Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself 9780192845566, 019284556X

Table of contents :
Dedication
Acknowledgements
Contents
Introduction
PART I: HOW THE WORLD MANIFESTS ITSELF TO US
1. The Causal Argument for Sense-data, ‘Philosophers’ Hallucinations’, and the Disjunctive Response
2. Naïve Realism and the Argument from Illusion
3. Intentionality and Perception (I): The Fundamental Irrelevance of Intentionality to Phenomenal Consciousness
4. Intentionality and Perception (II): Attempts to Articulate the ‘Content’ and ‘Object’ Distinction
5. Singular Reference and Its Relation to Intentionality
6. Objectivity: How Is It Possible?
7. Semantic Direct Realism, Critical Realism, and the Sense-Datum Theory
8. Building the Manifest World
PART II: WHAT THE WORLD IS, IN ITSELF
9. The Problematic Nature of the Modern Conception of Matter
10. Two Suggestive Berkeleian Arguments
11. Bishop Berkeley and John Foster on Problems with Physical Realism about Space
12. Mentalist Alternatives to Berkeleian Theism, and Their Failure
General Conclusion
Bibliography
Index
Пустая страница
Пустая страница
Пустая страница

Citation preview

Perception and Idealism

Perception and Idealism An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself HOWARD ROBINSON

Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Howard Robinson 2022 The moral rights of the author have been asserted First Edition published in 2022 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2022939629 ISBN 978–0–19–284556–6 DOI: 10.1093/oso/9780192845566.001.0001 Printed and bound in the UK by TJ Books Limited

John Foster (1941–2009)

Acknowledgements This book is a successor to my Perception, (Routledge, 1994) and to the last chapter of Matter and Sense (Cambridge University Press, 1982), and various things that I have written on perception and Berkeleian idealism since those books. ‘Two Berkelian arguments about the nature of space’, in T. Airaksinen and B. Belfrage (eds), Berkeley’s Lasting Legacy: Three Hundred Years Later, Cambridge Scholars Press, 2011, 79–90, is published with permission from Cambridge Scholars Press. Large parts of Chapter 1 are taken from ‘The failure of disjunctivism to deal with “philosophers’ hallucinations” ’, in Fiona Macpherson and Dimitris Platchias (eds), Hallucinations, MIT Press, 2013, 313–30, and I am grateful to MIT Press for permission to use this material. Extensive material from ‘John Foster’s and Michael Dummett’s arguments for idealism’ in (eds) Paul Gocke and Ralph Weir, From Existentialism to Metaphysics: the Philosophy of Stephen Priest, Peter Lang, 2021, 123–36, is used in 10.2 and 11.4. I am grateful to Peter Lang for permission to use this material. I also made extensive use of material from ‘Idealism’, in B. McLaughlin, A. Beckermann, and S. Walter (eds), The Oxford Handbook to the Philosophy of Mind, Oxford University Press, 2009, 186–205, when discussing the ‘powers’ conception of matter, and when considering Berkeley’s so-called ‘master argument’. I have drawn on many other things I have published, but not so directly as with these. Personal thanks are more complicated. Two OUP readers made very useful comments, which led me to make various changes, most seriously, to significantly expand the final chapter, and Peter Momtchiloff has been, as usual, very supportive. Michael Pelczar also read a complete draft and made very useful comments. I have been discussing perception with Mike Martin since the late 1980s, and he also read a complete draft of the book, with comments. Certain chapters were read to an NYU group hosted by David Chalmers and Ned Block, and to the CUNY seminar run by David Rosenthal. I am very grateful for comments from both these groups. What little understanding I have of quantum theory comes from attending David Albert’s wonderfully lucid graduate seminars at Rutgers, and from bringing all my problems to Barry Loewer, who patiently tried to get me to understand. I cannot remember how much of it has been read to the constant yet fluid group of philosophers who have met in whatever was my residence in Oxford ever since

viii



the late 1970s, but I know that Penelope Mackie, Bob Frazier, Anita Avramides, Ralph Walker Lesley Brown, Julie Jack, Adrian Moore, and others have made many valuable contributions. My debt to Central European University, both to colleagues and students, can hardly be measured. I am grateful to Philip Goff, for our discussions when he was with us; Katalin Farkas, who has supported and challenged me ever since my arrival in Budapest in 1994, and continues to do so in Vienna; and Tim Crane, who subjected my arguments on intentional theories to deep and vital criticism. Thanks to the fact that my wife—Jocelyn Wogan-Browne—held a distinguished chair in the English Department at Fordham University, I was able to work for part of each year from 2011 to 2019 in Fordham’s wonderful university library—the most comfortable working library I have ever known. I am grateful to the philosophy department there for making me a Visiting Scholar, and to the Jesuits of Fordham, for the ethos they maintain at their university. My deepest personal debt is to Fr Patrick Ryan. I am also grateful to Blackfriars Hall, Oxford for making me a Research Fellow, and to Brian Davies, O.P., for making that possible. I had the privilege of teaching advanced and graduate courses at Rutgers University for four semesters during my time in America, and of being a Senior Fellow of their Center for Philosophy of Religion, thanks to Dean Zimmerman, to whose support and friendship I owe a great deal. If one holds views which have been, at their best, unfashionable, and, worse, are often thought downright weird, support from philosophers who are nearer to the mainstream, such as those I have thanked above, makes a major contribution to one’s sanity. Nevertheless, close friendship with someone whose views are no less weird than one’s own, especially if they have the uniquely powerful philosophical intellect of John Foster, is even more central to maintaining one’s morale. Therefore, I owe most to the late John Foster, with whom I discussed the issues treated in this book from the late 1960s until he died in 2009. He spent his whole career as a Fellow of Brasenose College, Oxford. He died after more than a decade of painful cancer, during which time he wrote three books. And I could not count the number of dinners I shared with John and Helen at their home in Headington from 1970 until Helen’s death in 2014. I dedicate this book to John Foster, but remember them both with the deepest gratitude.

Contents Introduction 1 The Aim of This Book 2 Part I: Chapter by Chapter 3 Part II: Its Rationale through the Chapters 4 The Nature of Sense-Data

1 1 2 3 4

P A R T I : HO W T HE W O R L D M A N I F E S T S IT S EL F TO U S 1. The Causal Argument for Sense-data, ‘Philosophers’ Hallucinations’, and the Disjunctive Response 1.1 Philosophers’ Hallucination: Introductory Remarks 1.2 Preliminary Thoughts on the Role of Causation in Perception 1.3 Philosophers’ Hallucinations: The Argument 1.4 Strategies for Opposing the Causal-Hallucinatory Argument 1.5 Disjunctivist Accounts of Hallucination: Introductory Remarks 1.6 Disjunctivist Accounts of Hallucination; (i) Martin’s ‘Indiscriminability’ Account 1.7 Disjunctivist Accounts of Hallucination; (ii) Fish’s ‘Belief ’ Account 1.8 Disjunctivist Accounts of Hallucination; (iii) Soteriou and ‘Seeming to Experience’ 1.9 Naïve Realism and Philosophers’ Hallucinations: Conclusion 2. Naïve Realism and the Argument from Illusion 2.1 Introduction 2.2 Disjunctivism and Illusion 2.3 French and Phillips’ ‘Austere’ Naïve Realism, and Why It Is Not as Austere as They Hope 2.4 Brewer, Campbell, and Perspectivalism 2.5 The Perspectivalism of Fish and Kalderon 2.6 Genone and the Doxastic Theory 2.7 Conclusion 3. Intentionality and Perception (I): The Fundamental Irrelevance of Intentionality to Phenomenal Consciousness 3.1 Introduction 3.2 Searle’s Appeal to Intentionality in Perception, and the Illuminating Contrast with Crane 3.3 Presentationality and the ‘Blocking Function’

11 11 12 14 17 22 22 26 28 30

32 32 35 40 42 45 47 48

49 49 51 58

x

 3.4 Crane’s Own Account of Intentionality 3.5 The Intentionalist’s Dilemma and Its History 3.6 How Appeal to Transparency Helps No-one

4. Intentionality and Perception (II): Attempts to Articulate the ‘Content’ and ‘Object’ Distinction 4.1 Introductory Remarks 4.2 Modern Responses (i): Smith: ‘Phenomenal Objects’ Are Not Objects in the Relevant Sense 4.3 Modern Responses (ii): The Contents of Subjective Experience as Abstractions: Dretske, Lycan, and Jackson 4.4 Modern Responses (iii): Contents as Abstract: Johnston and Schellenberg 4.5 Modern Responses (iv): Schellenberg on Discriminatory Capacities 4.6 Conclusion

61 63 66

69 69 70 72 74 78 79

5. Singular Reference and Its Relation to Intentionality 5.1 Introduction 5.2 Brentano’s Howler 5.3 Mill as Supposed Proponent of Direct Reference 5.4 Intentionality and the Distinctive Character of Thought: Having a Conception of an Object, Mental Files, and Mental Maps 5.5 A Note on ‘Content’ and ‘Object’ 5.6 A Different Model of Intentionality for Sensations? 5.7 ‘Representation’ in a Reductive Sense 5.8 Conclusion: World Maps and Perception

80 80 81 82

6. Objectivity: How Is It Possible? 6.1 Introduction 6.2 Direct Realism and Objectivity 6.3 The Causal-Semantic Account of Objectivity 6.4 Burge on Distil Causes and the Experience of ‘How Things Look’ 6.5 The Transition to Hume 6.6 David Papineau and the Manifest Image 6.7 Constancy and Coherence: the Humean Account of Objectivity 6.8 Conclusion

91 91 92 97 100 102 105 107 114

7. Semantic Direct Realism, Critical Realism, and the Sense-Datum Theory 7.1 The Situation So Far 7.2 How We Might Understand Directness 7.3 SDR and Intentionalism 7.4 SDR and Relationalism 7.5 Critical Realism 7.6 The Sense-Datum Theory and SDR

115 115 116 117 122 123 125

84 87 88 89 90



8. Building the Manifest World 8.1 Introduction 8.2 The Role of Judgement in, and Its Integration with, Perception 8.3 The Sense-Datum Theory Is Not an Error Theory 8.4 Our Spatial World and Visual Experience 8.5 Perceiving Objects, Not Just Qualities 8.6 Availability and Phenomenology 8.7 Sense-Data, Direct Realism, and the Common-Sense Understanding of Perception 8.8 Conclusion

xi

126 126 127 132 134 135 141 142 143

P A R T I I : W H A T T H E WO R L D I S, I N I T S E L F 9. The Problematic Nature of the Modern Conception of Matter 9.1 Introduction 9.2 Sensible Qualities, the Nature of Matter and the Regress of Powers 9.3 Contemporary Discussion of the Powers Regress 9.4 Grounding Basic Powers 9.5 Quiddities, and Similar Devices 9.6 The Humean Account of Causation: Against the Primitiveness of Regularity 9.7 Scientific Realism about Quantum Theory, and Common-Sense

147 147 147 152 156 160 162 165

10. Two Suggestive Berkeleian Arguments 10.1 Introduction 10.2 The Sense-Dependence of Qualities 10.3 The Physical World and the Nature of Thought

174 174 174 178

11. Bishop Berkeley and John Foster on Problems with Physical Realism about Space 11.1 Introduction: Two Berkelian Arguments about the Nature of Space 11.2 Mites, Men, and Objective Space 11.3 Newton’s Thought Experiments and Absolute Space 11.4 John Foster on Spatial Topology and Empirical Reality 11.5 Conclusion

187 187 187 192 198 205

12. Mentalist Alternatives to Berkeleian Theism, and Their Failure 12.1 Introduction 12.2 Hume–Mill Phenomenalism 12.3 Panpsychism 12.4 Idealism without God 12.5 God as the Source of the Laws of Nature 12.6 Conclusion

207 207 208 213 220 221 224

General Conclusion Bibliography Index

225 227 233

Introduction 1 The Aim of This Book It is a standard feature of modern philosophy, at least from Locke, to tie together the questions of how we perceive the world and what we have reason to think the world is like in itself. This is a natural connection, because the questions of how we perceive it, and what kind of conception of it we can best form on the basis of that mode of perception, are obviously intimately linked. If one is a naïve realist, this would seem to license a belief that the world is broadly the way it appears, for if it were not how could perception be a straightforward awareness of the way the world is, as the naïve realist claims? But then the naïve realist faces the problem of explaining the difference between the world as he directly apprehends it, and the picture of the world that science seems to deliver. Locke was not a naïve realist but a representative realist, and that cleared the way for him to support the contemporary scientific theory that secondary qualities are not, themselves, intrinsic to the physical world, but are features of how the world affects us. Hume responded to this by claiming that, once naïve realism has been abandoned, we have lost all reason to hold on to any common-sense conception of the world, including Locke’s representative theory. [representative realism] has no primary recommendation either to reason or the imagination, but acquires all its influence on the imagination from [naïve realism] (Hume, 1739/2000: Treatise, I. 4, 2. 47)

Once one has decided that the world itself is not the immediate contents of sense experience, but ‘ideas’, or, in more modern jargon, ‘sense-data’ or ‘qualia’, one must surely seek for evidence other than, or in addition to, perception to decide what the world in itself is like. Mainly because of his views on causation and explanation, Hume reached sceptical conclusions about what we could rationally believe about any world beyond our own sense-data. This scepticism developed into the phenomenalism articulated by J. S. Mill, Russell at certain periods, the positivists and recently revived by Michael Pelczar. According to this theory, physical objects are just, in Mill’s phrase, ‘the permanent possibility of sensation’: this ‘possibility’ is not grounded on any further reality.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0001

2



Most philosophers have not reacted to the refutation of direct or naïve realist theories of perception in such a sceptical or reductionist way, but, like Locke, try to give a positive account of what lies behind the world of appearance. Part I of this book concerns the philosophy of perception, and seeks to defend a version of the sense-datum theory. The conclusion, in Chapter 8, is that in perception we are aware of the way physical objects manifest themselves to creatures like ourselves. It contains a direct realist element in the judgemental or informational component in perception, but is a sense-datum theory in relation to the phenomenal component. So we perceive the things out there, but in the way they manifest themselves to us. This leaves open the question of the nature of physical objects themselves, as against the way they manifest themselves to us. This latter topic is the concern of Part II.

2 Part I: Chapter by Chapter Part I is an attempt to bring up to date the arguments in my Perception of 1994. Chapter 1 presents the causal-hallucinatory argument against naïve realism, or relationism, as it is often called, and argues that the disjunctivist strategies against this argument are grossly implausible. In Chapter 2 I argue that relationism cannot cope with the much more prosaic phenomena invoked by the ‘argument from illusion’. These chapters dispose of naïve realism/relationism. Chapters 3, 4, and 5 argue against what are called intentionalist or representationalist theories of perception. Chapter 3 states the basic argument against the theory, and Chapter 4 looks at the various strategies that intentionalists have used to respond to objections, and finds them wanting. Chapter 5 looks more generally at how singular reference and intentionality are related, and does so in a way that reinforces the conclusion that intentionality is not a property of phenomenal content. Chapter 6 looks into what it is about our perception that makes it seem to be about external things and not just subjective feelings. I consider, and find inadequate, the argument of McDowell and others that the only way such a notion could be acquired is if we were in fact directly aware of external things. I also reject Tyler Burge’s argument, recently supported by David Papineau, that causal semantics explain objectivity. Finally I defend the Humean view that it is the way experience is structured and organized that endows us with the notion of objectivity. In Chapter 7 I argue that intentionalists and even relationalists can only achieve what I call semantic direct realism, and not phenomenal direct realism. That is, it is the judgemental or informational component that refers to external reality, not directly the phenomenal content. This conclusion is somewhat modified in Chapter 8, in which I try to explain how our manifest world—the world of common experience, or the way the world

 :     

3

manifests itself to creatures like us—can be built up from our sensory input and our innate resources.

3 Part II: Its Rationale through the Chapters While Part I considers how the world appears to the likes of us, Part II tackles the question of what it is like in itself, pressing the case for a Berkeleian idealism. But I work towards this by first looking at what might be considered orthodox scientific conceptions of matter, in Chapter 9. First I discuss the powers or energy conception of matter that dominates what one might call our post Newtonian conception. This domination came thanks both to the work of Roger Boscovitch, the eighteenth-century Croatian Jesuit who formulated the first coherent field or powers (as opposed to the traditional ‘hard billiard ball’) model of the atom, and to the role the concepts of energy, field, and force play in later physics. I take up the argument that such a conception of the physical world only makes sense within an idealist framework. Then, in the same chapter, I consider some features of David Lewis’s quiddities, and Michael Esfeld’s ‘matter points’. The ‘powers’ conception arises primarily from science, Lewis’s and Esfeld’s are more purely philosophical in motivation, but are also meant to elaborate a minimum ontology for the physical world, in the context of modern science. Finally in Chapter 9, I try, rather gingerly, to consider some aspects of the quantum conception of matter. My objective there is to show that, although science is usually presented as a sophisticated development of common sense, quantum theory either falls into a form of idealism, in the form of the Copenhagen theory, or resorts to world views far more bizarre than Berkeleian idealism. Having tried to show the shortcomings of what might roughly be called modern scientific, or science based, forms of physical realism, I move on in the next two chapters to consider arguments for idealism that are more purely philosophical and more closely based on traditional philosophical texts, particularly those of Berkeley. Three of the four arguments in these chapters draw on topics in Part I. One of them develops the intuition that qualities are essentially creatures of sensation, the other two on how the contents of experience determine the nature of physical space. The fourth, which is in the second part of Chapter 10, concerns not perception, but thought, and can be more easily aligned with the Hegelian than the empiricist tradition. Chapter 10 considers two suggestive, if not conclusive, arguments in Berkeley. One concerns the idea that qualities are necessarily mental. I draw on arguments found in Robert Adams and Michael Dummett. The other derives from Berkeley’s claim that we cannot form the conception of a world that is not dependent on mind. I look at the claim that our conception of the world draws on features which are necessarily contributed by the understanding.

4



It derives from what is often called Berkeley’s ‘master argument’, to the effect that we cannot conceive the unperceived. I consider various arguments that suggest that there are vital features of our conception of physical reality that presuppose the presence of mind. This argument could be taken in a Kantian, or in a neo-platonic manner, though I prefer the latter. Chapter 11 looks at arguments that claim to show that physical space is minddependent. I first deal with two arguments that Berkeley himself uses, then John Foster’s wholly original proof that the topology of physical space is essentially empirical. By this point, I have presented as strong a case as I am able that the world depends essentially on something mental. It has been my objective to defend Berkeleian idealism, but that is not the only mentalist picture available, because there is also panpsychism and bare phenomenalism. It is my aim in Chapter 12 to show that non-theistic phenomenalism is not defensible, despite Michael Pelczar’s recent subtle efforts to reconstruct it: nor panpsychism, a theory once regarded as bizarre, but which now, thanks to the work of Galen Strawson, Phillip Goff, and others, is almost fashionable. I end the chapter by defending John Foster’s argument that we cannot avoid laws of nature, and that these entail the existence of a law-maker, whom it is natural to identify with God, thus supporting the Berkeleian option. I hope I have shown that the two parts of this book are naturally connected. One way in which they are different is that most of what is in Part I about the philosophy of perception is material that will not be dismissed as so eccentric as to be beyond the pale. Part II runs the risk, simply because of its conclusions—not, I hope, because of its style of argument—of being dismissed as too eccentric. If someone is tempted to that reaction I would ask two things. First, that that would not lead them to be dismissive of Part I, by a ‘guilt by association’: second, that they take Part II as raising real problems facing the concept of matter, and so is, at least, a legitimate raising of aporia, even if one thinks these problems must have a more realist solution. There remains, however, a rather different introductory issue that must be addressed before I plunge into the arguments.

4 The Nature of Sense-Data This book is divided into two parts. Part I concerns the philosophy of perception and in it I defend what I label as a sense-datum theory of perception. It might, therefore, help if I start by saying how I understand the term ‘sense-data’ for the purposes of the arguments in this book. Both proponents and opponents of the theory often insist that the data are particulars or individuals. This can lead to problems if one follows the slogan, ‘no

   -

5

entity without identity’, for answering questions like ‘how many visual sense-data are you having now?’ or ‘as the blue patch in your visual field alters shape and fades slightly, how many sense-data are involved?’ The correct answer to these challenges is to say what C. I. Lewis said when discussing the data of experience. There are recognizable qualitative characters of the given, which may be repeated in different experiences, and are thus a sort of universals; I call these ‘qualia’ (1929: 121)

I do not think that there is any significant difference between Lewis’s use of ‘qualia’ and mine of ‘sense-data’. One can talk of them having instances, which suggests particularity, but that merely recognizes the fact that they can be presented at different times and to different subjects. If you think ‘instance’ commits one to individuality, then take is as shorthand for ‘occasions of presentation, or phenomenal presence’. The crucial point is that they involve the actual presence of the quality in question, and not merely representations of such qualities, as the intentional theory claims. How one individuates them is not important, any more than it is to decide how many images there are on a television or cinema screen at any given time. What matters is that there is a qualitative array, the features present in which can be picked out and identified. In the course of the book I will be dealing with standard objections to the theory, including the claim that sense-data constitute a ‘veil of perception’, but there is one line of objection that I will deal with here, because it relates directly to the topic we have been considering, namely the nature of sense-data. This is the claim that they are indeterminate and contradictory, and so cannot be genuine entities. Hardin argues as follows: The empirical evidence in favor of the indeterminacy of visual shape and color under various conditions of seeing seems overwhelming; we should take it to be a phenomenological fact . . . But accepting indeterminacy into phenomenology is one thing and incorporating it into ontology is another . . . The sharp sensory-interpretative distinction seems to be consequent on a strict distinction between act and object of perception . . . Drawing this distinction appears to be a core tactic of sense-datum theorists and leads in turn to the most damaging problem of all, the pressure to assign incompatible predicates to the same sense-datum . . . It may be possible to find and articulate appropriate pairs of phenomenological predicates that will capture the relevant sensory differences and similarities between the way that the rocks in the waterfall illusion seems to move and the way in which they do not seem to move, and likewise for other paradoxical sensory presentations. (1988: 108–9)

6



So, according to Hardin, there are two serious objections to the sense-datum theory. They are both objections to the reification of phenomenal content on the grounds that it fails to meet the standard necessary for being an entity, namely that entities must be determinate and must not possess inconsistent properties. My response to both these points is essentially the same, namely that there is no such thing as an indeterminate or inconsistent sensation; there are only ways of characterizing sensations in physical object terms which are indeterminate or inconsistent. Hardin, like others, makes a lot of phenomena such as the waterfall illusion but there is an example of putative contradiction much nearer home, namely feeling dizzy. If you feel dizzy, things seem to go round without actually getting any further. The experience is difficult to describe in a way that does not turn out to be incoherent, but it would not occur to many people to think of it as an incoherent experience because we think of it as just a sensation ‘in the head’ and not a representation of an apparent physical state of affairs. I suggest that in neither case is there anything contradictory about the experience per se, only that there is no physical interpretation of the scene that does not involve inconsistency. The situation is, I think, analogous to that of an Esher painting. What is there is not a contradiction; only if you see it as a three dimensional scene does the contradiction arise. Something similar applies to sensations: in themselves they are just as they are, but attempts to render them as representations of physical processes sometimes leads to contradiction. A similar approach works for indeterminacy. First it might be helpful to distinguish between indeterminacy and vagueness, where vagueness is something that allows the generation of sorites paradoxes. Sorites vagueness is a phenomenon of concepts; nothing real can be vague in this way. Indeterminacy is another matter. It seems to me to be unclear what absolute physical determinacy is, and absolute determinacy for a secondary quality such as colour is particularly unclear. We have an absolute geometrical notion of shape and size, but whether anything quite fits it, is another matter. If I take off my glasses, I have fuzzy images of things. It would be hard, if not impossible to create physical things like that, but it does not follow that sensations cannot be like that, or if they are, then they can only be ‘phenomenologically’ real and not ‘ontologically’. The view that the sense-datum theorist should take is this. Our common-sense or Manifest Image conception of the world fundamentally rests on the nature of our sensations. It is fashionable to talk of the transparency of sensations or qualia. This means that their qualitative nature is merely the other side of the coin of how we take the world to be. This is clearly true for many of our sensation. What red looks like, is what we take ‘naïve’ (or in Chalmers’ expression, ‘Edenic’) red actually to be like. The same goes for visual squareness and actual physical squareness, as the normally sighted hold it to be. But there are some sensations—such as dizziness— that cannot be treated in this transparent way. This is because, even though these

   -

7

phenomena apparently involve the external world (e.g., it seems to spin) and is not like a simple bodily sensation, such as a headache, they do not fit into the organized pattern of how the world works—they are not part of the ‘constancy and coherence’ that Hume pointed out was what leads us to take our experience as being of an objective world.¹ I think that this indicates the need to distinguish two conceptions of sense-data. The generic conception is: (a) Sense-data are entities instantiating the qualitative nature of how things appear. But there is an ambiguity here depending how one understands ‘qualitative nature’. Are comparative judgements—like comparisons of size—part of the qualitative nature, or is a cognitive process, however primitive, something added to the basic qualities sensed? This gives rise to the two interpretations of (a). (b) Sense-data are the qualitative content of ‘mature’ experience, after certain spontaneous cognitive processes have taken place, including judgements of similarity. (c) Sense-data are raw qualitative contents sensed prior to any cognitive processes beyond bare awareness itself. The former might crudely be labelled the empiricist account and the latter is more like a Kantian approach to the given. I do not see why the sense-datum theorist cannot say that you can use the term either way, depending on purpose and context. The essence of the sense-datum theory is that there could not be sense-experience of the kind with which we are all familiar without subject-dependent occurrences of standard phenomenal qualities. This applies paradigmatically to colour. As H. H. Price (1932: 3) pointed out: When I see a tomato there is much that I can doubt . . . There is one thing I cannot doubt: there exists a red patch of a round and somewhat bulgy shape . . .

¹ I was able to see Adam Pautz’s (2021) too late to include adequate discussion (but it is thanks to Pautz that I saw it in time at all!), but the objections to sense-data that he presents, p. 52ff, are open to the same answer. He cites cases of peripheral indeterminacy, and confused perceptions. These are just cases of sensations that need interpretation and modification to be informative about the world. I also reject his characterization of the sense-datum realm as a kind of virtual reality produced by the brain. Virtual reality is wholly independent of the subject’s immediate environment: that is not the case of sense-data involved in normal perception, so I think that this characterization is an unfair case of persuasive definition! Also, the argument in Chapter 8 does more to show how unjustified I think this characterization to be. Sense-data constitute the manifest world, that is the way that objects manifest themselves to creatures like us, not an independent realm of reality.

8



The reality of certain qualitative features that show up through whatever further cognitive processing that there might be is something without which there could be no experience. Dennett (1991) is not wrong to say that fully mature experience involves various layers of processing: he is wrong to try to ignore the phenomenal nature of that which undergoes this refinement. The line between data and further processing might be uncertain, without doubt being thrown on the need for qualitative data if there is to be experience at all.

PART I

H O W T H E W O R L D MA N I F E S T S ITSELF TO US

1 The Causal Argument for Sense-data, ‘Philosophers’ Hallucinations’, and the Disjunctive Response 1.1 Philosophers’ Hallucination: Introductory Remarks In this chapter I shall present what I believe to be the strongest argument against naïve or direct realism, namely the causal-hallucinatory argument. This argument depends on the possibility of ‘philosophers’ hallucinations’. After I have explained what philosophers’ hallucinations are, and set out the argument, I’ll draw attention to the role of what I call the non-arbitrariness of philosophers’ hallucinations in supporting this argument. I will then introduce the various ways in which one might try to resist the argument. Of those responses, this chapter will mainly be devoted to disjunctivism, the strategy favoured by modern supporters of naïve realism—or the relational theory, as naïve realism is also called nowadays. This discussion will naturally lead on, in the following chapter, to a discussion of relationalism and the phenomenon commonly known as ‘illusion’, namely when real objects look in ways different from how they actually are. This follows naturally from the discussion of disjunctivism because some disjunctivists apply their theory to ‘illusions’, as well as to hallucinations. They do this because neither hallucination nor ‘illusions’ can be understood in a simple, ‘naïve’ relationist way, because both involve experiences which do not present the world as it is; so it is naturally tempting to try to apply the same account to both. The name ‘relational theory’ for naïve realism comes because its proponents want to stress their difference from their main contemporary rival, the intentional or representational theory. Intentional mental states are not relational, in the usual sense, because they can have intentional objects that do not exist: for example, someone can worship Zeus even though he does not in fact exist. Normal relations require both (or all, if there are more than two) relata to exist: A can be to the left of B only if both A and B exist. According to naïve realism, a perceptual experience is not an intentional or representational state, but a genuine relation between the perceiving subject and the object he perceives; the object perceived is a constituent factor of the experience itself, not merely its cause. Because of this, the naïve realist must provide a different account of hallucinatory experience from that provided

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0002

12

    -

of perception, because the experiential state itself cannot consist in a relation to something external: if I hallucinate a pink elephant there is no pink elephant to which I am experientially related. The need for a different account for genuine perception and hallucination, however similar the experiences may seem, is what gives rise to the disjunctive theory, the details of which we shall see below.¹

1.2 Preliminary Thoughts on the Role of Causation in Perception Before stating and defending the argument itself, I want to make some preliminary and general remarks about the role of causation in perception, which, I hope, set a background of prima facie plausibility of the causal-hallucinatory argument. If one is unimpressed by these initial thoughts, it should make little significant difference to the one’s judgement of the soundness of the argument proper. In Robinson (1994, 4) I pointed to a deep tension that faced direct or naïve realist theories. On the one hand, a little reflection – that is, thought that does not resort to any science that goes beyond common experience – shows that perception involves some sort of physical influence running from the external object to the sense organ of the perceiver. On the other hand, the essential nature of experience seems to be that the subject reaches out to, and makes conscious contact with, the external object. The directionality of the physical process and that of the lived experience seem to be in direct conflict. How can a process in which the subject is the passive recipient of a stimulus be the physical realization of a process in which the subject reaches actively and consciously out into the world?²

The issue for the direct realist is whether they have managed to reconcile these two ‘directions of action’. The sense datum theory accepts the role of the causal process in allowing that it causes in the subject a phenomenal state. The intentionalist (as we shall see) does something similar, but claims that the intentionality of the state ¹ There is also the issue of whether intentionalism can properly be characterized as a form of direct realism. We will see in the next chapter that there is an important ambiguity between what might be called internalist and externalist forms of intentionalism, and the claims of the former to be a kind of direct realism are, at best, dubious. ² I might seem to have given a hostage to fortune in claiming that the perceiver is a ‘passive recipient’ of the stimulus: perception involves an active response to the input. But this is irrelevant, for the responses involved are the ways the stimulus is processed, in the light of already existent mental state of the perceiver. The fundamental point of the directionality of the causal process and the apparent direction of consciousness remain. Also irrelevant is the question of whether the role of the causal process enters into the concept of perception, as Grice, for example claimed, and Snowdon denied. The fact of the causal role is all that is needed to generate the problem.

     

13

produced internally allows it to constitute a reaching out to the external object. But the naïve realist or relationist seems to have no worked-out account of how the causal process could give rise to the experience. The jargon sometimes used is that, for the naïve realist, the process is selective, meaning that it picks out which features of the world one is to be aware of; whereas for the sense datum theorist and the intentionalist it is generative because it produces the phenomenal state. But I am not aware of any scientifically clear account of how such selection is supposed to work. It is not clear, however, that any other sense than vision even presents itself as reaching out into the world in this way. The contrast I want to make is between the ‘torchbeam’ sense of consciousness reaching out to the world, which is how vision initially strikes us, and the way in which other senses strike us as the world making an impact on us. That is to say, for touch, taste, and smell, I think the idea that we are being impacted or affected is actually how the experience strikes us, whereas for vision, it is as if we reach out to the world, and we do not have a sense of being on the receiving end of a causal process. Sound, it seems to me, starts off seeming to be in the same category as sight, as we hear noises ‘out there in the world’, but quite ordinary experience put this into question. We see the batsman hit the ball, and hear the sound a fraction of a second later. What has happened? Has the phenomenal sound travelled through the air until it reaches us, as one might interpret Aristotle as claiming? Or are the sound waves just setting up a sound in our ears, or are we hearing the sound made as and when the bat hit the ball, but somehow directly hearing something now in the past, as some realists claim we directly see stars that are no longer there? The view that sound is in the ‘impact’ category seems the only non-contrived answer.³ It seems to me that this leaves the naïve realist about vision in a difficult position. The metaphor often used to express the idea that in vision consciousness reaches out and encompasses the relevant part of the external world is to compare it to a torch-beam, which beams out and lights upon an object. This image, while nicely capturing the way the relationist understands sight, is embarrassingly unhelpful with the role of the causal process in vision. The naïve realist says that the external object constitutes the content of the experience, as the object on which the beam falls is the content of the illuminated scene; but the beam of light is not itself modified by the object—there is no feed-back to the light source. The object picked out does the whole job of determining the illuminated content. But

³ Alex Moran has suggested an ingenious response to time-lag arguments. The general problem for the direct realist with time-lags is ‘how can you directly perceive something that no longer exists?’ Moran’s response is that, if one is a four dimensionalist about time, then past events ‘still’ (in some nonor hyper-temporal sense) exist. So the sound is ‘still’ at the point of bat and ball contact and that is what you perceive. I cannot evaluate that suggestion here, as it would involve a discussion of the plausibility of four dimensionalism.

14

    -

this is not so in vision. What we seem to see is essentially dependent on what is impressed on the retina and what processes follow therefrom (and also what memory and learning are also stored). If you say that the torch-beam image is only a metaphor, it can be pointed out that what is true of it is also true of actual cases of an agent ‘reaching out’ into the world to apprehend or grasp something. Exactly the same reaching out can grasp an apple or an orange. What I grasp need not affect my grasping action. Similarly, if pure relationism were correct, would not one expect that its mere lighting upon the object itself constituted the object’s entering into consciousness, but that is clearly not how it works. It looks as if vision has to reconcile itself somehow to being another ‘physical impact’ process. This conclusion will be reinforced in Chapter 2 when we discuss the argument from illusion, for such phenomena are the result of the different ways that objects make a physical impact on our sensory apparatus depending on our relation to them, and on the condition of our senses.

1.3 Philosophers’ Hallucinations: The Argument Philosophers, when they discuss hallucinations, are mainly interested in what I have called ‘philosophers’ hallucinations’. These are not, as far as we know, hallucinations as they actually occur, but they are, it is argued, the hallucinations that would occur if the perceptual system and brain were stimulated in the just the way it is stimulated in genuine perception, but directly and not by the usual external objects. This would give, it is supposed, a hallucination indistinguishable to the subject from the corresponding perception, which is not the case, at least in general, for hallucinations as they actually occur. A belief in the possibility of such hallucinations is taken as grounds for rejecting naïve or direct realism, as follows. (1) Possibility of Philosophers’ Hallucinations It is theoretically possible, by activating some brain process which is involved in a particular type of perception, to produce a hallucination which is subjectively indiscriminable from that perception. (2) Same Proximate Cause, Same Immediate Effect It is necessary to give the same account of hallucinations and perceptual experiences when they have the same neural cause. Thus it is not possible to say, for example, that the hallucinatory experience involves a subjective image or sense-datum, but the perception does not, if they have the same proximate— that is, neural—cause.

