 201]).

118  The No-Self View Now, the objection I am interested in here claims that such a modal counterpart theory leads to difficulties precisely because, to do the job it promises, it must be a world-bound individuals theory, which allows individuals to inhabit just one single world and no more. It is this feature of Lewis’s theory that yields objections saying that it misses the target: counterpart theory does not provide a proper account of what may happen to us, because what are to be our counterparts are just different individuals who resemble us. So, what might happen to Jerry is not about Jerry at all, it is a story about someone else, Jerry*, living in some other world which, in Lewis’s view, has neither causal nor spatio-temporal relations with the actual world. This is Saul Kripke’s worry (Kripke [1972, p. 45]): we have a strong common-sense belief that we have many properties only contingently and that there are many different ways for us to be different. If, for instance, I just avoided a deadly fall while climbing up the Matterhorn, it is crucial, in order to justify my sensation of relief, that I have the belief that it was me who could have fallen and died. But according to counterpart theory, my idea that I could have died is not really about me: if I say this, I’m speaking about someone else – someone who resembles me and who is my counterpart but who is not me in any way. Following counterpart theory, we then get a concept of de re modality that is quite different from our common-sense beliefs about what might happen. Alvin Plantinga (1973, p. 164) raised the same point in an analogous fashion: if I say I’m able to perform a certain action A (“It is possible for me to do A”.) then, according to counterpart theory, I’m speaking about someone who does A but who is not me – and of what importance and interest could this possibly be to me? How can actions done by someone else in some other world be of any help to me to know what I can do or not? Indeed, it seems that, according to counterpart theory, I (I myself) cannot perform A, for when I say “It is possible for me to do A” I’m not speaking about myself but about some other individual. Yet a different way to put the worry is Trenton Merricks’s (2003, p. 522): That objection charges counterpart theory with changing the subject. When I ask whether I might have been happier, so the objection goes, I am asking whether I – this very person – might have been happier. It is simply not to answer my question to say that an other-worldly someone else is happier, even if he is very much like me, even if we call him ‘Merricks’s counterpart’. There is a perfectly good answer to these worries. The objectors are right to insist on the fact that all these things – being smaller, dying on the Matterhorn, being happier – happen to somebody else than the individual we are concerned with in the actual world. Jerry is not 175cm tall, I did not die on the Matterhorn, and Merricks is not as happy as some of his counterparts. Some other guys are (smaller, dead, happier). So, in what sense does modal

Eliminativism, Life, and Death  119 counterpart theory relate what happens to these other persons to Jerry, Merricks, or me? One way to answer the question is simply to insist that, for instance, “being possibly 175cm tall” just is “having a counterpart that is 175cm tall”, the latter being the analysis of the former. But such a way to put it will hardly convince the objectors. The better – now quite standard – way to put it is in terms of representation. The idea is exactly the same as in the story about me (in the actual world) and Jerry (in the actual world). I – I myself – feel happy for me, because Jerry (not me) managed to climb up the Everest. Why? Because given how similar we are, the fact that he did it means that I could. This is the most common sense and natural thing to say, and we often do think about de re modality in this way. There is a situation that is extremely similar to my own situation, and something happens in this situation. So, it could happen in my situation as well. And that’s all. The main counterpart-theoretic claim is no more loaded than that: the fact that this other very similar situation exists represents the possibility for my situation to be such as well, exactly as Jerry’s success (in the actual world) represents the possibility that I could succeed (in the actual world) as well, given how similar we are.

§3 Let us now compare four ‘Humphrey objections’: Modal counterpart theory à la Lewis

Modal realism à la Lewis

Modal ersatzism (and similarly for modal counterpart theory under ersatzism)

Temporal counterpart theory

That I can possibly climb the Everest is one thing, and that a counterpart of me climbed the (counterpart of) Everest is another. I myself do not have the property of possibly climbing the Everest, since only the counterpart did it. What the counterpart does is irrelevant to what properties I have.

That the actual world could be different is one thing, and that there exist other spatiotemporally and causally disconnected universes is another. These universes are just there, but they have nothing to do with modality, it’s just things that exist.

That the actual world could be different is one thing, and that there exist some abstract entities is another. These entities are just there, but they have nothing to do with modality, it’s just things that exist.

That I climbed the Everest yesterday is one thing, and that somebody else did it is another. If a counterpart of me – who is not me – did it, this does not mean that I did.

120  The No-Self View We have seen above the first objection, and we have seen that it can be satisfactorily answered by appealing to the notion of representation. What we realize here is that the way something has its modal properties is a very different story from the way it has its non-modal intrinsic properties. Jerry is 185cm tall. This is an actual intrinsic property he has. Modal properties are different. The property “being possibly 175cm tall” is not something that Jerry has in any comparable way. He does not instantiate, say, the trope of being possibly 175cm tall. There is no such trope. Rather, under modal counterpart theory, the better way to put it is to claim that it is correct to say that Jerry could have been 175cm tall, because there is a situation very similar to his where an individual very similar to him is 175cm tall – in the way we have seen at the end of the preceding section. Modal properties are things of a different kind than actual intrinsic properties, and modal counterpart theory acknowledges this fact very well – indeed, it requires it. But this is no defect of the theory, it is a virtue. To say that Jerry “has” a “modal property” is nothing more than to say that something is possible for him, and this is represented by the existence of a situation where this something is actual for somebody else, suitably similar to Jerry – thus, it is consistent/ compatible with the way Jerry is. Let us now move on to the other Humphrey objections. The kind of worry that lies behind the Humphrey objection to modal counterpart theory can be raised in a similar way against modal realism itself, and against ersatzism itself as well. Modal realism à la Lewis is, in short, the claim that there are many possible worlds as concrete and flesh-and-blood as the actual world. All possible worlds (including the actual world) are things of the same kind, and all possible worlds (including the actual world) are spatio-temporally and causally disconnected from each other – they are isolated – this is how they are distinguished one from each other. The notion of actuality is then given an indexical analysis: to be actual is no more than to be here, in this world. Possibility and necessity are then analysed in terms of possible worlds in a standard way: it is possible that P iff there is at least one world where P, and it is necessary that P iff in all possible worlds P. The way Merricks formulated (see above) the Humphrey objection against modal counterpart theory was that it changes the subject. A similar objection can be raised against modal realism itself: that the actual world could be different is one thing, but that there exist other spatio-temporally and causally disconnected universes is another; these universes are just there, but they have nothing to do with modality, it’s just things that exist. So, even if we grant Lewis all the worlds he claims there to be, he still did not say anything about possibility and necessity, he only made a claim about what there is, and the claim is that there are much more concreta than what we thought. But the fact that there are more people, more planets, and more donkeys has nothing to do with what is possible or necessary, it only has to do with what there is. Ersatzism is not very different from modal realism in this respect, although it is a very different theory of possible worlds. Ersatzism goes

Eliminativism, Life, and Death  121 around by different names like “actualism”, “moderate modal realism”, or “abstractionism”. “Ersatz modal realism” is the label given to this family of theories by David Lewis (1986, chapter 3), since according to this view possible worlds just are not worlds. Instead of a plurality of ‘genuine’ concrete spatio-temporal worlds à la Lewis, the ersatzers postulate a plurality of abstract non-spatio-temporal entities that play the theoretical role of possible worlds. This is why the view is also properly called “abstractionism” (see, for instance, Van Inwagen [1986]) since it makes extensive use of abstract (i.e. non-spatio-temporal) entities, and makes them to be a central piece of modal metaphysics. According to this view, there really is only one world (quantifying unrestrictedly), the actual world (that is absolutely, and not relatively, actual), and it contains abstract entities that play the role of possible worlds. Thus, on this view, possible worlds really are actual entities – and that’s why the view is also often called “actualism” since the claim is here that only what is actual exists, and merely possible non-actual things don’t. So, here we have a theory-schema: instead of a plurality of concrete possible worlds, we have here a plurality of actual abstract entities that play the same theoretical role. The question that arises now is – how? How do those entities manage to play the role they are intended for? And, in the first place, what are they? The answer here is not simple because many different versions of ersatzism are on the market. Abstract possible worlds have been said to be (maximal and consistent) states of affairs, sets of propositions, or special properties – ways things might have been. One of the best-developed version of ersatzism is Mark Heller’s (1998a, 1998b) linguistic ersatzism. In Heller’s (1998a) view a possible world looks roughly like this: {, , , …} a point in the manifold

a set of numbers representing properties

properties being exemplified at a certain point of the manifold the representation world = a set of sets/ordered pairs

This Quine-style view takes possible worlds to be complex sets that represent a consistent distribution of properties in a space-time manifold (only two-dimensional in the schema above). At this point it should be obvious how a Humphrey-style objection can arise here as well. Indeed, exactly as we have seen above in the case of

122  The No-Self View

{Jerry is 120cm tall., …}

represents correctly ‘is actualized’

{Jerry is 175cm tall., …}

represents falsely

represents falsely

{Jerry is 185cm tall., …}

concrete actual world

abrstract possible worlds

modal realism, one can worry here that it is one thing to say that the actual world could be different, but that it is an entirely different thing to say that there exist some abstract entities or complex sets. “The sets just sit there”, as Merricks (2003, pp. 535, 539) nicely puts it. As before, the worry is that these entities (sets or other) are just there, but they have nothing to do with modality, it’s just things that exist. So again, the objection goes, ersatzism is not a theory about possibility and necessity, it is a theory about what there is, and it claims that in addition to the concrete entities of the actual world there are also many abstract (actual) entities, like sets of numbers, propositions, or other. But, as before, the fact that there are such abstract entities has nothing to do with what is possible or necessary, it only has to do with what there is (in the actual world). Ersatzism can also come in a 2-in-1 package, combined with modal counterpart theory, as illustrated in the figure below where the abstract possible worlds are sets of propositions (to make the schema easier to read than Heller-style sets of numbers):

According to this view, the counterpart relation holds between the different abstract representatives of individuals from different possible worlds – and not between individuals themselves. Similarity also plays an important role here, like in the case of Lewisian modal counterpart theory. The idea

Eliminativism, Life, and Death  123 is that it makes sense to claim that some abstract entity can resemble some other abstract entity. The explanation of how it is that two abstract entities can resemble each other will depend on what kind of abstract entities one chooses to make up possible worlds. Perhaps abstract entities can have similar internal structures. In Heller’s (1998a) view, where as we have seen worlds are sets of ordered pairs of numbers representing properties, counterpart relations are understood as relations between properties instantiated at points of the manifold – the relevant similarity involved in the counterpart relation is then the similarity between these sets, which is roughly the question of how many members they have in common (see Heller [1998a, p. 301] for more details). This does not make things better, in the eyes of the objector, when it comes to the worry that this view has nothing to do with modality – that it changes the subject. As Merricks (2003, p. 536) puts it, “My being possibly 40 feet tall is not the same thing as there being a particular set of ordered pairs of ordered quadruples of numbers and sets of numbers”. Things seem here to be even worse than in the case of a Lewisian modal counterpart theory, for at least people were involved in the Lewisian analysis of what is possible for Jerry. Here, it’s just sets of numbers. But to say that there are such sets is one thing, and to say that Jerry could have been smaller is another – and a very different thing to say. To sum up, the family of worries arising here against modal realism, against ersatzism, and also against a modal counterpart theory under ersatzism is very similar to the initial worry we had with modal counterpart theory à la Lewis. When it comes to the two theories of possible worlds (modal realism and ersatzism), the worry is that they have nothing to do with modality but only with what there is, and when it comes to an ersatzist modal counterpart theory the worry is the same as the initial Kripkean one supplemented with the additional worry that sets are even weirder entities than other people to play the role ersatz counterpart theorists want them to play.

§4 In the preceding section we have seen three different theories (modal realism, ersatzism, ersatz modal counterpart theory), and we have seen how a Humphrey-style objection can arise against all of them. Now, it is time for answers. There is a perfectly good answer, very similar to the answer provided to the initial Kripkean objection in §2 above. There are two crucial components that constitute the answer: first, to “have” a “modal property” is a different thing than to have an intrinsic actual property (see above), and second, the best way to understand how our three theories work is in terms of representation. It is true that there is nothing modal in the possible worlds themselves or in the counterparts themselves. They are just things

124  The No-Self View there are. But the fact that they are out there means something. It means at the very least that it is possible for something or some world to be suchand-such, since it is such-and-such. Take Lewisian modal realism first. If it is the case that there is a spatio-temporal universe exactly like ours except that Jerry is 175cm there, it means that there is no (logical, metaphysical, and perhaps also nomic and biological) contradiction in this situation – otherwise it would not exist (at least under the assumption that there are no contradictions instantiated in a concrete world). This means something. It means that the existence of this world represents the possibility that our world could have been such. Think again about the story of Jerry (in the actual world) and me (in the actual world) from the beginning of this article. The fact that Jerry exists, that he is so much like me, and that he managed to climb up the Everest means to me that I could do it as well. So, we don’t even have to look in other possible worlds to get to modality. Modality is to be found in every situation similar to mine. Yes, Jerry (the actual Jerry) is “just a thing that exists”. But his existence and the way he is are enough for me to say that something is possible for me. It’s exactly the same story with possible worlds, especially in the case of Lewisian modal realism. Indeed, to be able to say that I could climb up the Everest I do not need to look in other worlds – in my own world, there is the actual Jerry, and that’s enough. But for other possibilities that concern me there is no such suitable doppelganger to be found in my own world. For instance, say that it is possible for me to climb up the K2. But, Jerry didn’t climb up the K2 and nobody else similar enough to me did. So, this is how the simple fact that there are more concreta than those that inhabit my own world is useful, for in some world or another there is someone, suitably similar to me, who did climb up the K2 (or not, in which case it is not possible for me to climb up the K2). So, the existence of all the concreta Lewis postulates has everything to do with modality.4 The answer is similar in spirit but technically a bit different in the case of ersatzism like Heller’s and in the case of an ersatz modal counterpart theory. Of course, one can say, following Merricks, that Jerry’s being possibly 175cm tall is a completely different thing than there being a set of sets of numbers. But these numbers represent something. They represent properties, instantiated in a space-time manifold. (To be able to do this, Heller-style worlds need to be interpreted – a number in itself does not represent anything, of course [see Heller (1998a, p. 296)]). If the distribution of properties in a space-time manifold is such that it makes up for the existence of someone very much like Jerry except that he is smaller, we get all we need in the same sense we have just seen in the case of modal realism. The fact that there is a (maximal and consistent) representation of a distribution of properties in a space-time manifold where there is a person like Jerry with such-and-such properties is not but means/represents that Jerry could have been smaller. This is how, again, the simple existence of these sets of numbers has everything to do with modality.

Eliminativism, Life, and Death  125

§5 I hope that I succeeded in providing some support to the claim that Humphrey-style objections to modal realism, ersatzism, and to modal counterpart theory (both under modal realism and under ersatzism) do have good answers. I will now try to show why a Humphrey-style objection raised against temporal counterpart theory is stronger and why the good answers from the preceding sections are not available here. Temporal modalities and alethic modalities will turn out to be relevantly different – which will be relevant for the case of personal identity. In short, temporal counterpart theory says that “an object will be F iff it has a future temporal counterpart that is (tenselessly) F, and that an object was F iff it has a past temporal counterpart that is (tenselessly) F” (see Sider [2006, p. 14]). According to this view, a person like Jerry exists at only one time (he is time-bound, in a similar sense in which he is world-bound according to modal counterpart theory): he is an instantaneous stage that persists through time not by existing at other times but by having different temporal counterparts at other times. Thus, there is a series of stages interrelated by a counterpart relation, and ordinary objects such as Jerry are conceived of as being the stages rather than the whole composed of them (that would be the perdurantist ‘worm view’). While persistence through time is thus understood as the having of temporal counterparts at different times, temporal counterpart theory does not deny the existence of temporally extended objects – the four-dimensional entities that are aggregates of stages – they exist as well as the stages do. It’s just that, according to this view, the objects we ordinarily name and quantify over are stages rather than four-dimensional worms. To get to see where a Humphrey-style objection comes from here, we can start by asking how this view accounts for there being change in intrinsic properties. Change, according to all perdurantists (friends of temporal counterpart theory or not), is simply the having of different properties at different times. The trouble with temporal counterpart theory is that there is no one thing that ever has the different properties. Since all objects only exist at one time, there just is nothing in the stage view theorist’s world that can undergo a change – it slips out of existence before it can change. Change takes time. But the counterpart theorist’s stages do not last long enough. The friend of temporal counterpart theory will of course reply here that a given stage at t1 is F and will be ¬F at t2 in virtue of having as a temporal counterpart another stage existing at t2 that is ¬F. But, the objector says – and here starts the Humphrey objection – this is only an appearance of a solution for these two stages are just two completely different things. Mellor (1998, p. 89) claims that “change needs identity as well as difference” but there is only difference in the stage view, there are only different things with different properties – and nothing that undergoes any change at all. A similar way to put it is this: temporal counterpart theory does not allow

126  The No-Self View ordinary objects to do the things they typically can do. People are stages. But stages are instantaneous entities, they do not have temporal extent. The unwelcome consequence of this is that people cannot do many of the things we would expect them to be able to do. For instance, it seems that a person should normally be able to utter a sentence. But, on the stage view, this turns out to be, strictly speaking, impossible: the utterance of a sentence takes some time and a stage does not last long enough to make such a performance. Or, normally, we would say that a person is able to run, but again, not according to the stage view; strictly speaking, nobody can run because a person is an instantaneous entity and running takes time. What is involved in such objections to the effect that temporal counterpart theory does not provide a good account of change and persistence is a version of a Humphrey objection, applied to temporal counterpart theory: if Jerry says now that he climbed up the Everest last week, the sentence turns out to be true iff he did climb up the Everest last week. But this is, according to temporal counterpart theory, simply impossible, because the person who says now that he climbed up the Everest is a stage, a momentary entity that did not itself exist last week and thus was not able to climb up anything. Jerry, the person who is doing the speaking (or perhaps not even that, see above), is simply not identical, in any sense, to the person who’s done the climbing. Granted, Jerry has a counterpart last week that climbed up the Everest. But whatever the counterpart relation is, it is not identity. So, the objection goes, if Jerry says he climbed up the Everest, he is not telling the truth – why should he be allowed to boast that someone else, similar and causally connected to him but still someone else, climbed up the Everest? If one generalizes this objection, one can claim that temporal counterpart theory denies persistence altogether, for it only provides us with different instantaneous entities. Sider defends temporal counterpart theory against this objection as follows: [It] is wrong to say that the stage view [i.e. temporal counterpart theory] denies that “You will do it” means that you will do it. “Ted was once a boy” attributes a certain temporal property, the property of once being a boy, to me, not to anyone else. Of course, the stage view does analyse my having this property as involving the boyhood of another object, but I am the one with the temporal property, which is the important thing. The stage view is consistent with stages having temporal properties; it’s just that temporal properties are given a counterpart theoretic analysis. (Sider [2001, p. 195]) But this reply is not likely to give satisfaction to the objector. Granted, the stage view is consistent with stages having temporal properties, but not the ones we want. To take Sider’s example, if we say “Ted was once a boy”, we are ascribing a temporal property to Ted (who exists now). But if we