 ’ :  

15

(3) Hallucinations do involve some subjective image or sense-datum. Therefore (4) Perception involves some subjective image or sense-datum. There have been two main responses. Naïve realists, or relationists generally opt for the disjunctivist, but others prefer the representational, or intentionalist approach. Disjunctivism is supposed to show that one cannot argue from the nature of hallucination to the nature of perception, and so to undermine (2). It is not so much an argument against (2) as a simple denial of it.⁴ According to the disjunctivist, a proposition of the form: S seems to see something F is essentially generic, being disjoined into either (a) S sees something F or (b) S is illuded that he sees something F. The state denoted by (b) does not figure in the state captured by (a), even under another description. This is a straight denial of (2). A contrast has now been set up between disjunctivism and the common factor theory, which says that seeing and being illuded share a common component and is the core claim in (2).⁵

⁴ Dominic Alford-Duguid has objected to (2) as follows. ‘Genuine singular thoughts are essentially such that they are about a specific object. The same-cause-same-effect principle, if it applies at all, should apply to any type of mental state. But genuine singular thoughts could share a proximate/neural cause with empty singular thoughts. So either there are no empty singular thoughts—apparently empty singular thoughts would have to be about specific objects—or there are no genuine singular thoughts. But those results are absurd.’ This raises the question of how singular thoughts work, and, indeed, of what they are. I deal with these issues in my discussion of intentional theories, and especially in Chapter 5. There are two kinds of singular thoughts, the ostensive or demonstrative, as in ‘that F in front of me’ and thought about a specific object, where you know, in a broader framework, which object it is that is in front of you, or where you are referring to a specific thing, which may not be an your region at all—Julius Caesar or your best friend at school, for example. As I argue in Chapter 5, these latter depend on ‘mental files’. This would mean that the thoughts about these things would have proximate causes were different from those things that did not depend on these files, but, in the case of perception, the immediate content is not dependent on this kind of individuation. If I see someone just like Sue, my visual experience, in so far as it is relevant to the immediate debate, is the same as if it were Sue herself. The strong sense of individuality does not enter into the visual data, in the sense immediately relevant. The main upshot is that if you already think that mental states can have world-involving contents, and have them essentially, then even if you’re not a naïve realist or disjunctivist about sensory experience you’ll have reason to reject 2. ⁵ Disjunctivism is generally deployed to cope with hallucinations, but there is a division about whether it is supposed also to handle the kinds of perceptual relativities covered by the so-called

16

    -

Intentionalism, on the other hand, as very commonly understood, accepts the common factor theory, but denies (3), at least as the sense-datum theorist understands it. Both perception and hallucination share a common factor, namely that they both represent the world as being thus and so, but these representational states do not instantiate any phenomenal qualities, such as red or square, they merely represent such features—they are of red or of square. As we shall see in Chapters 3 and 4, whether this saves any kind of direct realism is a fiercely contested issue. Our concern in this chapter is with the relationists and disjunctivist strategy. Unfortunately, matters are somewhat more complicated than I have just made them out to be. I have said that disjunctivists oppose the common factor theory and repesentationalists (and sense-datum theorists) accept it. We will see in Chapter 3 that this is not strictly true of one major form of representationalism, which we might call the externalist version. Intentionalism in that version can be seen as breaking down the distinction between common factor and disjunctive theories. But the situation regarding a common factor is not so simple even with the disjunctive approach. Disjunctivists can hold that there is a common factor, provided that this is not such as to constitute the full phenomenal character of both hallucination and perception. There could be weaker common factor claims and these could, in principle, take various forms. For example, one might hold that hallucination and perception share a common factor, but that in hallucination there is that common factor plus something further (e.g. a mental image) and in perception there is the common factor, plus something else, presumably, in this case, the external object itself. On another version of an attenuated common factor theory, the common factor might be something judgemental or discriminatory, which contributes to the phenomenology. Another option might treat hallucination as this common factor alone, but perception is the common factor plus. This is the option Michael Martin countenances in his 2004: 74 and William Fish in his 2009. On Martin’s approach, the ‘inability to discriminate’ element in terms of which he explains hallucination would be common to both hallucination and perception, whilst perception also involves direct awareness of the object. For Fish, the common factor is the belief that one is having an experience when one is not in fact having an experience at all, only the belief that one is having one. For both Martin and Fish, these common factor surrogates for real experience are indistinguishable to the subject from the real thing.

‘argument from illusion’. If it is, and if such ‘illusions’ cover all cases of an object appearing other than it exactly is, then almost all—if not all—cases of perception will be cases of ‘being illuded’. This will be no help to the direct realist. If, on the other hand, disjunctivism does not touch these cases of illusion, then the direct realist still faces the challenge of accounting for them. My suspicion is that the disjunctivist would like to think that he has a cure-all which deals with all cases, but fears to look too closely at how it might be applied to non-hallucinatory cases. I think Chapter 2 gives support to this sense.

 - 

17

1.4 Strategies for Opposing the Causal-Hallucinatory Argument There are various ways in which the argument has been opposed. Perhaps the most direct response is to deny the first premise, namely the claim that, if the internal or neural processes involved in perception were activated autonomously, a veridical-seeming hallucination would be produced; hallucinations have a quite different origin. A very straightforward version of this stance can be found in John Campbell’s discussion of hallucination (Campbell and Cassam, 2014: 90–4). After citing some cases of actual hallucinatory syndromes and showing that they are very different from forms of perception, he says the following. The philosopher’s idea of a hallucination (as opposed to the empirical phenomenon of hallucination) is the idea of a mental state that is intrinsically just like seeing something, but without the external world being there. Recall Moore’s point about the transparency of visual experience . . . The implication of transparency, and the relational view of experience we have been developing, is that we do not have the conceptual materials even to formulate the idea of such a ‘hallucinatory’ state . . . The point about hallucination is this: you are asked to imagine a visual experience that is, for example, just like seeing an airport, only without the airport being there. But what is left of seeing the airport if the airport is not there? The philosopher’s answer has been: ‘the visual experience’. But the point about transparency is that your ordinary introspective knowledge of the experience of seeing an airport gives you no knowledge of any such internal state. Introspection of the experience of seeing an airport amounts merely to inspection of the airport itself. Subtract the airport and there is nothing left to introspect. (Campbell and Cassam, 2014: 92)

Campbell is here drawing what seems to be the logical consequence of relationalism; if perceptual contents are constituted by external objects, then, if there is no appropriate external object there is no content and so no experience. This is a straight denial of the first premise of the argument. Notice, however, that, although Campbell calls these ‘philosopher’s hallucinations’ he makes no mention of the fact that, in the classic argument, they are taken to be stimulated through the same brain states as are involved in perception: he has nothing to say about what one might expect from such stimulation. He seems simply to assume that no experiential, or seeming experiential, state could result from this, if the relational theory is correct, and that this is a conceptual truth. He is, therefore, also denying (2)—that same proximate cause in this case will produce the same effect. Campbell is denying the whole causal story on which the argument rests.

18

    -

That the argument simply begs the question against naïve realism has been claimed by a variety of philosophers. (2) talks about proximate causes and does so in a way that seems to assume that the proximate cause of a conscious state is a brain state which immediately precedes it. Whether or not the conscious state itself is physical, it is the last event in a series, the rest of which is a causal process leading up to it. The last event in this preceding process is the conscious event’s proximate cause. Various opponents of the causal/hallucination argument challenge this picture. They claim that in the case of genuine perception the ‘experience’ itself is not something which is produced at the end of the causal process—they agree that to speak in this way plays into the hands of the argument—but a genuine perception is something that depends on, supervenes on, is caused by or is constituted by (different philosophers put it different ways) the process from distal cause to brain state as a whole. This view is adopted for purposes of argument by John Foster (2000) and Paul Snowdon (2005), and more categorically by Mark Johnston (2004) and Michael Martin (2004, 2006). Johnston, for example, says: Seeing the object is not the next event after the visual system operates. Seeing the object is an event materially constituted by the long visual process connecting the object seen to the final state of the visual system . . . There is no . . . ‘last’ brain state that then causes the seeing. (Johnston, 2004: 139)

Something like this is necessary for naive realism, because that theory requires that the external object—the distal cause of the process—be a constituent of the fundamental perceptual state. On this view, the situation is as follows. The physical process of perception can be divided into three constituents. There is the distal object of perception, A; there is the causal process, B, linking that to the subject’s visual system, and there is the visual system/neural state, C, activated at, or towards, the end of this process. For the naive realist and disjunctivist, A, B, and C are individually necessary and collectively sufficient to constitute, or sustain, a direct perceptual experience of A. It is an assumption of the causal argument, however, that C alone is both necessary and sufficient for the production of a veridical-seeming hallucination indiscriminable from the perception that relies on A, B, and C together. The naïve realist seems to be committed to saying that C alone is not sufficient to produce the experience on its own, for if it were it would be the case that A and B were not constitutive of experience in the case of perception, but had a merely causal role. The only alternative would be to say that, although all three factors were essential to constitute an experience with that phenomenology in the perceptual case, C could do it alone in the case of an hallucination. So when stimulated under artificial, non-perceptual circumstances, C does something it does not do when activated in the normal perceptual process. C is then doing something that it does not do in perception. This latter contradicts what I shall call the non-arbitrariness of hallucinations.

 - 

19

I argued against such arbitrariness in Robinson (1994) in the following way: Given that [it is admitted] that hallucinations could be produced by stimulating just those brain states involved in perception, how are we to make sense of why this should be so if it is not by thinking of [such] hallucinations as cases in which the state is activated and performs its normal function – that is, has its normal causal upshot – in an abnormal context? Otherwise the production of hallucinations in this way would seem to cast the brain in a role something like that of a Cartesian demon, producing an effect specialised solely to the context of deception. (Robinson, 1994: 157)

This might be called the Non-arbitrariness (or the Non-ad-hocness) theory of philosophers’ hallucinations: NA for short.⁶ NA is important to the argument for at least two reasons. First, it gives support to (2) and, second, it brings out the fact that the direct realist ought not to accept (1) and therefore not to believe in philosophers’ hallucinations. This latter is, I think, embarrassing for the direct realist, who usually wants to combine his perceptual realism with a standardly ‘scientific’ view of the causal powers of the brain. We shall see below that this is difficult. NA supports (2) because it gives support to the application of the ‘same proximate cause, same immediate effect’ principle to this case. What NA says is that, if it does not apply here—if the relevant brain state is not doing the same thing in both cases—then there should not be any philosophers’ hallucinations. So if you believe in such hallucinations and in NA, then you should accept (2). Second, NA is important because, I claim, unless you accept NA, it is unreasonable to believe in (1) and philosophers’ hallucinations. Why should you believe that you can or could cause a veridical-seeming hallucination by directly activating the neural processes involved in the corresponding perception, except because you think that you would be managing to do ‘by the back door’, as it were, what nature does for and to one in the ordinary case of perception? Now if we accept NA, and agree that C is just doing its normal thing under abnormal circumstances in the case of the hallucination, then what it is

⁶ Keith Allen (2015) rejects this argument. He says: However, this objection depends upon assumptions about causation—and in particular the principle same proximate cause, same immediate effect—that, we have already seen, the disjunctivist is likely to reject. His own theory is The subject’s brain state provides a way of selecting between different absent things, such that the ensuing psychological state nevertheless depends both on the subject’s neurophysiological state and the absence of the object. (Allen, 2015: 300) This theory does seem to involve a bizarre arbitrariness. As Allen acknowledges, even Martin dislikes the apparent causal influence of such absences.

20

    -

contributing to, and in, perception itself must carry with it all the subjective baggage necessary to constitute a hallucination. It is difficult to see how this leaves any room for a direct realist grasp of the external object in the case of perception. What is more, any such further element would be redundant. Suppose that factors A and B were, as the traditional representative or indirect realist says, simply the causes of C and a subjective factor common to hallucination and perception, what would the subject be losing, either phenomenologically or epistemologically by deeming the roles of A and B to be purely causal and external to the conscious state that arises solely from C? It would seem exactly the same to him as the direct realist thinks it should seem, and he would be reacting in just the same way to all the same external stimuli. It looks as if the naive realist must do one of two things. He must either deny NA, saying that, in perception, C does not produce something as full-blooded as what it produces in hallucination, and this I have already argued is bizarre: or he must deny (1) and the natural possibility of philosophers’ hallucinations, thus making NA inapplicable. One possible way of taking this route would be to say that C alone, in contributing what it contributes to a perception, was phenomenologically only like an imagistic conceptualization of the experience. Indeed, the idea that hallucinations involve the neural mechanisms associated with imagery, rather than those associated with perception has been a common speculation. It seems, however, that at least some of the most strikingly veridical hallucinatory syndromes involve perceptual mechanisms, not imagistic ones. Hallucinations that are associated with the Charles Bonnet Syndrome (CBS) are very realistic, unlike most other actual hallucinations. Dominic ffytche summarizes the facts as follows: The [CBS] brain activity challenges our current view of hallucinations. It reveals similarities between the substrates of hallucinations, illusions and afterimages, but different from that of imagery, a finding seemingly inconsistent with current philosophical and clinical classificatory schemes. The activity also casts doubt on the assumption that hallucinations are reactivated memories . . . (ffytche in Macpherson and Platchias, 2013: 47)

If this is so it seems that C is generally sufficient to produce a state subjectively indistinguishable from a perceptual experience. It seems to me that, in general, naive realists and disjunctivists show little or no explicit awareness of NA and its rationale, but the various theories they produce are in fact attempts to comply with it. They all try to show how there could be a state which, both avoids subjective phenomenal content yet which is indiscriminable from the subject’s point of view from a real perception, (which does of course, possess phenomenal content, but content which is attributed to the external object). The examples of this approach which I shall discuss are provided by

 - 

21

Michael Martin, William Fish, and Matthew Soteriou. But before discussing these contemporary approaches I think that it is right to draw attention to another strategy, with which I have no sympathy, but which I think should be mentioned because it draws on a line of thought that has been influential, perhaps in a rather ‘underground’ way, in some Wittgensteinian approaches to mind. The argument presented at the start of this chapter, and NA, rest on the assumption that what the relevant brain states produce when stimulated is a straightforward matter of fact, and, hence, so is the question of whether they produce the same kind of state when stimulated in normal perception and in philosophers’ hallucinations. An alternative view might be that there is no such fact, independent of interpretation: whether what is produced is a direct perceptual relation or a subjective state is a matter of our (perhaps a priori necessary) conceptual framework, not of some empirical fact about the workings of the brain. Disjunctivism would then be a conceptual truth, untouchable by science or scientific speculation. This approach is reminiscent of interpretational accounts of propositional attitude states, except that it seems to me that whatever plausibility interpretationalism has for such states, it entirely lacks in the case of core sensory content. The contents of my thoughts may be a product of interpretation and social construction, but the fundamental nature of the sensations produced when the sensory parts of the brain are stimulated is not. Or so I shall continue to assume. Nevertheless, when Wittgenstein says that ‘My attitude towards him is an attitude towards a soul. I am not of the opinion that he has a soul’ (Philosophical Investigations II, iv) he was including in that thought all the contents of consciousness, not just ‘propositional attitudes’. I see no plausibility in this approach, but it seems to me that there is a tendency for a certain kind of anti-Cartesian thought to fall back to this position, when convenient, so it should be noticed. In the context of the current debate (and who knows when earlier views will be resurrected), I think that most direct realist disjunctivists would definitely affirm that the direct contact with external objects which they claim is the essence of perception is a solid matter of fact and not—or, at least, not only—a reflection of the structure of our concepts. Admittedly this may be less clear when the direct realists are Wittgensteinians, for sometimes in this case the line between a common sense realism and linguistic idealism is not always obvious. Some of what we have said so far suggests that a disjunctivist might deny (1), but, in fact, (1) seems rarely, if ever, to be straightforwardly denied. For the most part, the target is (2), but Martin also denies (3); indeed that denial is the fundamental point of his account, and is what gives rise to the (possibility of) the views to which I have just alluded. Intentionalists believe, I think, that the argument can be avoided provided that the subjective state involved in perception and hallucination is intentional. They think themselves not to be disjunctivists: whether they take their theory to be a version of direct realism, or just to be a way of avoiding sense-data, is less clear, as we shall see in the next chapter.

22

    -

1.5 Disjunctivist Accounts of Hallucination: Introductory Remarks I responded in Robinson 1985 and 1994 to the attempts of Hinton (1973) and Pitcher (1970) to show the inapplicability of the ‘same proximate cause’ principle to this case and I am aware of no direct replies to what I said specifically to these points. Since those early days, I can identify two other major disjunctivist approaches. John McDowell (1986/1998), John Campbell (2002) and William Child (1994), for example, although by no means in complete agreement, form a group which deploys what is often called the epistemological argument for disjunctivism.⁷ The common core to this argument is that, without direct realism—and, hence, given the possibility of ‘philosophers’ hallucinations’, without disjunctivism—it becomes a mystery how we could have a grasp on the physical world. This particular strategy for defending direct realism will be discussed in Chapter 6. The accounts which will concern us here are ones which try to explain what kind of contents a relationist can give to hallucinations, given that the external objects that standardly constitute perceptual contents are missing. The philosophers whose theories I shall be considering are Michael Martin, William Fish, and Matthew Soteriou.

1.6 Disjunctivist Accounts of Hallucination; (i) Martin’s ‘Indiscriminability’ Account The disjunctive strategy that has drawn most attention is probably Michael Martin’s. As the discussion above of NA showed, the problem for disjunctivism is how to allow a ‘full-blooded’ experiential nature to philosophers’ hallucinations without denying NA. One strategy, I suggested, would be to allow a common factor which fell short of being a full-blooded experience, but which we somehow mistook for one. The view that in hallucinations we do not have experiences at all, we only believe that we do, would be an extreme version of this. Martin’s approach has at least some similarities with this. With veridical-seeming hallucination in mind, Martin suggests that At least when it comes to the mental characterization of hallucinatory experience, nothing more can be said than the relational and epistemological claim that it is indiscriminable from the [corresponding] perception. (2004: 72)

⁷ Campbell accuses McDowell of being an intentionalist, not a relationist. I do not think that this dispute or difference matters for present purposes.

 : (  )  ’ 

23

He exemplifies this principle elsewhere: For certain visual experiences as of a white picket fence, namely, causally matching hallucinations, there is no more to the phenomenal character of such experiences than that of being indiscriminable from corresponding visual perceptions of a white picket fence as what is. (2006: 369)

This implies that in normal perception there is ‘more’ to the phenomenal character, so in the hallucinations, a certain facet of ‘full blooded’ phenomenal character is missing, though we fail to notice this fact. This way of characterizing hallucinations might seem to be similar to J. J. C. Smart’s treatment of after-images. The man who reports a yellowish orange after-image does so in effect as follows: ‘What is going on in me is like what is going on in me when my eyes are open, the light is normal etc. etc. and there really is a yellowish-orange patch on the wall.’ (Smart, 1963: 94–5)

Both Smart and Martin are trying to free themselves of non-veridical phenomenal contents by characterizing these delusive experiences simply and solely in terms of their likeness to proper perceptions, thereby, they believe, putting the burden of carrying the ontological weight of the content onto the perceptual case. Smart, however, recognizes and welcomes a consequence of this that Martin disavows. Smart continues the above quotation: In this sentence the word ‘like’ is meant to be used in such a way that something can be like itself . . . With this sense of ‘like’ the above formula will do for a report that one is having a veridical sense datum too. Notice that the italicised words ‘what is going on in me is like what is going on in me when . . .’ are topic neutral.

The consequence that Smart welcomes is that, if our introspective knowledge of the subjective character of non-veridical experiences is topic neutral, then our knowledge of the subjective character of genuine perceptions, which are subjectively like them, must also be topic neutral. And Martin’s account of our introspective knowledge of the subjective character of hallucinations is topic neutral, for we know nothing about it except that it is indistinguishable from something else, namely the appropriate genuine perception. Martin, however, does not want to generalize this topic neutrality to our grasp on perceptual experiences. He is a naïve realist, and what it is like for the subject to perceive an external object is a function of the accessible sensible features of the object he perceives. This creates an asymmetry between veridical and hallucinatory experiences of a kind that

24

    -

Smart avoids. The issue is whether this asymmetry makes phenomenological sense. Here is an argument to the conclusion that it does not. Martin, I think, is committed to the following premise. (1) A subject, S, recognizes a veridical perception, x, as being phenomenologically the experience it is, from the sensible qualities that the object perceived presents, and appears to present, to him in that experience. I think that he is committed to this, both because (1) seems common-sense: how else can you recognize how something appears (‘being phenomenologically the experience it is’) except from the sensible features it seems to have? And because the example of the real lemon and the lemon-shaped soap that he gives, and we see below, is obviously based on the idea that the fact that they both present a yellow and lemon-shaped experience is what makes them seem similar.⁸ The following seems to me to be a necessary truth: (2) If someone, S, recognizes some object, state, event, etc., x, by its exhibiting or seeming to exhibit some feature F, then (i) anything y that seems to S indiscriminable from x must also seem to S to present F, and (ii) it is because it seems to present just that feature that it is indiscriminable by S from x. From these it follows that (3) If a seeming veridical hallucination, x, had by S is indiscriminable by S from some veridical perception, y, that S recognizes as being a perception of some object that presents and seems to present sensible quality Q, then (i) the veridicalseeming hallucination x must seem to be of an object that presents Q, and (ii) it is because it seems to present an object with that feature that it is indiscriminable by S from y. If the argument just presented is correct, then indiscriminability cannot be a primitive notion, when one of the indiscriminable things is recognized by some feature, but must then rest on indiscriminability in respect of that feature.⁹ This is what (2) asserts. Martin recognizes this principle when the indiscriminable things are both real physical objects—his example is a real lemon and a very convincing soap lemon (2006: 386). But he denies that it applies when one of the experiences ⁸ One of the publisher’s readers has pointed out that I show, in the next chapter, that Martin deals with the case of objects looking other than they are—for example a white object looking blue—by saying that it strikes one as if it were blue, so that the resemblance to the experience of a genuinely blue object to a non-blue one looking blue does not involve the presence of a sensible quality. But I think that, just as the characterization of the ‘similarity’ of hallucinatory experiences to veridical ones is to be explained in a way that does not apply to the similarities between genuine perceptions, something similar applies to ‘illusions’. It is only in the non-veridical cases that something other than recognition of genuine sensible similarity is to be denied. ⁹ This contrasts with the case often invoked by David Armstrong, of chicken sexing, where the sexers recognize the sex of the chickens without knowing what sensible feature they found their judgement upon. They never recognize a feature as that by which they do it. Veridical perceivers in normal cases, on Martin’s account, recognize quite clearly what sensible features they are judging about, yet the judgements of similarity in non-veridical cases are not founded.

 : (  )  ’ 

25

is hallucinatory. This seems to me to be totally implausible. First, I think we can assume that he does not deny (2) (i) and (3) (i): that is, he cannot deny that if two things are indiscriminable they must seem to present the same features and that this must apply to the hallucination—it could not be indiscriminable from the perception of the white fence if it did not seem to be of a white fence. So the pressure must come on clause (ii): it must be the case that a hallucination’s seeming to be of a white fence is not to be analysed as its possessing an intentional or phenomenal object of a ‘white fence’ kind, but that its having that object is to be analysed in terms of its indiscriminability from a perception of such an object. Martin has a very narrow line to tread. On the one hand, his characterization of the similarity of hallucination and perception is entirely negative. On the other, the negative aspect of the discrimination does not rest on ignorance: it is not like my inability to discriminate the object in front of me from one the other side of the wall, which I cannot see and of the properties of which I have no idea. Martin tries to allow for this distinction by saying that the indiscriminability is ‘impersonal’—that is, it does not rest on some incapacity of the subject, rather on the fact that the experiences are indiscriminable per se (2002: 74). But it is not just that they are impersonally indiscriminable, one can also say in what respect they are indiscriminable—for example, in respect of seeming to be of a white fence. This is a positive feature: how is it to be accommodated? It is not, for Martin, because (as I believe to be a consequence for Smart) all sensible qualities are known only topic neutrally. The only other option this seems to leave Martin is to claim that the hallucinatory subject makes a swift unconscious inference of the form ‘this experience is just like one I would have seeing a white fence, so it must be an experience of seeming to see a white fence’. This seems very bizarre, even if intelligible, and topsy-turvy. I say ‘even if intelligible’ because I do not see how something could strike one as experientially similar to another experience, when the latter experience has a clear object, except by explicitly striking one as having an exactly similar object. Only if Smart is right, and all experiences, non-veridical and veridical, are introspectively neutral and free of positive character, could indiscriminability alone be the primitive feature.¹⁰ Indeed, it is hard to see how Martin can avoid falling into Smart’s reductive, topic neutral theory of all experience. If simple indiscriminability is enough to constitute the phenomenology of hallucination, how could it fail to do the same job for a perception indiscriminable from the hallucination? Smart’s claim that phenomenology consists in recognizing similarities and dissimilarities, without any further grasp on what those similarities consist in, would seem to be implied by Martin’s account of hallucinations. ¹⁰ The problem with Smart’s theory is that it involves a radically externalist account of sensory content. All experiences are identified simply as ‘what I get when facing objects of type F’, but one cannot individuate experiences indirectly in this way because one only knows what objects are in the environment on the basis of what the experiences one is having are like. I argue this at length in Robinson (1994: 136ff; 2016: 8ff).

26

    -

There is a strategy that Martin uses in ‘The Reality of Appearances’ (1997) which cannot plausibly be deployed here. In the article Martin says that a distinction must be made between what are the contents of our experience and what we can discriminate, because of the following two cases. First, one may not be able to discriminate colour sample A from sample B, nor B from C, though one can discriminate A from C. This shows, it can be argued, that A and B are different contents, but one cannot discriminate that difference. Second, with identical twins, one perhaps cannot discriminate between them, but they are still different contents of experience. One could dispute the claims that there are different contents in both these cases, but even conceding that, the disanalogy between these cases and hallucinations is so radical as to make the comparison irrelevant. In both these cases, the fact that we cannot discriminate between them result from the almost exact, or exact, similarity of the phenomena qualitatively. The failure of discrimination between seeing a pink elephant (which I hallucinate) and seeing nothing (which is what, in a sense, is going on) is not a matter of the minute difference between the two things: in no way does the phenomenal difference just outrun my discriminatory capacity. I conclude that Martin’s attempt to avoid all kinds of reification of the objects of hallucinations by treating indiscriminability as a primitive, fails.

1.7 Disjunctivist Accounts of Hallucination; (ii) Fish’s ‘Belief’ Account Fish’s ‘belief ’ account of hallucination faces problems similar to Martin’s. His definition of it is as follows: For all mental events, e, in doxastic setting D and with cognitive effects C (in its subject), e is a pure hallucination of an F, if and only if • E lacks phenomenal character, and • There is some possible veridical visual experience of an F, V, that has a rational subject who is in D, and produces C, and • C is non-empty . . . In plain English, it states that, for a mental event e, to be a pure hallucination, it must lack phenomenal character yet produce the same cognitive effects that a veridical experience, V, would have produced in a rational subject with the same background beliefs, desires, and other mental states (i.e., in the same ‘doxastic background’) as the hallucinatory. (Fish, 2009: 94–5)

One problem with this definition is that, as stated, it has no positive account of hallucination itself. It says that it lacks phenomenal character and has the same

 : (  )  ’ 

27

cognitive effects as a perception from which it cannot be distinguished, but what this thing is that gives rise to these cognitive effects has no positive account. I suspect that the phenomenology of hallucination—what it is like to have one— may be being taken for granted, but that is what is supposedly being analysed. It seems that having the hallucination consists solely in being in the cognitive state in question. This gives rise to what Martin and Fish call the screening off problem. If the minimalist account either of them provide of hallucination is adequate as an account of a phenomenology indistinguishable from real perception, why is it not an adequate account of perceptual experience itself? This is the same problem as I have just raised for Martin: if our knowledge of the contents of hallucinations are topic neutral, why cannot this be extended to perceptual experiences themselves? The answer they give is that hallucinatory states are essentially derivative from the case of real perception—they are, so to speak, imitating it. Fish cites an analogy used by Martin. The property of being an unattended bag in an airport will cause a security alert despite there being nothing intrinsically threatening about an unattended bag. So why does this property cause a security alert? Because of the relation it bears to the property of being a bomb in an airport. Whatever explanatory potential the property of being an unattended bag in an airport has in explaining why there is a security alert in inherited from the property of being a bomb in an airport due to the relation the two properties bear to one another. (Fish, 2009: 89)

He then quotes Martin’s conclusion: . . . cases of inherited or dependent explanatory potential offer us exceptions to the general model of common properties screening off special ones. (Martin, 2004, 70; Fish, 2009: 89)

The analogy seems to me to be a misleading one. Some unattended bags have contained bombs, so one infers that this one might. One does not seem to hear an explosion. The bag case is a simple case of a (very cautionary) induction. The analogy is meant to be, I suppose, that if some bags left in airports had not had bombs in them, then no-one would think an unattended bag had a bomb: and, similarly, if there had not been real perceptions, the artificial stimulation of the system would not seem to the subject to be like a perception. On this view, a brain in a vat would not seem to have experiences, for she had no original perceptions. And presumably Jackson’s Mary could not have been given chromatic colour experiences by stimulating her brain. If they would, then the screening off argument would work. The need to deny that they would have experiences with

28

    -

real phenomenal content shows how non-naturalistic naïve realism is. If Mary were having an artificially stimulated state, she would not be aware of any sensible qualities, but if an appropriate external object took over the causing of that internal state, she would be, though the external object made no difference to what was going on inside her! Fish says that what is activated in hallucination are only higher order belief states, but if everything remotely relevant in the brain is activated, it would follow that the first order states are wholly external. In fact, if naïve realism were correct one would expect the hallucinatory case, with only the cognitive system functioning, to be rather like blindsight: the subject would be disposed to say ‘I don’t seem to see anything but I have an irresistible inclination to believe that there is (e.g.) a pink elephant in front of me.’ The rationale of my point here is that, if it is the external object that constitutes the phenomenal content, and causes the cognitive up-take, then the absence of the object should provoke the response ‘I’ve got these beliefs/dispositions to believe, but I can’t tell why.’ If it is not like this superblindsight, then I think that the claim that there is no phenomenology becomes empty. What is the difference between believing that there are subjective phenomena and there really being such? Suppose that S has a partial hallucination, hallucinating a large opaque red patch in the middle of his visual field, but perceiving normally in the periphery around it. The belief content ‘without real phenomenology’ still blocks access to the features of the world S should be seeing. If this is the nature of the state produced by relevant brain activity which is common to perception and hallucination, then its seeming content will play the same ‘veiling’ role when it is generated by perception as it does in the case of hallucination. There is no difference between real, private subjective phenomena and the subjective contents of a belief that one is undergoing such phenomena. (If you soberly and reflectively think you are in intense pain, you are.) So this tactic does not avoid ‘screening off ’. The conclusion is that neither Martin nor Fish can give an account of philosophers’ hallucinations that makes them both subjectively indistinguishable from the corresponding perception, but, at the same time, avoids the screening off problem: if one cannot avoid the latter, the account of hallucination given would be equivalent to the common factor account that disjunctivism is designed to avoid.

1.8 Disjunctivist Accounts of Hallucination; (iii) Soteriou and ‘Seeming to Experience’ Matthew Soteriou defends disjunctivism in a discussion that is subtle but which seems to me to be ultimately not very illuminating. Like Martin and Fish, he

 : (  )  ’ 

29

denies that hallucinatory experience is an experience with genuine phenomenal qualities—it only seems to be an experience. If a subject is hallucinating, then it is not just her epistemic position with respect to her environment that is compromised. Her epistemic position with respect to her own experience is also compromised; for given that she isn’t perceptually aware of anything, and so given that she doesn’t have the kind of perceptual access to the world that she seems to have, she doesn’t have the kind of introspective access to her experience that she seems to have. (Soteriou, 2013: 204)

One would naturally wonder how it could be that seeming to have an experience, for example, of a bright red patch differs from actually having an experience qualified by red, however one wants to characterize that (perhaps as ‘sensing redly’). Soteriou insists that a ‘real’ experience is a genuine perceptual one and so an hallucinatory experience isn’t a genuine experience. There is not necessarily anything wrong with that as a stipulation about how one intends to use ‘experience’, but that does not give an account of what is going on when one hallucinates. Both Martin and Fish have a go at this, the former in terms of ungrounded indistinguishability, the latter in terms of merely believing that one is having an experience with phenomenal character when one is not. Soteriou does not seem to opt directly for either of these. He only insists on the derivative nature of hallucination. In the case of hallucination, the conscious character of the conscious sensory experience one undergoes needs to be characterized in derivative terms, and this is because the kind of introspective access to one’s experience that one seems to have needs to be characterized in derivative terms; and this, in turn, is because the kind of intentional behaviour one engages in when one introspects one’s experience needs to be characterized in derivative terms. What one can know of the conscious character of the conscious sensory experience one undergoes when one hallucinates is that it introspectively seems to one to have the conscious character of a genuine perceptual awareness of some entity or entities. (Soteriou, 2013: 204)

This can be summed up, I think, in the following form. 1. Visual experiences are characterized in terms of what external objects one takes oneself to be seeing. 2. Hallucinatory experiences do not have the same phenomenal character as perceptions—they only seem to us to have it

30

    -

and 3.

Hallucinations are essentially derivative from real perceptions.

The crucial question is how one should understand and defend (3). If it is understood as a straightforward fact about hallucinations, presumably it must mean or at least imply that one could not have an hallucination with a certain character unless one had had a perceptual experience of at least a roughly similar character. But this would not, on its own, undermine the causal hallucinatory argument. It would imply that once one had had a sufficient number of experiences of a roughly general kind, the brain was trained or primed to produce the kind of data needed for an hallucination. But then, according to the non-arbitrary principle, it would be producing that data under normal perceptual circumstances. The result would be that, though perception might be direct or naïve initially, as soon as the brain had been conditioned, it would be direct no longer. The suspicion must be that the priority of perception over hallucination is a conceptual point, not a straightforwardly empirical one, but this seems to push us to the ‘Wittgensteinian’ position I discussed in $3, namely that the ontological status of sensations is a matter of conceptual interpretation, not of fact. This seems absolutely implausible.

1.9 Naïve Realism and Philosophers’ Hallucinations: Conclusion The naïve realist has three options. First he could simply deny that stimulation of the appropriate internal state would produce a state subjectively indistinguishable from the corresponding perception. This seems to go against what evidence there is, as presented, for example by ffitch (2013).¹¹ ¹¹ Keith Allen (2015) supports the ‘imagination’ view of hallucinations and responds to the empirical evidence as follows: The case for treating Charles Bonnet Syndrome hallucinations as kinds of perceptual experience depends upon identifying mental states and events in virtue of the neurophysiological mechanisms that produce them. But whether mental phenomena are natural kinds that ought to be classified in this way is controversial. At the very least, classifying mental phenomena by appealing to underlying neurophysiological mechanisms is likely to have a number of revisionary consequences for our common-sense classificatory practices. For instance, the mental events currently classified as hallucinations will form a disjoint set if, as seems likely, it turns out that their causal aetiologies differ: perhaps some hallucinations (such as schizophrenic hallucinations) will turn out to be degenerate kinds of imagination, whereas other hallucinations (such as Charles Bonnett hallucinations) will turn out to be degenerate kinds of perception. It also seems likely that this taxonomic approach will militate against the view that dreaming is a form of imagination and, more controversially, collapse the intuitive distinction between hallucination and illusion [ffytche, 2013]. There may be some perspectives from which this way of taxonomizing mental phenomena is advantageous. For example, from a clinical perspective where diagnosis and treatment is of

̈     

31

Second he could deny that, in the case of hallucination, the proximal brain processes have the same immediate effect as they do in perception, which denies the non-arbitrariness of hallucination. Third they can try to devise a common factor that the brain process produces that meets both the requirement that the subject cannot tell it from a real perception, but which in fact falls short in a way that leaves a role for the external object featuring in veridical perception to constitute the actual content of the experience in a way that meets naïve realist standards. I have examined all the attempts to do this of which I am aware and argued that none are remotely plausible.

crucial importance, it is necessary to be clear about whether hallucinations are the result of specifically psychological problems: because hallucinations are often considered to be a sign of madness, there is a tendency to misdiagnose, and therefore mistreat, subjects with Charles Bonnet Syndrome in particular [Menon et al., 2003: 59–60]. However, the clinical perspective is not the only perspective, and here, as elsewhere, accepting a theory that is revisionary of our common-sense classificatory scheme is a theoretical cost. Besides, this way of taxonomizing mental phenomena is one that naıve realists, and disjunctivists more generally, are already likely to reject. (Allen, 2015: 296–7) I do not see much force in these arguments. Why should all hallucinations be one natural kind? An hallucination is more like a symptom or syndrome rather than a specific disease. And the idea that naïve realists will not like the idea does not seem a powerful consideration!