Eliminativism, Life, and Death  127 endorse temporal counterpart theory, it is not the property of “once being a boy”, but rather, the property of “once there being a counterpart of Ted that is a boy”. If the counterpart theorist allows these two properties to be equivalent then, the objector claims, he is mistaken – for if it is the former that we ascribe to Ted, it is solely about Ted that we are speaking, but if we ascribe him the latter, we are speaking about Ted and someone else, and that makes all the difference. Sider’s response can only appear to be satisfactory if one takes the expression “once being a boy” to be a suitable paraphrase of the expression “once there being a counterpart of x that is a boy”, but such a strategy, objectors like Sally Haslanger will claim, “strains the limits of credibility” (Haslanger (2003, p. 337)). I can boast that I can climb up the Everest even if I did not – only my doppelganger Jerry did. Yes, I can boast that I can, but I cannot boast that I did. Somebody else did. This somebody else’s achievement represents the possibility of me being able to do it, as we have seen. Representation is fine and enough when it comes to alethic modalities. But it’s not what we want in the temporal case. In the temporal case, I want to say that I did it, and not that although I didn’t do it something represents me as having done it. If I didn’t do it, then I didn’t do it, and no representation can be of any help. That I climbed the Everest last week is one thing, and that somebody else (Jerry) did it is another. The latter is true, the former is false. And even if a temporal counterpart of me – who is not me – did it, this still does not mean that I did. In the case of alethic modalities it is enough to say that even if I did not climb up the Everest, I could, because there is a suitably similar situation where somebody else does. But in the temporal case, a stronger claim is needed. It is not enough to be similar and causally connected to someone who did something to say that I did it (given that temporal counterpart theory explicitly states that I did not). Mere representation is not enough here. In the temporal case we need the real thing, for the story here is about my past and my future, not just mere possibilities. I don’t want to have a merely represented past and future, I want to have a past and a future, full stop.5 As is apparent in the quote above, Sider will simply reply that this is just the way a counterpart-theoretic analysis works. I cannot refute this response, but I tried to say why in my mind it indeed “strains the limits of credibility”. To repeat it: representation is enough in the case of possibility and necessity, but not in the temporal case. This being said, we will see in the following sections how an eliminativist conventionalist strategy can be of help here.

§6 What we learn from the preceding sections and from the worries that surround temporal counterpart theory is that in the case of diachronic personal identity, we seem to need a sense of a “me” as being the same person at different times. The no-Self view denies that there is anything like a Self that

128  The No-Self View would be numerically identical at different times – indeed, there simply is nothing like an ontologically reified Self at all, at any time. Is that a problem for the no-Self view? Does this metaphysical claim have any undesirable consequences, when it comes to our lives? In general, what is the relationship between metaphysical theories and moral or other practical issues? To start, let us consider – rather quickly, mainly in order to have a clear and illustrative case in mind – Mark Heller’s argument against David Lewis’s modal realism from his article “The Immorality of Modal Realism, or: How I Learned to Stop Worrying and Let the Children Drown” (Heller (2003)). According to modal realism, possible worlds are as real and concrete as the actual world is, and so are their inhabitants. Thus, there are many more people than the people in our world – more people to “figure into our moral calculations than we had thought” (p. 20). In his article, Heller devises a sophisticated case (which I will not report here in detail) where it becomes morally permissible for someone to let a child drown while she could have saved her, because, to put it shortly, she knows that if she does not save the child in this world, someone else will save the child in some other world, and that if she does save the child in this world, it means that somebody else will let the child drown in some other world (thus, by saving the child here, she will make a child in another world drown). So, she has to choose between the two children (i.e. the two counterparts). But, under modal realism, both children are equally flesh-and-blood real. So, she is not morally obligated to save the child in this world – she can let the child drown. This is a crude summary of Heller’s argument, but it is sufficient for our current illustrative purposes. Simply suppose, for the sake of the methodological discussion here, that we accept Heller’s argument and his conclusion. Heller then claims that a modus tollens can be drawn: modal realism has a morally unacceptable consequence, and so it should be rejected. But, how exactly was the conclusion derived from the truth of modal realism? It is easy to see that it was not derived from this metaphysical theory alone. To get to the moral consequences Heller draws, at least these two additional moral claims have to be made: it is not permissible to let children drown when it would be easy to save them (see Heller [2003, p. 3]) • mere spatial proximity should not have any moral significance (when applied to the case of modal realism, this means that being located in the same world should not have any moral significance [see Heller (2003, p. 4)]) •

These two claims are very sensible, the first being – ceteris paribus – even probably obviously true, and the second being at least very defensible. But this is not the point here. The point is, Heller’s moral consequence of modal realism is not a consequence of modal realism. It is a consequence of these two moral claims (and perhaps others), in a situation where they become

Eliminativism, Life, and Death  129 practically significant – it is only in this sense that modal realism becomes relevant, namely, in the sense that it creates a context in which these two moral claims become relevant. In his article, Heller actually provides a situation entirely located in one – our – possible world, where a similar conclusion is drawn, namely a situation where it would be permissible to let a child drown, knowing that this will make another child saved by somebody else, further away on the beach (again, this is an oversimplification of Heller’s case, described in detail in Heller [2003, pp. 5–6]). Modal realism is merely a generalized case of the same reasoning. The objection to modal realism can then be that it gives rise to a framework in which this unwelcome consequence is of practical significance very often in many situations – since there are many more people to take into account than we usually do. This is perhaps what Heller had in mind. But for our present purposes it is important to see that it is not from modal realism that the unwanted moral consequence is drawn. It is drawn from moral/practical principles. It’s just that if modal realism is true, it has to be drawn more often.

§7 With the case of Heller’s argument against modal realism in mind, let us now focus in more detail on the case of practical conclusions one can allegedly draw from some views about personal identity. When it comes to personal identity, one can find in Buddhist metaphysics both a form of reductionism and a form of eliminativism about the Self. I shall not attempt here any exegetical or interpretational work, and the understanding of Buddhist thought I will have in mind here will be mostly based in how the Madhyamaka tradition can be understood from Siderits (2003, 2007), perhaps in a simplified way, although most of what is relevant to the issues I discuss here can also be found in, for instance, Abhidharma texts and in other interpretations. The latter is closer to some forms of reductionism while the former – the Madhyamaka – is perhaps closer to a full-fledged eliminativist view. In general, in this chapter, I will not focus on the differences between eliminativism and reductionism6 – the point I want to make applies equally to both types of views, and it’s the claims common to both types of strategies that will be relevant for our present purposes. This kind of rejection of an ontologically strong conception of the Self can also be found in Western thought, in many places, famously in Hume (1738), and in Parfit (1984), as we will see below. The basic common idea that can be found in all these traditions and views is that there is a succession of impermanent psychological (and sometimes physical) events/states, which are linked by causal relations (and perhaps also by similarity, and temporal/spatial continuity and/or contiguity), and – crucially – that’s it. There is no owner or bearer of these events/ states – there is no Self. In Buddhist thought, as we have already seen in Part I, Chapter 1, there is a distinction between two kinds of truths: conventional truths (samvrtisatya)

130  The No-Self View and ultimate truths (paramarthasya). Conventional truths are about conventional objects, which include ordinary material objects, and, relevantly to our concerns here, the Self. All of these entities are conventional, they are useful fictions. It is useful for us to say and think that there is a chair I am sitting on now, but a chair is merely a conventional object – exactly as in Heller’s eliminativist view. The concept of a chair is useful in everyday life, and this is why we have this convention. Ultimate truths are about the ultimate entities that make up the world. The chair is ultimately made of such entities (whatever they are). When arranged in a relevant way, it is then useful for us to say that such ultimate entities compose a chair, but again this is only a convention we have – the chair is a useful fiction. Thus, in ordinary language we can say that it is true that I am sitting on a chair, while this is false in the language of ultimate truths (in some versions of the view, it is meaningless rather than false, since the word “chair” does not refer to anything in the domain of ultimately real objects). Buddhists apply the same strategy to the Self. The truth – the ultimate truth – about the Self is that it does not exist. What there is, is a succession of impermanent psychological states/events. These are not persons, nor do they make up a person. They are ‘just there’, and succeed each other. But, if they succeed each other in a proper way, it makes then sense to speak of this succession as of a person – a person is then a useful fiction, in the same way a chair is a useful fiction. Persons, in this view, are conventional entities. As Siderits (2007, p. 57) remarks, the Sanskrit word samvrti, which is usually translated as “conventional” literally means “concealing”. Indeed, words like “chair”, used in ordinary language, conceal the truth – the truth that there is no chair. “Chair” seems to be a singular term, while it actually refers to a plurality of ultimate entities arranged chairwise. The same is true about the word “I”. We conventionally and usefully use it as a singular term, referring to a person, but there is nothing – no one thing – that this word refers to in the domain of ultimately real objects. This strategy could thus be understood as being able to overcome the worries surrounding temporal counterpart theory that we have seen above. It denies the existence of a persisting Self, since it denies the existence of a Self altogether, but it appeals to a conventionalist strategy to be able to provide a notion of a Self qua conventional/ fictional entity. The convention that there is a person7 to which I refer by the word “I” is a useful one, and in general it enhances my overall happiness. Indeed, if I identify with a person that lasts for a longer period of time, it then makes sense for me to act in ways that help me to promote my future well-being. For instance, I avoid drinking too much beer, I work hard to become a better philosopher, I exercise, and so on. What good can then come from the truth that, after all, I do not exist? The trouble is, in the Buddhist view, that if I take the convention that I am a person too seriously, it then leads to existential suffering. If I believe that I am more than a useful fiction, I then face the fact that I will die, which makes me face a situation that leads to suffering. This is why it is

Eliminativism, Life, and Death  131 important to realize that, in reality, I do not exist, which then liberates me from this kind of suffering. If I do not exist, then, to put it crudely, I cannot die. I cannot be frustrated by a possible lack of the meaning of my life,8 since there is nobody who could be the subject of such a frustration. If I do not exist, then my life cannot lack a meaning – and, it cannot have a meaning either. The question of the meaning of “my” life simply does not arise. This understanding of the reality of “my” situation helps me then to live more fully the here and the now, and to be more serene about “my” future. Parfit (1984, pp. 13, 95) also finds the truth of his brand of reductionism about the Self “liberating and consoling”. He understands the truth that there is no Self as entailing that there is less difference between his life and the lives of other people. After all, there are no other people – there only are successions of psychological states/events, related in various ways. Thus, his connection to other people is, to some extent, similar to the connections he has to “himself”. Similarly as in the Buddhist tradition, he then feels closer to others, and more concerned about their well-being, than under a nonreductionist view of personal identity – a view that makes him feel more isolated, “imprisoned” in himself. Parfit distinguishes between the ‘Extreme Claim’ and the ‘Moderate Claim’. The ‘Extreme Claim’ amounts to say that we have no reason at all to be concerned about our future, since in the future there will be nothing identical to what “I” am now – this is what the objection to temporal counterpart theory above was about. But there is also his preferred ‘Moderate’ way to see this situation. In this view, Parfit claims, reductionism (as well as eliminativism, we can add) helps to cope with the fear of death. Indeed, instead of seeing death as a destruction of myself, I can see this situation as one where my present experiences will no longer be connected in some relevant ways to some other future experiences. There will be other experiences, but they will not be causally and psychologically connected to my present experiences in such a way that I could conventionally say that they are “mine”. But there will be other experiences, some of which will be connected to my present experiences in other, less strong, but still interesting ways. For instance, there will be memories of me, or thoughts influenced by my work as a philosopher, or the experiences of someone who learned how to ski from me, and so on. As Parfit (1984, pp. 13, 95) puts it, “My death will break the more direct relations between my present experiences and future experiences, but it will not break various other relations. . . . Now that I have seen this, my death seems to me less bad”.

§8 In §9 below, I will discuss the practical conclusions that Buddhists and Parfit draw from their metaphysics, but before I do that, I will in this section examine the general methodology that underlies the steps they take. The Humean claim I want to put forward is this: as in the case of Heller’s

132  The No-Self View argument from §6, the practical conclusions of the kind we have seen above do not follow from the metaphysical claims. To be more precise: they do not follow from the metaphysical claims alone. Additional (practical) premises and steps are needed in order to derive any practical conclusion from a metaphysical claim (this is something that Parfit and Buddhists would probably accept). And, it’s these additional premises that actually do the main work. So, there is no direct path from a metaphysical theory to practical conclusions. Practical conclusions always require practical premises. The metaphysical theories involved merely provide a context where the practical premises become significant. In this sense, the metaphysical theories do contribute to the practical conclusions that can be drawn, but only in a very indirect way – again, only by creating a context where the practical premises become relevant and can do their work. Parfit wants to go from his reductionism about the Self to the conclusion that one can be less concerned about one’s own future suffering, or about one’s death. But, again, this route can only be an indirect one. Reductionism about personal identity is a claim about what there is, or more exactly about what there is not: there is no Self. Suppose that this is true. Even if it is, it is not enough to derive the desired practical conclusion. As in the case of Heller’s argument, connecting claims are required. Depending on which exact conclusion one wants to draw, these can take different forms; perhaps an example of such a connecting claim could be: “one should care less about appropriately causally connected future experiences to my current experiences than about future experiences had by the same bearer than the bearer of my current experiences”. This can be put in different ways, but the general idea here is that there has to be a claim of the form “one should care more/less about X than about Y”. Right now, I do not want to object to this claim (I will discuss it in §9 below). What I want to point out is that if one wants to derive a conclusion where we can be liberated from the fear of death, or from future suffering, or to feel closer to others, it is this (type of) claim that is doing the work, and not the metaphysical claim that the Self is to be reduced or eliminated. This latter claim only states what there is or what there is not. It does not say anything about what we should care for or how we should feel about it. It is not the absence of an ontologically reified Self that is important. It is the absence of caring in a too strong way about the Self that is the crucial point here. A way to see this is to realize that (i) even if one accepts reductionism or eliminativism, one can have an entirely different reaction/conclusion (as Parfit himself recognizes, bearing in mind the ‘Extreme Claim’) and that (ii) one can have the same reaction/conclusion even if one accepts nonreductionism.

Eliminativism, Life, and Death  133 When it comes to (i), as Parfit (1984, p. 13, 95) himself mentions, one could find the truth of reductionism depressing, rather than soothing. Parfit himself finds it liberating, but somebody else could very well have the exactly opposite reaction, since instead of thinking that “one should care more about X than about Y” she could be thinking that “one should care more about Y than about X”. This is where the crucial point lies – in what you care about and why. We have to argue about what counts more than what, and why, and what we should care for, how much, and why, and only then, and only from these practical considerations, can we derive any practical – soothing or depressing – conclusions. Reductionism, qua metaphysical claim, is compatible with numerous and different conclusions, and so is the rival non-reductionist metaphysical claim; nothing practical follows from them. Coming to (ii), one could very well have the same ‘conclusion’ as Parfit has, even if one rejects his reductionism and embraces a non-reductionist view. Indeed, it is very well possible not to be subject to existential suffering and the fear of death if one is a non-reductionist, for the same reasons Parfit provides in his case. Suppose I am a non-reductionist. As such, I could very well embrace the view that I care mostly for the experiences I have, and those that I will have, and for the quality of these experiences. Also, I can be someone who cares a lot about sharing experiences with others. For instance, I can care about having present and future experiences of freeride backcountry skiing in deep fresh perfect powder snow, and about sharing this experience with my son. As a non-reductionist, I believe that I exist, and that there is a Self that is the bearer of such experiences, and similarly for my son, but this is not what I think about when we go skiing. What I think about and what I care about is the (shared) experience, not the metaphysics of the bearer of the experience. I do believe, qua non-reductionist, that there is a bearer, and that in general the existence of experiences requires the existence of a bearer, similarly as in the case of properties that cannot ‘float free’ and that require the existence of a substratum (say, for the sake of the example, that in addition to being a non-reductionist, I am also a non-bundle-theorist). But in my practical considerations, that’s not what is important and what I care about. Thus, in the end, I care about the same things Parfit does care about, when it comes to his/my fear of death. I – and Parfit – care for there to be good future experiences. That’s what counts for me, and not the fact that there is a metaphysical glue gluing these experiences together in some way. In the same way, I care about the future experiences of a person – say, my son – who learned how to ski in deep powder from me. In this way, given my (causal) influence on the existence of future experiences – that is, my influence as a ski teacher, or (hopefully) my influence as a philosopher, etc. – there will be, because of me and my actions, interesting future experiences, and I do care for that, even if they will not be my experiences. Thus, after my death, there will be some interesting experiences linked in an interesting way to me. To paraphrase Parfit: now that I have seen this, my death seems to me less bad.

134  The No-Self View Thus, from (i) and (ii), we see that the metaphysical issue concerning reductionism and non-reductionism about personal identity, and the practical issue concerning, for instance, the fear of death, are entirely orthogonal. I have focused here on Parfit, but the same is true when it comes to Buddhist thought. Here as well, it is not the ‘emptiness’ of the Self that provides any practical conclusions. Rather, it is the way one interprets it and the way one faces this truth that will lead to practical – positive or negative – consequences. The great job done in the Buddhist tradition is then to show that such a metaphysics can be understood as having a positive impact on our lives, on our existential suffering, on the way we interact with others, and so on. This is where the great insights of Buddhist thought lie. But these are practical insights, relying on practical principles of the form “one should care more/less about X than about Y”, as well as others. So again, their practical conclusions follow mainly from their practical premises and principles, and not from their metaphysics. (And, as a consequence, such positive practical claims cannot serve as arguments in favour of the metaphysical claims. To repeat what we have seen above: the metaphysical theories involved can at most provide a context where the practical issues become salient.)