2 Naïve Realism and the Argument from Illusion 2.1 Introduction What is generally called ‘the argument from illusion’ does not mainly concern cases that would popularly be called ‘illusions’, rather it concerns the quite banal fact that objects look different from different points of view, or under different circumstances, and to people with different powers of perception (e.g., being shortsighted), although the object itself in no way changes under these circumstances. So if an object is seen, for example, from different perspectives, or different distances, or without one’s glasses, the content of the experience—‘how it seems to you’—changes, but the object itself has not changed at all. The challenge that this is taken to pose for the naïve realist is for him to explain how appearances can change, when the object itself does not, if the content of the experience is constituted by some feature of the object itself. The argument can be more formally expressed as follows: 1. In some/many/most/all cases of perception, we are aware of something that possesses different sensible qualities (however slightly) from those possessed by the physical object we take ourselves to be perceiving. 2. By Leibniz’s Law, if two things possess different qualities, they cannot be the same thing. Therefore 3. That of which we aware in the cases cited is something other than the object purportedly perceived.

In case one is not persuaded that in every case things appear at least very slightly different from the way they intrinsically are, one can add a further premise to the effect that one does not want a different account of the objects of perception in the case when one has not got one’s glasses on from when one has: that is, the continuum between supposedly perfectly veridical cases and slightly deviant cases requires that they both be treated as falling under the same account of perception.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0003



33

In Perception (1994) I said that the soundness of the argument from illusion depended on something I dubbed ‘the Phenomenal Principle’ (PP). (PP) If there sensibly appears to a subject to be something which possesses a particular sensible quality then there is something of which the subject is aware which does possess that sensible quality.¹

There is nothing original about (PP), but it has generally been seen as an accurate expression of the rationale behind the argument from illusion. This principle, if correct, would entail that one was aware of something other than the external object in cases of ‘illusion’, then the argument is generalized to perception in general by claiming that objects rarely, or never, look exactly how they really are, and so more or less everything is a case of illusion. The standard way of rejecting (PP) is to say that if something looks, seems or appears, for example, red, it does not follow that there is anything red in the offing, because these idioms all take an intentional object: it does not follow from something’s seeming red that there is anything red, anymore than it follows from someone’s believing in fairies it follows that there are any fairies. The tussle is between those who emphasize the intentionality of the idiom and those who insist on what seems to be the blatant nature of the phenomenon—of course when I seem to see something red I am actually aware of red.² It is now clearer to me that the issue is not that simple, and this greater clarity has come about because opponents of the sense-datum theory have become clearly divided between two theories, namely the relational and the intentional (otherwise known as representational) theories. Relationists, who also admit to the label naïve realists, do not want to allow that there is in perception any internal state that represents the external world. So the situation is that (PP) is, indeed, necessary if the argument from illusion is to establish the existence of sense-data, but is it necessary for the argument to establish the falsehood of the naïve realism that relationism avows? It would not follow from the fact that it is not a red quality-instance or quality-presentation that we are aware of when a white object seems red, that one instead is aware of the white object itself. For that conclusion one would need the further premise that one is either aware of a red sense-datum or of the white object—that there is no ¹ I said in the introduction that it was not an essential part of what I mean by ‘sense-datum’ that it is a particular. What matters is that the sensible quality in question is genuinely presented or present, whether this is thought of as an instance of the quality, or that the sensible quality qua quality is presented, as the term quale suggests. If one takes the latter route, in PP the words ‘—then there is something of which the subject is aware which does possess that sensible quality’ would be replaced by ‘—then the quality of which the subject is aware is genuinely present and presented’. For convenience, I shall stay with the original wording. ² In Robinson (1994: 38–40) I quote at length G. E. Moore’s struggle with himself over whether the idiom ‘seems F’ could replace the idea that one is actually aware of something F in cases of illusion.

34

¨      

other option. What is the situation if one’s awareness consists in being in a representational state, rather than being in a direct relation to something external? The opponent of a genuinely direct realism might be satisfied with a modified version of the Phenomenal Principle, as follows: (mod. PP) If there sensibly appears to a subject to be something which possesses a particular sensible quality then there is a subjective content of experience (that is, a representation) which is of the quality in question. By ‘subjective’ is signified the idea that nothing salient need exist outside the subject’s mental state for that content to occur.

We will find when we discuss intentional or representational theories in the next two chapters that there is disagreement about whether one can describe the subject as being aware of the content of a representational state and whether this makes a significant difference. For present purposes, however, what matters is that relationists want to accommodate or explain illusion without relying upon intentional idioms, whereas representationalists are very happy to deploy them, as they are an essential part of their theory. Paul Snowdon, in private communication, has argued, in effect, that the ‘looks’ idiom is a piece of perfectly ordinary language and does not carry theoretical baggage, as seems to be implied by using jargon such as ‘representationalism’ and ‘intentionalism’. It is true that talking of something looking—or seeming, or appearing—red, either when it is, or is not actually red, is very normal talk. But it seems to me to be unquestionable that (a) that such verbs are intentional, in the simple grammatical sense that they can be used truly whether or not anything red is present; and that (b) red, or redness, in some way figures in the perceptual experience, in that seeing something looking red differs, in some way, from seeing something looking green, or from seeing nothing, and ‘red’ enters into that difference. It is also true that (unless one accepts a pure ‘belief ’ theory of perception) how it enters in is different from how it does so in the case of beliefs, thoughts, and other propositional attitudes, which are not sensorily experiential as seeing is. It is this fact which leads to its presence in seeing being labelled ‘phenomenal’, in contrast to its conceptual involvement in the other cases. Dubbing the conceptual involvement as ‘representation’ seems to me to be fairly neutral, and the acceptance that, in the perceptual case, it is not merely representational, but presentational, as the features seems actually present and not merely an object of thought, also seems to me not to be unduly theoretical. The importance of the difference between ‘representation’ and ‘presentation’ will become salient in our discussion of intentional/representational theories in the following chapters. Snowdon’s very Austinian approach attempts to avoid the need to give anything that could be thought of as a philosophical explanation or analysis of

  

35

illusion, no doubt suspecting that, once one gets drawn down that path and leaves behind the sufficiency of ordinary language, one will become trapped in one of the theories that the naïve realist hopes to avoid. Other naïve realists, however, have allowed themselves to be ensnared into theoretical explanation, and there are a variety of ways that opponents of the ‘argument from illusion’ have sought to explain the phenomena in question. What they have in common is that they all deny that that the apparent quality that appears in the ‘illusion’ is a straightforward presentation of that quality. The approach which we will consider first, because it follows most directly from our discussion in the previous chapter, is the attempt to apply the disjunctivist strategy to ‘illusions’. According to this approach the apparent quality is replaced by the relation of bare indiscriminability. Second, I will look at what has been called the ‘austere’ approach, which asserts that, in the case of illusions, it is still the actual quality of the object that is presented, even though that is, seemingly, not how it looks. Third is the ‘perspectival’ approach, according to which there are special relational qualities, such as ‘red-from-here’ or ‘rectangular-from-there’, in addition to the monadic, qualities that the object possesses intrinsically.

2.2 Disjunctivism and Illusion The discussion of disjunctivism in the previous chapter concerned hallucination, and, indeed, all the detailed discussion by disjunctivists has seemed to concern itself with hallucination. But, as Martin points out, it must also cover ‘illusion’— things looking other than they really are. According to naïve realism, the actual objects of perception, the external things such as trees, tables and rainbows, which one can perceive, and the properties which they can manifest to one when perceived, partly constitute one’s experience. This talk of constitution and determination should be taken literally; and a consequence of it is that one could not be having the very experience one has, were the objects perceived not to exist, or were they to lack the features they were perceived to have. [Italics added.] (2008: 93)

Martin clearly thinks that illusion, like hallucination, falls under the disjunctive account. But this is not an arbitrary choice. It seems to be forced on the naïve realist because he cannot say that a simple extensional relation exists between the object and its properties in a case where the object is not as it is perceived or seems to be. Maybe the naïve realist can use a different strategy for these cases than for hallucinations, but Martin provides none. He also makes the following significant concession.

36

¨       I bracket here a concern whether there are in fact any actual examples of veridical perception. (93)

This is a strange remark, because the view that the number of actual cases of true veridical perception might be vanishingly small, or zero, is one deployed by the opponent of direct realism, when using the argument from illusion. Martin seems to think that veridical perception is the paradigm against which all perception must be measured or defined, even if it has no instances. This would seem to mean that, for example, a white object’s looking blue to a subject is the experience being simply indistinguishable from seeing something blue, without involving any blue phenomenal content, and even if there are no cases in which there is genuine phenomenal content. Perhaps we should take Martin as assuming that there are in fact cases— perhaps very many cases—of veridical perception, nevertheless this forces the relationist to make clear his approach to the argument from illusion. On the traditional account, either all, or virtually all, cases of perception involve perceiving things as being, at least ever so slightly, different from the way they actually are. This can be because of the effects of perspective, the unreality of secondary qualities, the differences between the perceptual acuteness of even normal perceivers, for example. If these have to be assimilated to the category of ‘being illuded that’, then, like the sense-datum theorist, the disjunctivists will be assimilating the contents of all, or almost all, perception to the same category as hallucinations. And this defeats the whole purpose of naïve realism. The relationist must, therefore, find a way of including almost everything we think of as ordinary perception within the ‘non-illuded’ category of his disjunction. If he cannot do this, then his attachment to naïve realism will be, at best, very attenuated. There is one obvious strategy that does not seem to be available to the relationists. The natural thing to say when trying to avoid the sense-datum response to the argument from illusion is that if something looks red, but isn’t, or looks square but isn’t, is that it seems red or seems square, but nothing in the situation actually is either of these things. This is the line explored by Moore when anguishing about whether, when it seems to him that there is something red, there might not actually be any red thing present. And, as I say above, it is a standard way of denying the so-called Phenomenal Principle—that if there seems to be a certain phenomenal quality of which one is aware, there must be such a quality present. But the ‘seems’ idiom avoids the reification of sense contents only because it is an intentional idiom and takes an intentional object. This puts it firmly in the intentional/representational camp. A relationist cannot, therefore, resort to it as a way of handling the argument from illusion, but must find some way of affirming the objective reality of the contents of perception, even in the cases traditionally classified as illusion. In this way, the development of the debate which has seen the opponents of the sense-datum theory divided into relationists and intentionalists

  

37

has made it much harder for direct realists to marshal their resources behind one banner. The standard Austinian line is to claim, as far as is possible, that objects are not looking other than they really are in the cases standardly classified as illusions. So objects really are coloured; a round object viewed from an angle does not look elliptical, it looks like a round object, as viewed from an angle; a stick in water does not look bent, it looks like a straight stick in water etc. Sometimes a causal explanation of how the object comes to look the way it does is given, as if this validates the appearance as veridical. For example, Fish argues: For example, to explain why the stick in water appears bent, we need only appeal to the natural physical phenomenon of light being refracted as it passes through materials of different refractive indices. In a sense, then, the explanation of the illusion . . . is complete by the time we get to the subject. Even if our perceptual processing of this visual information is unimpeachable, as often it is, we still suffer from an ‘illusion’ because of the particular way that things out there in the world affect the patterns of light incident upon our retina. (Fish, 2009: 148; cited in Soteriou, 2016: 185)

Even a determined direct realist like Snowdon can see that this argument is no good. The argument [from illusion] is sometimes met by the reply: ‘Nothing follows because we can easily explain what is going on in the case. There is nothing mysterious about that being the way things appear in such circumstances.’ The comment, probably true though it is, quite fails to engage with the options that any critical assessment of the argument must choose between. A critical reply is restricted to saying either it is invalid, or that one of the premises is false. (1992: 71)

So what one would need would be some reason to reject PP in both its forms. The reason why the causal ‘explanation’ is irrelevant is that the causal story tells us how the ‘illusory’ experience comes about, but it does not tell us anything about how to accommodate the phenomenological content of that experience into the naïve realist picture of the world. Indeed, the very fact that it is a causal story suggests that the phenomenology is the resultant of the causal process, not a feature out in the world, at the beginning of the process. No disjunctivist would suggest that an explanation of how hallucinations might be produced would thereby make them unproblematic for the direct realist, so how can such a genetic account work in the case of illusion? One might have expected a direct approach to this topic from Martin’s article ‘What’s in a look?’, but this is a strangely elusive work. He admits at the end of the article:

38

¨       The semantic account sketched is a very minimal one; on this, the main point of most of our talk about the looks of things and of how things look to us is to invite comparison between objects, or between psychological states that are what we take to be paradigm examples. So on this account, little can be read off about the nature of the psychological states that we are reporting on. And indeed, the point of the account here sketched is not to offer grounds for preferring one theory of sensory experience over another. The minimal semantics I have offered is quite consistent with a representational or intentional theory of sense experience and equally with a sense-datum approach . . . The aim [has been] rather to indicate reasons for why we should not look for evidence in favour of one of these views over any of the others in the way we talk about appearances. (223)

Martin’s most general view about ‘looks talk’ is that it is comparative, but he accepts that there is also a residual, essentially phenomenal component, even in vision. So the question is how to handle this component. So he says in the latter part of the essay he will . . . return to the most pressing problem: How should we understand correct uses of (19) (‘The stick looks bent’) where the object itself is not itself bent? (208)

The bent stick example stands for all the standard paradigms deployed in the argument from illusion. Martin concedes that what one is reporting is something psychological and subjective. One might point out that at least in some circumstances, an utterance of (19) behaves more like a psychological self-report than a description of the shape of the stick . . . So, one may suppose, this shows that (19) at least on a common interpretation comes to report the speaker’s psychological state, and is a subjective report, a report of the speaker’s perspective, and not a claim about any object in the environment. (209)

This, of course, leaves open the question of what one is doing when one reports a psychological state, especially if one wants or needs to avoid postulating the standard forms of phenomenal content, as a direct realist must in the cases of objects not looking quite as they really are. One go that Martin makes looks worryingly Wittgensteinian. In turn, if we understand (19) as somehow reporting a psychological fact, then it should be no surprise that our attitude to such an assertion is parallel with avowals of many other psychological states, such as first-person reports of belief, sensation, or desire. (209)

  

39

The standard use of the term ‘avowal’ treats it as a ‘criterion’ of a mental state, but not a report: it is a piece of linguistic behaviour that functions in the behavioural repertoire of language-using creatures as a sophisticated substitute for some more primitive behavioural response. This may not be Martin’s intent, but what follows is not reassuring. What now are we to make of uses of [(19)] that indicate some subjectivity and therefore seem close in impact to the use of (21) [‘The stick looks bent to me.’]? The answer lies, I suggest, in recognizing how in general there is a potential for subjectivity in the act of comparison and judging things to be similar.

He quotes with approval Evans: The relation ‘looks like’ [is] what we might call a secondary relation – on analogy with the secondary qualities of traditional philosophy, which hold in virtue of the effects they have upon human beings. According to this view, something will objectively look like something else if it strikes people as like that other thing; or, rather more usefully, b is objectively more like a than c only if b strikes people as more like a than c does. (Evans, 1982: 292) (215)

Martin seems happy to stick with the ‘strikes me’ idiom. The stick is similar to bent things simply with respect to how it strikes me, or the subjective bearing it has on me. So we can imagine that (19) could be uttered as a way of expressing how I find things similar, rather taken as directly reporting on ways that they are. (215)

This bare ‘striking’ looks very similar to his account of hallucinations, in the sense that it is as if one cannot avoid assimilating the cases, but there is no explicit phenomenal content on the basis of which one does this. Martin, however (in discussion), denies that this account is meant to be an extension of his treatment of hallucination. As we shall see below, Brewer does make this extension. Whether or not the ‘strikes’ and ‘finds’ account are meant to be equivalent to the ‘bare indistinguishability’ account, the argument I use above against the latter seems appropriate here: namely that the ‘striking’ and ‘finding’ are not intelligible except in terms of how it strikes one, and what one (thinks one) finds there, which can only be articulated in terms of the relevant phenomenal qualities. The intentionalist and the sense-datum theorists both have accounts of how these qualities figure: the naïve or direct realist seems to be committed to opacity.

40

¨      

2.3 French and Phillips’ ‘Austere’ Naïve Realism, and Why It Is Not as Austere as They Hope Putting aside those approaches to ‘illusion’ which seem to be close to disjunctivism, let us begin again with an approach that presents itself as the most defiant naïve realist response to the argument from illusion, namely the Simple, Austere Naïve Realism defended by Craig French and Ian Phillips. They characterize its components as follows: Simplicity: The character-constituting presented elements of ordinary veridical experience are just ordinary physical features of the mind-independent world: the objects we ordinarily take ourselves to see and their basic visible properties. Austerity: Illusions do not differ from veridical cases (as understood in Simplicity), neither in relational nature, nor in the kind of character-constituting presented elements to which they are relations. (..) ‘Simplicity’ is just a standard statement of naïve realism or relationalism, but ‘Austerity’ is radical, in that it claims that the very same properties as are presented in veridical perception, and only they, are what are presented in the case of illusion, even though they look different. The radical nature of this proposal consists in the fact that one might take looking different from F to be equivalent to having a different presentational character from F. Of course, it is agreed that the property ‘out there’ that is being physically presented does not change, but the presentational character of the experience, one might claim, must change if something looks different: that’s what ‘looking different’ is. In the light of this latter intuition, what positive account of ‘illusions’ do French and Phillips give? They take as examples of illusion two typical and uncontroversial cases, namely a red car’s looking orange under certain street lighting, and seeing a rectangular window as square because it is viewed from a certain angle. They deal with these cases as follows. Take Car Case. In day light, the car looks red to S; under sodium streetlights, orange; and perhaps, at night, grey. Despite this, nothing other than the car and its redness need be presented to S. For there is no unique way of perceiving these elements. The car and its redness can shape experiential character in many different ways. There is no need to appeal to different aspects of redness to account for the variation (as in Kalderon), or to different relational ‘shades’ (as in Fish), or to different relational appearances (as in Genone). Certainly, there is no need for represented colors or colored sense-data. Likewise, take Window Case. Looked at from one angle, the window looks rectangular to S, from another

‘  ’ ̈   

41

angle, square; and perhaps, through distorting lenses, oval. Despite this, nothing other than the window and its rectangularity need be presented to S. For these elements can shape character in many different ways according to different circumstances of perception. Again, there is no need to appeal to different perspectival shapes or relational appearances, let alone represented shapes or shaped sense-data. [Italics added]

What are we to make of the formula I have italicized, ‘shaping the character of an experience’? One natural way of reading this expression is ‘causing the experience to have the character it does’. Perhaps the use of the word ‘cause’, with its suggestion that the character it causes is an effect produced in the subject might seem incompatible with the directness involved in the naïve, relational theory. Nevertheless, it is difficult to see how something could ‘shape’ something without bringing it about that it is as it is, and this seems more or less a causal notion. But the main issue, I think, resides with the notion of ‘character’. The character of the experience may be shaped by the external quality, but it is different from it, so what constitutes this character which the external quality ‘shapes’? Somehow, this very question would have to be rejected; the character is the external quality ‘shaped’ in a certain way. At first, these idioms seem radically unexplanatory. How do you ‘shape’ red into orange? In fact they have a perspectivalist explanation of this, and so are quite close to Kalderon (2011) and to Fish (2009: 153ff). Elements can be presented, and so shape character, in many different ways, due to variation in perspectival factors.

But they differ from other perspectivalists, I think, in the following way. In general, perspectivalists think of the different perspectives as constituting different properties possessed by the object which can be viewed in these various ways, and so multiply the ontology. French and Phillips, by contrast, think that perspectives allow the intrinsic property to ‘shape experiential character’, but that this does not constitute a ‘perspectival property’. They think that the notion of ‘ways of perceiving F’ can do duty in a way that renders unnecessary the reification— even as an intentional object—of the way it is perceived. The appeal of this is that one does not intuitively think of the ways that objects can appear from different perspectives as further properties of the object itself. But neither do French and Phillips want to see them as subjective states—qualitatively characterized states of perceivers—so the character of how the object appears has, so to speak, no home, nowhere to belong. My problem with this is that the only way I can make sense of this character is as a form of intentional object. Thus, I perceive the red as orange, but there is no orange; the orange is an intentional object of the perception. Another way of making this point is that the notion of ‘the character of the experience’ cannot be rendered independent of what the experience seems to be

42

¨      

of, that is, of its intentional object. This suggests that French and Phillips cannot escape what they call ‘the Difference Principle’, which is the idea that, in illusion, one is aware of some quality different from that instantiated in the object perceived, for the intentional object of the experience of seeing a red car looking orange does involve an intentional object of awareness different in quality from the external object. We are back with the problem of finding a way of characterizing the content of the ‘illusory’ experience, without giving it any kind of positive object/content different from the property intrinsically possessed by the external object itself. We must look at the way that others who appeal to perspective try to cope with this problem.

2.4 Brewer, Campbell, and Perspectivalism Bill Brewer’s account of illusion has three conditions. The first two are as follows: (i) A subject, S, is directly acquainted with an object, o, (ii) and this acquaintance is ‘from a given spatiotemporal point of view, in a particular sense modality, and in certain specific circumstances of perception (such as lighting conditions in the case of perception)’. (96) The second condition does not explain how having a point of view brings to bear the sensible properties under which the object is perceived. Brewer tries to meet this requirement in the third condition. (iii) ‘in a case of visual illusion in which a mind-independent physical object, o, looks F, although o is not actually F, o [meets conditions (i) and (ii) and] has visually relevant similarities with paradigm exemplars of F although it is not itself actually an instance of F’. (105) Condition (ii), as Brewer admits, follows a similar line to Campbell, who says that the object is perceived from a standpoint to explain differences in appearance. The natural way to begin on a substantive characterization on the notion of a ‘standpoint’ is to proceed sensory modality by sensory modality. The ‘standpoint’ from which you are observing an object will in the first instance be given by specifying a particular time and place. But the significance of location . . . will depend on the sensory modality being used . . . In carrying out this exercise of articulating the notion of a standpoint, we show how we can characterize a way of experiencing an object without appealing to

, ,  

43

either the idea that ‘ways’ are characterized by associated representations or the idea that ‘ways’ are characterized by the idiosyncrasies of the mental paint involved. Rather than either the idea of an intervening level of mental representation or the idea of an intervening level of mental paint, we can simply appeal to the notion of experience as a three-place relation between an object, an experiencer and a standpoint. In this way we can do justice to Russell’s notion of acquaintance as a knowledge of things more basic than knowledge of truths.

Mark Kalderon takes a similar view, at least with regard to colour vision. The most straightforward objection to perspectivalism is to the idea that there is a fact about how an object is from a certain viewpoint (however objective or subjective this latter is taken to be). But objects do not exist from a viewpoint and they do not have properties from a viewpoint. They—the middle sized dry goods that we normally perceive—exist at given locations and so do their perceivable properties. There is a fact about how an object appears or would appear from a certain viewpoint, there is no fact about how it is, from a viewpoint, except in the sense of ‘appears’. At least, not on our normal conception of what an object is. This thought is no doubt part of what motivates Martin’s rejection of the perspectival talk as a way of explaining ‘looks’, for he says, when rejecting the perspectival talk: Accommodating within our ontology the looks of objects does not require that we posit additional features of these objects over and above those properties we are otherwise committed to supposing them to have through what we can know of them through perception. (2010: 161)

Peacocke expresses similar anti-perspectival sentiments. . . . a way in which a shape property may be perceived is to be sharply distinguished from a way of being shaped. A way of being shaped has to do with shapes themselves – it is a way of occupying space – and does not have to do with the way in which shapes are perceived. (Peacocke, 2001: 248: quoted in Fish, 2009: 159)

This, of course, is what motivates French and Phillips’ denial that they are multiplying properties, but we see they failed, and motivates talk of how the object ‘strikes’ one, to evade the obligation to give character to the ‘striking’. But even if one were to concede, for purposes of argument, that perspectivalism has a prima facie relevance to external matters, such as perspective and lighting conditions, it is not clear how it can cope with something subjective, like shortsightedness: it is not natural to treat having minus 12 vision as a ‘point of view’ from which the object is actually blurred. Brewer does discuss colour blindness

44

¨      

and double vision resulting from pressing the eye-ball, and calls these cases of degenerate acquaintance. Brewer does not, I think, attempt to explain this latter expression, but it is fairly plain what he means: these are cases in which the system is not working in the way that its proper functioning would require. But, first, it is not clear to me how deploying the adjective ‘degenerate’ explains anything philosophically—though, obviously, it is, in a sense, true. Second, this can hardly be made to fit ordinary short sightedness, for this is a matter of degree, and I doubt whether anything counts as perceiving the object absolutely and perfectly as it is, including as it is from a perspective. There are plenty of people whose eye-sight is good enough for them not to need glasses, but how even they perceive objects to be is relative to human visual ability and cannot be said to be a relation to the object as it is in itself, in abstraction from the degree of accuracy normatively normal for humans. The anthropocentricity of even paradigm perception will be relevant to the first objection given below against (iii). The point is that ‘point of view’ or ‘standpoint’ talk, if taken as something objective, does not deal with the ubiquitous imperfection of all perceptual experience. Fish, following Tye, tries to treat blurring as a version of loss of visual acuity, as loss of information; so one is aware of what is out there, but of less of it. This could perhaps be made to fit some cases of visual inadequacy, but not generally. Poor eyesight distorts the scene, it does not merely miss some things out. When you see an edge as blurred, you do not just see part of it, you see it as distributed over an area where it is not. There are other problems for Brewer beyond those he shares with Campbell. The account in (iii) of how sensible features enter into experience has a variety of problems. First, the explanation of how an object which is not F comes to look F is explained by reference to the visual similarities the situation bears to one involving an object that is actually F. But the fact that the similarities are visual means that they relate to how an F object veridically looks, not just to how it is. So we need an account of how sensible qualities enter into supposedly veridical perception in order to cope with illusion by comparison. ‘Transparency’ suggests that they are contributed simply by the object, but the above discussion of (ii) casts doubt on whether the notion of an absolutely veridical perception makes sense: nothing ever looks quite the way it actually is, only with the degree of accuracy that human senses allow, an idea which is itself not fully determinate. Second, the invocation of similarity may not seem very helpful, if you think that a situation which is visually similar to one in which an F object looks F is one in which F figures in the account of how it appears—how else could it be similar? At this point there is a connection between Brewer’s treatment of illusion and his handling of hallucination. For the latter, he follows Martin’s version of disjunctivism (110–11). This is the theory that hallucinations are not recognized by their objects, as are perceptions, but are simply indiscriminable by introspection from the corresponding perception. They do not have an object in any further

     

45

sense than this. This is very similar to saying that illusions are simply similar to certain paradigm perceptions, but not by specifying the similarity in terms of similarity of object. (Brewer makes reference to the causal basis for the similarity, but this is clearly not that which is the object of experience, in the same sense of ‘object’ as that in which external things are held to be the object of veridical experience.) (‘Similar to’ might seem to differ from ‘indiscriminable from’ because the former seems positive and the latter negative, but when ‘similar’ means ‘similar, but not in terms of similar objects of awareness’, it ceases to differ from ‘indiscriminable’.) So Brewer falls into the class of those disjunctivists who place ‘illusion’ into the ‘illuded’ disjunct, like hallucination. We have already seen that this leaves little or nothing of naïve realism remaining. His position here is not that different from Martin’s. Both of them have to resort to some notion of finding or being struck by a similarity to genuinely veridical case, whilst remaining bizarrely opaque about what phenomenon constitutes that similarity. Brewer’s appeal to perspective does not seem to be doing any extra work.

2.5 The Perspectivalism of Fish and Kalderon Fish begins with cases similar to those used by French and Phillips, namely colour and shape illusions. In the case of colour, he seems to follow Armstrong’s suggestion that illumination really changes the colour of objects. Armstrong says: It seems to me that we must admit that a real change in quality occurs at surfaces that, as we say, ‘appear to change’ when conditions of illumination are changed. I can see no ground for saying that such changes are in any way illusory or merely apparent. (Armstrong, 1968: 284; Fish, 2009: 153)

This contradicts French and Phillips’ claim that only the original colour is presented. Furthermore, there seems to me to be a devastating reason for thinking that Armstrong is right, especially in certain cases, and it difficult to see why the rationale does not generalize. Suppose I shine a red light on a white sheet; am I presented just with the white, as I am said by French and Phillips to be with the red in the case of the car when a yellowish sodium light is shined on it, making it look orange? Suppose the white sheet is the backdrop in a theatre, and the colours projected on it are part of the scenery, am I not presented with any of the colours or pictures projected? Suppose the sheet is operating as a screen, and a two hour long technicolor film is projected on to it, do I not see any of the colours or be presented with them? You might say that these are not cases of illusion, and that no confusion or error could result, but these considerations are a matter of

46

¨      

intention or purpose, and these contextual factors seems to me to be irrelevant to how one should analyse the basic phenomenology. These are all cases of a light falling on a surface and making it appear to be a different colour. Whether that light is deliberately projected, and for what purpose, cannot be relevant at the level of a fundamental account of what the result is when light of a certain colour falls on a differently coloured surface. Suppose, for example, that one purpose of sodium lighting was to make things look different, because it was easier on the eyes if they looked that way in the dark, would that make any difference to what was presented? If not, why would projection done for theatrical purposes alter the presentation? One might try to use the fact that a screen is white, and so neutral, as a reason for distinguishing these cases from the sodium lighting: then a pure red light covers and occludes a white surface; no colour ‘alchemy’ takes place. But the white plays an essential role. The beam of the film has no clear form until it falls on the surface. And if the projector beams two different colours from two places—say a blue and a green to make a yellow circle on the backdrop, isn’t it yellow that is presented? So the context affects how it is suitable to think of the situation—as obfuscating the real colour or presenting a new one—but the basic perceptual phenomenology cannot be different. So Armstrong is right; a naïve realist must accept that coloured lights ‘paint’ surfaces, ephemerally and superficially, in different colours. Fish, however, combines this phenomenon with another, namely the way that colours close to each other can influence the way each is perceived, and takes these phenomena as showing that colour is ‘a relational physical property’ (Fish, 153). He continues on from the quotation given from Armstrong above as follows. If the quality in question here is the object’s color, then Armstrong claims that objects really do change their color under changes of illumination. On such a picture, the color an object exhibits on a given occasion is not a matter of the object’s intrinsic reflectance properties, but also of the brightness and spectral distribution of the illuminant. What is more, simultaneous contrast illusions show that even under fixed illumination conditions, the precise color a particular exhibits is also dependent upon the nature of its surrounds. So we might bring the threads of this kind of response together in the claim that the specific color a particular exhibits on a given occasion depends not only upon its intrinsic surface properties, but also on facts about its ‘setting’ – facts about the illuminant and the object’s surrounds. (153)

It seems to me, however, that the natural way of understanding Armstrong’s suggestion is not to treat the colour property as relational, but as the production of a monadic property by mixing two others, as might be the case with paints. But this monadic model does not work for the colour interference case, because there the colours do not actually affect each other, or mix externally, but only in how

    

47

their proximity effects a normal human observer. So Fish is back to a pure perspectivalism. We can conclude that there are many objections to perspectivalism. (i) It is fundamentally counterintuitive, because objects do not have properties from a standpoint, but in themselves, where they are. (ii) Attempts to avoid the multiplication of properties by saying that the one, intrinsic property can be perceived in a variety of ways fails to give an account of these ‘ways of perceiving’, and the character of such perceptions, without invoking how things look, and, hence, intentional objects. (iii) Talk of how things ‘strike’ one, or invoking the ‘bare similarity’ of ‘illusions’ to reality, as if this avoided imputing the qualities involved in the ‘striking’, or in grounding the similarity, either sink back into the ‘indiscriminability as primitive’ deployed by Martin in the case of disjunctivism about hallucination, or are mysterious and metaphorical. (iv) Attempts to relativize perception to the capacities of the perceiver do not preserve the objectivity that naïve realism requires.

2.6 Genone and the Doxastic Theory Genone adopts what he calls a doxastic theory of illusion. If I am fooled by the lighting, on the other hand, this might lead me to conclude that the tomato is orange. In this case, the fact that the tomato appears orange to me is analysed in terms of the fact that I would judge it to be so on the basis of my experience. On the doxastic account, appearances are explained in terms of the judgements one would make on the basis of perceptual experience, and appearances are misleading if they lead one to make false judgements. Perceptual experiences themselves are not erroneous—perceptual error belongs to the mistaken judgements one forms on the basis of them. (2014: 353)

This is a puzzling theory. It is, of course, true that, if one judges a red object to be orange because it looks orange under a sodium light, then one is making a misjudgement, but this does not explain what it is for the object to look orange, because it is the fact that it looks orange that leads one to make the mistake. To avoid this problem, it would seem to be necessary to identify the belief with the experience, which would seem to be to adopt something like Armstrong’s theory of perception, where experience is analysed simply as the acquisition of beliefs. But Genone does not want to do this, because veridical perceptions are treated as direct, relational acquaintance, and only misperceptions are identified with

48

¨      

judgements. In the lines immediately preceding those I have just quoted, Genone says the following. Consider . . . the case of a red tomato in yellow lighting: due to the presence of the lighting, it appears to me as though the tomato is orange. It seems correct to say that things are not as they appear to be. Had I known about the lighting, however, the way the tomato appeared would not have struck me as the appearance of an orange tomato. I would instead have taken the appearance to be that of a red tomato in yellow lighting. So how things appear can be understood in terms of the way one is inclined to judge things to be on the basis of experience. If I am fooled by the lighting, on the other hand, this might lead me to conclude that the tomato is orange. In this case, the fact that the tomato appears orange to me is analysed in terms of the fact that I would judge it to be so on the basis of my experience.

The implication appears to be that, once one knows that the orange appearance is caused by the sodium light, the tomato no longer looks orange. But this is surely false. If one is seeing an ‘orange’ tomato, and someone tells you that it is the light doing this, it does not immediately, or at all, cease to look orange; you know that it is an orange tomato in yellow light, and this affects your judgement about its real colour, but that does not alter how it looks. Cognitive penetration is not that powerful! A similar objection can be made to Kalderon’s appeal to the comparative use of ‘looks’ in the case of the bent stick. But if the look statement is comparative, ‘bent’ does not qualify the stick, but the way the stick looks. Not only is having a bent look consistent with being straight, but having a bent look is consistent with looking to be straight. The stick looks to be straight even though it looks like a bent stick. (2011: 767)

I remember when I first noticed that my arms looked bent in a way that looked as if they were broken when washing myself before going to bed: I asked my parents why it did not hurt. They look no less like having a sharp break now that I know how it works; I know that they are straight but they don’t look it.

2.7 Conclusion It looks as if the naïve realist/relationist cannot give a plausible account of how objects so often (or always) look somewhat different from the way they intrinsically are. It seems that the only way open to the opponent of the sense-datum theory when it comes to illusion is to adopt the intentionalist/representational theory. The objective in the next two chapters is to show that this approach to perceptual experience is either confused or a merely verbal variant on the sense-datum theory.

3 Intentionality and Perception (I) The Fundamental Irrelevance of Intentionality to Phenomenal Consciousness

3.1 Introduction In this and the next chapter I shall be discussing the relation of intentionality to perception, and in Chapter 5 I shall look at how intentionality, in the form of singular reference, works on its ‘home ground’, in thought and propositional attitudes. This latter discussion will help to fill out the discussion of why intentionality does not do the work that many assign to it in perception. In Section 3.2 I will give John Searle’s argument that is supposed to show that intentionalism overcomes the classic arguments against direct realism, and how this leads him into conflict with another intentionalist, Tim Crane, who understands his theory in terms that at least appears to be less a form of direct realism. I shall use this conflict to illustrate what I claim is a deep tension within intentionalist theories, namely between its commitment to a ‘common factor’ theory and, hence, opposition to disjunctivism, and its belief that it is, to say the least, more fundamentally direct realist than the sense-datum theory is. The rest of this chapter and the next will concern various attempts by intentionalists to reconcile these commitments, and why these attempts cannot succeed. It is first worth remarking that the term ‘intentionality’ is used to signify at least three things that blur into one another. First, it is the name for an undoubted phenomenon of at least some mental states, namely their so-called aboutness—the fact that states with propositional content and perhaps others point beyond themselves to a topic, subject matter or ‘object’. No-one, I think, denies that thought, for example, has a feature that can be characterized at least roughly in this way. Second it is the name for a problem (or perhaps more than one) centred round the issue of how we are to understand or explain this aboutness. It does not appear to be a standard physical property, so the naturalist has to explain how he is going to treat it, and even a non-naturalist will probably want to throw some further light on its logical and ontological features. Third, it is proposed as an answer to a problem. Intentionality is invoked as an explanation of how mental states can be about things—it is because they possess intentionality. Once you build this in as some kind of primitive notion, various problems are supposedly

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0004

50

   (  )

solved or alleviated, particularly problems concerning the ontological status of mental contents. The name of the problem becomes the name of its answer. Furthermore, the volume of literature on intentionality is enormous and so is its sub-category, the literature on intentionality and perception. But the complexity of the issue does not stem mainly only from the quantity of what is written, but from various confusions concerning the nature of intentionality, and what consequences flow from this about how the notion can be applied to perceptual experience. As if that were not enough, the notions of intentionality and of representation overlap, sometimes harmlessly and sometimes as indicating very different—roughly, non-reductive and reductive—approaches to the same issues. In the case of perception, appeal to intentionality is usually presented, within the analytic tradition, at least, as an alternative to the sense-datum theory. If I hallucinate a pink elephant, it is agreed by all that there is no instance of elephant of which I am aware, but the sense-datum theorist thinks that there are instances of pinkness and elephant-shape that are objects of my awareness. The intentionalists I am considering, by contrast, put the qualities of shape and colour in the same category as the substantive elephant—all are ‘intentionally inexistent’, which is interpreted by analytical philosophers to mean that they are somehow represented by something mental, but are not instantiated.¹ This way of putting things is meant to be the answer to problems concerning the ontological status of sensory contents. This intentionalist view about perception claims two things. (i) That intentionality is a primitive phenomenon, in the sense that, though there may be a naturalistic reduction (we will hear more about that later), within the scope of our normal psychological concepts, intentionality is basic and a sui generis property of individual mental states. (ii) That intentionality ‘goes all the way down’: it applies to the sensational or qualitative contents in perception, as well as to the objectual. These two theses are connected in the following way. It is fairly clear that the having of qualia or sensations cannot be further analysed, at least with a mentalistic framework, so it would appear to be the case that if these states are intentional, intentionality must be a basic property, not further analysed. The view I shall end up defending in Chapter 5 is that intentionality can be analysed or elucidated within mentalistic terms and that this analysis cannot possibly be applied to phenomenal contents. The implication is that phenomenal contents are not intentional and they must, therefore, be instances of the sensible qualities involved.

¹ M. G. F. Martin has rightly persuaded me that I should emphasize the fact that many philosophers who are classified as intentionalists—for example, Husserl—are not intentionalists about basic sensations. I pointed this out in (Robinson, 1994: 27) My own theory, as developed later in Part I, has intentionalist elements, but not in a way that contradicts the sense-datum theory, where the core of the latter idea is that data instantiate, and do not merely represent, phenomenal properties.

 ’    

51

My conclusions in Chapter 5 will also have consequences for our understanding of the terms ‘intentionality’ and ‘representation’. These two terms are often used interchangeably. This is because they both express the idea that there is a psychological state the essence of which is to be of or about something other than itself. But, on the other hand, the associations of the two terms are often quite different. The intentionality of mental states has been cited by Brentano and many others as a reason for rejecting naturalism—physical states stand in spatio-temporal and causal relations to each other, but they are not ‘about’ anything; only the mental can be that. Theories of representation, by contrast, have generally been developed as attempts to give a purely naturalistic interpretation of intentionality: one state represents another if it stands in some (hard to articulate precisely) causal relation to that which it represents. This naturalistic account makes representation/intentionality a self-standing property of individual states, given their standard causal connections, and is often applied to sub-personal cognitive processes, or ‘informational transfer’, whereas intentionality is treated as paradigmatically part of conscious processes. In Chapter 5 I shall argue that intentionality, as it occurs in conscious thought, is not subject to the atomistic understanding often associated with ‘representation’.

3.2 Searle’s Appeal to Intentionality in Perception, and the Illuminating Contrast with Crane I said at the outset that intentionalist theories of perception were subject to various confusions. This can be brought out by considering Searle’s account of intentionality and its relation to perception, and contrast it with Crane’s theory, which outrages Searle. This dispute is not merely between the two philosophers I am citing, but it shows how one way of understanding intentional theories seems to make them very close to direct or even naïve realism, and another puts them almost in the sense-datum family. One emphasis comes from stressing the way the external world is supposedly transparent to the subject in perception, the other from concentrating on the importance of the common factor that perception and hallucination share. Intentionalists, broadly speaking, affirm both, but we shall see that there is a problem about how to hold them together. As we move on, we will see that both parties among intentionalists—those who emphasize the direct realism and those who emphasize the common factor—hope to make the theory consistent by the way they deploy the notions of content and object in the case of perception. The content is something internal and a ‘common factor’ of perception and hallucination, and the object is the external (or seemingly external—see below) thing perceived. Somehow, both are in consciousness, at least in real perception, but are not in competition for this position in consciousness. (Nor are we aware of two things, content and object, manifesting the same qualities.)

52

   (  )

In the rest of this section I shall use the Searle–Crane conflict to illustrate the potential content–object conflict. In Section 3.3 I shall argue that the presentational nature of experience makes it difficult for the intentionalist to distinguish his theory from the sense-datum theory, at least as far as the ‘veil of perception’ problem is concerned. In Section 3.4 I shall look at the foundation of the problems with ‘content’ and ‘object’. In Section 3.5 I’ll argue that the doctrine of transparency is unhelpful to the issue, and in Chapter 4 I shall consider a variety of contemporary attempts to sort out the content—object distinction in a way that would justify intentionalism. John Searle (2015) has a very direct and simple account of what is wrong with the philosophy of perception and how to put it right. What is wrong is that people have been taken in by the Bad Argument, which is any traditional argument of the causal or ‘illusion’ kind for sense-data: the remedy for this error is to realize that the intentionality of perception rests on the distinction between the content and the object of perception. Once this distinction is observed, no-one will confuse the subjective content of experience with the object—what we are aware of—and it is on this confusion that the Bad Argument rests. Step One: In both the veridical (good) case and in the hallucination (bad) case, there is a common element – a qualitative subjective experience going on in the visual system. Step Two: Because the common element is qualitatively identical in the two cases, whatever analysis we give of one, we must give of the other. Step Three: In both the veridical case and the hallucination case we are aware of something (are conscious of something, see something). Step Four: But in the hallucination case it cannot be a material object; therefore it must be a subjective mental entity. Just to have a name, call it a ‘sense datum’. Step Five: But by step two we have to give the same analysis for both cases. So in the veridical case, as in the hallucination, we see only sense data. Step Six: Because in both hallucinations and in veridical perceptions themselves we see only sense data, then we have to conclude that we never see material objects or other ontologically objective phenomena. So Direct Realism is refuted. (22–3) One could object to various steps, but the crucial step is number three, which says that in both the hallucination and the veridical case we are ‘aware of ’ or ‘conscious of ’ something. But this claim is ambiguous because it contains two senses of ‘aware of ’. . . . (24) Applying this lesson to the Argument from Illusion, we get the following result. In the case of veridical perception I am literally aware of the green table, nothing more. But what about hallucination? In the sense in which I am aware of the

 ’    

53

green table in veridical perception, in the case of hallucination, I am not aware of anything. In the ordinary sense, when you are having a total hallucination, you do not see anything, you are not aware of anything, you are not conscious of anything. But the source of the confusion is the following: In such a case you have a conscious perceptual experience, and ordinary language allows us to use a noun phrase to describe that experience and to treat that noun phrase as the direct object of ‘aware of ’. In that sense I am aware of a visual experience, but this is totally different from the intentionalistic sense, because, to repeat, the visual experience is identical with the awareness itself; it is not a separate object of awareness. In the case of the hallucination, there was an intentional content but no intentional object; there was an intentional state where the conditions of satisfaction were not satisfied. (25) The Bad Argument is an instance of a very general fallacy about intentionality, and it results from confusion about the very nature of intentionality. It is a confusion between the content of an intentional state and the object of the intentional state. In the case of hallucination, the visual experience has a content, indeed it can have the very same content as the veridical experience, but there is no object. The assumption that some authors make is that every intentional state must have an object, but this is confusion between the true claim that every intentional state must have a content and the false claim that every intentional state must have an object . . . Some authors even postulate an ‘intentional object’ as a special kind of object for unsatisfied intentional states . . . I hope it is obvious that this is a confusion . . . Right now I just want to emphasize that the Bad Argument does not stand alone; it is a result of a more general confusion between content and object. (27–8)

This ‘Bad Argument’ is straightforwardly a version of the causal hallucinatory argument with which we started Chapter 1. Has Searle found a very simple response to that argument? If not, how has he gone wrong? The answers to these questions are to be found in how Searle deals with the difference between representation and presentation. Though all intentional states with propositional content and direction of fit are representations of their condition of satisfaction, some of those representations are presentations. When I think about something, my thoughts are representations of whatever it is I am thinking about. But when I directly perceive it – when, for example, I see it – then my visual experiences are actual presentations of the object and state of affairs seen. (41)

This is a point that has often been made; that thought of or about something involves a conceiving of it, but in perceptual experience you do not merely conceive of it but you are as it were presented with it face to face.

54

   (  ) Searle also invokes this presentational quality in his 1983. . . . [visual perception] does not just ‘represent’ the object, it provides direct access to it. The experience has a kind of directness, immediacy and involuntariness which is not shared by a belief I might have about an object in its absence. It seems therefore unnatural to describe visual experiences as representations . . . Rather because of the special features of perceptual experiences, I propose to call them ‘presentations’ . . . [but] presentations are a special subclass of representations. (1983: 45–6)

In these passages Searle seems to suggest that this presentational quality is a feature of veridical perception, rather than of perceptual-type experience per se, but this cannot be correct. An hallucination is, phenomenologically, as if the object were immediately before one. But if the presentational nature of perceptual experience belongs, not just to the facticity of veridical perception, but to the phenomenology of perceptual experience as such, veridical or not, and if the presentational nature is what ensures that such experiences are awarenesses of something, then the content-object distinction does not work in these cases. Searle clearly does accept that hallucinations are, phenomenologically, presentational. When challenging disjunctivism, he says: On the face of it, what is common between the hallucination and the indiscernible identical good case is that they share an identical phenomenology, and, therefore, at least for a certain range of features, an identical intentional content. (169)

If hallucinations can ‘share an identical phenomenology’ with genuine perceptions, it follows that if presentationality is a phenomenological feature—that is, something that shows up to the subject in experience—then hallucinations are presentational. But if they are phenomenologically presentational, then, phenomenologically, there is something presented. Searle might deny that having a presentational nature is sufficient to ensure that an experience is an awareness of something: only the actual existence of the thing in question can do that. But this would be a vain move, for what does allowing that an experience is phenomenologically presentational mean if not that phenomenologically it is as of something, that is, that it has an object? Searle tries to illustrate the distinction between content and object by appeal to the experiences one has if one pushes one’s hand hard against a table and has a painful sensation. (a) I am aware of the table. (b) I am aware of a painful sensation in my hand.

 ’    

55

Both of these are true and though they look similar, they are radically different. (a) describes an intentional relation between me and the table. I had a sensation where the table was its intentional object. The presence and features of the table are the conditions of satisfaction of the sensation. In (a) the ‘aware of ’ is the ‘aware of ’ of intentionality. But in (b) the only thing I am aware of is the painful sensation itself. Here the ‘aware of ’ is the ‘aware of ’ of identity or the constitution of experience. The object I am aware of and the sensation are identical. I had only one sensation: a painful sensation of the table. I was aware of (in the sense of identity or constitution) the sensation, but I was also aware of (in the sense of intentionality) the table.

The point here must be that the sensation is the content of the experience and the table the intentional object. This explanatory picture seems to fit ill with the visual case. It is at least plausible that one can discriminate phenomenologically the sensation in the hand and the feeling of the table, but if one (in Searle’s favourite example) sees San Francisco Bay, Searle would not want to say that we can distinguish the sensational elements from the objective—at least, not whilst affirming (as he does) that the hallucinatory version of the experience is both indistinguishable phenomenologically from the veridical and consists solely of the content (i.e. sensational or subjective) element. If the content of the hallucination is experientially indistinguishable from the presentation of the real object then phenomenologically the content is the object of awareness. That presentationality is not just a feature of veridical experience is generally recognized by intentionalists. For example, Harman says: . . . what Eloise sees before her is a tree, whether or not it is a hallucination. That is to say, the content of her visual experience is that she is presented with a tree . . . (1990: 36)

So it would seem that the Bad Argument is not bad at all, because presentationality is a feature of experience whether veridical or not and presentationality is a legitimate ground for calling such experience awareness of, or consciousness of something. Searle denies that being presentational and having an object go together; The most important mistake to avoid is the confusion of content and object. Two perceptual experiences can have type-identical contents, but one has an object and the other one does not. This, as I have said, is true of the perception of an object and the corresponding hallucination. The perception is satisfied; the hallucination is not satisfied. They can have exactly the same content, but have an object in one case and not in the other case. (37)

56

   (  )

We shall see later, when we discuss Smith, a version of this strategy. Smith agrees that hallucinations have phenomenal objects, but denies that this is the relevant sense of ‘object’. I call this a version of Searle’s strategy for it is surely impossible to deny that there is a straightforward sense of ‘phenomenal object’—namely that in which the experienced is individuated in terms of what it seems to be of—in which hallucinations have phenomenal objects. If I am right, then Crane is right when he says something that outrages Searle. In a sense, then, critics of intentionalism are right when they say that on the intentionalist view, perception ‘falls short’ of the world, and in this sense creates what Putnam calls an ‘interface’ between the mind and the world. The essence of perception – perceptual experience itself – does fall short of the world. But, according to the intentionalist, this is not something that should create any metaphysical or epistemological anxiety; it is simply a consequence of a general aspect of intentionality as traditionally conceived. (Crane 2006: 141; Searle 2015: 185)

Searle, incredulous, says ‘This is a stunning passage’. It undermines for him the whole point of intentionalism. Searle expresses puzzlement about in what sense perception falls short of the world for the intentionalist. He picks up on Putnam’s use of ‘interface’, which Crane quotes. Anyone who says this should tell us exactly what is meant by ‘interface.’ Is there supposed to be some entity that gets between me and the object when I see it and is this the ‘interface’? It is perhaps not surprising that Crane, who is not a Disjunctivist, does not tell us exactly about the nature of the interface. He does not tell us why perception does not actually reach what it is supposed to be reaching, but only by way of an ‘interface.’

John Campbell seems to have an explanation of why there is an ‘interface’ on the intentional theory: On what I will call a Representational analysis . . . perception involves being in representational states, and the phenomenal character of your experience is constituted not by the way your surroundings are, but by the contents of your representational states. (2002b: 116; my italics)

Searle denies this way of looking at intentionality: This conception of intentionality is totally mistaken. And one of the main aims of this book is to present an account of perception that definitely refutes it as far as

 ’    

57

perceptual experiences are concerned. Perceptual experiences are direct presentations of their conditions of satisfaction, and they are experienced as caused by their conditions of satisfaction.

Campbell could surely reply it makes no sense to call perception ‘direct’ if the object of perception does not enter into the phenomenology, whereas Searle seems to be confining it to the intentional content, which is the common factor in perception and hallucination. To say that the object sometimes does enter into the phenomenology and sometimes does not, is not a common factor theory, for it makes the constitution of the ‘seeming’ state different between the good and bad cases, and this is disjunctivism. In Chapter 7 I shall distinguish between what I call ‘semantic direct realism’ and ‘phenomenological direct realism’, and argue that intentionalists are restricted to the former. Whether Searle would be happy with this, I do not know. If the object does not enter into the phenomenology, then Crane is right about ‘falling short of the world’. But Crane’s victory might seem to be Pyrrhic, for it would seem that Searle is right that unless intentionalism can reach fully out to the world and what is presented is, because it is presented, the object of the experience, then the theory fails to be any kind of direct realism. The apparent clash between Searle and Crane illustrates a tension within intentionism; and I say ‘apparent’ because their actual accounts are not substantially different, nor do they differ from other intentionalists, nor from Critical Realists, as I will now try to show. In general, intentionalists, representationalists and critical realists claim (a) that there is something subjective (call it a ‘content’) which is adequate to constitute a full-blown experience, whether perceptual or hallucinatory; and (b) that this content is not an object of experience, but it provides the illusion of an object in hallucination, and facilitates direct perception of a real object in perception. The ‘home ground’ for intentionality is thought, and if one wants to see how intentionalist theories of perception are meant to be a version of direct realism and so to avoid the ‘veil of perception’, one must first consider the features of intentionality in thought that they are trying to carry over into perception. This means that the issue for the intentionalist is whether the model that fits thought can be successfully transferred to perceptual experience, that is, to the case where something is not merely conceived of, or ‘represented’, but is also seemingly presented. In the case of thoughts, taking them to involve representation, one might characterize it as follows: (i) There is something internal to the subject – a representation or representations – which have as content something beyond itself which it is about, represents, intends etc.

58

   (  )

So, in the case of thought, if one thinks, for example, about the Eiffel Tower, the Tower is the object of one’s thought. If one thinks of the Yeti, then, if there is, indeed, a Yeti, that is the object of one’s thought, and, if there is not, one has a conception of it which is not answered or satisfied. The crucial point is that, though there is a vehicle of the thought—usually a sentence, or one might prefer to say the proposition expressed by a sentence, or just a concept—the conception that one has of the object of one’s thought does not stand between one and the object if it exists. Therefore in this case what one thinks of is the thing itself, if it exists, and one merely conceives of it if it does not. The question for the intentionalist/representationalist is whether this model can be preserved when the object is not merely represented in the mental state, but (seemingly) presented. It is when one builds the presentational feature of experience into one’s account that (ii) and (iii) following come into play. (ii) There is something subjective (call it a ‘content’) which is adequate to constitute a full-blown experience, whether perceptual or hallucinatory. (iii) This content is not an object of experience, but it provides the illusion of an object in hallucination, and facilitates direct perception of a real object in perception.

If this can be made to work, then the model of thought might apply, because the vehicle—the sentence/proposition in thought and the experiential content on perceptual-type experience—would not stand between the subject and the object. The question is whether the difference between representation and presentation makes the directness of thought inapplicable in the case of perception; thought is, after all, not direct in the same way as the presentational nature of perception is. How must apparent presence affect how one conceives of the mental content in question? So the distinction marked by the disagreement between Searle and Crane is between that which emphasizes the role of content as the vehicle which puts one in touch directly with the world, and that which emphasizes the phenomenological self-sufficiency of content. Unfortunately, however, the situation is not that straightforward, because, for the most part, both sides agree in wanting to articulate the view that content is not object, but an intermediary entity that puts one in contact with something that is more really called the object. The dispute is about how to walk this narrow line. I claim that it is not possible to pull this off because of the presentationality found in hallucination. The next section gives what I believe to be a powerful argument for this conclusion.

3.3 Presentationality and the ‘Blocking Function’ Fred is perceiving a white wall, but he hallucinates an opaque blue circle in the middle of his visual field. Suppose that in the area of the white wall at which he is

   ‘   ’

59

hallucinating the blue circle there are marks that would normally be visible. Fred cannot see those marks because of his hallucination. The part of the wall that is occluded corresponds exactly to the geometrical character of the hallucination. Someone who says that visual phenomena involve the actual presence of sensible qualities—as does the sense-datum theorist—says that Fred is aware of some blue that occupies his visual field at a point where it ought, if he were not hallucinating, to have been occupied by the marks on the wall. The intentionalist says that Fred is in a mental state that involves representing blue at a location where he should, if he were not hallucinating, have been representing those marks as located. But then it is beginning to look as if it is the internal (‘immanent’) content of the conscious state that constitutes the content of Fred’s purely sensory awareness cognitively blocks off an area of the world from Fred. Isn’t this as good a ‘veil of perception’ as sense-data are held to constitute? The formal reification of the sensible qualities involved in the content seems irrelevant. Suppose, now, that he were actually seeing a blue patch on the wall. Can the intentionalist say that it simply is the external blue patch that occupies his visual field? To make this clearer, imagine the case in which Fred has an induced hallucination of a blue patch at just the point at which there is in fact a blue patch on the wall. The intentionalist would still agree that Fred was not seeing the external blue patch, but was aware of the content of his representational state. Given the intentionalist’s common factor theory, the very same internal state as that which performed the blocking function in the case of the partial hallucination will also be present in the normal perception of a blue patch. As John Campbell puts it in the quotation I gave above On what I will call a Representational analysis . . . perception involves being in representational states, and the phenomenal character of your experience is constituted not by the way your surroundings are, but by the contents of your representational states. (2002b: 116; my italics)

The point is that once one is forced away from the purely representational account that is adequate for the intentionality of thought, and adopts a presentational account for all perceptual-type experiences, then something that cannot, with a straight face, be denied to be an object of experience is made immanent in the experience. But whatever you call it, this ‘content’ constitutes as much (or as little) of a veil as do or would sense-data. Only when the vehicle—the content—is a mere conceiving can the external thing itself be the immediate object of the mental act. The problem for the intentionalist is how to avoid the ‘blocking’ argument whilst denying that what constitutes the phenomenal character of hallucination is a different kind of thing from what constitutes the phenomenal character of perception: that is, whilst avoiding disjunctivism. The ‘blocking’ argument shows how attempts to introduce a sui generis notion of primitive sensory intentionality will not help. Introducing such a notion is,

60

   (  )

indeed, the line taken by both Schellenberg and, tentatively, Crane. The former propounds what she calls an ontological minimalism about phenomenology (Schellenberg (2011). She explains it thus: The common element between hallucinations and perceptions is constituted by the concepts the subject employs in a sensory mode regardless of whether she is hallucinating or perceiving. The relevant sensory modes are modes such as seeing, hearing, touching, smelling, or tasting. How can such a view explain the phenomenology of hallucinations? I will argue that employing concepts and analogous nonconceptual structures in a sensory mode grounds the phenomenology of experience. So any experiences in which the same concepts are employed in the same mode will have the same phenomenology. (13, italics added)

Crane, too, appeals to the notion of sensorily conceiving to explain the phenomenology of hallucination, and what makes it subjectively indistinguishable from perception. Concerning hallucinations he says, though tentatively: The intentionalist can say something like this: it is the state of sensorily representing that things are a certain way. In our example, it is the state of visually representing that there is a snow-covered churchyard outside the window – a state which someone could be in whether or not there is such a churchyard. (2005: 33 [online version])

I think the issue with this approach is whether stipulating or postulating (whichever you prefer to see it as being) a sui generis form of sensory conceiving really solves any problems or simply redescribes the situation in a way that leaves all the same problems, even if at first sight it seems to avoid some of them: the same lump reappears at a different point in the carpet. What is ‘conceiving in a sensory mode’ if it is not the apparent presence of a sensory quality—that is what the sensoriness of the mode consists in! As an analogy, Christopher Peacocke (1983) characterizes subjective perceptual states as possessing such qualities as red ‘because he wants to reserve red for physical objects, but, at least assuming that visual sensations also possess primary quality analogues, such as square’, then it is not clear how any of the problems of the sense-datum theory are avoided in this move. Is it clear that characterizing the subjective state as representing red or sensorily representing it really advances the situation over an account which has it instantiating red? This is the force of the ‘blocking argument’. Crane accepts the ‘blocking argument’, and it is a consequence of his view that ‘on the intentionalist view, perception ‘falls short’ of the world, and in this

 ’     

61

sense creates what Putnam calls an ‘interface’ between the mind and the world’, quoted above.²

3.4 Crane’s Own Account of Intentionality In view of the fact that Crane accepts the ‘blocking argument’, and believes, in my view correctly, that intentionalists are committed to it, we need to look at how he handles the special nature of perceptual representation. Crane analyses intentionality on the following model. They consist of four features: Subject – mode – content – object. The subject is the person in the state, the mode is what, in other contexts, is called the ‘attitude’, i.e. believing, desiring, loving, hating, etc., but also the forms of perceiving, seeing, hearing, feeling, smelling, etc. The content is what is desired, believed, seen, smelt, etc, and is usually in the form of a proposition, or an object; you believe that it is raining, but hate Fred. You can see that it is raining, or just see the rain. The object is what the whole intentional state is about or directed on. It is not part of the psychological state at all. If you claim to see a lion, and there is no lion, there is no object: if you believe in Zeus, and Zeus does not exist, your belief has no object. This does not mean that these states have no content; you seem to see a lion, even if there is none. In the case of Zeus, we can say you have a Zeusconcept, or that you believe that there is a god called Zeus, and this constitutes the content. But Crane rejects the notion of the intentional object as an immanent entity or factor which somehow enters into the constitution of the content. So, from the psychological point of view we can drop the object and simply have the trio: Subject – mode – content. The nature of the subject is not an issue here, and can be taken in an entirely neutral sense to signify whatever it is that is taken to be in this psychological state, so the issue concerns the nature and relation of mode and content, with it as given that it is of no help to say that the content is ‘the object immanent in the mind’, although this is probably the original, scholastic, sense of ‘inexistence’ in the expression ‘intentional inexistence’ which is often applied to intentional objects.

² That Crane accepts the ‘blocking’ is a consequence of the ‘interface’ is something he has affirmed in personal communication.

62

   (  )

At one level, this distinction between mode and content is uncontroversial. Everyone agrees that there are ‘attitudes’ or activities such as believing, desiring, hating, seeing, hearing, etc. and that there are things that these things are of or about or which are propositions which can be introduced with a that. What has, I think, emerged from the discussion of Searle above is what it is that makes the difference between representation, as it occurs in what are normally characterized as propositional attitudes, and presentation, which characterizes perceptual experience. Crane does not deny that this difference obtains, but he simply puts it down to the difference of mode between perceptual modes, and the rest. Again, in a sense, no-one will deny this, in the sense that it is a defining feature of perceptual experience that its object (or the appropriate sensible qualities, if one is talking of smell or hearing) are actually present to the subject, whereas you can, and usually do, the things signified by the other modes in the absence of their objects; one does not just believe in, desire or hate what one is confronting, but one must be in its presence to see, hear, smell it. Does one need to say more, or can one rest satisfied with saying that this is how perceptual modes represent their contents—that is in a presentational way? Crane thinks that this is as far as one can go without falling into error, by hypothesizing bogus entities such as sensedata. What reasons could there be for choosing one theory over another, between the sense-datum theory and intentionalism, as it has been developed in the account of Crane above? I argued in the introduction to the book that there was no incoherence in the notion of a sense-datum, so, assuming both theories are prima facie coherent, how can we choose? Are they merely verbal variants, as the naïve realists quoted above argue? Crane argues that whether one postulates sense-data or not makes a substantive ontological difference, and that representationalism is less rebarbative from a naturalistic point of view. For my part, I cannot see that representing red, in an irreducibly sensory mode, is any less dualistic than being aware of a red sense-datum. The advantage of the sense-datum theory is that it is a less elusive way of characterizing experience than intentionalism. If someone hallucinates a red patch, the sense-datum theorist says that something red is involved, and the representationalist denies this, but agrees that red is apparently present. Red, therefore, comes into the characterization of the phenomenology; it is not, for example, an hallucination of yellow, but of red, so I do not see how the intentionalist can deny that red gives the character to the experience that it has, and it does that by being what it seems to be of, which is equivalent to saying that is constitutes the phenomenal (i.e. experiential) object (i.e. what it is of). (If you still don’t like ‘object’ here, ‘phenomenal content’ would suit the sense-datum theorist. But the distinction between ‘content’ and ‘object’ will be the primary concern of Chapter 4.). This is quite different from the way red constitutes the object/content of a thought about red, because then it is the content of the concept

  ’     

63

or the word meaning that constitutes the phenomenology, not the phenomenal nature of red itself. I think that this makes the SD theory the more transparent account: it explains what it is to represent red experientially, in terms of the presence of red as an object or content of experience, whereas ‘representing red visually’ or ‘in the visual mode’ contains no explication of what this involves. From Chapter 6 onwards, I shall be trying to show how the sense-datum theory can put us in as direct cognitive contact with the world as intentionalism and even naïve realism can achieve.

3.5 The Intentionalist’s Dilemma and Its History I have tried to present a picture according to which, once one clarifies what the intentionalist is saying, his theory seems to resolve itself into either a version of naïve realism, or something not very different from the sense-datum theory. In fact, the intentionalist fights hard against being pushed towards the sense-datum theory by trying, in various different ways, to distinguish the internal element— the content—from the object of perception. In this way they hope to avoid being classified with classical indirect realists. And, as we have seen, even those, like Searle, who emphasize transparency, also insists that there is internal content sufficient to constitute hallucination, but not such as to be a bar to direct realism.³ One very generic way of putting what, in general, all these philosophers hope to achieve is to say that they want to classify the internal component as that by which one becomes aware of the object itself: the internal content is emphatically not that of which one is aware. The next chapter is about the various—not mutually consistent—ways in which different theories have tried to achieve this. The generic feature of all these theories might be expressed as follows. There is in perceptual-type experience a discreet conscious internal feature which, in genuine perception at least, facilitates direct awareness of external things as objects of perception.

I use the word ‘discreet’ because, though this internal factor is conscious and not ‘sub-personal’ in the way most of the physical machinery of perception is, it does not rival the external world in its role as the object of perceptual awareness. Most intentionalists have tried, vainly I believe, to say something helpful about this discreet internal element, and we shall look the gestures towards this task of certain contemporary philosophers, but first let us look at the roots in Brentano and Husserl. ³ Something very similar applies in the case of the theory called critical realism, famous amongst American philosophers in the early twentieth century and recently revived by Paul Coates.

64

   (  )

The back-and-forth between ‘concept’ and ‘object’ has a long history. The central problem in understanding the intentionality of mental states comes in giving a coherent account of their objects or contents. That there is disagreement on this—and even confusion—is clear from what has been shown above. Trying to look at the issue from first principles, we can say that it is uncontroversial that intentionality is characterized by its objects, or what it is directed upon. So the nature of an intentional state appears to come from its object. But this object need not exist, for one can think of, believe, seem to see, fear, etc things that do not exist. Because the object need not exist, it would seem that intentional objects cannot be treated as ‘things’ and that intentionality is not really a relation. In response to this problem, the object itself is explained as no more than an expression of the directedness of the act itself, which may be characterized as the content of the act, in contrast with its object. But directedness makes sense only in terms of what it is directed at, namely its object, so neither ‘content’ nor ‘directedness’ seem to be notions independent of ‘object’. Ever since Brentano introduced the jargon, there has been a back-and-forth between explaining objects in terms of contents or acts, and explaining the contents in terms of the objects. Originally Brentano, followed most extremely by Meinong put the emphasis on the object. Others, for example Husserl, emphasized the ‘immanent’ object is really a content, not a relatum, which turns it into a kind of internal accusative. This idea when developed by Chisholm, ends up as a form of adverbialism—the line now followed by Kriegal. The version of this theory as it has been—more or less—settled on in contemporary philosophy of perception goes something like this. There is an internal content to sense experience of which the following are true. (i) (ii) (iii) (iv) (v)

It is not sub-personal, as are neural causal mechanisms, because it is a feature of consciousness. It is sufficient to constitute veridical seeming hallucinations with the same content as the perception in which it normally plays a role. It is the vehicle by which one is directly aware of external objects. In order to satisfy (iii), it is not the object of our awareness, for that is the external thing, though it is the content. It is genuine common factor and so is not to be treated disjunctively.

The question is, can anything satisfy all these conditions. A reason for thinking it cannot, can be given as follows. 1. Anything presentational has objects of awareness; there is something that the experience is seemingly of. 2. Perceptual experiences are presentational phenomenologically.

  ’     

65

3. Anything phenomenologically indistinguishable from a perception is presentational phenomenologically. 4. Hallucinations are phenomenologically indistinguishable from perceptions. Therefore 5. Hallucinations are presentational phenomenologically; that is there is something that the experience is seemingly of. 6. If there is something an experience is seemingly of, then, qua experience (that is, phenomenologically) it has an object. Therefore 7. Hallucinations have objects. 8. Hallucinations consist, it is said, only of conscious contents. Therefore 9. Conscious contents are or include objects of awareness.

This is a valid argument and 1, 2, 3, and 4 are not in dispute, so the conclusions at 5 and 6 must be deemed sound. As we shall see below, Smith attempts to avoid this conclusion by denying that 7 follows, because ‘phenomenal object’ and ‘object’ do not mean the same thing and the former is harmless. The only other way of avoiding the conclusion is to deny 8, and with that condition (ii) in the list of desiderata. This would be possible if one argued that hallucinations consisted of contents plus some other factor. The natural candidate for this extra factor is some construing or interpreting the contents, so as to ‘see them as’ the phenomenological or intentional object of the experience. In fact this is the way that some intentionalists tend to talk. Smith, as we shall see, distinguishes between the non-intentional quale and the intentional object. Crane attributes the sensible quality to the mode, not the intentional object. Critical realists tend to talk of conceptual interpretation being what transforms content into a seeming object. Notice that I have withdrawn to using ‘seeming object’ here because it is part of the intentionalist confusion that sometimes they insist that there is an intentional object only when there really is an object there, so there is no intentional object in hallucinations; this is because they don’t want to admit a strange class of objects. This is what gives rise to the temptation to talk as if contents were sufficient alone—you need not mention the object to give an account of the ontology of hallucinations. On the other hand they do not deny that hallucinations seemingly or phenomenologically have objects, and ‘intentional object’ often does duty to cover this fact. This is part of the back and forth between concept and object. ‘Seeming object’ or ‘phenomenal(ogical) object’ maintain a neutrality on this. Critical realists at least seem to be free of this dialectic, having a relatively clear distinction between content or given, and conceptually constructed interpretation.