§9 In the preceding section, I have argued for the independence of metaphysical claims and the practical ‘conclusions’ that were supposed to be derivable from them. Now, I would like to say a bit more about some of these practical conclusions (this will also make clearer and provide perhaps even more support for what I said above). Often, the argument involved in Buddhist thought is formulated in terms of the meaning of life, as for instance Siderits (2003, p. 99 and p. 101) does it: Only when we think of ourselves as persons does the question arise what meaning my life can have in the long run, given that in the long run there shall be no me. . . . [T]he demand that a life appear to display some sort of unifying telos sets the stage for suffering. For this is the demand that my life be made up of projects that contribute to the meaning, value and purpose of this life. And how can this life be seen as meaningful when in the long run its central character will disappear from the stage? The problem of existential suffering is here – as well as in many other places in the literature on Buddhist thought – understood in terms of a meaning and purpose of life. I think that this is not the best way to understand what is at stake here. The problem of a meaning of life is a genuine and interesting problem of course, but it does not come from the alleged truth that we exist and that we are ontologically strong mortal Selves. To

Eliminativism, Life, and Death  135 put it shortly, the problem of the meaning of our lives is a problem that we would have even if we were immortal. Thus, the alleged truth that we are Selves that will one day die – that is, the alleged truth of the transitoriness of the Self (the fact that “in the long run there shall be no me”) – does not in itself lead to existential suffering understood in this way. This type of existential suffering, understood in terms of a meaning of our lives is an independent affair, affecting mortal Selves, as well as immortal Selves, and reductionists as well – indeed, even for a reductionist it is possible to ask, albeit in a different way of course, what meaning do the actions and experiences there are have. The trouble that (can) come from the fact that we are mortal and transitory Selves, and the existential suffering that can come from this, is, I believe, rather a type of regret directed towards the future – the regret that we will have no more future experiences. Suppose again, for the sake of the current example, that I am a non-reductionist. The type of regret I have in mind here is similar to the regret of having missed some past interesting opportunities. For instance, yesterday I was not able to go out skiing while the weather and snow conditions were absolutely perfect, because I was sick. I regret this, it’s something that will never come back and the day I ‘lost’ yesterday is lost for me forever, from this point of view. The same applies to how I (can) think of my death. When I will be dead, there will be no more skiing, no more philosophizing, no more drinking wine with friends, no more interesting conversations, and so on. Given that I know that my time is limited, and that the existence of my Self is transitory, I know that at some point I will not be able to have any more experiences whatsoever. This is the kind of existential suffering that really comes from the fact that our existence is transitory. Meaning is an independent issue, arising for everybody. Now, if we set the problem of the meaning of life aside, and if we have in mind an understanding of existential suffering in terms of a type of regret directed towards the future, would it help me to be a reductionist/eliminativist? Not really. Under a non-reductionist understanding of the situation, the suffering comes from the fact that I will not have any more experiences, that I will never see my loved ones again, that I will not be able to go skiing ever again, and so on. Granted, we don’t have this problem under a reductionist/eliminativist account. But we have an equivalent worry: there will be no more future experiences that will be appropriately (causally, and perhaps otherwise) related to ‘my’ current experiences. Yes, as we have seen in the preceding section, there will be future experiences that will bear a certain degree of causality and connectedness to ‘my’ current experiences and actions. For instance, my son will go skiing, some philosophers will continue working on issues in metaphysics and metaethics, taking into account (hopefully) my current work, and so on. Thus, there will be future experiences linked in interesting ways to ‘my’ current experiences. But this does not, pace Parfit and the Buddhists, take away my existential suffering. Yes, it’s better than nothing, and one can certainly find these ideas soothing – and

136  The No-Self View so can the non-reductionist, for the same reasons. But the caring relation is not of the right type to alleviate genuine existential suffering. When I care about my (past, present, and future) experiences, I care about them in a way that is different from the way I care about the experiences of others. (Some) Buddhists deny that, or at least they claim that I should care about all experiences in the same way. But from the point of view of a practical life, this reaction is not appropriate, as we can see for instance in the case of legal and moral responsibility. Even under reductionism/eliminativism, we need a notion of responsibility for past actions. I can be rewarded for some things I did, and punished for some others. Not because I did those things (unlike under non-reductionism) but because these past actions are appropriately (causally and otherwise) linked to ‘my’ current actions and experiences. It would be inappropriate to punish my children or my students for my crimes, if I commit any and die right after. Indeed, the relation between ‘my’ current actions and experiences and ‘their’ future actions and experiences is not of the right type. For the same reason, the existence of future experiences, after ‘my’ death, is not enough to alleviate existential suffering. It’s comforting to know that there will be interesting future experiences, in the same way it is nice to know that there currently are interesting present experiences, but we cannot get more than that. In the case of legal and moral responsibility, we have the fiction or convention that I exist (in agreement with Buddhists, Heller, and Parfit), even if ultimately there is no such thing as a permanent subject of experiences that can be held responsible for his past deeds. But then, by the same reasoning, when we think about the practical issues of our lives (and our deaths), and not about the underlying ontology, the reductionist/eliminativist is in the same position as the non-reductionist: here again, existential suffering of this type is independent on the ontology of the Self. Existential suffering of this type is a practical issue, and as such requires a practical approach, exactly as issues concerning legal and moral responsibility – thus, it is not solved or even affected by metaphysical issues. Again, the best way to see this is that both the reductionist and the non-reductionist can fight the same battle against the fear of death, only expressed in different ways.

Notes Le Bihan (forthcoming) defends a similar view in an interesting way. 1 2 The label ‘Humphrey objection’ comes from the original formulation of Kripke’s objection to Lewis: The counterpart of something in another possible world is never identical with the thing itself. Thus if we say “Humphrey might have won the election (if only he had done such-and-such)”, we are not talking about something that might have happened to Humphrey but to someone else, a ‘counterpart’. Probably, however, Humphrey could not care less whether someone else, no matter how much resembling him, would have been victorious in another possible world. Kripke (1972, p. 45)

Eliminativism, Life, and Death  137 3 According to Lewis, it would be enough to say that (ii) (X is possibly F) ↔ (at least one counterpart of X is F) because of postulate P6 he introduces in Lewis (1968, p. 111): “Anything in a world is a counterpart of itself” (that is, the counterpart relation is reflexive). My (i) is only more explicit about this. Similarly for (ii). 4 A notable defect of modal realism à la Lewis is that perhaps even the huge number of concreta he postulates there to be is not enough to account for all possibilities. This is Bricker’s (2001) objection which claims that ‘island universes’ are possible – this is the possibility of a single universe containing spatio-temporally and causally disconnected parts. If this possibility is genuine, as Bricker argues, Lewis’s theory cannot account for it since it would count such a universe as two (or more) and not as one, given that causal and spatio-temporal isolation is Lewis’s criterion for distinguishing possible worlds one from each other. 5 Schechtman (1996) is among those who stress this point. 6 Remember the discussion from Part I, Chapter 3 where I discuss the intimate relationship between reductionism and eliminativism in a general way. 7 For a discussion of conventionalism about persons see Miller (2004, 2013). 8 In §9 below, I discuss the role that the notion of the meaning of life can (or cannot) play here.

Part III

Aesthetic (and Other) Objects

9 ‘Upper Level’ Objects, Musical Works

§1 In the next two chapters, I will focus on aesthetic objects. But before we discuss in detail the case of aesthetics objects, let’s talk a bit about money. Since Searle’s The Construction of Social Reality (1995), there has been a rich discussion of the nature of social objects. Searle himself prefers to talk about social facts rather than about social objects. Think about money: a bill in your hand is a material object, a piece of paper. But this is not money. Money exists only relative to attitudes that people, as a community, take towards it – those attitudes, in Searle’s view, create new facts. This ‘collective intentionality’ is necessary for the existence of money. What there is then, in this view, are material objects (or representations of those, in the case of digital currency) and collective behaviour/attitudes/intentionality. For the eliminativist-minded metaphysician, who has in mind the discussion from the previous chapters, this is then a perfectly natural place to embrace a strong reductionism or straightforward eliminativism about money – as well as about social objects in general – in virtue of the existence of the material objects, the behaviour, the collective attitudes, etc. In Searle’s view these are lower level facts that exist independently on human beings, and the facts about money (and other social objects) are upper level facts, associated with ‘collective intentionality’. For the eliminativist, the natural thing to say is then that the upper level can be eliminated in virtue of the existence of the lower level. The conventionalist eliminativist (remember the discussion from I.1.§5) would say here that money is a conventional entity – which just seems, in the case of social objects, the natural option. Of course, not everybody is an eliminativist, or even a reductionist, about social objects, but eliminativism or reductionism are a clearly plausible starting point. This prima facie plausibility of eliminativism/conventionalism is not so straightforward in the case of aesthetics objects, where various forms of realism have been defended and where an eliminativist view is not the first thing that comes to mind. To my mind, given the views I have defended in Part I and Part II about ordinary objects and about the Self, it will come

142  Aesthetic (and Other) Objects as no surprise that I believe that an eliminativist strategy can also be successfully applied to aesthetics objects as well. But as in the case of the Self, and as in the case of ‘life’ (see II.5.§3–5), the point here is not to straightforwardly apply the overall general eliminativist claim from Part I, in the sense that aesthetic objects can be eliminated in virtue of the existence of simples arranged such-and-such-wise. This might be the case for some types of aesthetic objects, as we shall see shortly, but the cases I will focus on, namely musical works and photographs, since they typically have multiple instances, require a more specific approach where specific arguments relevant to these kinds of (allegedly existing) entities need to be put forward. In the case of the Self, and in the case of life, I have used such topic-specific reasons to argue for an eliminativist claim, and I shall approach the case of aesthetic objects in the same spirit (of course, the general eliminativist strategy from Part I will play a role as well). Paintings and sculptures can perhaps benefit directly from lessons learned in Part I, since they are macroscopic material objects. Statues have indeed been at the heart of the controversy about material constitution – remember the statue vs. lump of clay debate – and can be said to be eliminated in the virtue of the existence of simples arranged statuewise. Similarly, paintings can be thought of as distributions of paint on a canvas, and conceived of as ordinary macroscopic objects, and can thus be eliminated following the recipe from Part I. But while this might perhaps work for sculptures and paintings, such a strategy will clearly not – at least not so directly – apply to other kinds of aesthetic objects, and so it will be more interesting for us to focus on these. Especially photographs provide a genuine challenge for the aesthetician/metaphysician, as we shall see in Chapter 2 below. But before we get there, let us start with an introductory discussion of musical works, for its own sake and also as a preliminary discussion for the case of photographs. The general strategy will be, in both cases, this: (i) no traditional ontological category can account for such allegedly existing entities, (ii) introducing a new specific ontological category for each type of such entities seems arbitrary, unnecessary, and clearly too unparsimonious, since (iii) an eliminativist view behaves very well in these situations.1

§2 Let us start with musical works. This is a widely controversial and largely discussed issue, and I will not here do justice to the full complexity of it. But I will try to say enough to motivate an eliminativist approach. The arguments that will show up here will then arise also in the case of photographs, and can also be applied to other types of aesthetic objects, such as novels. Think of Rachmaninoff’s Tarantella.2 Metaphysically speaking, what kind of object is this? What core features does it have? Let us not focus now on its aesthetic features, on how strong and deeply beautiful the piano piece is. Let us focus on its core features qua an allegedly existing entity. First, it is

‘Upper Level’ Objects, Musical Works  143 an object that has been created by Sergei Rachmaninoff in Italy in 1901. It came into existence at a specific time (more precisely, over a specific interval of time), in a specific place. Its existence can thus be spatio-temporally located. Its first public performance can also be spatio-temporally located.3 Also, it is an entity that can be heard. At the first performance (and even before), people have heard it. And, not only some people have heard it, but they also heard the same thing. The Tarantella is shareable. Furthermore, it can be (and has been) performed many times over and over again. Now, with all this in mind, the metaphysician is in a difficult position.4 Indeed, it seems that we need to account for (at least) four features of the Tarantella, and of musical works in general: (i) Musical works are created (they come into being at some spatio-temporal location). (ii) Musical works can be heard (which means that they are able to causally interact with us, in space-time). (iii) Musical works can be performed. This is a feature Levinson insists on strongly: “Musical works must be such that specific means of performance or sound production are integral to them” (Levinson [1980, §III]). (iv) Musical works are not private, they are shareable. Intersubjectivity needs to be accounted for. Also, relatedly, musical works can be multiply instantiated, they are repeatable. All of these points are very controversial – or at least it is very controversial to see how we can account for them. Let us examine some possibilities.

§3 (i), (ii), and (iii) seem to suggest that musical works are concrete entities. As we have seen, some aesthetic objects like sculptures or paintings seem perhaps to be such, so it would be nice to have a metaphysical theory that would equally apply to all kinds of aesthetic objects. What kind of concrete objects could musical works then be? As a first candidate, sounds come naturally to mind. Sounds can be heard and they can be heard by a large audience, so they are in this sense shareable. But the sounds that are involved in a performance of the Tarantella are not created by Rachmaninoff in the relevant sense. They can be said to be caused by him if he is one of the two performers (the Tarantella being a piece for two pianos), but the fact that he can be one of the performers is only contingent, since most performances are performed by other performers, and – most relevantly – ‘creating’ the sounds is not the right kind of creating when it comes to the creation of the Tarantella. Also, the sounds die out and exist only for a brief interval of time – the time of the performance – while the musical work itself continues to exist: it is not destroyed once a performance is finished. Sounds are thus bound to a specific region of space-time, which is limited and does not correspond to

144  Aesthetic (and Other) Objects the span of existence of the musical work. Neither sounds qua vibrations of air nor performances qua spatio-temporal events thus qualify as being the musical work, since a musical work survives their end (and usually comes into being before they do). Furthermore, performances of the Tarantella may include wrong notes, while the Tarantella itself does not. Concerning the desiderata listed above, (ii) and (iii) are accounted for by the view that musical works are concrete objects like sounds or performances, but (i) is not accounted for in the right sense of creation, and (iv) is not accounted for since sounds and performances are by their very nature not multiply locatable in space-time, and so they cannot be multiply instantiated. What about the score? The score certainly is a concrete object and it can outlast a performance. Could the score be identified with a musical work? Quite clearly, it cannot. The score is usually a piece of paper with ink marks on it, or some other material object (say, pixels on a computer screen, or a piece of granite if the score were engraved in a huge granite slab), but the existence of such an object is not necessary for the existence of a musical work, and most people who are familiar with a work like the Tarantella have never seen its score. (Jimi Hendrix, who learned music by ear, never read nor wrote music at all.) Pieces of papers, computer screens, or pieces of granite cannot be heard, and they can hardly be said to be performed. They can perhaps be shared, since they can be multiply printed/realized and since they can be seen by many people, but (iv) is then the only desideratum on the list above that can be, in this limited sense, accounted for – none of the other points can be satisfied. As in the case of sounds, such material objects can be created, but again, not in the relevant sense of creating a musical work (it would be wrong to say that Rachmaninoff composed a score), and they can be destroyed while the musical work itself continues to exist. Trying to overcome these worries, musical perdurantism is an interesting recent addition to the existing list of options that take musical works to be concrete entities. Such a view is developed and defended in Caplan and Matheson (2006, 2008), as well as in Tillman (2011), and Tillman and Spencer (2012). In what follows, I will mainly focus on Caplan and Matheson’s version of the view, but the worries that I am going to raise apply equally to all versions of musical perdurantism. In this view, musical works are concrete objects understood as mereological fusions of performances. It is in this sense that this is a perdurantist view: musical works persist by perduring, that is, they persist in virtue of having temporal parts at every time at which they exist. These various temporal parts – the various performances – are related to each other by “the appropriate continuity relation for musical works” (Caplan and Matheson [2006, p. 60]). In addition, such a brand of musical perdurantism also requires an extra modal component. Indeed, the Tarantella could have had different performances than it actually has, and it could have had more or less performances. Thus something like a modal counterpart theory or a modal perdurantism5 is needed to supplement the musical perdurantist view, in order to account for

‘Upper Level’ Objects, Musical Works  145 the contingency of the (number of) performances that are the temporal parts of a musical work. In order to explain what musical perdurantism amounts to, Caplan and Matheson claim that the situation is similar to perdurantism about persons. In this view, persons have temporal parts at different times, related by an appropriate relation of continuity, and they persist in virtue of having these temporal parts. In the case of musical works, Caplan and Matheson do not say what the nature of the appropriate continuity relation is. They add that it is not necessary to say more in order to be able to embrace musical perdurantism. But as we shall now see, this poses a serious problem for musical perdurantism, and consequently, as it stands, this view just does not seem to work. To start to see this, we should realize that musical perdurantism is not analogous to perdurantism about persons. In the case of persons, we have a situation like this one, where Marty persists through time in virtue of being composed of successive temporal parts, in this way: Marty

T-part-1of-Marty

T-part-2of-Marty R

t1

T-part-3of-Marty R

t2

T-part-4of-Marty R

t3

T-part-5of-Marty R

t4

t5

The R-relation that holds between the different temporal parts is a relation of causality and spatio-temporal continuity. Causality is the core component here, it is the ‘glue’ that unifies the different temporal parts. The various temporal parts constitute a causal chain. Spatio-temporal continuity is typically a feature of persistence through time of persons, but it is not a necessary feature, since it can be absent in cases of time-travel. Let us make this thought-experiment: Let us say that Marty travels back in time, from t5 to t4, and is thus able to meet his younger self. In this situation, spatio-temporal continuity is broken, which is normal in the case of time-travel – indeed, this is exactly what time-travel is.6 But, crucially, the causal connectedness and continuity is maintained. To understand this situation well, Lewis (1976b) distinguishes between ‘external time’ and ‘personal time’. These are not two times in the sense of two temporal dimensions, or in the sense of a regular time and a hyper-time. These merely are two ways of ordering events. External time is simply the ‘normal’ time measured by the observer