66

   (  )

If one modifies 8 and (ii) by allowing that contents are not the only constituents of hallucinations, one might also have reservations about (i): perhaps contents can be, in a sense, sub-personal if they partially constitute consciousness but are not sufficient to constitute an object of awareness.⁴ We will see in the next chapter some of the ways that intentionalists try to get round this problem. Perhaps most significantly, we will consider Smith’s claim that phenomenal objects are not objects in the relevant sense; in other words that seeming or intentional objects do not allow one to move to any existential assertion, that there are things of which we are aware in these contexts.

3.6 How Appeal to Transparency Helps No-one The doctrine of transparency is common to both intentionalists and relationists. Its crux, in the way it is used in contemporary debate, is to emphasize that it is the object that we are really aware of. Tye, an intentionalist, describes his standing on the beach looking out at the ocean as follows: I found myself transfixed by the intense blue of the Pacific Ocean. Was I not here delighting in the phenomenal aspects of my visual experience? . . . It seems to me that what I found so pleasing in the above instance, what I was focusing on, as it were, were a certain shade and intensity of the colour blue. I experienced blue as a property of the ocean not as a property of my experience. My experience itself certainly wasn’t blue. Rather it was an experience that represented the ocean as blue. What I was really delighting in, then, were specific aspects of the content of my experience. It was the content, not anything else, that was immediately accessible to my consciousness and that had aspects I found so pleasing. (1992: 160)

Nudds, whose general theory follows Martin’s, considers the experience of a vase:

⁴ There is a view that I discussed in (1994: 179–80), which runs as follows; Ontologically or metaphysically content is non-objectual, but it seems objectual. This itself can be taken in either of two ways (a) In the case of hallucination, we misinterpret what our experience is actually like (i.e. we misinterpret the phenomenology)—it is not objectual at all. (b) Phenomenologically it is objectual, but really it is not. (a) cannot be right if hallucinations are phenomenologically presentational—as they obviously are. If (b) can be given any sense at all, it must be the one Smith gives to it in the account we will discuss in the next chapter.

     -

67

Suppose that you are looking at a vase of flowers on the table in front of you. You can visually attend to the vase and to the flowers, noticing their different features: their colour, their shape and the way they are arranged. In attending to the vase, the flowers and their features, you are attending to mind-independent objects and features. . . . [And] your experience seems introspectively to involve those mind-independent objects and features. (2009, p. 334)

‘Transparency’ is a buzz-word, but that does not guarantee it a clear sense. It is generally invoked by opponents of the idea that perceptual consciousness involves any internal or essentially mental objects of awareness. On that interpretation we have the following account; (1) ‘Transparency’ is the claim that the features that characterize perceptual experience are actual property-instances of external objects. This is, in effect, naïve realism. This definition does nothing to answer any objections there may be to naïve realism—it merely shows that they are also objections to transparency construed in this way. But, in common with naïve realism, it purports to give an account of the ontological status of perceptual contents, namely that they are just features of the external world. It would be possible to hold that some features of experience are transparent, but others are features of the experience itself; for example, perhaps the blurriness in an experience is a feature of the ‘act’ not the object, but then there will remain the issue of how to deal with the ontology of the blur—in addition, of course, to the problem of how to explain how the external feature and the blurriness are integrated in what seems to be a unified experience. A more neutral account is: (2) ‘Transparency’ is the claim that the features that characterize perceptual experience are features of the kind that we attribute to physical objects in our ‘manifest image’ conception of the world. This is fairly clearly what Moore meant when he originally used the term, and it is not very controversial, at least if it is not applied to all features of experience. It seems to be true that unless the basic features of experience were transparent in this sense, we could not form any conception of the manifest world. But this version says nothing about the ontological status of phenomenal properties. Berkeley, for example, certainly agreed with (2). I find it hard to believe that the proponents of (1) are not cashing in on the intuitive obviousness of (2) when they appeal to transparency as a way of discrediting traditional empiricist approaches to the subject. It is striking that intentionalists who appeal to transparency seem to be favouring (1), but this takes us back to the issue between Searle and Crane. I believe there is in fact another thought in the background of the fascination with transparency, namely:

68

   (  )

(3) Anything we consciously experience is, in some trivial sense, eo ipso, an object of awareness, so that we are ‘transparently’ aware of it. In this sense, the act-object theory of experience is trivially true; something cannot be part of one’s conscious experience without one’s being conscious of it. But intentionalists in general, whether in line with Searle or Crane, want to get heavy philosophical work out of the distinction between objects and contents of experience, and that issue is what we shall move on to next.

4 Intentionality and Perception (II) Attempts to Articulate the ‘Content’ and ‘Object’ Distinction

4.1 Introductory Remarks There are a variety of strategies adopted by intentionalists which can be seen as responses to the argument which was presented in Section 3.4 to show that presentationality is inconsistent with the claim that the contents of experience are not objects of awareness, and, indeed, to respond to the whole problematic that the last chapter articulated. Most generally, these responses take the form of trying to deflate the sense of ‘object’ in which there is an object—that is, something we are aware of—in the case of hallucination. These strategies are as follows: (i) Line 6 of that argument was 6. If there is something an experience is seemingly of, then, qua experience (that is, phenomenologically) it has an object. It has been argued, for example, by A. D. Smith, that the sense of ‘object’ occurring in 6 is harmless, because it only means phenomenal object and this is not equivalent to object in the sense in which intentionalists need to deny that they occur in hallucinations. (ii) The sense of ‘object’ in the conclusion of the argument 9. Conscious contents are or include objects of awareness.

is intentional object and they do not exist when the perception is not veridical, so they cannot be any ontological threat. This is Lycan’s view. (iii) The kind of object involved in 6 and the conclusion is an abstract entity and this cannot figure as any kind of barrier between us and the world. In slightly different ways, this is the theory of Dretske, Tye, Johnston, and, I think, at one stage, Schellenberg.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0005

70

   (  ) (iv) Saying that a mental state has an object is only to say that it has a kind of directedness, and this does not reify the object. This is Crane’s theory. (v) Perceptual contents are a form of mental act and this defuses any threat. This is Schellenberg’s developed theory.

One thing that should be borne in mind when discussing these strategies is whether, even if one were to accept one of them, it would avoid the ‘blocking function’ argument presented in $3.3. If it does not, then it is of no help to any kind of direct realism of a standard kind, and would seem to be only a verbal variation of the sense-datum theory. I argued in Robinson (1994: 174–80) that this is the fate of adverbialism. If that which constitutes our conscious state in the case of an hallucination is a mode of our mental activity, and if this a common factor with perception, and, therefore, a state of our mental activity constitutes our consciousness in perception, then naïve realism is not merely mistaken, but is a category mistake, for how could a mode of our mental activity be an external object?

4.2 Modern Responses (i): Smith: ‘Phenomenal Objects’ Are Not Objects in the Relevant Sense A. D. Smith, in his (2002) claims to be a direct realist and therefore should be on Searle’s side in the divide. He does not merely agree that hallucination possesses the same phenomenal character as perception, but is happy to talk of qualia constituting the content. Taking the case of hallucinating a vivid green patch on a wall, Smith says . . . there is, actually in your sensory experience, something corresponding to the greenness that you see on the wall: namely an instance of a chromatic quale. Neither this quale, nor the sensory experience of which it is a characteristic, is, however, the object of awareness . . . Your object is a patch on a wall. It is only that that doesn’t exist. (237)

So the common contents of perceptual-type experiences, whether veridical or hallucinatory, are qualia, according to Smith, but he claims that what hallucination and perception have in common—the qualia—is not the object. That qualia and sensory qualities are not objects of awareness is a vital part of Smith’s theory. He says that there is the possibility of a defence of Direct Realism in virtue of regarding sensory qualities, or qualia, as inherent features of sense-experiences themselves, rather than of some object – a sense-datum – of which we are aware. (61)

  (  ) 71 Qualia are non-intentional entities, which is probably why he does not generally use the term ‘content’ to describe them. One could just as well call them ‘nonintentional content’, when the primary point of the term ‘content’ is to indicate a subjective component that contrasts with ‘object’, which is (putatively) external, and so not ‘subjective’. Most uses of ‘content’ in the philosophy of mind treat it as the internal intentional state—the one that represents an external state of affairs. In Crane’s terms, it is the state which is directed to its object. Whichever way one looks at content, the same problem arises from its relation to the phenomenology of the experience. It seems that content, whether intentional or qualia, is sufficient to constitute an experience with all the subjective character of a perceptual experience, so how can it fail to be presentational and not just representational, and presentational phenomenology is, phenomenologically, awareness of an object. And the sense of ‘object’ which is at stake here is the phenomenological one, not the ‘independently existing thing’ one. Smith does not think that that is enough to embarrass the intentionalist. He puts great weight on the difference between the phenomenological and ontological senses of ‘object’. Talk of merely intentional objects is, however, an invocation of ‘nothing’ only in an ontological sense. We do, indeed, need an ontologically reductive account of intentionality. Non-existent intentional objects supervene on intentional experiences. We should, however, sharply distinguish this from any attempt at psychological reduction. Talk of the awareness of an object is inescapable if we are adequately to characterize certain psychological states as they are lived. Reference to intentional objects is not just ‘a way’ of talking about perceptual experience, but the phenomenologically necessary, only adequate, way . . . What is thus described is an experience. It is crucial to the present position that one is able to combine ontological conservatism with the claim to the phenomenological indispensability of reference to such objects. Remember that ‘object’ does not mean ‘entity’: the former term is here given an exclusively phenomenological interpretation. (244–5)

This quotation contains a very important claim, namely that the fact that visual experiences are phenomenologically objectual does not mean that what one is aware of subjectively or in an hallucination counts as the object of one’s awareness, because it is not an entity, only a phenomenal object. This seems to me to be deeply confused, because it is the analysis of the phenomenology that is at stake, not whether ‘phenomenological content’ is some sort of independently existing entity. One cannot help but hear echoes of Moore’s misconceived ‘refutation of idealism’ (and Hegelian idealism merits refutation), when he claims that an actobject analysis of experience entailed the possible existence of the object independently of the act, thus esse est percipi is incompatible with an act-object

72

   (  )

account. Smith agrees that this is not true, if the object is a phenomenal object. But that, surely, is the sense of ‘object’ that is relevant, for it concedes that ‘the object’ is what we are aware of in the experience. But Smith does not seem to be the only philosopher who thinks that phenomenal objects do not count as objects, in the sense relevant in controversies in the philosophy of perception. Schellenberg says the following. This structural difference between sensory awareness and perceptual consciousness is analogous to the distinction between relational and phenomenological particularity. A mental state is characterized by relational particularity if and only if the experiencing subject is perceptually related to the particular perceived and the perceived particular constitutes the mental state. A mental state exhibits phenomenological particularity only if it seems to the subject as if there is a particular present. Every experience that is subjectively indistinguishable from a perception exhibits phenomenological particularity. After all, it is unclear what it would be to have an experience that seems to be as of an external, mindindependent particular without it seeming to the subject that there is a particular present. Since the hallucinating subject is not perceptually related to the mindindependent particular that it seems to her is present, hallucinations are not characterized by relational particularity. More generally, we can say that if a subject has an experience that is intentionally directed at a particular, it will seem to her as if she is experiencing a particular—regardless of whether there is in fact such a particular present. If this is right, then any view on which a perception, illusion, or hallucination is intentionally directed at the experiencer’s environment is committed to saying that such experiences exhibit phenomenological particularity. (2018: 156–7)

One must wonder how an experience can possess phenomenological particularity without being characterized by the seeming presence of an object. The only possible answer to this that I can see is that the apparent or phenomenological presence of an object does not entail the actual presence of one, but, in the relevant sense of ‘object’ this is self-contradictory, for it is the phenomenology that is being analysed in these debates.

4.3 Modern Responses (ii): The Contents of Subjective Experience as Abstractions: Dretske, Lycan, and Jackson Dismissing the ontology of hallucination etc. on the grounds that it is abstract seems to be the line many representationalists take. The idea seems to be that, in so far as an experience is veridical, its intentional object is a thing, or facet of a

  (  ) 73 thing, in the external world: and in so far as its intentional object is non-veridical, though phenomenologically quite real, any question of its ontological status can quite simply be dismissed. Dretske, for example, explains the status of the contents of hallucination or misperceptions in the following way. He says that ‘the properties we are aware of in achieving this awareness (being universals) exist nowhere’ ((2000) 160). He follows this up by saying: Awareness of colors, shapes, and movements, when there is no external object that has the property one is aware of, is not, therefore, a violation of [the principle that experience involves no internal phenomenal properties]. A measuring instrument (a speedometer, for example) can (when malfunctioning) be ‘aware of’ (i.e., represent) a speed of 45 mph without any object’s (inside or outside the instrument) having this magnitude.

The idea that a malfunctioning speedometer and an hallucinating person both ‘hallucinate’, in a similar enough sense for the analogy to be of any use in providing a physicalist model for hallucination, is bizarre enough. But it also seems to carry the implication that crude analogue representing devices do actually experience what they represent—they just have rather limited experiences. This seems to me to be a reductio of this approach. Lycan’s (1987) view is similar: I take the view...that phenomenal individuals such as sense-data are intentional inexistents a la Brentano and Meinong. It is, after all, no surprise to be told that mental states have intentional objects that do not exist. So why should we not suppose that after-images and other sense-data are intentional objects that do not exist? If they do not exist then – viola – they do not exist; there are in reality no such things. And that is why we can consistently admit that phenomenal-color properties qualify individuals without granting that there exist individuals that are the bearers of phenomenal-color properties. (1987: 88)

The cavalier attitude to intentional objects does indeed, seem to be Jackson’s approach, as is shown by what Jackson says. Intentionalism means that no amount of tub-thumping assertion by dualists (including by myself in the past) that the redness of seeing red cannot be accommodated in the austere physicalist picture carries any weight. That striking feature is a feature of how things are represented to be, and if, as claimed by the tub thumpers, it is transparently a feature that has no place in the physicalist picture, what follows is that physicalists should deny that anything has that striking feature. And this is no argument against physicalism. Physicalists can

74

   (  ) allow that people are sometimes in states that represent that things have a nonphysical property. Examples are people who believe that there are fairies. What physicalists must deny is that such properties are instantiated. (2004, 431)

There are at least three problems with this cavalier approach to intentional objects. First, to say that they are abstract and so not part of the ontology of the empirical world leaves untouched the question of how, for a physicalist, an abstract, immaterial entity is supposed to constitute the content of a physical state: what is it for the brain to be aware of an abstract object? Second, the emphasis on the possible non-existence of intentional objects does not seem to be a fair account of the actual nature of a mental state—absence is not what gives a state the phenomenological nature it has. (This sounds rather too reminiscent of Sartre’s characterization of intentionality as ‘nothingness’.) Third, the difference between intellectual states and sensory ones is not given enough attention. The idea that phenomenal redness might simply not exist, in the way that fairies do not exist, hardly seems adequate. You can stop believing in fairies, but you cannot stop things looking red. It is a bed-rock fact that that is how things appear. Even the psychological event of someone’s exercising their belief in fairies does not consist wholly in an absence. Something actual and real—say, saying some words to oneself or entertaining a proposition—constitutes the phenomenology of it. The phenomenology of seeming to see something red must be constituted by something empirical and actual. One could add the following in the case of reductionists of a functionalist bent, as are Lycan and Dretske. What does presentationality consist in, within such a framework? Invocation of transparency seems to presume that they can help themselves to a notion of direct experience which seems to ignore the need to give a functionalist understanding of what such a thing could be. As usual, Armstrong seems to me to be the only causal theorist who tackles this head-on. For him, experience, as against thought, is characterized by the intensity of the acquisition of beliefs—or, in more modern jargon, inflow of information. Giving this a functionalist interpretation, it must be the richness of the functional response in which it consists: just believing that the Eiffel Tower is in Paris is a less dramatic or varied prompter to response than seeing it in detail. Hallucinations are the occurrence of a similar rich response readiness in the absence of the normal stimulus. It is surely interesting that, unlike Armstrong, functionalists like Dretske and Lycan feel the need to reinstate an object of experience and will not rely on the resources that their own ideology makes available to them.

4.4 Modern Responses (iii): Contents as Abstract: Johnston and Schellenberg Johnston is typical of those who treat sensory contents as a special kind of abstract object, a sensory universal.

  (  ) 75 Johnston expresses his theory as follows. Consider the sense-field or scene before your eyes . . . It is a scene type or sensible profile, a complex, partly qualitative and partly relational property, which exhausts the way the particular scene before your eyes is if your present experience is veridical . . . The sensible profile . . . involves more than the layout. For example, it includes the further condition that the relational layout be filled in with some particulars or other that have such and such qualities. But again, this way the scene is could be instantiated by many groups of particulars. This way is not particular but universal, not a token but a type. . . . The suggestion is that in the corresponding case of a subjectively indistinguishable hallucination you are simply aware of the partly qualitative, partly relational profile. This means that the objects of hallucination and the objects of seeing are in a certain way akin: the first are complexes of sensible qualities and relations while the second are spatio-temporal particulars instantiating such complexes. The visual system is adapted to put us in contact with scenes or visual instantiations around us. When the normal system misfires, as in hallucination, it presents uninstantiated complexes of sensible qualities and relations, at least complexes not instantiated there in the scene before the eyes. (Johnston, 2004: 134–5)

The issue is what Johnston thinks that the neural processes—what he calls ‘the visual system’—constitutes or causes. It is difficult to make sense of it constituting or causing something which is a universal. It is also difficult to make sense of the idea that a universal that it constitutes or causes (especially constitutes) should be identical to the instantiation of the complex of qualities out there in the world, in the case of perception. There are three things in play. The distal particulars, the complex of qualities and relations out there instantiated in the world, and whatever relation it is that obtains between the visual system and that complex considered as a universal. If the visual system merely put one in touch with the universals in their instances in the external world, it is wholly unclear why, in the absence of those or similar instances the visual system should generate a veridicalseeming hallucination, for there is nothing for it to put us in touch with. It seems it must generate awareness of the universals in their Platonic heaven, but then why should it direct our consciousness out into the world in the case of perception? It is, in fact, not correct to say that the sensible profiles are universals, when they are experienced, though they do contain universals. They are not universals because they present the qualities and the apparent particulars as present at certain places at the time of experiencing them: they are all putative particulars. The situation appears to be as follows. The visual system generates a common factor which is a profile. This is endowed with all the sensible features appropriate to a physical scene. If produced in a hallucinatory manner, say as a partial

76

   (  )

hallucination, it would occlude objects in that part of our visual field. Johnston claims, despite this, that there is no ontological problem about the status of these profiles . . . sensible profiles that happen not to be instantiated raise no more ontological difficulties . . . than manners of presentation that happen to have no referent. Sensible profiles are themselves manners of presentation that are themselves presented in sensing. Indeed this is a distinctive aspect of sensory experience, one that marks it off from belief or thought. Sensory manners of presentation are themselves sensed. (Johnston, 2002: 141)

First there is a confusion here. As we saw above, Lycan and others have tried to treat sensory objects as abstract or intentional and therefore possessing no ontological weight (Lycan, 1987: 88; see Robinson, 2008: 236–7 for a discussion of this and similar theories). But we are dealing here not just with abstract objects but with psychological events. If I think of the King of France and there is not one, the King of France has no ontological weight. But my thought is a real event and possesses introspectable properties: these vehicles of thought cannot be abolished as unreal. Furthermore, the fact that the contents of sensory modes of presentation are features of real events, namely our experiences, gives them an empirical reality which is weightier than mere universals or abstract objects: their qualia-ness must be real. So, on both grounds, they cannot be dismissed as raising no ontological issue. Yet such a profile is supposed to have real particulars inserted into its ‘argument places’ in veridical perception. There are in fact two accounts in the offing here. According to one, the profile is like a sensible version of an unsaturated Fregean sense, into which a particular is inserted in real perception, like a name in a predicate. When transferred to perception this model, already somewhat metaphorical, becomes more elusive. Is what is inserted a ‘bare particular’—a logically proper name, or is it a physical object with all its sensible properties? If the former, it seems to fail to be an experiential presence at all. If the latter, its qualities are competing with those of the sensible profile generated by or within the brain. It is no use to say that these, being universals, can ‘slip into’ the instantiation which is the particular, for they are not just universals but empirical phenomena. (There is also the fact that this version faces the problem raised by objects looking other than they are: if the profile just is the full-blown propertied object, how does it fit with the profile that is its experience?) At one stage, Schellenberg seemed to have a similar theory. Schellenberg says the following. So if I perceive a white cup o, the content of my perception will be

  (  ) 77 where MOPr (o) is a de re mode of presentation of the cup o and MOPr(P¹) is a mode of presentation of the property that this object instantiates. If I hallucinate a white cup and thus am not related to any white cup, the content of my hallucination will be

where MOP¹r(_) is the object place in an empty object-concept and MOP²r(_) in the property place is an empty property-concept. (2011: 28)

Now Schellenberg is clear that A perception and a hallucination can (seemingly) have the same phenomenology and thus be subjectively indistinguishable. (4)

So the absence of the object makes no subjective phenomenological difference, as that is entirely taken care of by the mode of presentation (MOPr). The difference, it seems, must be a sematic one—as one might expect when she describes these frames as ‘potentially gappy Fregean contents’ (27). Schellenberg distinguishes her theory from Mark Johnston’s, but I have to admit that I cannot see any important difference between them. She quotes the following passage as summing up Johnston’s view. When we see we are aware of sensory profiles. When we hallucinate we are aware merely of the structured qualitative parts of such sensible profiles. Any case of hallucination is thus a case of ‘direct’ visual awareness of less than one would be ‘directly’ visually aware of in the corresponding case of seeing. (Johnston, 2004: 137; quoted in Schellenberg 2011: 8)

Schellenberg’s objection to this theory is The property-cluster view has it that hallucinating subjects stand in an awareness relation to properties that are not instantiated where the subject experiences them to be instantiated. Since these properties are not instantiated where they are experienced to be, they are conceived of as universals. This view is phenomenologically controversial since universals are abstract entities. (8)

Schellenberg acknowledges towards the end of her paper an essentially similar problem for her own theory. One could object that concepts are abstract entities: the thesis that experiencing subjects employ concepts [including in hallucinations HR] implies that they are related to these concepts.

78

   (  )

Her response is: I have argued rather that the phenomenology of experience can be identified with employing concepts and nonconceptual structures in a sensory mode. (29; italics added)

So, in my words, Johnston has us being visually aware of a property-cluster, either instantiated (in the case of perception) or not (in the case of hallucination)

and Schellenberg says we employ a Fregean content in a sensory mode, either with the argument place occupied (in the case of perception) or ‘gappy’ (in the case of hallucination).

In both cases universals are present in a sensory, not purely intellectual, fashion (‘sensory mode’ and ‘visually aware’), either as attached to a subject, or not. The difference seems to me to be largely verbal. The common point is this. The phenomenology appears to belong to what is generated by the brain; it is a common factor to both perception and hallucination. The difference between perception and hallucination consist in whether there is an object to which these phenomenal states somehow belong. I will consider in Chapter 7 how this leads to what I shall call ‘semantic direct realism’. But that is an issue for later.

4.5 Modern Responses (iv): Schellenberg on Discriminatory Capacities Schellenberg’s most recent work (2018) is founded on two claims. (1) Perceiving is the exercise of a capacity. (2) The capacity in question is of discriminating particular external objects. She complains that the ‘capacity’ approach to perception has been generally ignored, but the problem seems to me to be that calling perception a capacity is trivial: it is uncontroversial that someone who is not blind is able—has the capacity—to see, and when they are actually seeing something they are, trivially, employing this ability. It is also a commonplace that we usually take ourselves to be perceiving particular external things. The philosophy of perception consists in trying to say what is involved in this activity. Schellenberg says that what is involved is discriminating these particulars. Again, there is a sense in which this



79

is not controversial, but I cannot see any clear indication that she distinguishes between two kinds of discrimination. One is the behavioural sense: S can discriminate red apples from green if he sorts them into boxes correctly. But if this is the sense of discrimination being employed, we have a very reductive account of perception in play, and I am sure that this is not what Schellenberg intends. The other understanding of discrimination is not overtly behavioural, but involves something like simply noticing or being able to tell the difference in colour between the apples. This is a subjective sense of discrimination and seems to presuppose the occurring of perceptual experience—it is features of objects as experienced that one is discriminating. So the problem with Schellenberg’s fully mature ‘ontologically minimal’ theory is that it is either a version of behaviourism or presupposes the very thing that she is supposedly analysing, namely the having of a perceptual experience.

4.6 Conclusion The argument of this chapter and the previous one might seem to have become convoluted and even muddy. This, I claim, is because intentionalism is essentially a confused theory, and following it through its tergiversations involves convolutions. The essential problem showed up in the controversy between Searle and Crane. This concerned the conflict between the desire to preserve ‘transparency’ and a form of direct realism, on the one hand, and the commitment to a ‘common factor’ theory on the other. Another way to put this dilemma is that the intentionalist theories we have been discussing in these chapters are designed to avoid the Scylla of naïve realism with disjunctivism, and the Charybdis of sense-datum theory. They try to do this by postulating a content which is both (i) unlike a sense-datum, in not being an object of awareness, and (ii) sufficient to constitute the complete phenomenology of experience. If this can be pulled off, one has a common factor—the content—and an object which is what one is (transparently?) aware of. The driving force behind my argument against these theories is that these two objectives are incompatible; anything sufficient to constitute the phenomenology of experience must be, or include, the phenomenal objects of awareness that characterize that experience. And ‘phenomenal object’ is the sense of ‘object’ relevant to the analysis of perceptual-type experience: a sense-datum theorist does not think that such data are objects in the ‘something out there’ sense in which, when we see, for example, a tree, the tree is an object ‘out there’. Nevertheless, there are issues still to be further dealt with concerning the integration and interaction of the phenomenal and cognitive components in perception. These will be finally dealt with in Chapter 8, but 6 and 7 will prepare the ground.

5 Singular Reference and Its Relation to Intentionality 5.1 Introduction In Chapters 3 and 4 I argued that appeal to intentionality does not help us to understand, or somehow demystify the subjective or sensational aspect of perception. In the subjective aspect of sensory experience, qualities are really presented, and not merely represented. In this chapter I hope to reinforce that conclusion by explaining how intentionality works in the case of singular thought and this account will make it clear how it is not a candidate for explaining basic experiential content. The point of this chapter is to show how we can represent individuals, irrespective of whether those individuals exist. It is, therefore, an account of intentional objects, and how they are sustained in our mental framework. Much of what follows is material I first put together when thinking about reference. It was, until recently, a universally agreed principle in the theory of reference—a topic in the philosophy of language—that one cannot refer to things that do not exist.¹ On the other hand, philosophers working on propositional attitudes all agree that someone can fear, love, believe in Zeus (e.g.) even though he does not exist. Reference was treated as extensional, but the attitudes as intentional, so, somehow, you could believe in, etc. Zeus, but you could not refer to him. Reference is extensional, but ‘aboutness’, intentional. But one must surely wonder how one could have a mental state which is about some individual (or utter a sentence expressing such a mental state) and yet one not be referring to that individual? This is surely a compartmentalization of philosophy of mind from philosophy of language of schizophrenic proportions.² The conflict in theory of reference concerning how we can pick out individual objects has classically been between theories of direct reference, associated with, for example, Kripke, and historically attributed to Mill, and descriptive theories, associated with Russell’s theory of definite descriptions and Frege’s account of names. On the side of intentionality, the idea that intentionality as a feature of

¹ For an early refutation of this doctrine, see Jonathan Barnes’ unjustly neglected (1971). ² Mark Sainsbury in his important book Reference without Referents (2005) defends the view that one can refer to non-existent objects, but ‘intentional object’ does not occur in the index. This is because it is a work of philosophy of language.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0006

 ’  

81

individual mental states is what enables thought to be about individual objects comes from Brentano. Before moving on to the positive account of how intentionality works, I want to draw attention to two moments in the history of philosophy that have played a part in our misconceptions of what is involved in picking out individuals in thought or reference. One is what I call ‘Brantano’s howler’, the other shows how Mill has been misinterpreted by those who invoke him as a proponent of ‘direct reference’.

5.2 Brentano’s Howler It is clear that there is at least an incipient confusion between intentions as concepts or universals and as putative particulars. How has this come about? It has its roots in the reading—or misreading—of Aristotle and the scholastics. One interpretation locates the essence of thought in the universal, the other in its ability to reach out to the world and particulars. I shall pursue the historical point because I believe it is directly relevant to the mistaken use to which intentionality is now put in the philosophy of perception. Within the analytic tradition of interpreting Aristotle, thought is seen as thoroughly general, and the problem is how thought of particulars is possible. Brentano, however, interprets Aristotle as saying that thought captures the particular. He says: Aristotle says in chapter 12 of De Anima 2 that the sense receives the sensible form without matter, and he illustrates this with the metaphor of the wax which takes on the form of the seal without receiving any gold or iron into itself. Of course, this metaphor is not absolutely perfect in that the formed wax does not individually bear the same form as the seal, but only one like it . . . (1977: 55)

Clearly what is happening here is that Brentano is following the (probably correct) scholastic interpretation of Aristotle, according to which forms in objects are individuals (or individualized forms) and assuming that when the sense receives the form without the matter it is the individualized form in question that it receives.³ He, therefore, complains that the metaphor is misleading, for only the shape as a type is transferred, not the individual instance. Aristotle, however, whatever his view of the individuality of forms in objects, is clear that in the sense, as well as in thought, they are universal, because they are without their matter.

³ For one account of this much discussed topic, see Irwin (1990).

82

  . . . the sense is that which is receptive of the form of sensible objects without the matter, just as wax receives the impression of the signet ring without the iron or gold . . . ; so in every case it is affected by that which has colour, or flavour, or sound, but by it, not qua having a particular identity, but qua having a certain quality, and in virtue of its formula (kata ton logon) (424a18–25)

That Aquinas is committed to a doctrine of individualized forms in external objects is uncontroversial, but he does not attempt to explain how thought can be of particulars by appealing to the reception of forms qua individualized. His account is as follows: Hence our intellect knows directly only universals. But indirectly, however, and as it were by a kind of reflexion, it can know the singular, because . . . the intellect, in order to understand actually, needs to turn to the phantasms in which it understands the species . . . Therefore it understands the universal directly through the intelligible species, and indirectly the singular representation by the phantasm. (S.T. Pt. I, Qu. 86, art.1)

There is no suggestion that the individualized form itself is the kind of thing that can enter the soul; the individuality has to be provided by something else. Just as matter provides the individuality to the universal form in external things, the phantasm provides it in the intellect. But how it does so is mysterious, especially given that it would seem to be required to provide the individuality for the object of the mental act, not for itself, unlike normal matter, which constitutes the individuality of the thing of which it itself is a part. If the phantasm were to represent the individuality of the object that caused it, it—the phantasm itself— would have to possess intentionality, which is quite contrary to what Brentano intends, where the form is what provides the intentionality. To say that the intellect ‘turns to’ the phantasm does no more than say that somehow the particularity of sense-content contributes individuality, it does not explain how. As we shall see briefly at the end of this chapter the sense-datum theorist has a better account of this than Brentano, and it is one which, as far as I can see, a Thomist would be free to deploy.

5.3 Mill as Supposed Proponent of Direct Reference Mill is normally cited as having a directly referential, not a descriptive, theory of names and so of reference in general. This is because he says names have denotation but no connotation. Modern exegetes assume that something with only denotation must function like a name in Frege–Russell logic and be in no way dependent on a descriptive background. This is clearly not Mill’s

     

83

view. For him, although proper names are mere labels, they presuppose background information. If, like the robber in the Arabian Nights, we make a mark with chalk on a house to enable us to know it again, the mark has a purpose, but it has not properly any meaning . . . When we impose a proper name, we perform an operation in some degree analogous to what the robber intended in chalking the house. We put a mark, not indeed upon the object itself, but, so to speak, upon the idea of the object. A proper name is but an unmeaning mark which we connect in our minds with the idea of the object . . . . . . By [a name’s] enabling [the hearer] to identify the individuals, we may connect them with information previously possessed by him; by saying This is York, we may tell him that it contains the Minster. But this is in virtue of what he had previously heard concerning York; not by any thing implied in the name. (1949: I. ii. 5, p. 22)

Those who cite Mill on this topic tend to emphasize the fact that names are marks with purpose but no meaning, and mere labels, and to ignore the claim that their role is to label the object via the conception of that object the subject already has, and to connect new information with the information already held about the same object—in the more modern jargon, to put it into the right dossier. If York is to be the intentional object of my thought, I must have a conception of it, and it can figure in my thinking only because I can refer anaphorically to the object as specified in that dossier and conception. So for a thought to have an intentional object is for it to activate an objectconception, usually possessed by the thinker. That this is the case is not contradicted by the fact that one can pick up someone else’s reference, whilst knowing no more about the object of that reference other than it is the person just referred to. Kripke cites the possibility of hearing someone refer to the physicist Feynman, about whom the hearer knows no more than that bare reference, and then picking up that reference and using the name, and uses such a possibility to argue that reference need only rest on such a minimal picking up of a use, and that, therefore, anything further is not necessary for, so cannot enter into the analysis of, successful reference. It is not irrelevant, however, that though one can, as in Kripke’s Feynman case, deploy others’ greater expertise, this is not the central case. The legal maxim that ‘hard cases make bad law’ is often relevant in philosophy, and the philosophical concern with necessary and sufficient conditions can obscure the vital importance of central cases for understanding concepts. The practice of referring could not work if no-one knew more about Feynman—or anyone or anything else—than the hearer knew in the case imagined by Kripke. Parasitic cases cannot be central in showing how a practice or concept works.

84

 

Intentionality is, like reference, a complex phenomenon, not an intrinsic and primitive property of ‘pointing’ or ‘directedness’ belonging to individual mental acts considered in isolation, as I shall now explain.

5.4 Intentionality and the Distinctive Character of Thought: Having a Conception of an Object, Mental Files, and Mental Maps So we can now see that Brentano’s belief that the scholastic notion of intentionality explains how single mental acts can have a mental grasp on individual objects is false, and that Mill’s claim that names have denotation without connotation does not mean that we can individuate objects without informational background. Mill, indeed shows how having a conception of an objects works. In the writings of Strawson (1950a), Lockwood (1971) and more recently, Recanati (2012) and others we find the notions of a ‘knowledge map’ and of ‘mental files’. Although this idea is associated now with cognitive psychology, it has a more purely philosophical history. The idea is that we develop dossiers of putative information on all the individuals we believe to be in the world and put these together to form a kind of world map. Morris (1984: 59) reports the positions as follows. Strawson . . . has us picture to ourselves a sort of knowledge-map . . . On this map are many dots, each representing what he calls a ‘cluster of identifying knowledge’ . . . [T]hey each represent some propertied individual about whom we have some knowledge . . . Any proper name by which we refer to an object is written on the map adjacent to the dot which represents its bearer. From each dot radiate lines bearing predicate expressions. These represent all the propositions we are able to affirm of each object within the scope of our knowledge . . .

Strawson (1974) uses another image. A person is, amongst other things a machine for receiving and storing information. The machine contains ‘one card for each cluster of identifying knowledge in his possession’. New knowledge is entered on the appropriate card, and an identity statement, if accepted, leads to two cards being amalgamated into one. Lockwood has a similar idea of what he calls ‘mental files’. The notion common to these approaches is that of a world map. A complete world map would represent every spatio-temporal object that existed in the world and would track it—space-time worm fashion—throughout its history. Such a map would enable one to individuate objects with complete accuracy, in so far as the objects were themselves properly individuated—that is, did not overlap. Thus one could always tell whether the F at p, t, was the same F as that at p’, t’, by seeing

    

85

whether they were part of the same space-time worm. And if someone responded to ‘the first man to climb Everest’ with the question ‘Who was that?’ and one replied by providing all the information about Sir Edmund Hillary including relations, that are to be found on a complete world map, then it would make no sense to respond to that description with ‘But who was he?’ Presumably, only God could have a complete map of the world, but lesser beings might have lesser, but similar, maps. The world map of a human being would be restricted by being a map of the environs, both spatially and temporally, of a particular individual, so being regionally restricted, by running only to the present, and by having great gaps in the information for most of the objects represented in it. The primary information we possess about other objects is fragmentary, but, by augmenting this by common-sense assumptions about how the world works—for example, by assuming that the chair in the next room is not mysteriously replaced by an exact replica whenever I am not watching it—we can fill in many of the gaps. This pragmatic and implicit completing of maps is essential to give determinacy to our conception of, and hence our reference to, the objects in our map. With these assumptions of spatio-temporal locality and common-sense extrapolation in construction, we can imagine how a normal finite subject could possess what might be called a personally adequate world map. Such a map is sufficient to give its owner enough information about the different objects with which he has to deal for him to know which of them is which, and to which of them new pieces of information relate. So we are not in danger of confusing the major dramatis personae in our lives, even including the historical ones that we know about only by report. Our map also enables us to recognize new information that did not apply to anything already in our map, but which was not inconsistent with what he already knew, as indicating the presence of an object not so far recorded. We can then start up a new file or life-track. Thus world maps say which object our thoughts are directed on, for locating an object within such a map, in the standard case individuates it adequately. Does the existence of such maps depend on the existence of the objects they represent? That you cannot have reference without existence—that reference is an extensional, not intentional relation—is a dogma of analytic theory of reference, but there seems to be nothing in what I have said about the maps which requires that the world mapped be the actual world. Indeed, the notion of a complete map, with which I started, was equivalent to the description of any possible world; it seems very similar to what Plantinga calls a ‘world book’: or, perhaps, his world books are one possible form that a world map might take.⁴ There seems no reason why there should not be local maps and personally adequate maps of any world. There seems to be nothing to prevent the proverbial brain in a vat from having a

⁴ Plantinga (1974).

86

 

personally adequate world map. Having just the same kind of putative information that we have, he will be able to individuate the putative objects in his notional world just as clearly as we can, and will know which of them to attribute new information to in just the same way as we do. Given the possibility of non-existence, it would seem that these maps ought to consist of very complicated definite descriptions, with one for each entity, for such descriptions are not dependent on the entities they purportedly describe. But if descriptions lie at the root of all individuation in this way, does it not commit us to a descriptive theory of names, with all its notorious problems? The model of identificatory thought emerging, however, is not accurately represented as a set of Russellian definite descriptions. Although a map could be represented in this way at any given time, the dynamic process of its development must be conceived differently. If the map were thought of as consisting of descriptions through time, any new piece of information relating to an object already in the map would require that one identify the relevant description and ‘re-open’ it and add the information. As Sainsbury (2005: 125ff) has argued convincingly, the process of thought behind this could not be represented in Frege–Russell terms, for the subject would be recurring to the object anaphorically in a way that neither employed a logically proper name, nor was already within the scope of the brackets of a quantified definite description. A Fregean semantics models the language on a domain. Only if the speaker is acquainted with the domain does he fully understand the language. If the domain is the positive integers, or one’s own experience, then there is no problem, for the objects in the domain are given completely and, in a sense, a priori. But if the intended domain is the external world and if one’s understanding of what is in the world is in a state of permanent development, then the domain of one’s discourse is constantly changing. New names are introduced, but on the back of what had previously been descriptions. Suppose that at a certain point in the development of my world map, I get the information that there is something F at p, t. Later, as more information comes in, it becomes clear that this is not a phase of any of the other objects I have previously recognized, but is a new object of which I begin to build up a conception, and I finally reach a point at which it is natural to give it a name, say ‘Fred’. (This situation might be one in which I am coming to recognize and individuate animals in a wild environment.) If we think of the accumulated description as ‘the unique F’, we can say that the unique F is ð¼Þ Fred: But we say this as a way of introducing a new name and a newly grasped individual, not as a way of identifying the unique F with something already given. But neither am I saying that ‘Fred’, like Evans’ ‘Julius’ means whoever was

   ‘  ’  ‘  ’

87

the unique F.⁵ Objects are being discovered and names are being introduced on the backs of descriptions, but not just as going proxy for the descriptions, because the descriptive background is already enough to pick out the object on the speaker’s map. As Sainsbury shows, this process of development cannot be captured within a Frege–Russell conception of language. The language is modifying its domain or model as it progresses: formally speaking, this looks like simply stopping and beginning again with a new language. In fact the Fregean conception is analogous to Aristotle’s account of science: it is fine for expressing a completed theory but it cannot capture the rational structure of its development. The epistemological dimension is missing. One needs to replace the extensional approach to reference by one that treats it more like the other propositional attitudes, taking something much more like an intentional object. As Sainsbury (2005) says: . . . a referring expression is one that ‘purports to refer’: it needs to succeed if an unembedded occurrence of the expression is to express a truth; it may fail to refer without detriment to intelligibility; a correct semantic will associate it with a reference condition rather than with a referent. (130–1)

The reference conditions will, of course, be found in the dossier or file on the object in one’s world map. How this works is illustrated by what Mill actually says about the functioning of names, as we saw above. Modern exegetes assume that something with only denotation must function like a name in Frege–Russell logic and be in no way dependent on a descriptive background. This is clearly not Mill’s view. For him, although proper names are mere labels, they presuppose background information.

5.5 A Note on ‘Content’ and ‘Object’ We can return now to the topic of content and object which figured so centrally in our discussion of perception in the previous chapters. In contemporary debates in the philosophy of mind, ‘content’ is assumed to mean something like conceptual content, so asking whether a certain state has content is more or less equivalent to asking whether it is intentional. This seems to me to be an unfortunate use of the term. The content of a mental state is whatever constitutes its intrinsic nature. So if you are a sense-datum theorist, you will hold that such data are (part of) the content of certain mental states. If you hold that

⁵ Evans (1982), 31ff. Evans introduces the ‘descriptive name’ ‘Julius’ to designate whoever invented the zip.

88

 

mental states are all and wholly intentional, then all content will be intentional or conceptual, but that this is so is not part of what makes it to be content. The object of a mental state is something external to the state itself, which it is about, or putatively about. How you characterize the object may be a matter of choice. So if John occurrently thinks that Tom loves Mary, the content of his mental state is the proposition that John loves Mary. You could variously say that the object of his thought is John, is Mary, is loving, is John-and-Mary, is John’s loving Mary. A thought can be about a variety of things. The problems arise when or if one wants to say that the object in some way constitutes the content of the mental state. First, it is difficult to see how the variety of options that I have just illustrated, and which suggest that the answer one gives is interest relative, could apply if the factual matter of the intrinsic nature of the state were at stake. But this is not what the debate has focused on, which is how something external to the state can be a constituent of it. If the object exists then this can be answered by proposing a ‘direct realist’ theory of thought, according to which external things can be constituents of mental states; they become intentionally inexistent, where this last word signifies existing in the mind, the ‘in’ not being a negation and so ‘inexistence’ not being a kind of non- or shadowy existence. The real issue arises if the object of the thought—Zeus or the fountain of youth— does not exist. Then there both is and isn’t an object. There is because Zeus or the fountain of youth is what you are thinking about, and there isn’t because these things do not exist. The purpose of this chapter has been to show how our intentional states are supported by the furniture in our minds, without having to posit a primitive notion of intentionality which is attributed to individual mental states. Meaningful thoughts composed into mental files do the job by enabling one to have a conception of a putative individual, irrespective of whether it exists or not.

5.6 A Different Model of Intentionality for Sensations? Nevertheless, is it not possible that sensational states are intentional, but in a different style from the others? Sensible conceiving, it could be argued, is a sui generis kind of thing, analogous to intellectual conception. We have the intellectual concept of a colour—say ‘grey’—which is exercised in thought and a sensible version which is exercised in sense-experience. Is this impossible? This was essentially the thought of Crane and of Schellenberg, as we saw in Section 3.4, and we expressed doubt about what this strategy achieved, but now we can see a further problem, namely that sensory representation is always of putative particulars, but concepts are essentially general and individual reference is only achieved by a complex process. This is most easily understood, it seems to me, if what represents objects in sense-perception is itself a genuinely individual entity, an actual instance of the quality in question.

‘  ’    

89

So the idea of pure sensory conceiving, or representation in the intentional sense, misses a fundamental point. Intellectual conceiving of grey is essentially general, but sensory experience, veridical or not, represents putative particulars. If the grey patch in the centre of Fred’s visual field is the content of a representation, it is a representation of something directly in front of him in space; this is a putative particular, not a general content like the content of a conceiving. The following seems to me to be the most reasonable position. There are two ways in which a particular can figure in the content of a mental state. One is to be an instance of the kind of confabulated constructions that, I have argued, are the intentional objects of propositional attitudes. The other is to be an actual particular—a real instance of some property. Phenomenal contents are certainly not the former, so they must be the latter. Although the sense-datum theory remains unfashionable, I take it that both it and naïve realism are intuitively preferable to the intentional theory at least in the following sense. To say that, when I seem to see red then there is indeed some red of which I am aware, has a straightforward phenomenological and conceptual clarity not possessed by saying that I am in a state which visually represents red. But, as we have seen, Searle correctly recognizes, in perception things are presented and not just represented. Phenomenologically, then, visually representing red means representing it as being-present-to-me-in-a-way-that-makes-me-awareof-it. This seems to suggest that, phenomenologically the correct description of my experience, as a subjective phenomenon, is in terms of the experiential presentness of the sensible quality that characterizes my experience, namely red; but that, for the intentionalist, in some more basic sense, no red is really involved. As an account of experience, as opposed to an account of what is actually there in the outside world (we all agree that I can seem to see red and nothing ‘real’ and red be present) it is unclear what the disclaimer of the presence of red actually achieves. This seems especially clear if one takes into account the way in which contents, intentional or not, can occlude the actual features of the world, as argued above.

5.7 ‘Representation’ in a Reductive Sense In Chapter 3 I said that though the terms ‘intentional theory’ and ‘representationalism’ were often used interchangeably, there was a tendency for the latter to be exclusively employed if one were trying to give a naturalist or physicalist account of how intentionality works. There is nothing in this chapter so far that mentions explicitly physicalist accounts. But, if I have not mentioned explicitly physicalist theories so far, neither is there anything in what is above that explicitly says that any kind of dualism is involved in the arguments presented. What difference, if any, might bringing in physicalism make?

90

 

The account presented of how intentionality works for propositional attitudes is presumably not affected. It should not be controversial that bare sensation and perception does not depend on mental files in the way that thought does, and that it is presentational, not just representational. If physicalism cannot accommodate this latter, then it is inadequate to the phenomenology of perception, so let us assume that it can. But it was the fact that this presentationality is common to perception and hallucination that undermined the externalist or direct realist version of intentionalism that Searle, for example, espoused, and once the internalist version is established, then the rest of the argument seems to follow. So I think that the conclusion is that if physicalist theories can accommodate the phenomenology of perceptual experience, then the arguments presented in this chapter apply to them. If they cannot accommodate that phenomenology— then they are not worth considering in the first place!⁶

5.8 Conclusion: World Maps and Perception I have said nothing so far about the role of perception in the formation of personal world maps, as the discussion was entirely in terms of the kind of propositional information that goes into our mental files. But, of course, perceptual information is amongst the foundational material for any picture of the world we may have. As far as the idea of world maps in general is concerned, it might seem that any theory of perception might fill this role, but the putative perceptual information that there is a tree in front of me could be false, just as the putative perceptual information that there is a unicorn in front of me could—though not so certainly mistaken! The tree-belief could be part of an envatted brain’s world map. It is because of this possibility that philosophers who believe that we are ‘transparently’ aware of the external world are forced either to be intentionalists, or, if they are naïve realists, to delve into the opaque world of disjunctivism. We have seen, in earlier chapters, that both these theories fail. So the perceptual foundation has to be of the sense-datum or qualia kind. This leaves the challenge of showing how, from such a basis, our conception of an objective world, which lies at the foundation of our world map, can be developed. This task will be tackled in the following two chapters.

⁶ Of course, I do not think that they can cope with phenomenology, as I argue in Robinson (1982) and (2016), and elsewhere, but that is not the issue for the present context.

6 Objectivity How Is It Possible?

6.1 Introduction It seems reasonable to assume that our notion of an objective, public world, which exists independently of us, must somehow be derived from our perceptual experience. For what else tells us about the world outside of us? But what is it about experience that delivers to us the idea of an objective world? In this chapter I shall consider three accounts of this. Two of them are defended at length in the recent literature. The third has only recently been suggested—or perhaps, more accurately, revived—and has not yet, I think, been generally influential. It is this last account that is correct. The first account is found amongst a group of direct realists—who are also disjunctivists. Their claim is that it is only by being directly in contact with the physical world in normal perception that we can have a conception of such a world: if experience were in some way internal, then we could never even conceive of an external reality. John McDowell, Bill Brewer, and John Campbell are paradigms of this group. The second theory is that objectivity comes from the dependence of our concepts on the existence of a causal connection between events internal to the individual and the external world. Tyler Burge will be my stalking horse for this theory, but a theory of this kind is needed by anyone who believes in nonintentional qualia or ‘mental paint’ and who does not accept a version of the third—Humean—account. They think, that is, that the sensational or qualitative element in experience is connected purely contingently with the features of the world it refers to or represents.¹ One way of understanding the difference between these theories is that the former theory holds that connection with the world is essentially phenomenological; the second makes it essentially causal-semantic; the phenomenology plays either a minor role or no role at all.

¹ Contingency of this sort is easy to understand when one considers colour, because of the possibility of colour inversion, but in the case of shape and primary qualities in general it seems to make no sense. This will emerge later.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0007

92

:     ?

The third theory is that our conception of objectivity comes from the way our experience is organized: what David Hume called the coherence and consistency of our experience. I defended this theory in Robinson (1994 and 2013), it was taken up in Farkas (2013) and Cassam (2014). It is elegantly defended in Ayer (1963), but, as with most classical empiricist insights, it is wilfully ignored in much of contemporary debate.

6.2 Direct Realism and Objectivity The view that we can have a conception of an external world only if we perceive it directly fits naturally with the relationist theory, but, in fact, one of its strongest defenders is McDowell, who is an intentionalist. For present purposes, the intentionalist McDowell and the naïve realists/relationalists Brewer and Campbell are in the same camp because they agree in holding that it is only by being in direct contact with the world in perception that we can have any conception of such a world. They are all, as a consequence, disjunctivists. McDowell puts forward two arguments for the necessity of direct realism. The first is epistemological as follows: (a) Unless direct realism were true, we would have no answer to the sceptic about the external world, and so could not know anything about it.

There is a stronger and more important argument which is conceptual and not merely epistemological. (b) Unless direct realism were true, we could have no conception of the external world at all.

The former emphasizes the epistemic failure that is said to accompany the rejection of direct realism, the latter goes further and makes this into a conceptual failure. Despite the differences between McDowell, the intentionalist, and the relationists in this group, I think the rationale of their position can be summarized quite simply. The phrases in square brackets relate to interpretation (b) of how our grasp on the world would fail if direct realism were not true. The overall rationale is, I think, as follows: No inner event, whether conceived of as a physical ‘representation’ in the brain, or as a subjective sensation can be made to be [or present itself as being,] of or about something external simply in virtue of being caused (or standardly caused) by the

   

93

appropriate external thing. One might represent this thought by saying that the intentionality of our mental states cannot be simply reduced to an inner event plus a causal relation. On the other hand, these philosophers reject the idea that inner states can possess an intrinsic or ‘magical’ property of intentionality. So, they conclude, there is no way that mental events conceived of simply as internal could [even seem to] put us in touch with the world. Because these philosophers believe that our ‘intending’ the world can be treated neither reductively nor primitively, they conclude that the only way in which we can [even seem to] reach out to the world is if, paradigmatically, our perceptions encompass the world itself in direct realist fashion. McDowell states the case for (a) in the apparent absence of (b) most blatantly in his (2008). The fundamental claim is that the common factor theorist has no convincing reply to the sceptic, but the direct realist does. This is held to be so because scepticism concerning the external world is founded upon the assumption that we are never directly aware of anything except appearances in our own individual ‘Cartesian theatre’. What shapes this scepticism is the thought that even in the best possible case, the most that perceptual experience can yield falls short of a subject’s having an environmental state of affairs directly available to her. (McDowell, 2008: 378)

There is an obvious response to this claim, and Crispin Wright (2008) makes it very forcefully. It is that the foundation for scepticism is not the ‘veil of perception’, but our inability to tell whether a perception is of a real object or not. The fact that an experience is in fact directly of a certain object is no help against the sceptic unless we are in a position to tell that it is a real external object of which we are aware. When Descartes sets out the sceptical challenge he does not invoke the internal status of appearances, but only that we cannot tell whether we are really perceiving, or are dreaming or are the victim of an evil demon. Inability to distinguish the states is the core issue, not their ontological status and disjunctivism does nothing to enable use to distinguish when we are directly perceiving and when we are not. McDowell seems surprisingly oblivious of this, both as a historical point about Descartes and as a purely philosophical matter, even though the literature is replete with discussion of the issue. The typical example used to illustrate the point is the following. Consider a countryside in which there is apparently standing a large number of barns. Some of these apparent barns, however, are just very accurate facades, as on a film set. From the road, the passer-bye cannot tell which is which. Suppose that Tom is in fact looking at a real barn at the moment, he cannot be said to know that that is a barn, especially if he is aware that some of them are phoney. What theory of perception one adopts would not seem to make a significant difference to this situation.

94

:     ?

One response to Wright’s argument is to claim that there is an asymmetry between perception and hallucination of the following sort. Often when dreaming, one thinks that one is awake, so judgement about one’s cognitive state when dreaming is not reliable, but one’s judgement that one is awake when one is actually awake is entirely reliable—so the argument goes. Similarly, one’s judgement that one is perceiving when one is hallucinating is unreliable, but it does not follow that one’s judgement that one is perceiving when one is perceiving is similarly uncertain; it parallels the case of one’s waking judgement. The general principle is that if one’s cognitive state is disordered, than one’s judgement is unreliable, but if one’s cognitive state is not disordered, it is not unreliable. Sleep disorders one’s cognitive state and so does hallucination, whereas in a waking state or normal perceptual state, one’s cognitive state is not disordered. The efficacy of this response depends on whether one can assimilate sensory states to judgemental states. They are both, in a general sense, cognitive states, but the opponent of the assimilation of hallucination to dreaming must insist that, in the case of ‘philosophers’ hallucinations’ at least, one’s thought processes are not disordered in the way that they are when one is asleep. In many kinds of hallucination there is, no doubt, such disorder, but, even outside the philosophical case, there are cases where people have taken hallucinatory drugs and reported on their experiences in a way that shows they were fully in control of their thought processes, unlike a sleeper. That there is this kind of distinction, within one’s cognitive processes, between sensory states and thought is shown by cases that have nothing to do with hallucination. During an optical examination, one is asked to report on, for example, how blurred things look. If things look blurred, there is some disorder in one’s perceptual cognitive apparatus, but that one’s reporting of phenomena is just as reliable as normal is shown by the fact that the optician will modify the lenses, on the basis of one’s reports, and arrive at a prescription that gives one accurate vision. The relevance of this is that it shows that, though perception is part of one’s cognitive apparatus, an alert human can very easily be dysfunctional perceptually, and entirely lucid and in control of his intellectual faculties and will. One could just stipulate that an hallucination counts as a relevantly disordered condition, whereas short-sightedness does not, but in the imagined cases where the hallucination involves only interference with the perceptual apparatus, leaving the judgemental side untouched, this would be arbitrary, question-begging and disingenuous. The weakness of the argument for (a) alone suggests to me that the arguments in (1994) and (2008) are meant to be taken in conjunction with the earlier arguments for (b). The claim, therefore, is not just that direct realism alone defends us against the sceptic. A more indirect argument could be the following. The very fact that we have a conception of the external world shows that we must be directly in contact with it, and this implies that, paradigmatically, our

   

95

perception must be of the world and general scepticism false. Whether or not this latter conjecture is true, we must consider the argument for (b). Characterizing the common factor theory as the ‘Cartesian picture’, McDowell asserts his position clearly enough. The threat that the Cartesian picture poses to our hold on the world comes out dramatically in this: that within the Cartesian picture there is a serious question about how it can be that experience, conceived from its own point of view, is not blank or blind, but purports to be revelatory of the world we live in. (1986/1998: 243)

The main idea behind McDowell’s charge of ‘blindness’ is that an inner state cannot present itself as a representation of the outside world simply by dint of standing in some causal relation to that world. But why cannot it be taken to be or seem to be of an external world? According to McDowell it cannot. According to the fully Cartesian picture, it cannot be ultimately obligatory to understand the infallibly knowable fact disjunctively. That fact is a self-standing configuration in the inner realm, whose intrinsic nature should be knowable through and through without adverting to what is registered, in the innocuous position, by the difference between the disjuncts – let alone giving the veridical case the primacy which the innocuous position confers upon it. This makes it quite unclear that the fully Cartesian picture is entitled to characterize its inner facts in content-involving terms – in terms of its seeming to one that things are thus and so – at all. (1986/1998: 151–2)

I think that the argument is that, if perceptual contents are internal, it is unclear how they could even seem to be external. Why would they not just seem to be what they are—subjective images or sensations? Pains and itches don’t seem external, so why should—or how could—perceptions, if they are only internal images? One might reply to this by pointing out that hallucinations do not present anything actually external, but they can be exactly similar to perceptions, so apparent externality cannot be essentially tied to actual externality. The riposte to this is that hallucinations or misperceptions should be seen as essentially derivative: they can only occur by reactivating, so to speak, the kind of experience that has the job of putting us actually in contact with the world. We discussed this account of hallucinations in Chapter 1, and found it implausible, but it seems essential to any form of disjunctivism. A somewhat different line of argument to the same conclusion is presented by John Campbell. He takes as his starting point what he calls ‘Berkeley’s puzzle’,

96

:     ?

which is: ‘how can experience of an object explain our grasp of the possibility of existence unperceived?’ (2002a: 129). His answer is as follows: . . . I will argue that what makes it difficult to find the resolution of this puzzle is the difficulty of finding a clear view of the conception of experience that we need to understand how experience could be what explains our grasp of concepts. The two mistakes it is easy to make are, first, to suppose that experience is exhausted by its propositional content, and second, that it is only caused by the object it is of. I will argue that we have to think of experience of an object as a cognitive relation more primitive than thought about the object, that none the less makes it possible for us to think about the thing. I call this the relational view of experience. (2002a: 129–30)

By ‘experience . . . is only caused by the object it is of ’, he means those theories of perception which accept a common factor to perception and a corresponding hallucination, and which claim that what distinguishes perception from hallucination is simply the causal role of the perceived object in activating the common factor. The opposite of such theories is generally held to be disjunctivism and he first sets off to defend disjunctivism, along similar lines to McDowell. We cannot extract the conception of a mind-independent world from a minddependent image . . . It seems as though it ought to be possible, though, to extract the conception of a mind-independent world from an experience which has a mind-independent object as a constituent, which is what the disjunctivist view ascribes to us. (Campbell, 2002a: 135)

I think that McDowell is entirely correct in claiming that the fact that an inner state has certain causal relations cannot on its own constitute an adequate account of why or how the inner state is experienced as being of something external, but the opposite view is the one we shall be considering in the next section. It could be argued that the polemical rejection of intrinsic intentionality as ‘magical’, begs the question against intentionalist theories of perception. Campbell in his (2002b) has a different argument against the idea that sensory intentionality or representation (‘propositional content in a sensory mode’) could play the content. He argues as follows. The argument turns on the explanatory role of experience. Experience is what explains our grasp of the concepts of objects. But if you think of experience as intentional, as merely one among many ways of grasping thoughts, you cannot allow it this explanatory role. Suppose someone said: ‘Actually, reading newspapers is the fundamental way in which you understand the concepts of a mindindependent world. All your conceptual skills depend on your ability to read

 -   

97

newspapers.’ The natural response to this would be that reading newspapers does indeed involve the exercise of conceptual skills, but it is simply one way among many of exercising those conceptual skills. Just so, if all there is to experience of objects is the grasping of demonstrative thoughts about them, then experience of objects is just one way among many ways in which you can exercise your conceptual skills. At this point we do not have any way of explaining why there should be anything fundamental to our grasp of concepts about experience of objects. (2002b: 122)

This seems to me to be a very strange argument. It is that if perception is not the only intentional state, then any of the others could equally well fill the role of giving us the grasp on our concepts of objects, but no reason is given why this should be so. Different intentional states have different modes, according to their proponents, and an essential feature of the sensory mode is that it is presentational and not just representational, and it is its presentationality that makes it ‘fundamental to our grasp of concepts about experience of objects’.² In his debate with Cassam, Campbell seems to claim that perceptual intentionality would have to consist of propositional content plus an accompanying sensation, the latter part free of intentionality, and Cassam rightly points out that this is a question-begging assumption: why cannot the sensory content partake of the intentionality of the state? (2014: 48 and 171). The only argument that these direct realists present for their account of the foundations of objectivity is that something internal could never do the job, either on its own (by being essentially intentional or by being a sense-datum) or in virtue of the existence of a causal relation to the external world. We must look further at these alternatives, bearing in mind that, in the last chapter we found the disjunctivism to which these direct realists are committed is deeply implausible. If either the Burgean theory, or the Humean one works, McDowell and his cohort will be in trouble.

6.3 The Causal-Semantic Account of Objectivity Tyler Burge summarizes his view as follows: The objectivity of perceptual states is formed partly through patterns of nonrepresentational causal relations to a mind-independent environment, and partly by subindividual objectifying operations in sensory systems that distil the

² We saw this in Chapter 3 when discussing Searle and Crane. I also argued there that this feature of experience grounded the ‘blocking argument’, which cast doubt on the difference between intentional theories and the sense-datum theory, but this is irrelevant to Campbell’s argument.

98

:     ? environmental from the sensory effects of proximal stimulation. Both of these sources of objectivity are blind. The patterns of causal relations are part of blind nature impacting the lives of simple organisms and of the blind primitive agency of organisms on nature. The objectification is not an exercise of individual agency. The developmental, phylogenetic, psychological, and constitutive sources of objectivity in perception lie below the level of individual representation, control, awareness, or responsibility. . . . Individuals do not construct objective perception from subjective representation or consciousness. (2010: 547)

It is important that the source of objectivity is very primitive and sub-personal, but it shares features with both the direct realist theory we have been discussing and the Humean theory which we will discuss next. It shares with the former insistence on the role of the ‘mind independent environment’ which is ‘distilled’ from the ‘sensory effects of proximal stimulation’. In other words, by virtue of such phenomena as perceptual constancy, it is the distal cause, not the proximal cause that comes through as the object of perception. On the other hand, Burge agrees—unknowingly—with Hume that it is the patterns of the stimuli that enable the perceiver to take them as objective, not the mere fact of their externality. Where he differs from both the other theories is that ‘subjective representation or consciousness’ play no role; the grounds of objectivity are to be found in creatures too primitive to be conscious, and, even in our case, operate at a sub-personal level. I shall discuss later whether what Burge says need be controversial for those who hold a different theory, or whether the causal mechanisms he describes are only the background to the problem philosophers are usually discussing. First let us consider why he rejects the naïve or direct realist approach. Burge dismisses naïve realism because he has no patience with the disjunctivism that necessarily accompanies it. Disjunctivism is implausible. Not only common sense but . . . scientific knowledge . . . support this initial evaluation . . . Disjunctivism entails that token distal differences, in the causal chain leading to perceptual states, that make no relevant difference to proximal stimulation, or to other causal processes that provide input to our perceptual system, or to antecedent psychological states, determine differences in perceptual state types . . . This view is not only undermined by scientific knowledge. It controverts wellentrenched views about the form of causal explanation in psychology. (2005: 27)

This attack on disjunctivism is entirely consonant with the arguments above in Chapter 1.

 -   

99

Burge, consequently, is some kind of a common factor theorist. It might be thought that he, therefore, has two options. He might attribute to this common factor an intrinsic intentionality by which it represents, or purports to represent, the external world. Or he might be a traditional representative realist, like Locke or Mackie, according to which the common factor is a non-intentional sense-datum. In fact he explicitly rejects both these positions, on the grounds that the internal common factor, which he identifies with the basic phenomenology, is inadequate to explain how experience reaches out to the world, whether or not you are an intentionalist about it. . . . attempts to individuate perceptual states in purely phenomenological terms fail to provide any insight into the intentionality of perceptual states. Either they help themselves to representational power associated with our phenomenology, without explaining that power, or they offer a feature, phenomenology-withrepresentational-content-bracketed, that does nothing to explain representational power. (2005: 20–1)

This rejection is not surprising, because both the positions he is here rejecting are individualistic: they both allow perceptual experience to be conceptually independent of the external world, and, as Burge implies in the last quotation, he believes that individualistic theories fail to explain how representation works. But prima facie Burge is in a position which leaves him some explaining to do. Common factor theorists are normally or naturally representationalists of a sort that seems to require an internalist approach to perceptual experience. That is why perceptual externalists tend to be disjunctivists. It is interesting to follow those respects in which Burge goes along with internalism (or, as Burge prefers to say, individualism) in the nature of experience, and those respects in which he rejects it, and to see whether he manages to reconcile these two tendencies. Burge does not deny that something rather like a brain in a vat—which possessed a central nervous system like ours—could have a phenomenology like ours, but does deny that this would constitute an adequate basis for its having perceptual-type experience. Thus, when arguing against Brian Loar’s ‘brain in a vat’ scenario, he considers the following possibility. Suppose that we . . . imagine a system that is molecule for molecule homologous to our brains, but came together as a cosmic accident. I have little confidence about how to imagine such a being from the inside or outside. At least in its first moments, it would seem to lack most of the cognitive and perceptual systems that I have. I am inclined to think that it would have similar qualitative, phenomenal ‘feels’, since I conjecture that certain qualitative aspects of the mind depend purely on the underlying chemistry. But at least until it has

100

:     ?

interacted with its world, I do not think that it has any ‘outer-directed’ intentionality. I think that it does not even have a visual system until it has interacted with its world. One can be easily confused in phenomenal exercises. One can imagine that things would ‘look’ just the same to the homologous accident. Such an imagining would be corrupt. The notion of ‘look’ already depends on presumptions of perceptual and conceptual content that I believe are illegitimately imported into the envatted accident. The processes in this thing at first lack meaning and function. At most the individual would have similar phenomenal features. (2003: 444)

Thus Burge distinguishes between the phenomenological component in perceptual experience, and the representational. To put it in the common jargon, he distinguishes between phenomenology and content, and denies that there could be the latter without there being, paradigmatically, a genuine causal relation to the appropriate feature in the external world. The issue is to consider whether Burge manages to reconcile internalism about phenomenology with externalism about perceptual experience: does Burge manage to reconcile his claim that science does not allow distal causes to have an unmediated influence on our mental states with their externalist role in influencing ‘how things look’?

6.4 Burge on Distil Causes and the Experience of ‘How Things Look’ At the outset, I presented the following quotation from Burge. Disjunctivism entails that token distal differences, in the causal chain leading to perceptual states, that make no relevant difference to proximal stimulation, or to other causal processes that provide input to our perceptual system, or to antecedent psychological states, determine differences in perceptual state types . . . This view is not only undermined by scientific knowledge. It controverts wellentrenched views about the form of causal explanation in psychology. (2005: 27)

Later, I quoted what he said when discussing the spontaneous ‘brain in a vat’ case. One can be easily confused in phenomenal exercises. One can imagine that things would ‘look’ just the same to the homologous accident. Such an imagining would be corrupt. The notion of ‘look’ already depends on illegitimately imported into the envatted accident. The processes in this thing at first lack

   

101

meaning and function. At most the individual would have similar phenomenal features. (2003: 444)

These two quotations seem, at the very least, to be in tension. It is difficult to see how he squares the latter with the scientific realism he castigates the disjunctivist for ignoring. The difference between ‘phenomenal feel’ and perceptual experience seems to be both genuinely phenomenological and dependent on distal facts directly, that is to say not through any additional change they make in the proximal states in the central system. It is natural to take how things look to be something experiential, so if things look differently, or fail to look at all, then it is experientially different for the subject. But Burge says that how things look depends on distal factors—what has been the standard, typical, or appropriate external cause, whereas in the first quotation he says that the idea that distil causes that make no difference to proximate stimuli should influence the perceptual state goes against science and psychology. Burge is working with a distinction between ‘phenomenal features’, sometimes called ‘phenomenology’, and ‘how things look’, which is part of normal perception. (But notice that Burge says in the passage quoted criticizing disjunctivism that distil causes cannot directly influence the perceptual state, not just the phenomenal contents.) Ex hypothesi, the brain in a vat has the same proximal states as the normal subject, but Burge wants to claim that whether these give rise to a proper perceptual ‘looks’ depends on their having, directly or indirectly, the right aetiology. Burge might be thought to have the following way out of the problem. It might be argued that he is not a simple common factor theorist, in the following way. A straightforward common factor theorist would claim that the perceptual experience as a whole depended, for its phenomenology, entirely on the brain. This is what Burge appears to be affirming in the quotation from (2005; 27) above. But he appears to believe that less central factors do affect the phenomenology: a phenomenologically indiscernible hallucination, produced by direct stimulation of central areas of the brain, rather than through visual pathways, might not even count as a perceptual state.

His reason for this is that for a genuine perceptual duplication, it must be the case that relevant afferent and efferent internal processes that provide input to the perceptual system are the same. (2005: 25)

This means that we must imagine the BIV as being more than a brain—more like a total nervous system. I cannot see that this makes any difference: imagine the whole nervous system in a vat and the stimulation provided by the evil scientist to be peripheral or central, as required. This may not infringe the ‘same proximate

102

:     ?

cause, same immediate effect’ principle, provided that the extended process in the nervous system can be regarded as a complex proximate cause of the experience. But this does not help with the present problem, for it only extends the internal states that are relevant; it does nothing to accommodate the role of distal causes. The reason why ‘looks’ depend on the right aetiology is because how things look depends on intentionality, which is a semantic phenomenon. This might seem to force one to conclude that, for Burge, how things look is not a matter of phenomenology, experience, or ‘how it is for the subject’ at all, but of a more abstract, purely conceptual, feature of the experience. It is simply that a certain experience only ‘counts as’ being as of an external object if it has a certain aetiology, though this makes no difference at all to how it seems to the subject. Apart from being massively counterintuitive, it would also be a perverse use of ‘looks’, which is a ‘how it is for the subject’ term, if anything is. It would also go against the general ‘common sense’ tone of Burge’s exposition of his own theory. It is worth noting what seems to me to be a particularly blatant case of trying to have it both ways. When discussing a ‘Twin Earth’ type of case in which the same internal states are associated with different distal stimuli, called ‘O’ and ‘C’, he says of the counterfactual—C—case: They regularly obtain information about instances of C; and we may imagine that their physical movements and discriminative abilities are quite different from the ones they have in the actual circumstances. Only the protagonist’s body, non-intentionally and individualistically specified, need remain the same. (1986/2007: 204: italics added)

How can the agent’s physical movements be different if his internal physical states are all the same? His behaviour might (somewhat tendentiously and misleadingly, in my view) be differently characterized, but the sheer movements cannot differ if his internal states are the same and the external world is, as in a Twin Earth, only indiscriminably different. It is again unclear how Burge can hold the internal fixed and really alter the mental, externalistically conceived. Burge has, therefore, failed to make a plausible case for making a radical distinction between phenomenology and how the world looks. It follows that, if the phenomenology is internal, and if our conception of the world depends on how it appears in perception, then somehow phenomenology must have the resources to generate our conception of the external world.