146  Aesthetic (and Other) Objects Marty

T-part-1of-Marty

T-part-2of-Marty

T-part-3of-Marty R

R

T-part-4of-Marty R

T-part-5of-Marty R

Rp (causality) (personal time) T-part-6of-Marty

T-part-7of-Marty R

t1

t2

t3

t4

T-part-8of-Marty R

t5

t6

who does not travel in time. This external time corresponds to the times t1, t2, t3, t4, t5, t6, . . . . Personal time, on the other hand, corresponds to a different ordering of events, in the case of a time-traveller. In Marty’s case, for instance, according to his personal time, t4 comes after t5 and it also comes before, his sequence of events corresponds to: t1, t2, t3, t4, t5, t4, t5, t6, . . . . To repeat, this ‘second time’ is nothing like a different temporal dimension, it merely is a different way of ordering events located in time (in the one and only dimension of time there is). There are many ways to order and organize the successive events that take place in time, depending on the point of view of the observer, and all of them are equally valid – external time is the ‘normal’ time but this does not make it in any way ‘more real’ than personal time; these ‘two times’ are merely two successions of the same events, but ordered in different ways. By convention, we can keep the labels “external time” or “normal time” for the ordering that we know in the situation where there is no time-travel, and we can use the label “personal time” for the ordering that corresponds to the travels of time-travellers such as Marty. In the case of external time and in the case of personal time as well, causality plays the theoretical role of the glue that unifies the space-time worms – the persons, in the perdurantist view. But nothing like this is available in the

‘Upper Level’ Objects, Musical Works  147 case of musical works. Indeed, when it comes to musical works, the situation is very different, and is not analogous to the case of persons: First half of performance 3

Tarantella

R2

Performance of Tarantella 1

R1

Performance of Tarantella 3

R1

Performance of Tarantella 4

R3

Performance of Tarantella 2

t1

Performance of Tarantella 5

t2

t3

We have here a number of various relations that relate entities which are not continuous. Caplan and Matheson (2006, p. 64) say that “although one might want to know more about which relation, exactly, the continuity relation for musical works is, a perdurantist about musical works can say that that question need not be settled before one adopts the view”. But if we say nothing about the relevant relation(s), what we get is literally a cacophony. Again, compare to the case of persons. If we did not say anything about the R-relation here, this would mean that there would be a person composed of the following temporal parts: Jiri-1978-cum-. . .-cum-Jiri-1985-cum-DavidLewis-1986-cum-Jiri-1987-cum-. . .-cum-Jiri-2017. This would have the nice consequence that I could boast to have written On the Plurality of Worlds at the age of 8, but it would have the unwelcome consequence that the perdurantist view of persons would be a mess. And a mess it is if we say nothing about the various R-relations in the case of musical perdurantism. There is not one R-relation, there are many, and they are significantly different. We need to distinguish between R1 and R2. R2 is the relation between a temporal part of a performance and the performance (on the figure above,

148  Aesthetic (and Other) Objects the relation between the first half of Performance 3 and Performance 3). R1 is the relation between a temporal part of a musical work and the musical work. These are two very different situations. But if we say nothing about the nature of R1 and R2, we then cannot distinguish between these two cases. It is in a different sense that the first half of Performance 3 is a part of Performance 3 than the sense in which it is a part of the Tarantella. The first half of Performance 3 has a different relation to the performance and to the whole musical work – and the different performances have a different kind of relation between each other. All these cases need to be distinguished. There is also the case of Performance 4 and Performance 5 that both take place at exactly the same time, but in a different place. They are both temporal parts of the Tarantella as a whole, indeed they are the same temporal parts, but only temporally speaking. Spatially speaking, they are different. Here again, more needs to be said about what makes them to be different and what makes them to be the same, and how they are related. Performance 1 and Performance 2 partially temporally overlap, and the same question should be raised here as well. If the musical perdurantist only says that all of these elements are temporal parts of the Tarantella and nothing more – that is, if she doesn’t say anything sizeable about the different types of relations there are – there will be no reason to say, for instance, that the first half of Performance 2 is the first half of Performance 2 and not the second half of Performance 1, where Performance 1 would then consist of two qualitatively identical halves, and would thus not be a performance of the Tarantella at all, but only a performance of twice the first half of it. Similarly in the other situations mentioned above. Even in the most standard case where performances nicely follow each other, without any overlaps, something needs to be said about the nature of R1. In the case of R1, there is probably something like qualitative identity involved, but not exactly since performances typically differ, even if perhaps very slightly. Perhaps R1 should be understood in terms of resemblance, but resemblance is a context-dependent notion, and a very flexible one, perhaps too flexible, so resemblance only might not be enough. In the case of persons, there is causality (in addition to resemblance and, usually, spatio-temporal continuity), but causality is clearly not an option in the case of musical perdurantism because in no way one performance causes another one. Caplan and Matheson (2006, p. 62) raise a problem that they call “the problem of perception”, concerning the situation where I say “I have listened to the Tarantella this morning”. If musical perdurantism is true, I have not listened to the Tarantella. Instead, I only listened to a part of it. But this is not a problem of perception, and it is not a problem at all for a perdurantist. Any perdurantist about anything, persons, musical works, tables, or any other material objects will always typically say that we only perceive a part of an object at a given time – and that we perceive the whole object in virtue of perceiving a part of it. The perdurantist can say that this is the case anyway since we only always see spatial parts of objects that are in

‘Upper Level’ Objects, Musical Works  149 front of us – for instance, I only see the top surface of the table I am sitting at right now. The fact that the entity I listened to this morning is only a part of Tarantella and not the Tarantella itself is not a problem of perception. It’s a problem of understanding how musical perdurantism works. This view should be able to distinguish between different relations of the type “is a temporal part of”, as we have seen above. Perdurantists about persons provide causality, similarity, spatio-temporal continuity, and if needed the distinction between external time and personal time. But we have nothing like this in the case of musical perdurantism. In this sense, musical perdurantism is an empty shell that sadly avoids all of the really interesting questions – namely, what a musical work’s nature is, and how it is related to its performances. In the form in which musical perdurantism is put forward in Caplan and Matheson (2006) what we get is a claim that there is this spatio-temporally scattered whole, but it is not clear why there is a whole at all, and what the relations between the whole and its parts (and between the parts themselves) is. Thus it is not clear at all in what sense the whole is one entity that allegedly is the musical work. I submit then that the job of the musical perdurantist would be to distinguish all of the types of relations that are involved in the musical perdurantist picture and to say what they are – a likely uneasy job. (One can also mention other worries that the musical perdurantist needs to address. Indeed, musical perdurantism has the consequence that if one of the performances contains a wrong note, then the musical work itself contains a wrong note [see Tillman and Spencer (2012, p. 252)]. Also, another consequence of musical perdurantism is that the Tarantella does not last [roughly] 6 minutes but that it is much [much (much)] longer [see Kleinschmidt and Ross (2012, p. 131)].) If, then, musical works are not concrete objects like sounds, scores, performances, or wholes made of performances, could they be identified with some kind of mental entities? There are two possibilities here. First, one could say that a musical work is to be identified with thoughts in the composer’s mind. Second, one could say that it is to be identified with her thoughts as well as thoughts of other people who perhaps heard it, or who read the score. Both ideas, though, appear to be inadequate. Musical works can be heard, and performed, but thoughts cannot. Thoughts pop into existence and out of existence in a very short-lived manner and in a scattered way, while the Tarantella or Voodoo Chile exist continuously from, respectively, 1901 and 1968, until today. Furthermore, if musical works were identified with private entities like thoughts, they would not be shareable since we cannot access each other’s minds – it seems that musical works need to be public entities (like scores, sounds, or performances are).

§4 The lesson to be learned from the preceding section is that musical works are not spatio-temporal things (or not temporal, in the case of mental entities).

150  Aesthetic (and Other) Objects We are then left with the option that they are some kind of abstract, nonspatial and non-temporal, entities, perhaps universals or kinds. Well-known defenders of this brand of view include Kivy (1983) and Wolterstorff (1975, 1980). From the start, one could object here to the very existence of such abstract entities in a general anti-Platonist way, but let us focus on issues specific to the case of musical works. Indeed, problems arise here as well. The first worry arises from the fact that musical works are created (see (i) above) and that they can be destroyed. This appears to be a point that is quite central to the way we thing of musical works. Consider what Levinson points out: [The claim that musical works are created] is one of the most firmly entrenched of our beliefs concerning art. There is probably no idea more central to thought about art than that it is an activity in which participants create things – these things being artworks. . . . Musical works must be such that they do not exist prior to the composer’s compositional activity, but are brought into existence by that activity. (Levinson [1980, §I]) Concerning the destruction of musical works, things seem to be somewhat less clear: [A work of art] cannot be a universal because it . . . can be destroyed. . . . Clearly, sculpture, architecture, painting, and etching can be destroyed . . . hence, they cannot be kinds. Music and literature cannot, in this sense, be destroyed, simply because their properly formed examples can be generated by reference to a notation and a notation is not a work of art. All of the tokens of a notation may be destroyed, however, and the notation may cease to be remembered; in that sense, music and literature can be destroyed. (Margolis [1980, p. 22]) If we stick to the idea – as we should – that Rachmaninoff created the Tarantella and that Hendrix created Voodoo Chile, in the sense that they brought something into existence, and not that they discovered something pre-existing, we can then see that such entities cannot be abstract kinds or universals. The case of destruction is, as Margolis indicates, more complicated in the sense that a musical work can last as long as at least one score or one memory of it exists, but he is right to point out that at the end of the day, if all memories and all recordings and scores cease to exist, the musical work ceases to exist as well. Perhaps, the creatability requirement is here more straightforward and stronger that the ‘test of destruction’, but in both cases, independently, it appears that there is a serious problem with the view that musical works are abstract entities, unless of course one were ready to give up the idea that musical works can be genuinely created (and/

‘Upper Level’ Objects, Musical Works  151 or destroyed). In agreement with Levinson, as well as with many others, I suggest that giving up on these requirements would be giving up on musical works entirely. Furthermore, some difficulties similar to the issues raised in the preceding section arise. Indeed, as we have seen, musical works can be heard, while abstract entities cannot, since abstract entities do not possess any physical or perceptual properties at all. A possible reply to this worry might claim that it is true, on the abstractionist view, that we cannot hear musical works, but that we can hear their instances – say, their performances. After all, a friend of universals will systematically claim that we cannot be in contact with any properties (circularity, redness, . . .) whatsoever but only with their instances, so it would be the natural thing for the friend of abstractionism to say. However, in the case of musical works, this claim raises a specific problem, namely, the fact that a performance often includes, or at least can include, wrong notes, and it then becomes difficult – for the metaphysician – to claim that such a performance is an instance of an abstract structure that does not include such notes. This problem becomes perhaps even more salient in the case of even ‘less well-behaved’ performances where much more than just a few wrong notes does not correspond to the original composition – just think of the many performances by different performers of Voodoo Chile, that can quite strongly differ from Hendrix’s own original version. This problem becomes perhaps even more salient, if one takes on board the idea that at least some musical works are such that their instances are never instances of the type universal. Consider what Pouivet says about rock: When it comes to rock music, the function of the recording [in the studio] is not to simply reproduce an execution, for instance one of the executions of the song “How high the moon” by Les Paul and Mary Ford. A recording is most of the time an aggregation and a mix of several independent pieces of recording. In other words, what we hear when we listen to a recording has never been heard by anyone before the recording – that is, the result of studio techniques – was finished. By creating this recording, Les Paul creates a musical work. (Pouivet [2010, p. 12]) Pouivet’s own point here is to motivate the idea that we need a new ontological category for such musical works (I’ll come back to this in III.10.§4 below, where we will see the case of photographs and novels as well, and I will argue that such an ontologically costly proposal is an unnecessary burden). For our current purposes, what we realize here is that in many cases of musical works, there can be, and often are, very significant qualitative differences between ‘the’ musical work and all of its instances. In some cases, they are perhaps always systematically different. The metaphysical link between universals and their instances becomes then here even much more

152  Aesthetic (and Other) Objects problematic than what it already is in any general theory of universals. (The possible suggestion that musical works could be sets of which their instances would be members is not adequate either, since sets have their members essentially and are individuated by them, while musical works could have less or more occurrences than they do in the actual world.7) The issues I rather quickly raised in the two preceding paragraphs are a matter for a long controversy and would require a book of their own. I am not going to do that, since I have here a different book to write. But I hope that even though the discussion above is short, the problems are quite clear, and that it is quite clear that there is no easy solution to them.8 Before closing this part of the discussion, let me mention one last issue, which will also re-arise in the case of photographs. Levinson (1980, §II) discusses the situation where two composers, independently of each other, compose the ‘same’ musical work. If musical works were abstract universals, then the two composers would indeed compose the very same musical work, since they would compose the same universal – this universal being understood here as an abstract sound structure. But as Levinson points out, if two composers, independently and in different situations, composed the same sound structure, they would nevertheless not produce the same musical work, since the relevant features of a musical work do not depend only on their sound structure but also on the context in which they are created. This is true not only of musical works, but of all kinds of artworks in general. Indeed, the aesthetic value of a painting or a novel is influenced by its creative origins (the historic, social, political, etc. contexts determining the artwork’s originality or even meaning [for instance, in the case of a novel such as Orwell’s 1984]), and the same is true when it comes to musical works. For instance, the originality of a sound structure corresponding to Voodoo Chile is much greater if the piece is created in 1968 than if the same sound structure is created in 2017, given the historicmusical context in which it is created. The claim here is that in addition to first-order non-aesthetic properties such as sound structural properties, or in the case of other types of artworks colours and shapes, some relational properties also have to be included in the base on which aesthetic properties of artworks supervene.9 These include, then, the history and context of production of an artwork. In Walton’s (1970) terms, aesthetic properties of an object depend not only on its narrow non-aesthetic properties, but also, importantly, on broad non-aesthetic relational properties, like the process and history of production of an artwork as well as the context in which it was created. Two indistinguishable paintings, indiscernible in the sense that they are exact duplicates and exactly the same arrangements of paint on a canvas of the same size, shape, texture, and so on, would (or, at least, could) still possess different aesthetic properties depending, for instance, on the period when they were created – exactly as in the case of Voodoo Chile or the Tarantella if created in, respectively, 1968, 1901, or in 2017.

‘Upper Level’ Objects, Musical Works  153 This ‘broadening’ of the supervenience base solves a problem raised by Scruton (1974, p. 36), who criticizes the aesthetic supervenience thesis when he says that “different emergent ‘properties’ can depend on precisely the same set of ‘first-order’ properties”. What he has in mind here is that one and the same artwork can be context-dependently characterized as sad or as joyful, without contradiction. (For a discussion of this phenomenon, see for instance Pettit [1987], Zangwill [1994], and MacKinnon [2001].) We can respond to such an objection simply by pointing out that, once we include the context of production in the supervenience base, it is not anymore the case that ‘different emergent properties could arise from the same base’. What we see here, then, is that musical works cannot be only identified with sound structures understood as abstract universals. More is needed in order to account for their aesthetic properties.

§5 Now is the time to step back a little and consider what the list of worries concerning the claims that musical works are concrete entities, mental entities, or abstract entities amounts to. Again, each and every item on the list of desiderata and objections we have seen could be questioned and rejected or replaced by something else. But I hope that one thing is clear: whatever way one chooses to go, a very strenuous path lies ahead. Musical works just do not seem to fit into one of the three traditional ontological categories. (We will have the same situation in the case of photographs in the next chapter.) A very realist-minded metaphysician could then suggest that a new ontological category needs to be introduced in the case of musical works. But this would be making the same mistake that Dharmakīrti condemned in the case of the Self:10 we must not reify too much, we must resist our natural tendency to do so. As we have seen, it is a mistake to claim that there is an entity such as the Self that corresponds to our sense of subjectivity or persistence through time. In a similar way, it is also a mistake to demand that there must be an entity, a something, some thing, that can be identified with a musical work. As we have seen above, only trouble arises from such an ultimatum. If we take on board lessons learned in Part I and Part II, that is, if we consider what an eliminativist claim amounts to in the case of musical works, we can see how natural it actually is, and how many of the worries listed above are merely artefacts of our tendency to reify. Here is how things look from the point of view of the eliminativist about musical works: there are composers, there are pianos and electric guitars, there are sounds, there are perceptions and thoughts, there are memories, there are events such as performances, there are scores, and so on. Since we have all of that, why would we need musical works in addition? Composers, pianos, guitars, scores, sounds, thoughts, memories, and performances all fit in our ontological categories (or they can be eliminated further, if one follows the

154  Aesthetic (and Other) Objects recipe from Part I all the way down). Musical works then serve no real practical purpose – we already have everything we need to enjoy music. As aestheticians-metaphysicians, do we need musical works, understood as sui generis entities, to be the bearers of aesthetic properties? Lesson from Part I: instead of chairs and whales, we have simples arranged chairwise and whalewise. Lesson from Part II: instead of the Self, we have impermanent individual experiences arranged Self-wise. If needed, we can make property attributions to pluralities, as in the case of the crew of the USS Enterprise. We can say that “the crew” is brave, but we are not thereby talking about one entity being brave, we are talking about a plurality – the members of the crew – behaving in such a way. In the case of the no-Self view, the fact that there is no Self understood as an enduring bearer of experiences or psychological properties does not prevent us either to say things like “Ueli Steck is brave” – there are experiences and psychological states/properties arranged Steck-wise; that is, there are these individual impermanent mental entities relevantly causally connected, and this makes true (in ordinary language, not in Ontologese!) that Steck is brave. When it comes to musical works, the situation is similar. We want to say that the Tarantella is a powerful piano piece or that Voodoo Chile contains one of the greatest guitar solos of all time. Metaphysically speaking – that is, when we speak Ontologese – there is no entity such as the Tarantella or such as Voodoo Chile. But there are all the other things: performances, sounds, scores, composers, and so on. These can – in ordinary language! – be collectively (or selectively) said to be beautiful or powerful; they can be – loosely, ordinarily speaking – the ‘bearers’ of any attributions we may want to make, including attributions of aesthetic properties. But, one must not forget that it is only in a conventional sense (remember I.1.§5) that we can talk about musical works in this way. Musical works can be conceived of as conventional entities, in the same way chairs or whales can be thought of as being conventional. As soon as one forgets that, the natural tendency to reify comes back, and metaphysical worries come back with it.

Notes 1 Cameron (2008) defends a view that is similar to an eliminativist view about musical works. 2 Sergei Vasilievich Rachmaninoff, Suite No. 2 for Two Pianos, Op. 17, Movement IV. 3 24th November 1901. First performance, by Sergei Rachmaninoff and Alexander Siloti. 4 A difficult position that Dodd (2008, p. 1117) characterizes nicely in this way: our musical practice does not speak to us with one voice: we cannot hope to read off nascent accounts of works’ persistence and identity conditions from our everyday thought and talk; and, furthermore, it is unclear how we could arbitrate between philosophers with contrary intuitions on these matters. Bartel (2017) offers a detailed critical discussion of the role intuitive claims such as (i) or (iv) can play in the philosophy of music. Relevantly, remember the

‘Upper Level’ Objects, Musical Works  155 discussion from Part I, Chapter 2 above concerning the role intuitions can and cannot play in philosophy in general. 5 See Benovsky (2006a, Part II, especially chapter 6). 6 At least the kind of time-travel where the traveller steps into a time machine and is transported instantaneously to a different (spatio-)temporal location. The Terminator or Back To The Future exhibit this type of time travel. H. G. Wells’s time machine works in a very different way since it travels in time by going through all of the intermediate times (forward or backwards) between the time of departure and the time of arrival. 7 Dodd (2008, p. 1118): To be sure, it would seem to be a mistake to regard the one-many relation holding between a work and its occurrences to be either that obtaining between a property and its instances, or that holding between a set and its members. Whilst a property is repeatable inasmuch as it is an entity capable of multiple instantiation by particulars, Vaughan Williams’s Oboe Concerto looks categorially unsuited to be identified with a property of its performances and playings. Rather than being a mere respect in which performances or playings can be alike or differ, the work itself is the blueprint for such performances and playings: a thing in its own right. This being so, works of music look more akin to entities such as The Red Flag, The Daffodil, and the word ‘refrigerator’ – that is, types – than they do to properties. But neither, so it seems, could works of music be sets of their occurrences. For, given that sets have their members (or lack of them) essentially, no set can differ in its membership. Vaughan Williams’s Oboe Concerto, by contrast, does not have its occurrences essentially: it might have had more, fewer, or different performances than it has had actually. 8 Of course, one could try to answer these worries in various ways – for instance, to avoid the worry with the relationship between universals and their instances one could say that musical works are kinds (see Lowe [2006] or Hoffman and Rosenkrantz [1994, 1997] in I.2.§1 above) and that performances are instances of kinds, or one could say that musical works are functions from world/time pairs to the performances occurring at that time and in that world. My aim above was not to reject the idea that there are possible replies to the worries I raised. I merely wanted to highlight the fact that there is no easy and straightforward solution, and that any possible reply will require some heavy metaphysical machinery (like, say, an ontology including kinds) and a great amount of work to show that this machinery really can solve the series of problems I have sketched. 9 For the sake of brevity, I use here the term “supervene”, but in Benovsky (2012) I argue that we should rather be talking about grounding instead of supervenience here. 10 See II.7.§2.

10 Photographs

§1 The case of photographs is similar to the case of musical works, but it is perhaps even more complicated, and thus even more interesting. For the metaphysician, photographs are very puzzling entities indeed. And even from the non-philosopher’s intuitive point of view, it is not that clear what sort of thing a photograph is. Typically, if someone wants to purchase a photograph, she can mean very different things by ‘buying a photograph’: she can mean to buy a print or a number of prints, or she can mean to buy a negative (when traditional film photographs are concerned) or a file (when digital photography is concerned), or she can mean to buy a right to use a photograph a precisely determined number of times in a number of brochures or on a website, and so on. When meeting a new client, a photographer typically faces the problem of explaining to her what it is that she is actually buying – and it is not always clear that she is ever buying a photograph. As metaphysicians, we face a much more difficult challenge: find out to what ontological category photographs belong. Are they concrete spatio-temporal entities like prints, are they universals since there can be many ‘prints-instances’ of a same photograph, are they sets or aggregates of prints, or something else? In this chapter, I will examine all plausible metaphysical categories to which photographs could belong to, and try to see which one is the fittest. As we shall see, similarly to the case of musical works, in this ‘survival for the fittest’ competition between traditional metaphysical categories, there will be no real winner: several categories will reveal themselves to be enlightening and useful when describing features of what photographs are, but none will prove to be really satisfactory. Photographs, it seems, are a sort of borderline entities that share some but not all aspects of several traditional metaphysical categories. Is it then justified to postulate a new ontological category to which photographs would properly belong? On mainly methodological grounds, I shall argue that it is not, and I will suggest – with no surprise – that the best way out of this metaphysician’s trouble is to follow the eliminativist recipe. Thus, I will defend the claim that photographs do not exist – but, as before, I will also argue that this is not really a revisionary or anti-commonsensical claim.

Photographs  157

§2 Let us get started with two basic premises about what sort of entities photographs are – these will be my desiderata that a metaphysical theory about the nature of photographs must be able to account for. First, photographs can be seen – they are entities such that they can be visually perceived by normal human beings. This is of course only a necessary, not a sufficient condition, but it is an important one since for instance it immediately rules out the claim that photographs are mental entities. Indeed, as Levinson (1980, p. 63) rightly remarks about musical works, and as I think is true in the case of photographs as well, photographs (as well as other artworks) would become inaccessible and unshareable if they were ‘private’ mental entities, while it is part of our central conception of photographs that they can be seen by a number of different observers. Second, as in the case of musical works, photographs come into and go out of existence. Another way to put this is that they are created rather than discovered, they are human-made entities. We have seen this claim, defended by Levinson (1980, 1991) in the case of musical works. The main reason to endorse it lies in the fact that it is a deeply anchored and widely intuitively shared belief we have about photographs. Endorsing the opposite claim, namely that photographs somehow eternally exist without ever coming into and going out of existence, would lead to the conclusion that all photographs exist even before they are taken by a photographer, which teeters on the edge of absurdity. Bearing these two desiderata in mind, let us now examine the process of production of photographs. Indeed, a photograph is the result of a process that involves various (types of) entities, as the following figures illustrate. The figure on the left illustrates the typical and minimal process of production in traditional film photography, while the figure on the right illustrates the process of production in digital photography. I think it is important to emphasize that only at the end of the process, we can say that we have a photograph. Granted, there is a sense in which a photograph ‘is already there’ in former stages of the process of production (it is at least partly ‘there’ in the unprocessed film, in the negative, in the raw file, or in the image file), but the entities involved in these stages (unprocessed film, negative, computer files) are not photographs since the process of production is not finished, and since none of these entities satisfies the desideratum of being visually perceivable by normal human beings. True enough, one can see an image if one looks at a negative (unlike the other three entities, where nothing at all is there to be seen), but it is not the photograph, it is an unfinished image whose colours, for instance, are not yet determined properly and will only be after a development process will have taken place. In some ways, such seeing of this unfinished photograph is a bit like seeing an unfinished painting. Thus, on the one hand, these four entities are not good candidates for being the photograph, but on the other hand, we could say that the photograph is there in a dispositional state. While I will not say

158  Aesthetic (and Other) Objects

light

light

lens

lens

body + film

body + sensor

unprocessed film ("latent image")

"raw" file (not yet an image)

developer stop bath fixer

in-body or additional software

negative

image file enlarger developer stop bath fixer

image1

image2

image3

printer monitor … image1

image2

image3

that photographs are dispositions, we will see below that these stages of the process of production where the photograph ‘is there as a disposition’ do play a crucial role in the finding out of what ontological category photographs belong to. Another way to emphasize the fact that it is only at the end of the process of production that a photograph comes into existence, is to insist on the dark room’s tools and its digital counterparts to be essential parts of photographic systems. It would be a mistake to think that a photographic system is composed only of a lens and a camera body – rather, a photographic system genuinely does include various components that correspond to the stages of production of photographs such as chemicals, enlargers, software, printing devices, and so on. Without these tools no photograph could ever come into existence, and so it is only natural to claim them to be parts of any standard photographic system. The case of software is a bit special. Software used in digital photography can be actually built in the body of the camera (and so the raw file is processed there, which happens in the case of most amateur photographs – very often people who use this system are not even aware of there being two such stages involved since the process is fully

Photographs  159 transparent to the user), or it can be applied to the raw file only later, using manual settings on a computer, and produce an image file only there (which happens in the case of most professional photographs, and parallels more closely the process of production of a traditional film photograph). The important thing is that, built-in or not, software (as well as its film photography counterparts in the dark room) is one of the essential components of a photographic system that is necessary in order to bring about the existence of a photograph. There is a difference, though, between digital and film photography, that will be important below: raw files and image files are much easier to duplicate and they can be duplicated without any loss of qualitative properties (without loss of information), whereas unprocessed films and negatives are very difficult to duplicate and always involve loss (and gain) of qualitative properties, thus not preserving qualitative identity. This is, of course, a matter of available technology: if we had at our disposal replicators like those used on Star Trek starships, we could as easily duplicate unprocessed films and negatives as we copy a computer file – but still, as photographic technology actually stands, there is such a difference between film and digital and it must be taken into account, which I will do in due course below.

§3 Given the two desiderata from the preceding section, and given that I claimed that it is only at the end of the process of production that there is a photograph, it would seem now only natural to say that photographs belong to the ontological category of spatio-temporal material objects since they are prints (paper + ink) or computer screen images (arrangements of backlit liquid crystals, for instance). These objects can be seen, they come into and go out of existence in the way such concrete objects do, and they account for much of common sense talk about photographs (“I carry photos of my loved ones in my wallet”.). Thus, the ontological category of concrete (material spatio-temporal) entities seems to account well for some central features of photographs – some, but not all. For photographs, typically, are repeatable. Although it can be the case that there exists only one print of a photograph, typically a photograph has a number of prints.1 At this point, it may be suggested that photographs are sets or aggregates of all the prints, but such a claim cannot be satisfactory: first, sets are mathematical entities and as such are not there to be seen; second, one can destroy a set or an aggregate by destroying a member or a part of it while destroying one print does not destroy the photograph; third, this also works the other way around since, as Rohrbaugh (2003) remarked, sets or aggregates are individuated by their members or parts and could not possibly have other, while a photograph can very well have more prints than it actually presently has.

160  Aesthetic (and Other) Objects So it seems that while the category of concrete entities does do some good work, it fails to account for the repeatability of photographs – that is, for photographs to have multiple instances, very much in a way universals have multiple instances. Perhaps then photographs are universals rather than concrete objects? Such a suggestion is a widely discussed one, more so in the case of musical works than in the case of photographs, as we have seen in the preceding chapter (see, inter alia, Wolterstorff [1975], Levinson [1980, 1991], Kivy [1983]). The first distinction to be made here is the one between platonic non-spatio-temporal universals and immanent spatio-temporal universals (à la Armstrong [1978]). If photographs were said to be platonic universals, it would become impossible to account for the second desideratum from §2 above – such entities are not temporal and do not come into and go out of existence, and I take it that this is the main reason why photographs cannot belong to this ontological category, as in the case of musical works. Besides, somewhat less importantly, such a view would make it less straightforward to meet the first desideratum as well: only instances of photographs could be actually seen, but not photographs themselves. Thus, if photographs were to be universals, they would rather be immanent spatio-temporal Armstrongian universals (I shall just say “universals” from now on). One of the core ideas about the nature of such universals is that they exist iff they have an instance, which is highly relevant since it allows us to say that a photograph comes into existence when its first print is created, and then other prints can be created, and later if everything is destroyed (I shall say more on this later), the universal goes out of existence. This conception of universals also accounts for the fact that there is a good sense in which photographs are multiply spatio-temporally located which straightforwardly accounts for commonsensical claims such as “In my living room I have hung for display the same photograph you have in yours”. Immanent universals thus possess some central features required to account for the metaphysical nature of photographs; most importantly for their repeatability. I shall now examine four objections to this view, and see what force they have. Margolis (1980, p. 29) objects to such a view on the ground that if it were true we could never see a photograph, since we can never see universals. (And he shares the view that this is a desideratum that a theory about the nature of photographs should be able to satisfy.) As we have seen in the preceding chapter, Kivy (1983, p. 110) when speaking about musical artworks provides an answer, easily adapted to the case of photographs: by seeing a universal’s instance, we see the universal itself. Kivy’s reply seems to me controversial, and unnecessarily strong in order to face Margolis’s challenge. It is controversial because, at least prima facie, from the phenomenological point of view, we see circular objects but we never see circularity and it would take a strong argument to be able to claim the contrary, an argument that Kivy does not provide. But most importantly,

Photographs  161 it is unnecessary to embrace such a strong claim. Any friend of universals will usually be happy, and rightly so, to claim that, granted, we never see circularity itself, but that we see circular things, and that’s enough. Why would it be required to claim that we see circularity in order to say that it exists and that its existence accounts for claims of attribute agreement among circular things? Friends of universals typically don’t feel obliged to endorse any such strong claim, and when applied to the case of photographs they should feel exactly the same: namely, they could and should amend the first desideratum’s claim and say that if photographs are universals, it is not necessary to require that a photograph must be visually perceivable, it is only required that instances of photographs should be visually perceivable. A different objection can be found in Dorsch (2007, p. 10). Dorsch does not explicitly speak about photographs, rather the target of his analysis are novels, but since novels are repeatable very much in the sense in which photographs are, his objection is easy to adapt. He says: novels have more fine-grained individuation conditions than types or kinds. Again, if two independently working authors write exactly the same sequence of sentences, their novels should still count as numerically distinct. But if the non-temporal types or kinds in question are not defined in terms of sequences of words, it is unclear in terms of which aspects of novels they might be specified instead. In the case of photographs it seems perhaps even more natural to claim that if two photographers take two qualitatively exactly similar photographs, these should still count as two different photographs that are not the same universal. But it would become hard, on the universals view, to account for this distinction – universals are individuated by their qualitative nature in such a way that two instances of circularity or two instances of the exact same shade of red are by necessity instances of the same universal. Thus, I believe that the Dorsch objection stands (as it does in the case of musical works as well). There is I think an even stronger reason to reject the universals view. We have seen that an advantage of the immanent universals view is that it can account for the coming into and going out of existence of photographs. But it does not work quite as well as we could hope. Suppose I take a photograph using a digital photographic system and make two prints, so there exist two instances of it. Suppose then that I don’t like it and destroy both prints by burning them. Did I make the photograph go out of existence? If photographs were universals whose instances are visually perceivable entities such as prints, the answer should be affirmative – but it is not, since raw files and image files remain, and any number of prints can easily be re-made using those (this is the ‘dispositional aspect’ of photographs that I already mentioned in §2 above). As long as there remain these entities where the photograph is in a sort of dispositional state, the photograph

162  Aesthetic (and Other) Objects cannot genuinely be said to have gone out of existence – while the universal did, which shows that the universals view is not adequate. To avoid this uncomfortable situation, the friend of the universals view could claim that entities like raw files, image files, negatives, or unprocessed films also are instances of the photograph, but then we’re back to the problem that is behind Dorsch’s claim about novels: there is very little resemblance and no qualitative identity between entities like prints and entities like raw files (or unprocessed films) and it would become really hard to defend the claim that these so qualitatively different entities are instances of the same universal. Furthermore, for the same background reason, the universals view has difficulties even to account for the normal case where only prints are concerned, for even those are never exactly qualitatively identical due to technological reasons concerning imperfections of printing devices, computer monitors, and so on. Thus, even in the most straightforward case, it is not obvious for the universals view to account for the fact that, say, two prints are instances of the same universal, since they are not qualitatively identical (and sometimes are even qualitatively quite different). In what I said above, I made use of what can be called “the test of destruction”. Indeed, I think it is good methodology, when asking oneself what an entity E is, to look at what happens with E when some entities X, Y, Z are destroyed: if E is thereby destroyed then it very likely was identical with X, Y, Z, and if E is not destroyed then it is something different. I have already hinted in several places above at how this methodological principle can be applied to the question of ontology of photographs; let me now sum this up and consider the standard case where we have a photograph taken using a traditional or a digital photographic system of which two prints were made. Obviously, if one of the prints is destroyed, the photograph was not destroyed since another print remains. As we have just seen above, even if both prints are destroyed, there still is a very good (‘dispositional’) sense in which the photograph was not destroyed – so, it does not seem correct to claim that a photograph is ‘solely’ a print, a set or aggregate of prints, or a universal of which the prints are instances. It seems that the entities (negatives, unprocessed films, raw files, images files) where the photograph is in a sort of dispositional state somehow must play a crucial explanatory role in the ontological status and nature of photographs – but, it is true about these entities as well that even if they all were to be destroyed but prints were to remain, the photograph was not destroyed either. Thus, neither solely prints nor solely these other entities can be said to provide a full account of what photographs are – somehow ‘a bit of both’ is needed: if one truly wants to destroy a photograph, one has to destroy all prints, and negatives, and unprocessed films, and raw files, and image files. What is the right conclusion to draw from this situation? Is it that a photograph is an entity that is somehow made of all these very different types of entities? If so, photographs would be bizarre entities indeed, dispositional and

Photographs  163 non-dispositional, visually perceivable and not visually perceivable, repeatable or not repeatable (in the case of unprocessed films). What an uncomfortable position to be in for a metaphysician!

§4 The conclusion that stems from arguments in §2–3 above is that traditional ontological categories like concrete objects, sets or aggregates of those, mental entities, platonic universals, immanent universals, or even dispositions are such that no one of them is satisfactorily able to account for the ontological status and nature of photographs. Simply put, photographs seem to share a bit of each category (except perhaps the category of mental entities and platonic universals) – each category has something interesting to say, but none is capable of saying it all. Remember that we had the same situation in the case of musical works as well. But how can this situation be handled by the metaphysician? Are we to claim that photographs are some sort of trans-categorical entities? But how does that make sense? One possible reaction to this situation is to postulate a new ontological category that has all the features that we need to account for the nature of photographs. This is the line that Rohrbaugh (2003) takes and that Dorsch (2007) takes about novels as well.2 According to Dorsch, novels fall into a new category of “reproducible prototypes”, while according to Rohrbaugh (2003, pp. 34–35) photographs are “higher-level non-physical continuants” which stand in a relation of ontological dependence to causally connected series of physical particulars. I will not discuss here the details of these interesting proposals. At this stage, I am more interested in their overall strategy. Indeed, I believe that such a way to face the uncomfortable situation we find ourselves in is methodologically misleading – it amounts to postulating a new ontological category, that is, a fundamental category of being, to account for something humans contingently did. Postulating an ontological category is a claim about the very structure of what there is at the most fundamental level, while photographs do not seem to be anything like some sort of basic and fundamental components of reality. Thus, introducing a new ontological category solely for the purpose of accounting for the nature of photographs is methodologically (metatheoretically) ill-motivated. Both Rohrbaugh and Dorsch are sensitive to this worry, and both defend their strategy (their new ontological category) by trying to show that it can serve a wider purpose than just the one of solving a puzzle about ontology of photographs or novels. Dorsch, for instance, tries to show that his postulated category also accounts very well for the ontological status and nature of sounds, and Rohrbaugh claims his new category to be useful to account for “novels, musical compositions, species of animals, clubs, sorts of artefact, and words of natural languages” as well (see Rohrbaugh [2003, p. 35]).