6.5 The Transition to Hume I want to contrast this line of thought with the one I would attribute to the classical representationalist or Humean. It is useful to help ourselves to an expression of

   

103

Sellars’s, namely the Manifest World (MW). MW is the world of common sense and of the naïve realist. Colours are essentially as they seem, shape, space and time are as they seem. This contrasts with the Scientific Image of the world, which lacks secondary qualities and in which, if Russell is right, space and time are nothing like the way they seem: the scientific world and the manifest world share a formalizable structure capturable in mathematics and that is all. One can hold to the contrast between the manifest and the scientific without accepting Russell’s extreme position: one could follow Locke in thinking that there was qualitative resemblance on the primary quality level. With this jargon in place, the unBurgean argument goes as follows. (1) (2)

Our perception of the MW is grounded on a projection of our qualia. Qualia are internal/individualistic.

Therefore (3) Our perception of the MW is grounded individualistically. This argument seems clearly valid. Burge is an internalist about qualia and phenomenology, and so would accept (2), so, given that he would not accept (3), the disagreement must concern (1). Burge’s view runs as follows: (4) The same internal states—dispositions, phenomenology etc—can represent different external things in different worlds or scenarios. (5) A perceptual state is a perception of what it represents. Therefore (6) Perception is (in part, at least) externally constituted. It is in fact (1) and (4) that are in conflict, for if the-world-as-we-perceive-it is a projection of qualia, it cannot be systematically true that the same qualia could represent different features of a MW. We seem to be on familiar ground. This is the same territory as the claim that, if you see grass the way I see fire-engines, then your red qualia represent green. This would be the Burgean option. The alternative view is that we represent the MW as being significantly different, though we would never realize this. I think that the second option is compulsory, if one accepts that there is a MW at all, as I think we must. Why does Burge accept (4)? In Burge’s own exposition of the line of argument I have given to him, his first premise is: (7) I begin with the premise that our perceptual experience represents or is about objects, properties and relations that are objective. That is to say, their nature (or essential character) is independent of any one person’s actions, dispositions, or mental phenomena. An obvious consequence is that individuals are capable of having perceptual representations that are misperceptions or hallucinations . . . To put this consequences with some gesture of precision: for any given person at any given time, there is no necessary function from all of that

104

:     ?

person’s abilities, actions and representations up to that time and the nature of those entities that that person interacts with . . . (1986/2007: 198–9) From this premise Burge concludes as follows. My first premise [which is (7) in my reconstruction] gives us this much. The objectivity of the objects of perception entails that there is always a possible gap between the proximal effects of those objects on an individual’s mind or body . . . and the nature of the objects themselves . . . The same proximal effects, representations, thought, and activity could in certain instances derive from different objective entities.

He combines this gloss on (7) with his second premise, which is equivalent to (5). This is that perceptual representations ‘specify particular objective types of objects, properties or relations as such’. From (7), glossed as in the quoted passage, and the idea that representations are essentially of objective external features, he concludes that the same proximal effects could represent different entities—that is, (4) in my exposition of his argument. Burge’s transition of thought is that the objectivity of things that we perceive allows for the possibility of error and that this generalizes in a ‘Twin Earth’ kind of way to the possibility that the same internal representation could systematically represent different things. This transition might be challenged in two quite different ways. First, it is not obvious that the same basic dispositions to bodily movement could fit different primary quality features. (Bennett’s contrast between colour blindness or spectrum inversion, which can pass unnoticed, and size or shape blindness, which would be radically disruptive, is relevant here. You cannot map the body movement appropriate to getting through a large rectangular door onto climbing through a small round porthole. So, too, is the passage I remarked on at the end of the previous section, where Burge seems to think that the same internal state could give rise to different bodily movements.) Second, and more salient from my point of view, is the fact that Burge’s theory seems to omit the manifest world altogether. On Burge’s theory, how the world looks is a function of what it is objectively like, not of the nature of the qualia it gives rise to. Just as contrasting qualia could equally represent green, so it is not the qualitative nature of our visual or tactile shape qualia that condition our common sense conception of space, but what space is actually like, for that is what determines, according to Burge, how the world spatially looks—for remember that Qualitative or phenomenological features of perceptual states do not in themselves bear any explanatory relation to the environmental properties that perceptual states represent. (2010: 76)

     

105

It is as if Burge is combining a radical Russellian view that the world may share no qualitative components with our subjective experience, with a kind of direct realism according to which we perceive it as being the way it actually is. In defending his second premise, that perceptual representations ‘specify particular objective types of objects, properties or relations as such’, he has to insist that Perceptual representations do not all have the contents like those of ‘whatever normally causes this sort of perceptual representation’ – or ‘whatever normally has this sort of perceptual appearance’, where the description denoted some objective property. (1986/2007: 199)

Such expressions would be appropriate on Russell’s theory for denoting the properties in the scientific image of the world. Burge seems to think that our perception, through our semantics, captures these properties directly, although they cannot be thought of as mirroring the qualitative nature of our subjective experience. By contrast, the manifest world draws its qualitative content from the phenomenal content of experience. According to Burge, a world of experience cannot be constructed in this way, so the manifest world is lost. There are only phenomenal contents, which cannot constitute anything even seemingly objective, and the world of objective—presumably scientific—fact, which perception represents. This is blatantly not how experience is.

6.6 David Papineau and the Manifest Image David Papineau, like Tyler Burge, has a purely causal, contextual account of the representational aspect of perception; and, like Burge he believes that there are internal qualitative states that make perceptual experience conscious, but these have nothing to do with the representational work of perception. I have argued above that this combination entirely deprives one of the manifest world, and, without the manifest image, there is no way to the scientific image. But, in certain ways, Papineau’s position is more equivocal than Burge’s. On the one hand, he says the following. Imagine a cosmic brain in a vat, a perfect duplicate of my brain that coagulates by happenstance in interstellar space, together with sustaining vat, and proceeds to operate just like my brain for some minutes, with the same sensory cortical inputs, motor cortical outputs, and intervening neural processes. I take it that this being would share all my conscious sensory experiences. Yet its sensory states would represent nothing. They would lack the kind of systematic connections with worldly circumstances required for representational significance. They

106

:     ?

would no more have representational content than would the marks ‘Elvis Presley once visited Paris’ traced by the wind on some mountain on Mars. (2021: 6)

This account looks to be very close to Burge’s, making perceptual content dependent on relations to the external world. But later he says other things which seem to be in tension with this, and to connect with our discussion of intentionality, especially in Chapter 5. Katalin Farkas (2013) defends an account of objectivity essentially similar to the Humean account that I am about to defend. Sense contents are not individually intentional, but it is their ordering in experience, not the causal relations to external objects, that leads us to take certain of our experiences as presenting us with external objects. Papineau draws a comparison between this and the cosmic brain in a vat he had mentioned earlier. Speaking of his cosmic brain he says: I take it that this being would have just the same subjective experiences as I do. In particular, elements of its experience would display the kind of apparent mindindependence Farkas draws attention to. They would display all the constancy, predictability, and temporal variation that give rise to the appearance of mind independence. Yet in the case of the cosmic brain, there would be nothing further behind this apparent mind-independence. It would consist of nothing but happenstantial patterns in the intrinsic qualitative properties of the cosmic brain. Farkas does not disagree. The aim of her paper is specifically to identify a structure within experience that makes it seem to point beyond itself. She explicitly brackets any further questions about the relation of experience to things outside the experiential realm. (93)

Papineau seems to be saying that, from the point of view of the subjective phenomena, there seem to be public objects, but subjective experience cannot be said to represent these things. He characterizes these as ‘quasi-objects’ with ‘quasiproperties’, but does not want to call them ‘intentional objects’ because they are essentially internal, and so are not candidates for being things that perception is ‘directed at’ or ‘pointing at’ (93ff). It might seem that the structured phenomena are putatively of objects in our manifest world, the world of common experience, but the fact that they might not exist discourages Papineau from wanting to treat them as being represented as intentional objects. He seems to align himself with those who think you cannot refer to, and, hence, cannot represent, non-existent objects (101–2). This is a view that I believe we have shown to be implausible in Chapter 5. It might, nevertheless, seem that talk of ‘quasi-objects’ and ‘quasi-properties’ is just Papineau’s way of characterizing the ‘manifest image’ world of common

  

107

sense, for the manifest image is very different from the scientific image, which, one might think, is nearer to what the world is actually like, so the manifest world only ‘quasi-exists’. He does not allow this, however, because he denies that these quasi entities have the properties that common-sense attributes to objects. Experiences do not have the same properties as the physical objects they represent. They are not themselves round, or yellow, or to the left of each other, in the way observed physical objects can be. (95)

Papineau says that this is why he is not a sense-datum theorist, for sense-data are yellow, etc., and they do not fit into ‘any sensible ontology’—by which he means physicalism (29–30). This point could be a merely verbal one, asserting, that the proper semantics of, e.g. ‘yellow’ is not the way that it appears, which is what the naïve realist calls ‘yellow’, but whatever lies behind this phenomenon. This is the view of Reid, and Peacocke, who provides such label as ‘yellow*’, but who, I think, does mean this to be like the naïve realist yellow. But Papineau’s point is not as innocent as this. It is much more the case that the properties of the experiential states are things that we are not directly aware of at all, and this is because of Papineau’s physicalism. Papineau adopts the phenomenal concept strategy, albeit in a somewhat modified form to reconcile conscious experience with physicalism (129ff). This is not the place to follow up the physicalism issue, as I discussed the phenomenal concept strategy, including one version of Papineau’s deployment of it, in Robinson (2016; 73–92). In sum, I think that we can say that Papineau’s position is like Burge’s, as far as the nature of our perceptual relation to the world is concerned, but is made even more elusive by his attempt to give an explicitly physicalist account of the subjective qualitative states, which I do not think Burge tries to do.

6.7 Constancy and Coherence: the Humean Account of Objectivity Neither the direct realist account, nor the causal semantic one are at all plausible. How does the Humean account fare? The two contemporary theories completely ignore the most important classical account of how we come to take most of our data as external, namely David Hume’s. Hume’s account of how our conception of the world is built up from our experience of sensible qualities in ‘On scepticism with regard to the senses’. Hume argues that

108

:     ?

the opinion of the continu’d existence of body depends on the COHERENCE and CONSTANCY of certain impressions . . . (130) Objects have a certain coherence even as they appear to our senses; but this coherence is much greater and more uniform, if we suppose the objects to have a continu’d existence; and as the mind is once in the train of observing an uniformity among objects, it naturally continues, till it renders the uniformity as complete as possible. The simple supposition of their continu’d existence suffices for this purpose, and gives us a notion of a much greater regularity among objects, than what they have when they look no farther than our senses. (198)

In other words, given that the ordering of the sensible qualities is such as it is, then there is no other way of making sense of these patterns other than to take them as being of enduring, external physical objects. Whether this process be a rationally defensible one—an implicit argument to the best explanation, for example—or brute irrational habit and association, is another matter. Despite the title of the chapter in which Hume introduces this account, one does not have to be either a phenomenalist or a Humean sceptic to think that this is essentially a true account of how we come to take most of our experiences as objective. Hume’s scepticism and phenomenalism rest, not on his account of how we form our conception of the objective world, but on his sceptical approach to causation and explanation. Hume’s scepticism and seeming antiquated mode of expression does not validate the fundamental point which we can state as follows: Given the natural rationalizing dispositions of the human mind, (Hume’s ‘tendencies of the mind’) the presentation of sensible qualities in the kind of ordering in which they are presented, is both a necessary and sufficient condition for their being taken as being experiences of a physical world. That such an ordering in experience is necessary for objectivity can be seen by considering the problems facing the naïve realist theory we discussed above. The naïve realists argue that, if we did not perceive the external world directly, we could have no conception of such an external or objective world. In McDowell’s striking phrase, all would be ‘darkness within’. But that direct contact with the external world is not enough to ground a conception of it can be seen by considering the following case. Suppose a normally blind person very occasionally had a direct realist visual perception of the external world, but that these experiences were too fragmentary for him to connect them with his normal tactile and other perception of the world. As far as he was concerned, they would just be odd sensations that he had. The fact that the content is actually external would make no difference without the presence of the appropriate structure and order. So ordering within experience

  

109

seems necessary for that experience to be taken as objective. It is tempting, too, to regard it as pragmatically sufficient. How could a creature of roughly human construction and intelligence have experiences ordered in the way ours are, without coming to interpret them as being of a physical world? We cannot imagine any other way of rationalizing—rendering manageable and projectable—our experience, taken qualitatively, than the physical interpretation we give to it. This also shows the unreasonableness of the way Papineau plays down the Humean approach. Howard Robinson, following Hume, holds that humans would be unable to form any conception of an objective world beyond the mind were it not for the ‘constancy and coherence’ displayed by elements in their sensory experience (Robinson 1994: 232; 2013: 322–3). This is an interesting suggestion, but I take no view on this issue in developmental psychology. (94, n. 8: italics added)

The fact that such order is a necessary condition for the possibility of taking experience to be of an external world remotely like ours, and that, given certain capacities which it is hard to imagine human beings lacking yet surviving, it is sufficient for our forming that conception, it seems unfair to take it as merely a piece of speculative developmental psychology. The account is, in a sense, genetic, but it operates within philosophical constraints concerning what is necessary to make sense of how the process achieves its objective. It is not one speculative theory among others. It is, I think, worth emphasizing that almost no consideration was given in the contemporary debate about our grasp on the notion of objectivity to the broadly Humean approach, until very recently, despite its vintage and massive intuitive plausibility. I brought the theory up in Robinson (1994: 232–3) and in (2013) and Farkas developed the idea in her (2013). The latter source prompted Cassam to deploy it against Campbell in his (2014). Cassam sees Kant as having a similar notion, not seeing that the idea is fundamentally Hume’s. (Kant’s categories and the like seem often to be pompous reifications of Hume’s ‘tendencies of the mind’.) Farkas expresses it in the following way: I pointed out a difference between ordinary perceptual experiences on the one hand, and experiences that don’t seem to present experience-independent objects – afterimages, phosphenes and others – on the other. The difference was that the simpler phenomenal features of perceptual experiences are organized into a systematic, cross-modally coherent and predictable order. This order is what I call the ‘structure’ of the experience, and I shall suggest that this is responsible for the phenomenal appearance of an experience-independent object. (2013: 109)

110

:     ?

The role of constancy and coherence in establishing objectivity is nicely illustrated by Ayer who shows how different forms of organization characterize the private and the public: Tables are public; it makes sense to say that several people are perceiving the same table. Headaches are private: it does not make sense to say that several people are feeling the same headache . . . But if people were so constituted that they were communally exposed to headaches in the way that they are communally exposed to the weather, we might cease to think of headaches as necessarily private. A London particular might be a local headache as well as, or instead of, a local fog. (1963: 51) If it were an empirical fact that people in a given neighbourhood habitually felt the same, or, if you prefer it, similar, feelings on the same occasions, we might come to think of such a feeling as an object, analogous to a physical sound, which pervaded the region and existed independently of any given person’s feeling it, perhaps even independently of its being felt at all. (77)

Cassam, unlike Farkas, is an advocate of the intentional theory, but intentionalism and the Humean insight sit uneasily together, for, if experience was intrinsically intentional a single perceptual datum would convey externality, rendering appropriate structure unnecessary. Pautz, for example, condemns what he calls ‘inner state views’, which would include the Humean view, because they contradict the fact that experience is ‘essentially “externally directed” ’. There is, I think, however, confusion in his exposition of this idea. He defines it as follows: External Directedness. Necessarily, if an individual has the ball experience, then that individual has an experience as of a bluish and round thing in a certain viewer-relative place, even in hallucination cases where there is no such thing . . .

But the scenario as set up implies that it has a context of other experiences—this is how one is aware of a ‘viewer-relative place’. Everything Pautz says is consistent with externality and objectivity in experience being a product of structure and organization.³ It might be objected, on Pautz’s behalf, that it is a problem with the Humean theory that there is no way of specifying what the order in experiencing we are

³ It seems to me that in his struggle with Campbell, Cassam comes out clearly victorious, but his arguments against relationalism could be taken as arguments for the sense-datum theory, as for the intentional theory. All except, that is, this argument from structure in defence of objectivity, which manifestly makes the intentional theory redundant and favours sense-data.

  

111

talking about is, except as being that order which is appropriate to the physical interpretation, which would mean that the experience must be taken to be of something external in order to detect the structure. It seems to me, however, that this is not an objection, unless it is supposed to have the consequence that there could not be experiences qualitatively like the individual ones in that order unless they presented themselves as being of physical objects. And this has just been argued to be false, by showing that the ordering itself is necessary for the physical interpretation: that is, it is not intrinsic to experiences unsupported by others. The fact—if it is a fact—that there is no other way than via physical object concepts of specifying the structure does not show that the elements in the structure can only be identified in a way that already commits one to that interpretation. So, the burden seems to me to be on the anti-Humean to prove that order of the right kind could not be sufficient to suggest externality, especially as it is clearly necessary, which, it ought not to be on either the relationist or intentionalist accounts. I said at the start of the discussion of Burge that he shared with the Humean a belief in the essential role of patterns in the way we are stimulated. ‘Patterns’ and ‘constancy and coherence’ signify essentially the same thing. But for Burge this coherence is not a feature of conscious experience and for Hume it is. What is the significance of this apparent disagreement? I think both parties (and perhaps everyone, including the naïve realists) could agree about the following. (a) If it had not been for the kinds of mechanisms that Burge describes in primitive creatures and those subpersonal mechanisms we possess as an evolutionary result of them, then our conscious experience (if there were any at all) would not manifest Hume’s coherence and constancy. (b) If one wished to build a robot that ‘perceived’ and interacted with its environment, it would need to embody a structure of the kind Burge describes. I said in my formulation of the Humean theory that it followed given the natural rationalizing dispositions of the human mind, (Hume’s ‘tendencies of the mind’)

that appropriate ordering of experience would give rise to our conception of an external world. Burge’s theory merely fills in some of the details about the genesis and grounding of those ‘tendencies’ and ‘dispositions’. Let us suppose that the mechanisms Burge postulates exist, but that they happen not to give rise to a Humean constant and coherent conscious experience. Could we, then, as conscious beings, have a conception of an objective world? I do not see how. There might seem to be two questions, the first of which Burge answers.

112

:     ?

(i) What is the general structure of the underlying mechanism that makes perception of an external world possible? The other question is phenomenological. (ii) What is it about our conscious experience that makes it seem to us that certain features or parts of that experience are of external, mind-independent objects? Burge denies that there is any answer to (ii), because conscious experience does not have that role. Perhaps an argument of Child’s might seem to give the materials for a reply to the Humean strategy. Child starts by invoking the Russell-Evans account of singular thoughts, and remarks that . . . one would expect discussions of the issue between Russellian and nonRussellian conceptions of singular thought to bear more of less directly on the question about experience. (1994: 146)

He fills this out shortly afterwards as follows. If knowledge is based on experience, and the intrinsic character of experience is world-independent, then it is impossible to see how we could have knowledge of the world beyond our experiences. One might hope to bridge the gap between experience and world by appeal to a theory, or by inference to the best explanation of our experiences. But that hope is blocked when we consider the related line of thought about content. According to that line of thought, one cannot have thought about Fs unless: either one is (or has been) in direct cognitive contact with Fs; or one can construct a way of thinking about Fs from concepts of kinds of things with which one is (or has been) in direct cognitive contact. Now on the non-disjunctive conception of experience we are not in direct cognitive contact with the world, since the most basic mental characterization of experience is world-independent. But it is arguable that no concept constructed solely from world-independent contents can itself be a concept of an objective world independent of thought: if that is right, then no theory, or inference to the best explanation, could get us from experience conceived of as a highest common factor to thought about the world. (1994: 147)

The Humean strategy is not an argument to the best explanation, because it is not an argument. It is an explanation of how we come to take experience objectively, not a justification of so doing. That comes later and only in a philosophical context. But that is not a major point. (Child is not explicitly discussing Hume

  

113

in the passage I have quoted.) Child claims that it is a general truth that there is no reference without direct contact, so if direct realism is not true, our perceptual thoughts could never reach to the external world. There are two responses to this attempt to employ direct theory of reference in the philosophy of perception. First, the theory of reference is not an epistemological theory. It does not say that thought content fails if you do not know that there is an object there, only that it fails if there is not. The directness in question is not epistemic or conscious, but entirely de re. The direct theory of reference does not tell against a causal theory of reference. Second, and I think more interesting, is the fact that Child’s argument rests on the orthodox claim that there can be no reference without a referent—reference is extensional, not intentional. This dogma is more and more under attack, as we saw in Chapter 5. One can have a conception of an object—Venus or Vulcan—whether or not it exists and such conceptions are what sustain referential acts. The Humean ordering in experience gives us a conception of a world, whether or not it is there, and gives content to our referential acts, whether or not they are successful, in the sense of hitting an actual target. Given the existence of the right kind of coherence and consistency in the phenomena, there is no issue of how our experience manages to represent the world. In a sense, the situation is just the opposite: the world of common experience—the manifest image world or the world of naïve realism—is a creature of our phenomenology. What it is like to see red, and what red is (naively) like, what it is like to see, head on, a square patch and what squareness is visually like, these are two sides of the same coin. It could not be further from the truth to say, as Burge does above, that Qualitative or phenomenological features of perceptual states do not in themselves bear any explanatory relation to the environmental properties that perceptual states represent.

Beyond the issue of ‘the right kind of ordering’, I do not see why there should be any problem about how representation gets into perceptual experience. To suggest that there is a problem here is to ignore the difference between what one might call the iconic representation in perception and the digital or conventional representation in thought and language. An objection that one might put to the Humean account is that, unless it is combined with direct realism, it is an error theory. This is an objection made, for example, by Martin against non-direct realist, or internalist theories: if the phenomena are not themselves external, then our taking them to be so is an error. This is an argument that I shall try to rebut in Chapter 8, but it should be said at this point that it is not obviously true of intentionalism: if the internal experience can be characterized as being essentially a representation as of an external thing,

114

:     ?

then we would not be in any kind of error in taking our experience to seem to be presenting something external. This issue, and the limitations of the response to the charge of ‘error’, will emerge in the discussion of intentionalism in Chapter 8. There is still one important question that the Humean theory has not yet answered, namely, that of actual, as opposed to seeming objectivity. The Humean account says what can make it seem to one that there is an objective external world, but it does not guarantee that one is actually in touch with such a thing. The answer to this is, at one level, very straightforward: there has to be a non-accidental match between the naïve or phenomenal world that is constructed on the Humean principles and at least those features of the world itself that are structurally and practically important. A world without secondary qualities or with an inverted spectrum is manageable, but failure to capture the primary qualities on some level would be disastrous. It might seem that the expression ‘non-accidental match’ avoids the major issue of explaining what the relationship is—what kind of causal relation is involved, for example. I think that this is a mistake. For philosophical purposes, it is enough to say that it is whatever kind of dependence that in fact holds in the cases in which we would judge that someone is perceiving something standardly and correctly. There may not be one account that fits all cases, but that does not matter as long as we can recognize cases of normal perception and distinguish these from odd cases (such as seeing things reflected or distorted) and can say what is peculiar about the non-standard cases.

6.8 Conclusion We have now shown what was promised at the end of Chapter 5, namely how, from the basis of subjective data, we can begin to form a conception of an external physical world. This task is not yet completed, however, for the feeling that a Humean picture may be too subjective is difficult to overcome. Overcoming it is, broadly speaking, is the task of Chapter 8.

7 Semantic Direct Realism, Critical Realism, and the Sense-Datum Theory 7.1 The Situation So Far Amongst the variety of reasons for which the sense-datum theory fell into disrepute, the fact that it was supposed to cut us off from the world by constituting a ‘veil of perception’ was one of the most commonly given. As the natural alternative to that, the theories that were favoured to replace it were claimed by their advocates to articulate some form of direct realism, according to which our cognitive contact with the world was not by some indirect route. Nowadays, the rivals to the sense-datum theory (though the use of ‘rivals’ is generous because the sense-datum theory is generally treated as no longer a serious option) are, on the one hand, the intentional or representational theory of perception, or, on the other, what is called ‘relationalism’, which is propounded as a form of naïve realism. Given the forced retirement from the fray of the sense-datum theory, intentionalism/representationalism and relationalism became the contenders for being the correct account of perception. In Chapters 1 to 5, I tried to show the inadequacy of these theories. In Chapter 6 I raised the problem of how perception could achieve objectivity, that is, put us, or seem to put us, in touch with a world external to us. Naïve realism had already been rejected as a result of the arguments in Chapter 1, and in Chapter 5 we considered the causal semantic account, and found that inadequate, too. We saw that the Humean option was the only viable candidate. I want now to retrace our steps and consider what kind of direct realism intentionalism and relationalism could achieve, even if we accepted them. I will argue that what I call semantic direct realism is all that survives of realism within intentionalist and even naïve realist accounts, even on their own terms. This will prepare the way for the rehabilitation of the sense-datum theory. So it is my purpose in this chapter to investigate in just what sense the rivals to the sense-datum theory, in their own terms, really put us—or would, if correct, put us—in ‘direct’ contact with the world, and to suggest at the end that the form of directness they achieve is not significantly different from that available to the

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0008

116

  

sense-datum theory. A discussion of the wrongly neglected theory known as critical realism will help in this task.

7.2 How We Might Understand Directness One might define Generic Direct Realism (GDR) as follows. Perceptual experience is direct realist, in a generic sense, iff perception standardly puts the perceiver directly in touch with the external object perceived. This obviously leaves it open what ‘directly’ means here. There are two possible ways of filling it out. The usual one is Phenomenological Direct Realism (PDR) PDR is the theory that direct realism consists in unmediated awareness of the external object in the form of unmediated awareness of its relevant intrinsic properties. I want to contrast this with Semantic Direct Realism (SDR). This is the following: SDR is the theory that perceptual experience puts you in direct cognitive contact with external objects, but does so without the unmediated awareness of the objects’ intrinsic properties that are involved in PDR. One might explain this by saying that, in humans and many animals, perception enshrines what one might call a judgement about external objects. If you do not like the word ‘judgement’ when applied to animals, or to neat perception itself, you can say that it involves information, about external objects. It is a direct realist theory because the judgment or information encapsulated in the experience concerns a putative external object, and the information is consciously, not subpersonally, received. PDR is what most understand by direct realism, and certainly by naïve realism. My argument in this chapter is that, under pressure from the arguments from illusion and hallucination, defenders of representationalist/intentionalist theories, and even of relationists theories in fact retreat to SDR. I then argue that SDR is equally open to a sense-datum theorist and that nothing would be gained by adopting intentionalism or relationalism. The force of the word ‘semantic’ is that the directness is not fundamentally a phenomenological fact. For the naïve realist, the realism and the directness are enshrined in the fact that the phenomena are intrinsic features of the object itself—they are how the object is in itself. The semantic direct realist is agnostic about how the phenomena relates to the object, but asserts that the experience

  

117

constituted by the phenomena subjectively embodies information putatively about something external: that it does this is something of which the subject is conscious; it takes its experience to be of a tree—or, if it is a dog, a bone, or its owner. This leaves open the possibility of raising questions about the relation of the phenomenal qualities of which we are aware in perception, and the object about which we are directly informed in perception: are they intrinsic to it as PDR claims, or are they more remote from it? Could it be the case that the object itself is very different from the way that it looks, even though the judgment contained in the experience is firmly anchored on the object itself? This issue is not one I shall discuss here, but in the second part of this book. It is a metaphysical one, concerning the nature of the physical world, not one in the philosophy of perception, but I point it out here simply to show that if one follows SDR and not PDR—as I claim all theories in the end are forced to do—what one can say is that the sensible qualities one is aware of in perception are how the objects presents itself to us: the question of the relationship between what objects are like in themselves and how they present themselves to human perceivers remains a further question, and not one confined to the philosophy of perception, but also involving general metaphysics. Strange as it may seem, something like SDR can be found amongst both relationists and intentionalist philosophers of perception. I will use as my principle stalking horses, for the intentionalists, Susanna Schellenberg, and for the relationalists, Bill Brewer, but I hope that it will be, or become, clear that my strategy is not ad personam but has force against both the schools of which these are eminent defenders. I choose them because of the care and detail with which they developed their respective positions.¹

7.3 SDR and Intentionalism Schellenberg believes, as does the follower of SDR, that a sense experience is propositional in form, but sensory in mode, and that the difference between a genuine perception and the corresponding hallucination is that the argument place in the proposition is empty in the case of the hallucination. As we saw in Chapter 4, Section 4.4, using ‘MOPr’ for ‘mode of presentation’ and ‘o’ for the object presented and ‘P’ for property presented, Schellenberg says the following. So if I perceive a white cup o, the content of my perception will be

¹ It is pretty clear that, amongst representationalists, it applies to Tyler Burge, who thinks that objectivity has nothing to do with phenomenology or consciousness: and much of what I say about Brewer applies to John Campbell, as we shall see below.

118

  

where MOPr (o) is a de re mode of presentation of the cup o and MOPr(P¹) is a mode of presentation of the property that this object instantiates. If I hallucinate a white cup and thus am not related to any white cup, the content of my hallucination will be

where MOP¹r(_) is the object place in an empty object-concept and MOP²r(_ ) in the property place is an empty property-concept. (2011: 28)

Now Schellenberg is clear that A perception and a hallucination can (seemingly) have the same phenomenology and thus be subjectively indistinguishable. (4)

So the absence of the object makes no subjective phenomenological difference, as that is entirely taken care of by the mode of presentation (MOPr). The difference, it seems, must be a sematic one—as one might expect when she describes these frames as ‘potentially gappy Fregean contents’ (27). We also noted in Chapter 4 that Schellenberg distinguishes her theory from Mark Johnston’s, but I claimed that their theories are essentially similar and that I cannot see any important difference between them. In both cases universals are present in a sensory, not purely intellectual, fashion (‘sensory mode’ and ‘visually aware’), either as attached to a subject, or not. The common point is that the phenomenology appears to belong to what is generated by the brain; it is a common factor to both perception and hallucination. The difference between perception and hallucination consist in whether there is an object to which these phenomenal states somehow belong. Taken in this way, both Schellenberg and Johnston adopt a form of SDR, in that the contact with the external object depends entirely on its logical role in a way that does not manifest itself in the sensory phenomenology. This internalist interpretation of the phenomenology on their theories might be challenged. In the case of both theories, the question arises of what is the relation between the phenomenology and the object. I have just suggested that the phenomenology is naturally understood internalistically, leaving any direct contact to be only semantic (in Schellenberg’s terms, merely a matter of ‘content’)? As we have seen, Schellenberg says: A perception and a hallucination can (seemingly) have the same phenomenology, and thus be subjectively indistinguishable

And one might challenge an internalist interpretation by drawing attention to the ‘seemingly’ in this quotation. Is the implication that really they do have a different

  

119

phenomenology, but one that cannot be discriminated subjectively? Similarly, someone might want to interpret Johnston (or perhaps develop his remarks) by claiming that, when there is a genuine perception, the property-cluster of which one is aware is a cluster of the appropriate property-instances in the world before one. The internal neural cause only guarantees the presence of a property cluster somehow. It is a sufficient condition for the subject’s being aware of a cluster, but the ontological status of that cluster will vary between perception and hallucination.² The suggestion of this asymmetry between the status of the phenomenology in the case of perception and in the case of illusion might seem to impute disjunctivism to these philosophers, but in fact such a thought can also be a feature of intentionalism. In my view intentionalism is not, as is usually assumed, simply in opposition to disjunctivism, but, on some interpretations, its essence is to break down the dichotomy between disjunctivism and the common factor theory, as we shall see. The contrast between externalist (and, therefore, disjunctivist) intentionalists and internalists can be seen in the conflict between John Searle and Tim Crane concerning the status of phenomenology in the case of perception, which was discussed in Chapter 3. If one is a Searlean intentionalist, then in perception, it is the external object that enters directly into the experience. If one is, like Crane, an internalist, on the other hand, then it cannot. As was argued in Chapter 3, the importance of the ‘intentionalist schism’ between Searle’s PDR and Crane’s internalism cannot be overestimated. We saw in discussing Searle that his externalism cannot be made to hold, given the role of presentation, over and above representation, in the case of perceptualtype experience, including hallucination. Schellenberg’s MOPr has the force of presentation, being subjectively indistinguishable from a real perception, so she, too, is forced to be an internalist like Crane. Given this, the object does not enter into the phenomenology and can only be semantically present. Schellenberg in fact has a further argument that might seem to defeat this internalism. She can be read as accepting internalism about phenomenology, but defends externalism about content.³ What this means is that connection with the external world and with the individuals in it (she seems to conflate these two) stems not from the phenomenology but from the (conceptual) content. This is a very similar view to SDR, but she tries to work a compromise with relationism by committing such a theory to externalism. ² Construed in this way the theory becomes a bizarre version of disjunctivism. The internal cause only churns out a property cluster when the external world fails to provide one. For objections to this kind of theory, see Robinson, 1994 and 2013. ³ I say ‘can be read’ because, as the argument above intimates, it is not clear whether her claim that the phenomenology of perception and hallucination can be the same is merely an epistemological point (as Searle would accept) or is a ‘common factor’ claim about ontological status.

120

  

. . . I will argue that perceiving subjects employ concepts the possession of which depends on perceptual relations to the very objects or property-instances that the concepts pick out. Hallucinating subjects employ the same concepts that in a subjectively indistinguishable perceptual experience are employed as a consequence of the perceiving subject being perceptually related to the world. The concepts involved in perceptual experience are typically demonstrative concepts. We can understand the requirement for concept possession in the following way: to possess a concept is to have the ability to refer to the mind-independent objects or property-instances that the concept is of. This ability involves being able to discriminate between the things that fall under the concept and the things that do not. If concepts ground the ability to pick out the objects or propertyinstances that the concept is of, then a subject who possesses say the concept RED must be able to use it to pick out red things. This will involve discriminating red things from things that are not red. So on the notion of concepts in play, concepts cannot be analyzed independently of what it means to have the ability to pick out the objects and property-instances that the concept is of. In this sense, possessing a perceptual concept is analyzed in terms of perceptual relations to the very external, mind-independent objects and property-instances that the concept is of. (2011: 732–3)⁴

There seems to me to be a very straightforward fallacy in this as an argument for externalism, namely that it assumes that the ability to pick out something external presupposes direct experience of that (kind of) external thing. This is an argument common amongst disjunctivists and relationists, that only if one has had direct contact with an F can one have the concept of F. But there seems to be no reason why someone might not learn what ‘F’ is from a controlled hallucination and therefrom be able to pick out actual Fs. In principle, this ability could be possessed by a brain in a vat—or, less dramatically, by someone like Jackson’s Mary who had not so far experienced a certain colour but who was made to hallucinate it. At least, there is nothing in Schellenberg’s argument to show why it could not be so. Her claim that the concepts in question are demonstratives does not alter this situation, for ostensive definition under hallucination is perfectly intelligible. The fact is that acts of reference are intentional, not extensional—that is, you can perform them whether or not the object referred to exists or not. So semantic direct realism is not an externalist theory. A ‘brain in a vat”s sense experiences contain judgments or putative information about external things, just as much as do ours (Robinson, 2003; Sainsbury, 2005, and Chapter 5 above).

⁴ Schellenberg’s account of concept possession reinforces an internalist interpretation of her view of phenomenology. She holds that phenomenology consists in the exercising of a concept, and that possessing a concept is the possession of an ability, and abilities are internal things, even if they concern external objects.