164  Aesthetic (and Other) Objects If there is no good alternative, then maybe we should do what Rohrbaugh and Dorsch recommend. But, only if we have no good alternative – and we do.

§5 With no surprise, as in the case of musical works, I will now claim that photographs do not exist. Furthermore, I will argue that this is not a very revisionary nor a very strong claim. As before, the simple idea behind an eliminativism about photographs is that we already have all the entities we need without having to postulate photographs as sui generis entities. We have unprocessed films, we have negatives, we have raw files, we have image files, we have prints, we have computer screen images, and so on. Importantly, all of these entities have a clear ontological status: they are concrete material objects with quite straightforward individuation conditions (additionally, these entities could be eliminated even further, if one embraces the kind of eliminativism defended in Part I). Metaphysical worries arise only when we want to claim that in addition to all these things, there are photographs. But why should we feel the need to claim so? Remember the troubles I mentioned at the beginning of this chapter that a photographer has when a new client wants to purchase a photograph: it is not always clear what exactly she wants to buy. Often, when asking to buy a photograph, she means to buy a print. Very often as well, she means to buy a computer file and the right to use it to make an agreed-upon number of prints, or to display it for an agreed-upon period of time on a website, and so on. Rarely, she means to buy a negative or an unprocessed film. Never, she means to buy something that would not be a concrete object + a right to use it in a certain way.3 This commercial behaviour shows how we ordinarily think about photographs: mostly in terms of prints, negatives, or computer files. At no point in our ordinary understanding of photographs do we feel the need to appeal to any different sort of entities. Thus, our concept of a photograph simply (and often vaguely) supervenes on our concepts of metaphysically unproblematic ordinary material objects. My suggestion is that, as metaphysicians, we follow this ordinary way of understanding the nature of photographs. As metaphysicians as well as buyers, we do have all the entities we need, and we don’t need to postulate entities of a new ontological category. Talk (not just ordinary talk but theoretical talk as well) about photographs can be easily paraphrased and understood as talk about one or more of the unproblematic entities available. Most importantly for art critics and aestheticians, the attribution of aesthetic properties to allegedly existing photographs does not create a problem. If I want to say how beautiful a photograph is, I am not attributing the aesthetic property to a negative, to an unprocessed film, or to a file – I am usually attributing it to some of the visually perceivable objects like prints or computer screen images. On purpose, in the preceding sentence, I said “some” – indeed, very often,

Photographs  165 when discussing the beauty of a photograph we can appreciate how better this print or this computer screen image looks than another (less well made) print or another computer screen image (displayed on a less well-calibrated monitor). Thus, attributions of aesthetic properties only require the existence of these visually perceivable entities, and they often are attributions concerning one of those (as well as the context of creation, as we have seen in III.9.§4 above). They can, of course, also be attributions concerning a number of prints or even all existing prints of the same photograph (that is, all prints that are results of a process of production that contains the same unprocessed film or the same raw file). If needed, linguistically speaking, such attributions can then work exactly like plural attributions such as “The USS Enterprise crew is very intelligent”. Eliminativism does not force us to reject the two desiderata from §2 either. Granted, if photographs do not exist, they cannot be seen and they cannot come into and go out of existence. But, as we have just seen, the core idea behind the eliminativist claim is that for all practical and theoretical purposes we have all the entities we need: prints can be seen, computer images can be seen, and all of the entities involved in the process of production of these come into and go out of existence – thus, the two desiderata are easily taken as being about the nature of these entities, that (singularly or cumulatively) play the role of photographs for all purposes. Given eliminativism about photographs, strictly speaking, the desiderata have to be rejected of course since photographs don’t exist, but the motivation behind these desiderata and the core idea they cherish are preserved. Lessons from Part I should of course be taken on board here as well. The apparent ‘contradiction’ is easily taken care of: indeed, on the one hand, I claim here that it is false that photographs exist, while on the other I want to accept the truth of claims such as “There is a beautiful photograph hanging on the wall of my living room”. Such two existential claims seem to be incompatible – but they are not, because they are not uttered in the same language. Indeed, as we know from Part I, the former claim is a claim made in Ontologese – a fundamental language of the metaphysician who recognizes that strictly speaking there are no entities such as photographs. The latter claim is a claim uttered in ordinary English where it can be unproblematically asserted that there is a photograph on the wall. In this way, if needed, photographs can be conceived of as conventional entities, in the same way musical works, chairs, or whales can be understood. The idea common to both eliminativist strategies about photographs and about chairs is that without having to postulate such entities, we already have all we need to account for all phenomena that need to be accounted for: we do not need photographs because we have prints, negatives, files, and so on, and we do not need chairs because we have simples arranged chairwise. Let us think of the first desideratum and compare the two brands of eliminativism. As I am now typing this sentence on my computer, it seems to me that I see a table in my visual field. But, as we have seen in Part I, my visual experience would be

166  Aesthetic (and Other) Objects exactly the same even if there were no table and if there were ‘only’ simples arranged tablewise. Indeed, the visual experience I have is caused by light reflected by the simples and this reflection would be exactly the same if there were a table. The idea here is that our sensory experiences can be accounted for in terms of some more basic and genuinely fundamental (and existing) entities, and so there is no need to postulate a further entity (a table). In the case of photographs, the analogous (but not exactly the same) claim is that the desideratum that photographs can be seen can be satisfied by realizing that our visual experiences of photographs can be accounted for in terms of prints (or computer screen images, . . .), that is, in terms of other more basic entities, and so there is no need to postulate the existence of a further entity and a further ontological category. Were there to be photographs, our visual experiences would be exactly the same than experiences we actually have of prints. Were there to be tables, our visual experiences would be exactly the same than experiences we actually have of simples arranged tablewise. And of course, the two brands of eliminativism can be fruitfully combined. Once we realize that we do not need photographs because we have all the other entities like prints, we can then eliminate the existence of prints in terms of simples arranged print-wise, thus effectively embracing a full-blown eliminativism.

Notes 1 For simplicity, I will always speak about prints, but of course this also applies to computer screen images, or any other material spatio-temporal visually perceivable entity that can reasonably be taken to be a photographic image. 2 As already mentioned in the previous chapter, Pouivet (2010) makes a similar claim about rock music. Rock, in his view, is a type of musical work that is sufficiently different from other kinds of musical works to justify the claim that a new specific ontological category is needed to account for it. Indeed, in his view, rock is a type of artefact, a recording, made in such a way that it can be widely distributed. Pouivet (2010, p. 47 and p. 118) claims: Rock is an ontological ‘category’; not a stylistic or sociological category. . . . The claim I defend here is that rock is essentially an ontological novelty, a new type of thing in the world: musical works qua artefacts-recordings, musical constructions that did not exist before the middle of the 20th century. But Pouivet, unlike Dorsch and Rohrbauch, does not use the label “ontological category” in the same sense in which we use it when we say that there are fundamental ontological categories such as universals, substances, and so on (ibid, p. 47). He thus does “higher-order ontology” and not “fundamental ontology” (this distinction and the higher-order ontology approach are critically discussed by Brown [2011]). Indeed, as I understand his view, he merely wants to convey the idea that rock is a very specific type of musical work that requires a specific understanding, but when it comes to the issue of fundamental ontology rock is to be treated like an artefact, and thus I take it that in a general and fundamentally ontological sense his view is even compatible with the truth of eliminativism of the kind I defend here. 3 What ontological category rights belong to is a difficult question that falls outside the ambitions of my discussion here.

Concluding Remarks on Eliminativism and Monism

§1 The aim of this book was to show that eliminativism works, and that it works in a variety of cases. We have thus seen how eliminativism, understood as a set of ontological thesis, is of great service in various situations. These theses, qua ontological claims, are different from each other – they are topic-specific, and the arguments in their favour are topic-specific as well, giving rise to different eliminativisms. But we have also seen what they have in common: the general strategy and methodology. The opponents of eliminativism are the realist views of the topics at hand. I have argued against realist views concerning ordinary objects and the Self, and we have seen that eliminativism also is the best option around when it comes to aesthetic objects. Other types of objects might of course be concerned as well. On the “other side” of the spectrum of possible views lies monism. At a very first glance, a monist view seems to be the exact opposite of eliminativism since it claims that the only object that exists is the world, while eliminativism denies it and claims that there are only simples arranged worldwise. But is this really a very strong disagreement? After all, while the eliminativist claims that there are no chairs but that there are simples arranged chairwise, the monist also claims that there are no chairs, but that the world has chairish aspects. Maybe then, the disagreement is much less substantive than what might be thought. Indeed, the idea that monism is not at the “other side” of the spectrum of possible views, and that the two views are perhaps even somehow equivalent, is at least prima facie plausible. This ‘equivalence’ idea has been rather quickly put forward by Baptiste Le Bihan,1 and it has been carefully discussed by Schaffer (2007). Let us examine this idea in more detail.

§2 First, we need to distinguish between two types of monism, namely, priority monism and existence monism. Priority monism (see, inter alia, Schaffer [2010, 2014]) is clearly a view that is very different from eliminativism. It is

168  Aesthetic (and Other) Objects a type of monism, so it claims that there is exactly one object that exists, the world. For the priority monist, this is the one and only fundamental object. But, in this view, it has parts. The parts of the world are not fundamentally existing, they exist only in a derivative sense. Thus, priority monism is a hierarchical ontology – some things (i.e. one thing) are fundamental, while other things are derivative. But both types of things exist, albeit in different ways. It is hard to understand what ‘derivative existence’ means here (perhaps the relation of grounding can be of help), but whatever it means, we can be sure of one thing: nothing like this can be found in eliminativism. The eliminativist recognizes only one way of existing, the fundamental way, the one we talk about in Ontologese. Eliminativists, as we have seen in I.1.§5, can be conventionalists – they accept that chairs and whales are conventions in the sense of useful fictions. But such ‘conventional objects’ do not exist fundamentally, and they do not exist derivatively either (whatever that means). The eliminativist in no way reifies these conventions; indeed, this is one of the core ontological and methodological claims of eliminativism. A second important difference between priority monism and eliminativism lies in the fact that the former view accepts that there are relations of parthood (relations of decomposition), while the eliminativist denies any such ‘corresponding’ relation of composition, since composition never occurs, in her view (again, except in a conventional sense, not to be reified).

§3 The interesting type of monism here, as a possible candidate for being ‘equivalent’ to eliminativism, is existence monism. Existence monism is a rarely defended view, but a recent defence can be found in Horgan and Potrč (2000, 2008, 2012). I am not going to take on board here the semantic component of their particular view; I shall mainly focus on the metaphysics and on possible similarities with eliminativism. In short, existence monism claims that there exists only one object, the world, ‘the blobject’ as Horgan and Potrč call it, and that’s it. This object has a complex structure and is locally varied, but this does not mean that it has parts, not even derivatively. The existence monist does not help herself to any notion of derivative existence, and she does not let in any relation of parthood; there is no composition or decomposition relation in her view. Similarly to the eliminativist (see I.1.§5), the existence monist can appeal to a paraphrase strategy in order to be able to talk about ordinary objects. Thus, whenever the realist says that there is a chair, and the eliminativist says that there are simples arranged chairwise, the monist can say that the world is locally chair-ish. Similarly to what we have seen in the case of eliminativism in Part I, Chapter 2, the monist can deal in this way with a possible charge of contra-intuitiveness or craziness of her view – if needed, she can salvage ordinary talk in this way, without reifying it. This similarity

Concluding Remarks  169 also holds when it comes to perception. As we have seen, our visual (or otherwise perceptual) experience would be the same whether there were chairs around us or whether there are simples arranged chairwise. The monist can make a similar claim here: our experience would be the same whether the world contained chairs or whether it is locally chair-ish. These remarks suggest that the difference between eliminativism and existence monism might perhaps be merely terminological – precisely because they are so exactly opposed, that they actually mirror each other. It is the ‘middle views’ between these two ‘extremes’, namely the realist theories, that are genuinely different. Schaffer (2007) discusses in detail this idea, and concludes that (i) eliminativism and monism are very close to each other indeed (as he rightly points out, there is a sense in which existence monism is a species of eliminativism, since where the eliminativist holds that all concrete objects are simple, the monist agrees and adds that there is only one such simple), but that (ii) they are not equivalent, because (iii) existence monism is a better view than eliminativism. Schaffer provides three reasons to think that existence monism is superior to eliminativism. One of them is based on a way to understand the nature of quantum entanglement: in this view, there are emergent holistic properties of entangled quantum systems that are not derivable from the intrinsic states and spatio-temporal arrangements of their components (see Schaffer [2007, §VI]). In this situation, the eliminativist has an objection to face which does not arise for the monist. I am not going to focus on this objection, which would require a long discussion of its own, and which presupposes that we can fruitfully derive metaphysical conclusions from an interpretation of quantum mechanics, since quantum mechanics is an incomplete theory, with various interpretations, and one should be very careful when trying to establish metaphysical claims based on it. A detailed critical discussion can be found in Calosi (2014) who concludes by stating the importance of being careful: “a lot of work, in terms of both physics and metaphysics, needs to be done. Until this work is done the quantum mechanical case for Monism remains inconclusive” (Calosi [2014, p. 927]). Alternatively, the eliminativist could follow the lead of Miller (2014) who offers a possible reply for the eliminativist to this objection, based on Bohmian mechanics. A different kind of defence of eliminativism concerning this worry can be found in Caves (2015) who argues that simples taken together can, in a collective way, instantiate emergent properties. Another reason to favour monism over eliminativism that Schaffer puts forward concerns the possibility of gunk. In the case of such a possibility, Schaffer (2007, p. 183) claims, existence monism fares better than eliminativism since eliminativism, as Schaffer understands the view, contains the assumption that there are simples, and is thus incompatible with a gunky world. The monist does not make this assumption, or a similar assumption. But we can dismiss this point, given the discussion from Part I, Chapter 3, where we have seen that this assumption is not essential to eliminativism

170  Aesthetic (and Other) Objects and that eliminativists can easily accommodate the possibility of gunk or even other kinds of fundamental ontologies. The third reason Schaffer provides in favour of monism is ontological simplicity. Indeed, one can hardly imagine a view that would be simpler than the claim that there really is only one object. This is, as Schaffer (2007, p. 187) puts it, “the simplest sufficient ontology”, and he adds that the eliminativist’s “quest for the simplest sufficient ontology culminates in monism”. In his view, eliminativists should thus by their own strategic and methodological lights go even further and embrace monism. But it is not clear that monism really is a simpler view. As we have seen in the Introduction, there are various kinds of simplicity, and the claim that there exists only one object exhibits only one of these. Granted, from a quantitative simplicity point of view, such a claim is hard to beat. But this one object that the monist introduces has an internal structure, and is thus complex, even if it does not have parts. This complexity of structure corresponds to the eliminativist’s arrangements of simples, which are simpler than the monist’s one simple – they are not only simple in the sense that they do not have parts, but they are also simple in the sense that they have a simpler structure, a simpler nature. Monism is thus quantitatively more parsimonious when it comes to counting objects, but it is less simple when it comes to the nature of the objects (or object, in the monist’s case) that exist – on this point, eliminativism is the simpler view. Furthermore, we have also seen in the Introduction that even in a situation where one view is clearly simpler than another (which is not even the case here), this does not establish that the former view is true or that it should be embraced rather than the latter. Simplicity is not truth-conducive and it is not a decisive tool when selecting the best theory around. The point that is supposed to count here in favour of monism thus stands on a very shaky ground indeed.

§4 In Meta-metaphysics,2 especially in Part I, chapter 4, I have defended the view that what counts when it comes to claims of ‘metaphysical equivalence’ between theories is to see what they do, how they do their job. This is a functional meta-metaphysical view of metaphysical equivalence. In this view, in order to claim that two theories are equivalent, not only must they do the same overall job, but they must also do it in the same way in a sense that is relevant to an apt level of analysis (that is, the claim of metaphysical equivalence must not be too general and it must not be too local and detailed either). Let us have a look at eliminativism and monism from the perspective of this meta-theoretical background. Is there something that one of the views can do that the other cannot? How do these two theories do their job? Let us focus on the issues raised in Part I. Both views deny the existence of ordinary objects, and both views deny a genuine relation of (de)composition

Concluding Remarks  171 (unlike priority monism, and unlike realist views of ordinary objects). Both views thus avoid problems with arbitrariness and vagueness, and they both easily solve any puzzles concerning coincident objects in the same way: no objects, no worries. With this (important) respect, both views do their job in the same way, in the sense that their respective strategies exactly mirror each other. In this sense, they are functionally equivalent. In Part I, I also discussed in detail the relationship between eliminativism and common sense. The eliminativist claims that there are simples arranged chairwise and whalewise, while the monist says that the world is locally chair-ish or whale-ish. Perhaps one might claim that from the point of view or common sense and ordinary intuitions eliminativism sounds a bit less crazy than monism, since in everyday life we are used to building houses from smaller bricks and playing Lego games with our kids. It is thus more natural to claim that ordinary objects are eliminated in virtue of the existence of smaller components, than to claim that ordinary objects are local aspects of the world. The bottom-up approach fits better with our everyday practice than the top-down approach. The fiction of conventional objects in the eliminativist view is thus easier to take on board than the fiction that the monist appeals to in her paraphrase strategy (“the world is locally whale-ish”). But even if this were true, there is very little weight to be put on this ‘advantage’ of eliminativism over monism. As we have seen in Part I, Chapter 2 in detail, such intuitive claims of common sense should not be our guides when it comes to metaphysical theories, and even if it turned out that eliminativism really is less counter-intuitive than monism this should not constitute a reason to select it as being a better theory. As I already said in the Introduction, the reason to adopt eliminativism is not its simplicity (see above), and intuitions should not play any role here either – rather, the good reason to embrace eliminativism is its explanatory power and its capacity to solve and/or avoid problems with ordinary objects. It’s the way a theory does its job that counts. And in this respect, monism and eliminativism seem to be two sides of the same coin indeed – as we have seen above, the ways they work mirror each other. To my mind, the reason to prefer eliminativism to monism thus lies in the generality of the method eliminativism provides. Indeed, as Parts II and III of this book show, eliminativism is not only an ontological claim to the effect that there are no ordinary objects, but it is also a method that can be applied to a variety of other cases in topic-specific ways, like in the case of the Self or of musical works. Eliminativism is thus not only serviceable as a great theory in the field of metaphysics of objects, but it can also serve as a recipe to be applied in other, different, situations (the Self, philosophy of mind, aesthetics, ontology of social objects, . . .). Note that these two roles of eliminativism can be combined but do not have to: for instance, one can embrace eliminativism about musical works as defended in Part III without further eliminating scores, performances, etc. One can do this – combining in this way eliminativism about musical works with eliminativism about

172  Aesthetic (and Other) Objects objects like scores and performances, using the tools of Part I – but one does not have to (but yes, one should, of course). In this sense, eliminativism is more appealing than monism. The two views mirror each other when it comes to issues discussed in Part I, but the topic-specific arguments from Part II and Part III involve an eliminativist methodology applied in a topic-specific way. Existence monism would have to be supplemented with topic-specific arguments concerning these issues in order to be as serviceable as eliminativism, and it is not clear that this can be done – what would such a view look like when it comes to the Self or to aesthetic objects? This is not a defect of existence monism. This brand of monism was designed to deal with ordinary objects and it does its job as well as PartI-eliminativism here. It cannot be directly applied to other cases, such as those from Part II and Part III, but this is only to be expected – existence monism (and Part-I-eliminativism as well of course) is simply silent on these issues. Existence monism thus mirrors Part-I-eliminativism, but the beauty of eliminativism lies in the fact that its methodology can be used to produce topic-specific arguments in various situations, thus providing a general and unified worldview.

Notes 1  Eliminativism about ordinary objects can be an aspect of a top-down view of the universe in which what is real is the cosmos, or the stuff. Here, ordinary objects are substituted by proper parts of the cosmos, the stuff or spacetime, resulting from a mereological relation of decomposition. . . . Let us call this position ‘top-down eliminativism’. Eliminativism can also be construed as a bottom-up approach. According to bottom-up eliminativism, objects are substituted by collections of mereological simple entities, resulting from a relation of composition. . . . Actually, I believe that bottom-up and top-down eliminativisms are two faces of the same coin: the two views are different descriptions of the very same universe. (Le Bihan [2016, pp. 2155–2156]) 2 Benovsky (2016b).

References

Akiba, K. 2004. “Vagueness in the world”, Noûs 38 (3):407–429. Aristotle. 1961. De Anima. Edited by D. Ross. Oxford: Clarendon Press. Armstrong, D. M. 1968. A Materialist Theory of the Mind. London: Routledge. Armstrong, D. M. 1978. Nominalism and Realism. Cambridge: Cambridge University Press. Armstrong, D. M. 1984. “Consciousness and causality”, in D. M. Armstrong and N. Malcolm (eds.), Consciousness and Causality, pp. 103–191. Oxford: Wiley-Blackwell. Armstrong, D. M. 1997. A World of States of Affairs. Cambridge: Cambridge University Press. Barnes, E. 2010. “Ontic vagueness: A guide for the perplexed”, Noûs 44 (4):601–627. Bartel, C. 2017. “The ontology of musical works and the role of intuitions: An experimental study”, European Journal of Philosophy 25:1. Bennett, K. 2009. “Composition, colocation, and metaontology”, in D. Chalmers, D. Manley and R. Wasserman (eds.), Metametaphysics: New Essays on the Foundations of Ontology. Oxford: Oxford University Press. Benovsky, J. 2006a. Persistence Through Time and Across Possible Worlds. Frankfurt: Ontos Verlag. Benovsky, J. 2006b. “Four-dimensionalism and modal perdurants”, in Paolo Valore (ed.), Topics on General and Formal Ontology. Milano: Polimetrica Publisher. Benovsky, J. 2008. “The bundle theory and the substratum theory: Deadly enemies or twin brothers?” Philosophical Studies 141:175–190. Benovsky, J. 2009a. “Presentism and persistence”, Pacific Philosophical Quarterly 90:3. Benovsky, J. 2009b. “Eternalist theories of persistence through time: Where the differences really lie”, Axiomathes 19 (1). Benovsky, J. 2012. “Aesthetic supervenience vs. Aesthetic grounding”, Estetika: The Central European Journal of Aesthetics XLIX/V (2):166–178. Benovsky, J. 2013a. “Branching and (in)determinism”, Philosophical Papers 42 (2):151–173. Benovsky, J. 2013b. “The present vs. The specious present”, Review of Philosophy and Psychology 4 (2):193–203. Benovsky, J. 2014. “I am a lot of things: A pluralistic account of the self”, Metaphysica 15 (1):113–127. Benovsky, J. 2015. “From experience to metaphysics: On experience-based intuitions and their role in metaphysics”, Noûs 49 (4):684–697.

174 References Benovsky, J. 2016a. “Dual-aspect monism”, Philosophical investigations 39 (4):335–352. Benovsky, J. 2016b. Meta-Metaphysics. Switzerland: Springer. Bigelow, J. 1996. “Presentism and properties”, in James E. Tomberlin (ed.), Philosophical Perspectives: Vol. 10. Malden, MA: Wiley-Blackwell. Brenner, A. 2015a. “Mereological nihilism and theoretical unification”, Analytic Philosophy 56 (4):318–337. Brenner, A. 2015b. “Mereological nihilism and the special arrangement question”, Synthese 192 (5):1295–1314. Brenner, A. 2017. “Simplicity as a criterion of theory choice in metaphysics”, Philosophical Studies 174 (11):2687–2707. Brentano, F. 1874. Psychology from an Empirical Standpoint. Edited by A. C. Rancurello, D. B. Terrell and L. L. McAlister. London: Routledge, 1973. Bricker, P. 2001. “Island universes and the analysis of modality”, in G. Preyer and F. Siebelt (eds.), Reality and Human Supervenience: Essays on the Philosophy of David Lewis. Lanham, MD: Rowman and Littlefield. Bricker, P. 2006. “The relation between the general and the particular: Entailment vs. supervenience”, in D. Zimmerman (ed.), Oxford Studies in Metaphysics, Vol. 2, pp. 251–287. Oxford: Oxford University Press. Broad, C. D. 1923. Scientific Thought. New York: Harcourt, Brace and Co. Brown, L. B. 2011. “Do higher-order music ontologies rest on a mistake?” British Journal of Aesthetics 51 (2):169–184. Calosi, C. 2014. “Quantum mechanics and priority monism”, Synthese 191:915–928. Cameron, R. 2008. “There are no things that are musical works”, British Journal of Aesthetics 48 (3):295–314. Caplan, B. and Matheson, C. 2006. “Defending musical perdurantism”, British Journal of Aesthetics 46:59–69. Caplan, B. and Matheson, C. 2008. “Defending ‘Defending Musical Perdurantism’ ”, British Journal of Aesthetics 48:80–85. Carmichael, C. 2011. “Vague composition without vague existence”, Noûs 45 (2):315–327. Caves, R. 2015. “Emergence for nihilists”, Pacific Philosophical Quarterly 97:2. Chalmers, D. J. 2003. “Consciousness and its place in nature”, in Stephen Stich and Ted A. Warfield (eds.), Blackwell Guide to the Philosophy of Mind, pp. 102–142. Malden: Wiley-Blackwell. Chisholm, R. 1990a. “Events without times: An essay on ontology”, Noûs 24:413–428. Chisholm, R. 1990b. “Referring to things that no longer exist”, Philosophical Perspectives: Vol. 4, pp. 545–556. Churchland, P. M. 1981. “Eliminative materialism and the propositional attitudes”, Journal of Philosophy 78:67–90. Churchland, P. M. 1988. Matter and Consciousness. Cambridge, MA: MIT Press. Churchland, P. S. 1986. Neurophilosophy: Toward a Unified Science of the Mind/ Brain. Cambridge, MA: MIT Press. Cleland, C. E. 2012. “Life without definitions”, Synthese 185:125–144. Cleland, C. E. and Chyba, C. 2007. “Does ‘life’ have a definition?” in W. T. Sullivan and J. Baross (eds.), Planets and Life: The Emerging Science of Astrobiology, Ch.5, pp. 119–131. Cambridge: Cambridge University Press. Craig, W. L. 2000. The Tensed Theory of Time. Dordrecht: Kluwer. Dainton, B. 2000. Stream of Consciousness. London: Routledge.

References  175 Dainton, B. 2003. “Time in experience: Reply to gallagher”, Psyche 9 (10). Dainton, B. 2008a. The Phenomenal Self. Oxford: Oxford University Press. Dainton, B. 2008b. “Sensing change”, Philosophical Issues 18, Interdisciplinary Core Philosophy. Dainton, B. 2010. “Temporal consciousness”, in Stanford Encyclopaedia of Philosophy, online. https://plato.stanford.edu/entries/consciousness-temporal/ Dainton, B. 2012. “Selfhood and the flow of experience”, Grazer Philosophische Studien 84:173–211. Davies, P. 1995. About Time: Einstein’s Unfinished Revolution. Harmondsworth: Penguin Books. De Rosset, L. 2010. “Getting priority straight”, Philosophical Studies 149 (1):73–97. Dennett, D. 1978. “Why you can’t make a computer that feels pain”, in Brainstorms, pp. 190–229. Cambridge, MA: MIT Press. Dennett, D. 1988. “Quining Qualia”, in A. Marcel and E. Bisiach (eds.), Consciousness in Contemporary Science, pp. 42–77. New York: Oxford University Press. Deplazes-Zemp, A. 2012. “The conception of life in synthetic biology”, Science and Engineering Ethics 18 (4):757–774. Descartes, R. 1637. Discourse on the Method of Rightly Conducting One’s Reason and of Seeking Truth in the Sciences. Paris: Adam et Tannery. Dodd, J. 2008. “Musical works: Ontology and meta-ontology”, Philosophy Compass 3/6:1113–1134. Dorsch, F. 2007. “Novel sounds”, manuscript, RERO digital library link, http://doc. rero.ch/record/11857. Dupre, J. and O’Malley, M. 2009. “Varieties of living things: Life at the intersection of lineage and metabolism”, Philosophy and Theory in Biology 1 (1). Foster, J. 1982. The Case for Idealism. London: Routledge & Kegan Paul. Foster, J. 1992. The Immaterial Self. London: Routledge. Gallie, Ian. 1936. “Is the self a substance ?” Mind 45 (177):28–44. Garfield, J. and Edelgass, W. (eds.). 2009. Buddhist Philosophy: Essential Readings. Oxford: Oxford University Press. Gennaro, R. 1996. Consciousness and Self-Consciousness: A Defense of the Higherorder Thought Theory of Consciousness. Amsterdam and Philadelphia: John Benjamins Publishing Company. Gennaro, R. 2005. “The HOT theory of consciousness: Between a rock and a hard place”, Journal of Consciousness Studies 12 (2):3–21. Guillot, M. 2016. “I me mine: On a confusion concerning the subjective character of experience”, Review of Philosophy and Psychology :1–31. Haslanger, S. 2003. “Persistence through time”, in Michael J. Loux and Dean W. Zimmerman (eds.), Oxford Handbook of Metaphysics, pp. 314–354. Oxford: Oxford University Press. Hawthorne, J. and Cover, J. A. 1998. “A world of universals”, Philosophical Studies 91:205–219. Heller, M. 1990. The Ontology of Physical Objects: Four-Dimensional Hunks of Matter. Cambridge: Cambridge University Press. Heller, M. 1992. “Things change”, Philosophy and Phenomenological research 52:695–704. Heller, M. 1993. “Varieties of four-dimensionalism”, Australasian Journal of Philosophy 71:47–59.

176 References Heller, M. 1996. “Against metaphysical vagueness”, Philosophical Perspectives 10:177–185. Heller, M. 1998a. “Property counterparts in ersatz worlds”, The Journal of Philosophy 95:293–316. Heller, M. 1998b. “Five layers of interpretation for possible worlds”, Philosophical Studies 90 (2):205–214. Heller, M. 2003. “The immorality of modal realism, or: How I learned to stop worrying and let the children drown”, Philosophical Studies 114. Heller, M. 2008. “The donkey problem”, Philosophical Studies 140 (1):83–101. Hirsch, E. 2010. Quantifier Variance and Realism: Essays in Metaontology. Oxford: Oxford University Press. Hoerl, C. 2009. “Time and tense in perceptual experience”, Philosophers’ Imprint 9 (12). Hoffman, J. and Rosenkrantz, G. 1994. Substance Among Other Categories. Cambridge: Cambridge University Press. Hoffman, J. and Rosenkrantz, G. 1997. Substance: Its Nature and Existence. London: Routledge. Hollingdale, R. J. 2000. The Waste Books: Georg Christoph Lichtenberg. New York: The New York Review of Books. Horgan, T. and Potrč, M. 2000. “Blobjectivism and indirect correspondence”, Facta Philosophica 2:249–270. Horgan, T. and Potrč, M. 2008. Austere Realism: Contextual Semantics Meets Minimal Ontology. Cambridge, MA: MIT Press. Horgan, T. and Potrč, M. 2012. “Existence monism trumps priority monism”, in Philip Goff (ed.), Spinoza on Monism, p. 2012. Basingstoke: Palgrave Macmillan. Howell, R. J. and Thompson, B. (forthcoming). “Phenomenally mine: In search of the subjective character of consciousness”, Review of Philosophy and Psychology :1–25. Huemer, M. 2009. “When is parsimony a virtue?” Philosophical Quarterly 59 (235):216–236. Hume, David. 1738/1978. A Treatise of Human Nature. Edited by L. A. Selby-Bigge and Revised by P. H. Nidditch. Oxford: Oxford University Press. Husserl, E. 1964. The Phenomenology of Internal Time-Consciousness. Bloomington: Indiana University Press. Hyde, D. 2011. “Sorites paradox”, Stanford Encyclopaedia Online, http://plato. stanford.edu/entries/sorites-paradox. James, W. 1890. The Principles of Psychology. New York: H. Holt and Company. Jansson, L. and Tallant, J. 2017. “Quantitative parsimony: Probably for the better”, British Journal for the Philosophy of Science 68 (3):781–803. Kant, I. 1781. Critique of Pure Reason. Cambridge: Cambridge University Press, 1998. Originally published in 1781/1787. Kivy, P. 1983. “Platonism in music: A kind of defense”, The Worlds of Art and the World, Grazer Philosophische Studien 19:109–130. Kleinschmidt, S. and Ross, J. 2012. “Repeatable artwork sentences and generics”, in Mag Uidhir (ed.), Art and Abstract Objects. Oxford: Oxford University Press. Korman, D. 2009. “Eliminativism and the challenge from folk belief”, Noûs 43 (2):242–264. Korman, D. 2014. “Debunking perceptual beliefs about ordinary objects”, Philosophers’ Imprint 14 (13).

References  177 Korman, D. 2015. Objects: Nothing Out of the Ordinary. Oxford: Oxford University Press. Kriegel, U. 2003. “Consciousness as intransitive self-consciousness: Two views and an argument”, Canadian Journal of Philosophy 33 (1):103–132. Kriegel, U. 2009. Subjective Consciousness: A Self-Representational Theory. Oxford: Oxford University Press. Kriegel, U. 2013. “The epistemological challenge of revisionary metaphysics”, Philosophers’ Imprint 13. Kripke, S. 1972. Naming and Necessity. Cambridge, MA: Harvard University Press. Krueger, J. 2011. “The who and the how of experience”, in Dan Zahavi, Evan Thompson and Mark Siderits (eds.), Self, No Self? Perspectives from Analytical, Phenomenological, and Indian Traditions, pp. 27–55. Oxford: Oxford University Press. Le Bihan, B. 2013. “Why a gunk world is compatible with nihilism about objects”, Studia Philosophica Estonica 6 (1):1–14. Le Bihan, B. 2015. “No physical particles for a dispositional monist?” Philosophical Papers 44 (2):207–232. Le Bihan, B. 2016. “Super-relationism: Combining eliminativism about objects and relationism about spacetime”, Philosophical Studies 173:2151–2172. Le Bihan, B. forthcoming. “The no-Self view and the meaning of life”. Philosophy East and West. Le Poidevin, R. 1991. Change, Cause, and Contradiction. London: Palgrave Macmillan. Le Poidevin, R. 2000. “Continuants and continuity”, The Monist 88 (3):336–352. Le Poidevin, R. 2007. “Projecting the present: The shock of the now”, in R. Le Poidevin (ed.), The Images of Time. Oxford: Oxford University Press. Lehrer, K. 1990. Theory of Knowledge. London: Westview Press. Levine, J. 2001. Purple Haze: The Puzzle of Consciousness. Oxford: Oxford University Press. Levinson, J. 1980. “What a musical work is”, Journal of Philosophy 77 (1):5–28. Levinson, J. 1991. “What a musical work is, again”, in Music, Art, and Metaphysics, pp. 215–263. Oxford: Oxford University Press. Lewis, D. 1968. “Counterpart theory and quantified modal logic”, Journal of Philosophy 65 (5):113–126. Lewis, D. 1973. Counterfactuals. Oxford: Wiley-Blackwell. Lewis, D. 1976a. “Survival and identity”, in Amelie Oksenberg Rorty (ed.), The Identities of Persons, pp. 17–40. Berkeley, CA: University of California Press. Lewis, D. 1976b. “The paradoxes of time travel”, American Philosophical Quarterly 13 (2):145–152. Lewis, D. 1983. “Survival and identity”, in D. Lewis (ed.), Philosophical Papers, Volume I. Oxford: Oxford University Press. Lewis, D. 1986. On the Plurality of Worlds. Malden: Wiley-Blackwell. Lewis, D. 1991. Parts of Classes. Malden: Wiley-Blackwell. Lewis, D. 1993. “Many but almost one”, in Keith Cambell, John Bacon and Lloyd Reinhardt (eds.), Ontology, Causality, and Mind: Essays on the Philosophy of D. M. Armstrong, pp. 23–38. Cambridge: Cambridge University Press. Locke, J. 1975. An Essay Concerning Human Understanding. Edited by Peter H. Nidditch. Oxford: Clarendon Press. Loux, M. J. (ed.). 2001. Metaphysics Contemporary Readings. London: Routledge.

178 References Lowe, E. J. 1996. Subjects of Experience. Oxford: Oxford University Press. Lowe, E. J. 2006. The Four-Category Ontology: A Metaphysical Foundation for Natural Science. Oxford: Clarendon Press. Lycan, W. G. 1996. Consciousness and Experience. Cambridge, MA: MIT Press. Lycan, W. G. 2001. “A simple argument for a higher-order representation theory of consciousness”, Analysis 61:3–4. Machery, E., Mallon, R., Nichols, S. and Stich, S. 2004. “Semantics, cross-cultural style”, Cognition 92 (3). MacKenzie, M. 2008. “Self-awareness without a self: Buddhism and the reflexivity of awareness”, Asian Philosophy 18 (3):245–266. Mackinnon, J. E. 2001. “Aesthetic supervenience: For and against”, British Journal of Aesthetics 41 (1). Maclaurin, J. and Dyke, H. 2012. “What is analytic metaphysics for?” Australasian Journal of Philosophy 90 (2):291–306. Margolis, J. 1980. Art and Philosophy. Atlantic Highlands, NJ: Humanities Press. Markosian, N. 2004. “A Defense of presentism”, in Dean W. Zimmerman (ed.), Oxford Studies in Metaphysics: Volume 1. Oxford: Oxford University Press. Maudlin, T. 2007. The Metaphysics Within Physics. Oxford: Oxford University Press. McGrath, M. 2005. “No objects, no problem?” Australasian Journal of Philosophy 83 (4):457–486. Melia, J. 2005. “Truthmaking without truthmakers”, in Helen Beebee and Julian Dodd (eds.), Truthmakers: The Contemporary Debate. Oxford: Oxford University Press. Mellor, D. H. 1998. Real Time II. London: Routledge. Merricks, T. 1999. “Persistence, parts, and presentism”, Noûs 33:421–438. Merricks, T. 2001. Objects and Persons. Oxford: Oxford University Press. Merricks, T. 2003. “The end of counterpart theory”, Journal of Philosophy 100 (10):521–549. Miller, E. 2014. “Quantum entanglement, Bohmian mechanics, and humean supervenience”, Australasian Journal of Philosophy 92 (3):567–583. Miller, K. 2004. “How to be a conventional person”, The Monist 87 (4):457–474. Miller, K. 2009. “Defending vontingentism in metaphysics”, Dialectica 63 (1):23–49. Miller, K. 2010. “Three routes to contingentism in metaphysics”, Philosophy Compass 5 (11):965–977. Miller, K. 2013. “ ‘Personal identity’ minus the persons”, Philosophical Studies 166 (S1):91–109. Newton, I. 1687. Philosophiæ Naturalis Principia Mathematica. Cambridge: Harvard University Press, 1972. Ney, A. 2008. “Physicalism as an attitude”, Philosophical Studies 138:1–15. Nida-Rümelin, M. 2009. “An argument from transtemporal identity for dubjectbody dualism”, in Robert C. Koons and George Bealer (eds.), The Waning of Materialism: New Essays. Oxford: Oxford University Press. Nida-Rümelin, M. 2011. “The conceptual origin of subject body dualism”, in Annalisa Colliva (ed.), Self and Self-Knowledge. Oxford: Oxford University Press. Nietzsche, F. 1886. Jenseits von Gut und Böse: Vorspiel einer Philosophie der Zukunft. Trans. as Beyond Good and Evil: Prelude to a Philosophy of the Future, R. P. Horstmann and J. Norman, 2002. Cambridge: Cambridge University Press.

References  179 Oaklander, N. 1991. “A defense of the new tenseless theory of time”, Philosophical Quarterly 41:26–38. Olson, E. T. 2007. What Are We? A Study in Personal Ontology. Oxford: Oxford University Press. Parfit, D. 1984. Reasons and Persons. Oxford: Oxford University Press. Paul, L. A. 2010. “Temporal experience”, Journal of Philosophy 107 (7). Paul, L. A. 2012. “Metaphysics as modeling: The Handmaiden’s tale”, Philosophical Studies 160 (1):1–29. Paul, L. A. forthcoming. “The subjectively enduring self”, in Ian Phillips (ed.), The Routledge Handbook of the Philosophy of Temporal Experience. Oxford: Routledge. Pettit, P. 1987. “The possibility of aesthetic realism”, in Eva Schaper (ed.), Pleasure, Preference and Value. Cambridge: Cambridge University Press. Phillips, I. 2011. “Indiscriminability and experience of change”, The Philosophical Quarterly 61 (245): 808–827. Plantinga, A. 1973. “Transworld identity or worldbound individuals?” in M. Munitz (ed.), Logic and Ontology. New York: New York University Press. Pouivet, R. 2010. Philosophie du rock. Paris: PUF. Prior, A. N. 1970. “The notion of the present”, Studium Generale 23:245–248. Prior, A. N. 2003a. Papers on Time and Tense. New Edition. Edited by Per Hasle and Peter Ohrstrom. Prior, A. N. 2003b. “Changes in events and changes in things”, in Prior, 2003a, pp. 7–19. Prosser, S. 2012. “Why does time seem to pass?” Philosophy and Phenomenological Research 85 (1):92–116. Quine, W. V. O. 1960. Word and Object. Cambridge, MA: MIT Press. Rettler, B. 2018. “Mereological nihilism and puzzles about material objects”, Pacific Philosophical Quarterly, First published: 7 March 2018, online first. Rodriguez-Pereyra, G. 2002. Resemblance Nominalism: A Solution to the Problem of Universals. Oxford: Oxford University Press. Rohrbaugh, G. 2003. “Artworks as historical individuals”, European Journal of Philosophy 11 (2):177–205. Rosen, G. 2006. “The limits of contingency”, in Identity and Modality. Oxford: Oxford University Press. Rosen, R. 1985. Anticipatory Systems: Philosophical, Mathematical and Methodologicalfoundations. Oxford: Pergamon Press. Rosenthal, D. 1986. “Two concepts of consciousness”, Philosophical Studies 49:329–359. Rosenthal, D. 2002. “Explaining consciousness”, in D. Chalmers (ed.), Philosophy of Mind: Classical and Contemporary Readings, pp. 406–421. Oxford: Oxford University Press. Sagan, C. 1970. “Life”, in Encyclopaedia Britannica. 15th ed., pp. 985–1002, 22. Chicago, IL: Britannica. Sartre, J. P. 1936. La transcendance de l’ego. Paris: Vrin; Trans. as The Transcendence of the Ego. F. Williams and R. Kirkpatrick. New York: The Noonday Press, 1957. Sartre, J. P. 1943. L’être Et le Néant. Essai d’Ontologie Phénoménologique. Paris: Gallimard. Sattig, T. 2015. The Double Lives of Objects. Oxford: Oxford University Press.

180 References Schaffer, J. 2007. “From nihilism to monism”, Australasian Journal of Philosophy 85 (2):175–191. Schaffer, J. 2009. “On what grounds what”, in D. Manley, D. J. Chalmers and R. Wasserman (eds.), Metametaphysics: New Essays on the Foundations of Ontology. Oxford: Oxford University Press. Schaffer, J. 2010. “Monism: The priority of the whole”, Philosophical Review 119:1. Schaffer, J. 2014. “Monism”, Stanford Encyclopedia of Philosophy, https://plato. stanford.edu/entries/monism/. Schaffer, J. 2016. “Grounding in the image of causation”, Philosophical Studies 173 (1):49–100. Schaffer, J. manuscript. “Grounding as the primitive concept of metaphysical structure.” Schechtman, M. 1996. The Constitution of Selves. Ithaca and London: Cornell University Press. Scholl, B. J. 2007. “Object persistence in philosophy and psychology”, Mind and Language 22 (5):563–591. Scruton, R. 1974. Art and Imagination: A Study in the Philosophy of Mind. South Bend, IN: St. Augustine’s Press. Searle, J. 1995. The Construction of Social Reality. New York: Free Press. Shoemaker, S. 1997. “Self and substance”, Philosophical Perspectives 11. Sider, T. 1993. “Van Inwagen and the possibility of gunk”, Analysis 53 (4):285–289. Sider, T. 1997. “Four-dimensionalism”, The Philosophical Review 106:197–231. Sider, T. 2000. “The stage view and temporary intrinsics”, Analysis 60:84–88. Sider, T. 2001. Four Dimensionalism: An Ontology of Persistence and Time. Oxford: Oxford University Press. Sider, T. 2006. “Beyond the Humphrey objection”, Unpublished. Sider, T. 2013. “Against parthood”, Oxford Studies in Metaphysics 8:237–293. Sider, T. 2015. “Nothing over and above”, Grazer Philosophische Studien 91:191–216. Siderits, M. 2003. Personal Identity and Buddhist Philosophy: Empty Persons. London and New York: Ashgate. Siderits, M. 2007. Buddhism as Philosophy: An Introduction. Indianapolis: Hackett Pub. Co.. Simons, P. 2000. “Continuants and occurrents”, The Aristotelian Society 74:78–101. Skow, B. 2009. “Relativity and the moving spotlight”, The Journal of Philosophy 12:666–678. Smid, J. 2015. “The ontological parsimony of mereology”, Philosophical Studies 172 (12):3253–3271. Sober, E. 1981. “The principle of parsimony”, The British Journal for the Philosophy of Science 32 (2):145–156. Strawson, G. 1997. “The self”, Journal of Consciousness Studies 4.5/6:405–428. Strawson, G. 2003. “Real materialism”, in L. Antony and N. Hornstein (eds.), Chomsky and His Critics: Reprinted in Strawson, Real Materialism and Other Essays. Oxford: Clarendon Press, 2008. Strawson, G. 2006. “Realistic monism – why physicalism entails panpsychism”, Journal of Consciousness Studies 13 (10–11):3–31. Swain, S., Alexander, J. and Weinberg, J. 2008. “The instability of philosophical intuitions: Running hot and cold on truetemp”, Philosophy and Phenomenological Research 76 (1):138–155.

References  181 Tallant, J. 2013. “Quantitative parsimony and the metaphysics of time: Motivating presentism”, Philosophy and Phenomenological Research 87 (3):688–705. Tallant, J. 2014. “Against mereological nihilism”, Synthese 191 (7):1511–1527. Thomasson, A. L. 2015. Ontology Made Easy. Oxford: Oxford University Press. Thomson, J. J. 1983. “Parthood and identity across time”, The Journal of Philosophy 80:201–220. Thunder, S. D. 2017. “Mereological nihilism: Keeping it simple”, Thought 6 (4):278–287. Tillman, C. 2011. “Musical materialism”, British Journal of Aesthetics 51:13–29. Tillman, C. and Spencer, J. 2012. “Musical perdurantism and the inheritance problem”, Analysis 72:252–259. Tye, M. 1990. “Vague objects”, Mind 99 (396):535–557. Unger, P. 1979. “There are no ordinary things”, Synthese 41 (2):117–154. Van Cleve, J. 1985. “Three versions of the bundle theory”, Reprinted in Loux 2001. Van Fraassen, B. 2002. The Empirical Stance. New Haven, CT: Yale University Press. Van Inwagen, P. 1981. “The doctrine of arbitrary undetached parts”, Reprinted in Van Inwagen 2001. Van Inwagen, P. 1985. “Plantinga on trans-world identity”, Reprinted in Van Inwagen 2001. Van Inwagen, P. 1986. “Two concepts of possible worlds”, Reprinted in Van Inwagen 2001. Van Inwagen, P. 1990. Material Beings. Ithaca, NY: Cornell University Press. Van Inwagen, P. 1994. “Composition as identity”, Philosophical Perspectives 8:207–220. Van Inwagen, P. 2002. “What do we refer to when we say ‘I’?” in R. Gale (ed.), The Blackwell Guide to Metaphysics. New York: Wiley-Blackwell. Varzi, A. 2003. “Naming the stages”, Dialectica 57:387–412. Velleman, D. 2006. “So it does”, Amherst Lecture in Philosophy. Walton, K. 1970. “Categories of art”, Philosophical Review 79. Walton, K. 2008. Marvelous Images: On Values and the Arts. Oxford: Oxford University Press. Weinberg, J., Gonnerman, C., Buckner, C. and Alexander, J. 2010. “Are philosophers expert intuiters?” Philosophical Psychology 23 (3):331–355. Willard, M. B. 2014. “Against simplicity”, Philosophical Studies 167 (1):165–181. Williams, J. R. G. 2006. “Illusions of gunk”, Philosophical Perspectives 20 (1):493–513. Williams, J. R. G. 2008. “Ontic vagueness and metaphysical indeterminacy”, Philosophy Compass 3/4:763–788. Williamson, T. 2011. “Philosophical expertise and the burden of proof”, Metaphilosophy 42 (3):215–229. Wittgenstein, L. 1921. Logisch-Philosophische Abhandlung. Trans. as Tractatus Logico-Philosophicus, C. K. Ogden, 1922. London and New York: Routledge & Kegan Paul. Wolterstorff, N. 1975. “Toward an ontology of art works”, Noûs 9:115–142. Wolterstorff, N. 1980. Works and Worlds of Art. Oxford: Oxford University Press. Wright, C. 2009. “The illusion of higher-order vagueness”, in Richard Dietz and Sebastiano Moruzzi (eds.), Cuts and Clouds: Vaguenesss, Its Nature and Its Logic. Oxford: Oxford University Press.

182 References Zahavi, D. 2005. Subjectivity and Selfhood: Investigating the First-Person Perspective. Cambridge, MA: Bradford Book and MIT Press. Zangwill, N. 1994. “Supervenience unthwarted: Rejoinder to wicks”, Journal of Aesthetics and Art Criticism 52. Zimmerman, D. 1996. “Persistence and presentism”, Philosophical Papers 25:115–126. Zimmerman, D. 1997. “Chisholm and the essences of events”, in Lewis Hahn (ed.), The Philosophy of Roderick M. Chisholm, pp. 73–100. Peru, IL: Open Court.

Index

aesthetic objects 1, 6, 141 – 143, 167, 172 arbitrariness 9, 10, 14 – 23, 45, 53 – 54, 71 – 77, 142 Armstrong, David 49, 54 – 55, 100, 109 – 110, 160 Brentano, Franz 100 – 101 Buddhist philosophy 18 – 19, 92, 100, 107, 109, 112, 115, 129 – 136 bundle theory 30, 51 – 52, 65, 80 – 90, 92, 104, 108 – 110 cogito 109 coincident entities 20 – 22, 68 – 69, 171 composition 6, 9, 10, 15, 30, 41, 49, 59, 70, 168, 170 composition as identity (CAI) 49, 55 consciousness 2, 31, 33, 36 – 38, 41, 65 – 77, 92 – 94, 100 – 109 conventional objects 18 – 19, 130, 168, 171 counterpart theory 21 – 22, 50, 115 – 131, 144 Dainton, Barry 79, 81, 88, 108 – 110 Descartes, René 65, 78, 95, 107, 109 – 110 Dharmakīrti 92 – 93, 101, 103, 107, 111, 153 diachronic unity 83, 87, 93, 105 – 107, 127 dual-aspect monism 91 – 102; see also monism dualism 31 – 32, 34, 95 – 97

easy ontology 28 – 30 endurantism 30 – 34, 38 – 42, 44, 69, 83 – 87 equivalence (metaphysical) 79, 83, 87, 127, 135, 167 – 171 exceptionalism 17, 19, 65 – 77 existential suffering 115, 130, 133 – 136 explanatory power 5, 61, 83, 87, 171 four-dimensionalism 20 – 22, 31, 67 – 70, 77, 103, 125; see also perdurantism gunk 58 – 61, 169 – 170 Heller, Mark 114 – 115, 121 – 124, 128, 132, 136 Hirsch, Eli 41 – 42 Hume, David 65, 78, 80 – 81, 89, 93, 98, 103, 105, 108 – 110, 115, 129, 131 Husserl, Edmund 35 intuitions 16, 18, 23, 25 – 46, 59, 171 James, William 35, 104 Korman, Daniel 26 – 30, 45 – 46, 59 Kriegel, Uriah 92, 101 – 102 Lewis, David 3, 9 – 10, 54 – 55, 68, 70, 102, 107, 114, 116 – 124, 128, 145 Lowe, Jonathan 29, 107, 110, 126

184 Index Merricks, Trenton 1 – 2, 15 – 16, 19, 23, 30, 47, 49 – 51, 55, 61, 71 – 76, 83, 108 – 111, 118 – 124 mineness 92 – 107 monism 27, 60 – 61, 167 – 172; see also dual-aspect monism musical works 1, 61, 141 – 154, 156, 157, 160 – 161, 163 – 166, 171 Nagarjuna 18 – 19 Newton, Isaac 4 – 5 no-self view 78, 88, 91 – 112, 127 – 128, 154 nothing over and above 49 – 56 ontologese 18, 26, 44 – 46, 60, 76, 154, 165, 168 ordinary objects 9 – 19, 23, 26, 29 – 30, 42 – 43, 54, 65, 71, 75, 77, 92, 103, 111, 115, 125 – 126, 167 – 168, 170 – 172 paraphrase strategy 18, 22, 26, 58, 60 – 61, 168, 171 Parfit, Derek 92, 107, 114 – 115, 129, 131 – 136 parsimony 1 – 6 passage of time 31 – 34, 38 – 39, 44 Paul, L.A. 36 – 37, 41, 103 – 105 perception 15, 17, 30, 36, 44 – 45, 78, 80, 98, 101 – 102, 108 – 109, 132, 148 – 149, 153, 169; see also phenomenal experience perdurantism 29, 33 – 38 – 40, 44, 50, 83 – 84, 86, 100, 103, 125, 144 – 149; see also fourdimensionalism phenomenal experience 16, 17, 30, 33 – 34, 38, 40 – 41, 43, 94, 103,

107, 111, 132; see also fourdimensionalism photographs 1, 61, 142, 151 – 153, 156 – 166 physicalism 53 – 54, 91, 95 – 97 Quine, W. V. O. 53 – 54, 121 reductionism 49 – 57, 114 – 115, 129, 131 – 137, 141 Schaffer, Jonathan 55, 167, 169 – 170 self 1 – 2, 31 – 32, 34, 63 – 136, 142, 153 – 154, 171 – 172 self-reflexivity, self-reference, selfrepresentation 92, 100 – 103 Sider, Theodore 5, 16 – 17, 19 – 20, 25 – 27, 30, 44, 46, 55, 58 – 60, 68, 125 – 127 Siderits, Mark 129 – 130, 134 simplicity 2 – 6, 83, 170 – 171 social objects 141, 171 Strawson, Galen 89, 95 subject of experience 91 – 92, 103, 111, 136 subjectivity 91 – 94, 98, 100 – 107, 111, 143, 153 substance 29, 30 – 31, 52, 65, 78 – 90 substratum theory 30, 52, 78, 80 – 81, 133 Thomasson, Amie 28 – 30 vagueness 6, 9, 10 – 23, 45, 52 – 53, 71, 73 – 74 Van Inwagen, Peter 2, 6, 19 – 23, 25, 31, 52, 58 – 59, 65, 71 – 75, 77, 81 – 82, 85, 111, 121 Walton, Kendall 152