  

121

In fact Schellenberg’s error here seems to me to be very similar to Searle’s. Just as he mistakenly talks as if the presentational nature of perceptual-type experience belongs only to real perception, she talks as if the kind of acquaintance that can ground concept possession can only come from real perception. Both ignore the communality of phenomenology between perception and hallucination. On reflection, it seems that the available forms of Generic Direct Realism are not just Phenomenological Direct Realism and Semantic Direct Realism, but that the latter can be divided into externalist versions, with reference extensional, and internalist versions, with reference intentional. Only the last is actually available.⁵ Cassam’s version of representationalism is very similar to Schellenberg’s. He quotes McLaughlan’s account of representationalism: [O]ne visually perceptually experiences an object by having a visual experience with a representational content and the object’s bearing an appropriate causal connection to that visual experience. (McLaughlan, 2010: 246)

then continues: On this view, visually perceptually experiencing an object is indeed a relation that a perceiver bears to the object, the relation of seeing. This leaves it open that while the object is the cause of the visual experience it is a constituent of relational experience of visually perceiving it. The latter brings the object itself into view, and into the subjective life of the perceiver. This is something which representationalism would only be forced to deny if it is committed to indirect realism, on which the objects of perceptual experience are images of objects, but representationalism has no such commitment. It holds that one perceives an object by having an experience with representational content but that what one perceives is the object itself, not an image of the object. (Cassam, 2014: 139)

Nevertheless, he distinguishes representationalism from relationalism by the way they treat the qualitative content of experience. The relationist adopts what Cassam calls (CT): (CT) The qualitative character of sensory experience is constituted by the qualitative character of the objects and properties of the scene observed. (136)

But the representationalist accepts (RT):

⁵ Searle is indignant about Crane’s internalism, but it worth remembering that it was Roderick Chisholm who reintroduced intentionality to analytical philosophy, including the philosophy of perception, and his interpretation of intentional content was adverbialist. Nothing could be more internalist than that.

122

  

(RT) The qualitative character of sensory experience is constituted by its representational content. (139)

The rejection of (CT) is the rejection of phenomenal direct realism, and (RT) together with the claim above that ‘. . . the object . . . is a constituent of relational experience . . .’ constitutes the adoption of SDR.

7.4 SDR and Relationalism It might seem that relationists, who normally identify as naïve realists, must be PDRs not merely SDRs, as, indeed, Cassam’s imputation to them of (CT) above would suggest. The question is whether they manage to stick to this when faced with the phenomena knows as ‘the argument from illusion’. According to the relational theory, perception is a direct relation of acquaintance between a subject and an object in the external world. This is an ‘act-object’ theory, like the sense-datum theory, but the object is external and mindindependent, not mind-dependent, like an idea or impression. One major problem for such a theory is to explain how, if experience is simply a relation of acquaintance with an object, that object can appear in some way different from how it actually is: how to reply, that is, to the ‘argument from illusion’. My stalking horse here will be Bill Brewer, but I hope to show how others are no better off. The first thing to notice about Brewer’s theory is how it differs from ‘naïve realism’—and, from PDR. According to these theories, what one is acquainted with in perception are certain of an object’s sensible parts and properties—those that are suitably placed to be accessible to sight, or touch, or whichever sense. So, one is acquainted with an object by or in being acquainted with certain of its individual parts and properties. This is what makes the argument from illusion so powerful—how can one be directly acquainted with the sensible properties of an object if the properties of which one is aware are different from those that the object intrinsically possesses, as is standardly the case, even if only slightly so? J. L. Austin’s strategy was to deny that illusions generally took place. Sticks did not look bent in water, they looked like straight sticks in water, railway lines do not seem to converge in the distance, they simply look to be getting further away, pennies do not look elliptical from 45 , they merely look like round objects viewed from 45 , etc. I am going to assume that, though this trick might work for a few cases (perhaps the penny, given the oddities of depth) it cannot generally cope with the relativity of perception. This is also what Brewer and Campbell, for example, assume. We discussed the argument from illusion and how it is handled by relationists in Chapter 2, and we found that they were forced to handle phenomenal qualities, in cases where perception was not absolutely veridical (if there are any absolutely

 

123

veridical cases) in such subjective terms as how the object strikes one. If one puts this with Brewer’s first condition on relationist perception, namely, as we saw in Chapter 2; (i)

A subject, S, is directly acquainted with an object, o,

we have SDR in this condition, with the phenomenal content located in the subjective response, that is, ‘how it strikes one’. It seems that, under pressure from the argument from illusion, even the naïve realist/relationist has to retire to a semantic direct realist position.

7.5 Critical Realism If directness of perception is achieved, not by phenomenal directness, but is purely semantic, how are we to understand this sematic connection as working? One line of approach to this can be through the theory known as Critical Realism. Critical Realism is a theory of perception which was popular, especially in the United States, in the 1920s and 1930s. Some of its main proponents were R. W. Sellars and C. A. Strong and a crucial volume of essays was Essays in Critical Realism, edited by Durant Drake in 1929. Unfortunately discussion of the theory has disappeared from view until recently, when it was recovered and defended by Paul Coates (Coates, 2007). Critical realists have a ‘two component’ view: there is (i) a phenomenal side and (ii) an interpretation. As Coates says; ‘Seeing is also a cognitive process, whereby creatures represent their surroundings’ (Coates, 2007: 12). Coates explains critical realism as follows. The distinctive feature of the Critical Realist version of the causal theory of perception is its claim that the concepts exercised by the perceiver in perceptual experience refer directly to the physical objects perceived. When I look in normal circumstances at a red apple, I am non-conceptually aware of a red phenomenal quality which belongs to my inner phenomenal state. But I do not usually attend to my inner state, and it is not what I see. I see the apple, I directly classify it as such. As Mackie points out (Problems from Locke, p. 45) perception is judgementally direct, though causally moderated by my inner phenomenal states. (Coates, 2007: 12) . . . . there are good reasons for analyzing perceptual experiences as containing two contrasting components. Experiences contain some form of conceptual activity, which accounts for the fact that we grasp what we are experiencing, and also for the intentional nature or ‘directedness’ of perception. Experiences also have what

124

  

is variously called a phenomenal or sensory component. The latter component accounts for the fact that phenomenal qualities are immediately present, are actual elements of consciousness, in a manner that differentiates such perceptual experiences from purely representational states such as thought. (Coates, 2007: 148)

So the judgemental element brings in what we have been calling the ‘object’, and the phenomenal component is the content. We saw when discussing intentional theories in Chapters 3 and 4 that some intentionalists seem to treat the phenomenal content as itself representational or intentional (I think both Searle and Crane do this) and some, like Smith, treat is as non-intentional qualia, locating the intentionality elsewhere. Critical realists fall in the second camp: the intentionality comes from the judgement built into perception. The problem that faced Smith (see Section 4.2) faces the critical realist, namely, given that ‘phenomenal qualities are immediately present, are actual elements of consciousness’, in what sense are they not things of which we are aware, and, hence, objects of awareness? Coates says that we ‘do not usually attend to’ our ‘inner [phenomenal] state’, but this confuses not attending to the qualities we seem to perceive with not recognizing them as being ‘inner’. We do, generally, notice the colours, shapes, sounds, smells, etc. that are presented to us, and take them to be features of the external world. This latter claim may be challenged by those who emphasize the role of constancy in perception. The phenomenal qualities we sense vary in all kinds of subtle ways—shadowy lighting, angle of vision, for example—but we still see the object as the standard colour or shape that it is. But the fact that we correct for these things does not mean that we do not see them. I am perfectly aware that half the wall is in shadow, but that does not deceive me as to its standard colour— usually. It does not require a painterly skill to be aware of these facts. Nevertheless, I am far from wanting to deny the role of what has become known as cognitive penetration in the final form of perceptual experience, and we will return to this topic in Chapter 8. Critical realists, and, indeed, I think, Smith also get into trouble over how the phenomenal content contributes to our grasp on the external world, if that role is assigned entirely to the propositional element. We saw when discussing objectivity in Chapter 6 that both McDowell and Campbell made the point that it is difficult to see how the experiential factor could play its role in giving us a conception of the world if it was purely private or internal and only the thought-like component reached out to the external world. The representationalist in that debate, for example, Cassam, agreed that this was a good reason for building the phenomenal element into the representational, but it is unclear how the critical realists propose to do this.

 -   

125

7.6 The Sense-Datum Theory and SDR Once PDR is lost to both intentionalism and relationalism and they are forced into SDR, as I have argued above, at least with regard to the authors I have discussed, then nothing is gained by denying that the phenomena are constituted by genuine qualitative presence, provided that such presence can be taken as informing one of the (apparent) presence of the objective physical world—that is, of things which have more to them than they are manifesting at the moment. But there is no problem in this being so. None of the other theories put us in more direct contact with the world. Intentionalism’s pretence that it does falls with the failure of theories like Searle’s, and relationism abandons its naïve realist claims when it has to admit (a) that we perceive objects not as they are but as they appear from (an often subjective) viewpoint, and then (b) in order to accommodate the properties apparently involved in these perspectival appearances has to analyse them in terms of their supposed similarity to perfect perceptions that probably never take place. The opponents of sense-data cannot, it seems, be Phenomenal Direct Realists. It would seem to follow that they can do no better than to say that how things appear is how they manifest themselves to us, whilst leaving the question of what the objects are like in themselves as a further question. But the sense-datum theorist can say the same thing. We take the data to be how things manifest themselves to us, whilst awaiting scientific investigation to reveal—perhaps—what the objects are like in themselves. Nevertheless, it might seem to be a problem for the sense-datum theorist, as it is for the critical realist, to explain how the phenomenal and the propositional components are integrated. Showing how the sense-datum theory can really make manifest to us the world of common experience is the task of the next chapter.

8 Building the Manifest World 8.1 Introduction This chapter is the climax of Part I of this book. In it I attempt to show that all the ways in which the sense-datum theory has been set up as the bogy-man in the philosophy of perception are completely misconceived. This is mainly because the classical empiricist tradition had a crude account of thought and judgement, which did not allow it to grasp the interrelation of cognition and sensation. Once this is sorted out, one does not need to impute intentionality or representation to the fundamental nature of sensation in order to preserve both our natural intuitions about the directness of perception, and the dependence of perception on the particular nature of our sensory capacities. I believe I have shown that all the rivals to the sense-datum theory are radically flawed, and that none can achieve any stronger realism than the realism of reference that I have called ‘Semantic Direct Realism’. I have also argued that this realism is available to the sense-datum theory. In the chapter on objectivity, I argued that organization of the data, as captured by Hume’s principle of ‘coherence and constancy’, was the foundation of objectivity. Nevertheless, this leaves a set of serious tasks to be accomplished. Semantic direct realism involves the presence of something that can be called ‘judgement’, in a sense that I explain below, in perception, so the first of these tasks is to give an account of the relation between the phenomenal and the judgemental components in experience, and to integrate them in a plausible manner. The second issue is to show that the version of the sense-datum theory that I am defending is not an ‘error theory’ and that it can preserve the idea that we really do perceive external objects as external, and in a common spatial world. Thirdly, there is the issue of how, on the foundation of the perception of mere sensible qualities, we find ourselves perceiving whole objects: on the basis of having black, cat-shaped sense-data, I naturally take myself to be perceiving a cat. Fourth, I shall show how the primacy of direct realist discourse is preserved within the sense-datum theory, given the resolution of the above tasks. These issues can all be illustrated by contrasting our normal perception of objects with someone who is expertly looking at a radar screen, or similar diagnostic instrument. They can maybe recognize various kinds of different entities by the traces they make, but it would be exaggerating to suggest that it

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0009

    

127

seems to them that they seem to see the things themselves—they know that they are reading and interpreting the data, however ‘lost’ they might get in the experience. They would not be tempted to be ‘naïve realists’ about their perception of the objects they are indirectly perceiving, whereas naïve realism seems to be the default in normal perception. The radar operator’s experience is not how standard perception seems. So normal perception is not as if one were reading off from some evidential input that was manifestly distinct from the objects one took oneself to be perceiving. This, I think, is what philosophers usually mean when they say that our access to the external world is not inferential, from some more basic data.¹ The example of straightforward cases of interpreting data, like reading a radar screen, brings out the four issues I have just mentioned. First, such interpretation involves acts which seem clearly intellectual and conceptual, which are plainly separate from the perception of the dots on the screen in a way that perceiving and judging are not so separate in ordinary perception. One is definitely interpreting what one sees. Second, there is the issue of externality: the radar signal does not present itself as being ‘out there’, beyond the instrument screen: is a sense-datum theorist committed to allowing that the data either present themselves as internal to the mind, analogously to the dots on a screen, or are they unconsciously and erroneously projected onto the external world? If one could show that the sensedatum theory does not present the data as internal, nor commit one to an error theory, one could perhaps meet the fourth condition of showing how the direct realist mode of discourse is consistent with the theory. It could never be completely natural to report the radar experience in direct realist terms. It is clear that in the radar experience one does not seem to see the object detected in its full, three-dimensional richness, and we must be able explain how such richness is possible on the sense-datum theory, as required by the third condition.

8.2 The Role of Judgement in, and Its Integration with, Perception Before discussing the integration of data and judgement, we must defend the idea that perception involves anything as seemingly intellectual as judgement at all.

¹ Someone might claim that an expert machine operator might come to take what he sees as the object itself. (Robert Frasier suggested this to me.) I doubt if a radar operator really thinks he sees the true shape of an aircraft, even if something about the experience enables him to judge the exact type. But all I need for my illustrative is that a less adept user would be aware of the kind of distinction I am pointing to, and to the idea that that is not what normal perception is like.

128

   

For some philosophers, the term ‘judgement’, seeming to involve propositions and hence the deployment of concepts, seems too intellectual. A. D. Smith, for example, says: Concepts are simply irrelevant to perception as such . . . For when I say that concepts are irrelevant to perception, what I mean is that they are irrelevant to what it is that makes any sensory state a perception at all: they are irrelevant to the intentionality of perception, to its basic world-directedness. (2002: 95)

This is also Charles Travis’s idea, as found, for example, in his (2004), ‘The silence of the senses’. He approvingly reports J. L. Austin as follows: Austin’s idea is that, rather than representing anything as so, our senses merely bring our surroundings into view; afford us some sort of awareness of them. (64)

On this view, conceptualizing or judging is a further step. Despite Smith’s certainty, I find it difficult to discern exactly what the debate is about. Its roots seem to be in the Wittgensteinian idea that there can be no thought (and possibly even no consciousness) without language. It is clear that there are no concepts without thought and if there is no thought without language, then there are no concepts without language. As it is very natural to think that non-linguistic animals, including very young human beings, are perceptually conscious, then perception must be non-conceptual. But there are various ways to see that this cannot be both true and relevant in this context. I do not think anyone could deny that in perception creatures gain information about the world around them. Now, the notion of information can be regarded as pretty abstract, but it seems hard to deny that, in humans and many animals, this information is consciously appreciated or assimilated. In a more than purely behavioural sense, the dog recognizes its master, sees the bone, etc. All that is required to justify the use of a term such as ‘judgement’ in this context, is that in perception information about the world is, or can be, assimilated in the creature consciously. Self-consciousness need not come into it. So for those who worry that this theory makes perception, as in animals or babies, too intellectual by invoking any notion like judgement, they should see it as just involving the following two points. (i) Perception involves the reception of putative information about the environment. (ii) This inflow of information is taken in consciously, in the sense it is not like a purely sub-personal process, nor like one in a mere machine. There is a sense in which the subject, whether animal or infant, takes it to be about its

    

129

environment, in a way in which a sub-personal or machine system does not take the information that might flow in, in any way at all. This does not require that one is reflectively noting that it is his environment, but that the experience is a conscious one. And so taking it is not only a matter of behavioural reaction, as a sub-personal one might be. That this is so for the core perceptual experience of any creature we believe to be conscious, should, it seems to me, to be entirely uncontroversial. Furthermore, there seems to me to be a weak and a strong form of conceptualization. Any mental act that is genuinely conceptual must involve an irreducible grasping of a universal in some form: concepts are general entities and therefore universals in some loose sense of that term. So if the kind of judgement involved in perception—including animal perception—is to be different from that involved in discursive thought, it must turn on the difference between the ways in which universals are grasped in these two cases. This difference, I believe, is between grasping a universal as it presents itself in a sensory or quasi-sensory (imagistic) form, and grasping it where it no longer depends on a sensory medium or instantiation for its apprehension: for example, as linguistic meaning or in ‘pure thought’, assuming there to be such. Grasping of universals only in their imagistic form is the weak sense of conceptualization. In the latter, linguistic, stronger, sense of ‘conceptualization’ it is intimately connected with that sense of ‘thought’ where thought happens naturally in the absence of instances of the universals that constitute its content. In the weaker form, there is still a grasping of universals, but they cannot be handled with the freedom and flexibility that ‘thought in absence’ allows. This is where nominalism is relevant. If you regard consciousness itself, whenever it involves even the slightest element of recognition, however fleeting, as involving the apprehension of a universal in re, then you will see the world as already, in a sense, conceptualized for us and apprehended as such. Put in another jargon, if perception is the reception of form, then it is at least protoconceptual. For most nominalists, on the other hand, universals only enter in at the point at which a certain fairly sophisticated kind of mental activity comes into play, and this is held to be far too intellectual a level to be essential to animal perception. Aristotle can be called to our aid at this point. It is through perception that the universal is initially established in us, ‘for though one perceives the particular, perception is of the universal’. Even animals can acquire universals in this primitive form. (See Post. An. 100a 16.)² The sense in which perception is a form of judgment, according to my account above, only involves the form of weak conceptualism that involves the handling of universals in their instances or quasi-instances. I do not think that this should be

² I am grateful to Bill Jaworski for pointing this reference out to me.

130

   

denied to animals or babies. On this conception, the idea that there is a sharp break between perception, which can be purely passive, and primitive conceptualization, which is active, involves a false dichotomy.³ So it cannot be held against those theories that build in something that we can call judgement, as I have explained it above, over-intellectualize perception.⁴ There still remains the problem of explaining how the phenomenal and the propositional components are integrated. I think that the most straightforward answer is that the judgement is realized in the phenomena, as its vehicle. If I think about the Eiffel Tower, that it is in Paris, I think using words with meaning, but no-one thinks that these vehicles of thought constitute some sort of veil between me and it. The thought is about the Eiffel Tower, simpliciter. If we can take perception as being a form of judgement, represented not in language but sensorily, I do not see why it cannot be thought to be just as much of its object as is a verbal thought. The phenomena present the apparent state of affairs naturally, and not via convention, as do words. If ‘clouds mean rain’ naturally, even more directly cloud-like data mean—present, inform one of the presence of—clouds. Conscious awareness of the data is assimilation of their cognitive import. It is important from the sense-datum point of view that the words which are the vehicle for the thought actually possess sensible qualities as well as signifying things for us. The phenomenal qualities of the data are perfectly real and in the context of the conscious cognitive role of perception inform us about the manifest world. This theory also restores an element of phenomenal direct realism, which I argued in Chapter 7 neither intentionalism nor relationalism could retain. This is because the sense-data are both the vehicles—the language—in which the perceptual information is expressed, and the actual features of objects as they manifest themselves to creatures of our kind. There is a parallel here with Susanna Siegel’s argument that any theory that allows that we perceive properties can accept that perception involves ‘content’ (Siegel 2010). The term ‘content’ is usually associated with representational, intentional theories, but Siegel argues that even a naïve realist can admit that perception presents itself as being of properties—the perception of which might be more or less accurate—and so can concede that perception has content. The weak conceptualization account that I have given can equally be described as involving ³ This account of the nature of perception, as the apprehension of properties, conceived as involving universals, shows the error of the belief that the distinction between sensation and judgement is a gap between the ‘space of causes’ and the ‘space of reasons’, as Sellers and McDowell claim. There are properties in the world, and these are assimilated in perception, and, as universals, are the natural material of thought. This does not involve the claim that causation and reasoning can themselves be assimilated, only that the material delivered in perception is already suitable for rational handling. ⁴ This account of weak conceptualization, which follows closely Aristotle’s idea that perception involves the reception of form, also faces a problem which it is sometimes said Aristotle’s own account faces. At the level of data, at least as understood by a sense-datum theorist, the forms or universal were acquainted with are sensible qualities, not directly substance concepts, such as ‘tree’ or ‘table’ or ‘horse’. This issue will be tackled in Sections 8.5 and 8.6.

    

131

content, but the relevance of this to what I say above is that ‘content’, because it implies some sort of conceptual grasp, makes brute experience and cognition very close companions. If you perceive, you discriminate one thing from another and apprehend similarities, so you recognize properties, in some primitive more-thanbehavioural sense, and this is enough to make perception ‘judgemental’, in a basic sense. My position on this could be compared to John McDowell’s. Using the Kantian word ‘intuition’ for the nature of perception, he says: If intuitional content is not discursive [that is, ‘not a result of our putting significances together’ as in ordinary thought], why go on insisting it is conceptual? Because every aspect of the content of an intuition is present in a form in which it is already suitable to be the content associated with a discursive capacity, if it is not – at least not yet-actually so associated. That is part of the force of saying, with Kant, that what gives unity to intuitions is the same function that gives unity to judgements. If a subject does not yet have a discursive capacity associated with some aspect of the content of an intuition of hers, all she needs to do, to acquire such a discursive capacity, is to isolate that aspect by equipping herself with a means to make that content – that very content – explicit in speech or judgement. The content of an intuition is such that its subject can analyse it into significances for discursive capacities, whether or not this involves introducing new discursive capacities or not, the subject of an intuition is in a position to put aspects of its content, the very content that is already there in the intuition, together with discursive performances. (2008: 7)

The similarity between this and what I have said above should be clear, so the question arises of the relation between his kind of direct realism and my sensedata direct realism. McDowell thinks that conceptualization must enter into perceptual experience at the bottom level. The ‘myth of the “given” ’ is the belief that something preconceptual is given, when, rather, the truth is that what is given is already encapsulated in our mode of understanding. He and I are agreed that conceptualization enters ‘harmlessly’, as it were, at the basic level. But McDowell takes this in a Kantian sense, namely that these concepts reflect the mode of our mental cognitive structures, not the objective properties or real resemblances of what is presented, which I have called an Aristotelian approach. It is my claim that only a radical nominalist—that is, someone who denies that there is a fundamentum in re or real properties or real resemblances that we are tracking and which are available to us—should put the onus on the perceiving subject, rather than on the sensible nature of what is presented. I am, however, claiming that we are only aware of the world ‘as it manifests itself to us’ and that it is a further question in what ways the world is similar to

132

   

these ‘manifestations’. (It must at least be in some sense structurally similar or we could not negotiate our way around it on the basis of how we perceive it.) Does this make my position closer to McDowell’s? Not really, because the ‘Kant versus Aristotle’ contrast remains. The objective similarities we register are not, on my view, features of our cognitive structures— the ‘categories of our understanding’—but of the nature of the phenomena that are given to us. This depends on how our senses work, not on our categories of thought. McDowell might reject this distinction, but, if he does, it can only be because he rejects the idea that the kind of conceptualization I discussed in connection with Smith is already present in the world, in the shape of the properties present. For McDowell, of course, what we conceptualize is the world out there, and not, in any sense, subjective phenomena, like sense-data. We saw in the chapter on objectivity the reasons he and the relationists had for thinking that we must be directly in touch with external reality, namely that if we did not, we could not have any conception of an external reality. But in that chapter I argued that this was not correct and that the Humean account of how we come to form the conception of an external, objective world, on the basis of the ordering of our experiences, was the correct one. If this is correct, and if it is correct that objective similarities of the right salient sort are present in our sense-data, then everything McDowell wants to say about the role of conceptualization can be imported into my account. I take this as showing that, when it comes to any important sense of whether our perception is ‘direct’, the sense-datum theory is not worse than any other.

8.3 The Sense-Datum Theory Is Not an Error Theory The theory can be said not to be an error theory if the data correctly presents objects as external, and it does not do this because we mistakenly treat as external something that is really just subjective and internal. In order to justify the claim that the sense-datum theory can meet these standards, one must show how, in normal perception, and taking sighted people as the paradigm, the visual phenomena—the sense-data—are structured in a three dimensional visual space that simply presents the objects as seeming to be where they actually are in the common-sense or manifest world. If you combine this with the primitive conceptualization that recognizing properties involves, you get the articulation, in a purely ostensive or demonstrative fashion, of the consciously presented information that things are presented thus and so. This is perceptual experience. I say that the sense-data must be structured in three dimensional space, and this, of course, raises the standard issue of how to understand the third dimension, visual depth. Is visual space fundamentally two dimensional, as Berkeley thought,

-      

133

and depth is to be explained in terms of the anticipation of further experiences or actions; or are H. H. Price (1932), and other modern sense-datum theorists, right to take depth as a primitive? Can one combine these and say that the way one is primed for action and anticipation ‘infects’ the phenomenology in the way cognitive penetration—the influence of acquired information or past experience on perception itself—can reasonably be held to do? Fortunately, one does not need to go into these questions. It is enough that (a) depth has a phenomenological manifestation and (b) that it thereby presents objects, fairly accurately, as being the distance they actually are away from us. That is, it is enough for there to be something about visual experience that naturally presents itself to us as being depth, and that it is adequately accurate. I think what I have said above shows that a sense-datum theory understood as I have just described it is not an error theory, because the information contained in a normal experience is broadly accurate, and the apparent structure of the phenomena corresponds to the spatial relations in which the objects stand. But aren’t the phenomena still subjective and private and don’t they appear to be in public space, so doesn’t it follow that projectivism, in a sense that involves an error, is true? Is the privacy of sense-data in conflict with common sense? I think that there is a tension in common sense when it comes to the privacy of experience. On the one hand, I think common opinion would accept that how something looks to me is private to me. On the other hand what I see is, in the normal case, the thing out there and its properties, which are not private to me. Furthermore, the ‘how it looks’ is constituted by the sensible properties that manifest themselves to me, although they are meant to be the properties of the object ‘out there’. I think that the best way of reconciling these factors is in the following way. (i) The sensible properties of the object appear to be at the distance that the object is from me, with a high degree of accuracy. (ii) The sensible properties that the object manifests are the way that objects of that sort look to creatures like us. So we see objects as being where they are, in relation to us, and looking the way they look to creatures like us. We naively assume that objects are intrinsically like the way they appear, and this may be wrong, but appearance is not wrong about where they are and how they appear, with appropriate and adequate kinds of accuracy. Nevertheless, data could be private in a slightly different sense. The fact that they are presented at a distance phenomenally does not show that the space of which that distance is a part is the same space as that in which the experiences of other people are located. We think of ourselves as seeing the same objects in the

134

   

same physical space as are seen by other people. The account I have just presented is consistent with each subject operating in a private three-dimensional space of their own. What is required to make this space and its occupants genuinely public? We are operating in a context in which it has been shown that what might be called a kind of representative realism is the best the realist can do. This leaves two options. The most naïve is that physical space is qualitatively the way it appears to our senses—say, for convenience, to sight.⁵ In this case, our visual space will map neatly into physical space. Whereas, in one sense, visual space has no spatial relation to physical space, informationally, it overlaps it. So, taking perception in its cognitive aspect—which, once it becomes perception and not just sensation, is its fundamental nature—its domain is physical space. The other option is that physical space is not qualitatively similar to the space presented in any of our senses. In this case, the manifest world of common-sense, in its spatial nature, must be a construct from our experience, for it has nothing else to constitute it. Is it, then, just a collection of separate spatial bubbles, one for each subject? First, it is united by the fact that different people’s experiences share common causes: your table-type experience and mine have, when appropriate, the same table as their cause. Second, the image of their being separate bubbles induces a picture in which they are spatially distinct and distant from each other, but in fact they have no direct spatial relation to each other, but, as with the more naïve option, they map into the physical world, and doing so, they overlap each other when we are viewing the same piece of the world.

8.4 Our Spatial World and Visual Experience Maybe there is the following problem. Even if it is agreed that visual space can be mapped on to manifest physical space, are there not different spaces for the different senses, and doesn’t this make subjective experience essentially private and non-objective? I deny, however, that there are different spaces for the different senses. For a sighted person, the world as he experiences it is primarily a visual world. One thing that this means is that his conception of space—what he takes space to be actually like—is fundamentally the way it visually appears. Into this primarily visual world, everything else that he experiences is located. Sounds are experienced as emanating from objects in that world, smells pervade regions of it.

⁵ Of course Molyneux’s famous problem arises here. If the qualitative nature of space in the various senses, principally sight and touch are not the same, then one is faced with the dilemma of deciding which captures the real nature of space, or of abandoning the idea that senses get its real qualitative nature at all. I shall not pursue this here.

 ,   

135

We do not perform (at the relevant level, at least) acts of correlating visual, auditory, tactile and olfactory realms and spaces, and mapping them onto one another. We find them all in one experiential world, which, for the sighted, is primarily visual in its structure. It is in this world that the invitingness of the water is located when we face a pool on a hot day, and it is this world which is experientially different for us because we are habituated to the availability of its hidden features and conceive it in the light of that fact. But it is not clear that it is correct to call the way that it is experientially different an explicitly visual one. Sounds and smells are not visual phenomena, although located in the visual world. Colours and visual shapes are paradigmatically visual. But there is no reason to think that all the features that go up to make our one, united experiential world each fall neatly into some one category—visual, auditory, tactile, emotional, etc. These are all post facto rationalizations—well founded on paradigm cases—of an experiential world which is, in itself, confusingly diverse and continuous, not neatly categorized. Someone might raise the question of whether this synthesis, so to speak, of the various senses is part of perception itself, or of some process ‘further down the line’. It is not clear to me exactly what this query, though a natural one, actually means. Is it a question of whether it happens in those parts of the brain that are classified as, for example, the visual centre, or in some area not directly concerned with perception? Or is it about whether a cognitive science functional map should place in a ‘perception box’, or in some other? Or does it concern whether it involves what is called cognitive penetration, according to which perception proper is modified by acquired beliefs or concepts? I think all that matters for present concerns is that it is part of the phenomenology of perceptual experience, however one might want, for other purposes, to break this down.

8.5 Perceiving Objects, Not Just Qualities It seems to be generally agreed that we perceive sensible qualities—in the case of vision, these are colours and shapes—and it also seems to be generally agreed that the total experience of perceiving is, in some sense, richer than this. We do not see only a kaleidoscope of colour patches. Why not simply say, therefore, that we perceive qualities and other things—at least physical objects, for example— alongside or in addition to them? It might seem that it is much easier for the naïve realist to accommodate this fact than for the sense-datum theorist to do so, because the naïve realist claims that it is the objects themselves that enter into the constitution of experiences. The matter is not that simple, however. It is an indisputable fact that one only perceives parts of objects: you cannot see the back or the inside of objects, for example, and yet there is a sense in which one takes oneself to be perceiving the thing itself, in some sense as a whole. Exactly

136

   

how this wholeness of perception comes about cannot, therefore, be solved simply by saying that objects enter into the constitution of experience, for certain features enter more directly into it than others, and it is, in some sense, through the presence of those features that one is enabled to see the object as a whole. Alva Noë expounds the case for saying we see objects as such, and the problematic nature of the relation between this fact and the fact that we see only certain parts and properties of things, as follows: When you see a tomato, you can’t see its back. When you see a cat behind a picket fence, you only see, strictly speaking, those parts of the cat that show through the slats . . . Nevertheless, one can hardly dispute that we take ourselves, when we see the tomato, or the cat, to have a sense of their presence – a perceptual sense of their presence – as wholes. (413)

Susanna Siegel makes a similar point but takes it further: Do sensory experiences represent any properties other than color, shape, illumination, motion, their co-instantiation in objects and successions thereof? I will . . . argue that some visual experiences do represent properties other than these.

Because at least some of these further properties are natural kind properties, she calls everything that goes beyond the above list ‘K-properties’, and her claim is: Thesis K:

In some visual experiences, some K-properties are represented.

She distinguishes three kinds of theory, starting from the sparsest. The first is: . . . akin to what David Marr called the 2 ½ D sketch. Roughly, the 2 ½-D sketch represents color, shape and illumination properties of facing surfaces, but does not represent which surfaces belong to the same object, or how those surfaces continue out of view . . . A slightly more permissive view is that visual experience represents that some surfaces and edges – for example, those making up a cup’s handle and the rest of the cup – are grouped together in full volumetric (3D) units. More permissive still is the view that visual experience represents colors, shapes, volumetric groupings, and objects. Thesis K is even more extreme: it allows that, in addition to all these things, visual experience represents properties such as being a house, and being a tree. (482–3)

Despite an interesting difference in the ways they express their views (Noe talks of what we see, Siegel of what is represented in visual experience) they both want to

 ,   

137

grant to visual experience itself (not vision augmented by ‘interpretation’) a wider competence than a classic empiricist approach would seem to allow. Nevertheless, simply talking comprehensively about ‘representation’ is in danger of not forcing one to come to grips with the need to explain the primacy of role that one’s perception of the visually accessible features plays. There is a deep-seated empiricist instinct to defend the view that in order to see objects and anything beyond qualities one need only perceive the qualities and take them in certain way. Whatever else is perceived is not perceived in quite the same direct and fundamental way as the qualities. This is illustrated by a comparison or analogy. Just as a painting can convey an ontologically rich view of the world whilst being only patches of colour seen in a certain way, the same trick can be performed for vision itself. But the empiricist intuition is not just motivated by the analogy with pictures. It is motivated by at least three considerations. One has already been mentioned. It is the intuition that if you perceive a certain array of qualities and take them in a certain way then there is no more involved in perceiving a fuller and richer world. Another is that those features presented over and above the qualities are not actually perceived in the straightforward way the qualities are. In seeing something as a tomato I see it as the kind of thing that has a rear and an inside, but I do not see the rear and inside. I can tell by looking whether it has mould on the facing side, but not whether it has mould on the hidden side or is rotten inside. There is a dependence of all perception on perceiving sensible qualities. A good account of perception will explain this priority possessed by the qualities and not merely add indifferently to the list of things perceived. It will explain how experience in general is sustained by experience of qualities. But the dominant influence behind the empiricist intuition historically is what one might call the sense-datum intuition, and it entails the other two, but they do not entail it. The essential feature of the sense-datum theory is that it regards experience, whether veridical or not, as involving the actual presence of sensible qualities, in a way that it cannot, in its ‘common factor’ nature, involve actual substances. So if I hallucinate a pink elephant, my visual field is genuinely characterized by pinkness and elephant-shape, whereas it cannot be characterized by an instance of elephanthood. This forces one to regard the sensible qualities as the building blocks of perceptual experience, and to give a different kind of account of the rest.⁶

⁶ An intentionalist, however, might be tempted to think that his theory provided a licence for blurring the distinction between qualities and other contents, for they are all intentionally inexistent, so the nature of the presence of elephanthood and greyness is not essentially different. I do not think, however, that this is a line down which the intentionalist should or would want to go. The danger for the intentionalist is that he find himself without the resources to distinguish perception from thought, and the seeming presence of objects that characterizes perception, and the way that the different senses are differentiated by their proper objects—the secondary qualities that distinguish them—requires even the intentionalist to give a foundational role to sensible qualities.

138

   

This still leaves us with the problem of how we are to explain and understand the way that substance concepts enter into the experience, although they are not explicitly present in the way that qualities are? One might try merely asserting that one brings the substance concept to bear, and that that ‘informs’ the experience itself. There is, indeed, evidence for thinking that the ability to see things as objects, not just qualities, is innate.⁷ One could, indeed, leave it at this, but one might also wish to push deeper, because one wants to know what it is to possess such concepts, or because one might want to know how we could have come to acquire such concepts, given the content of experience. Berkeley, Hume, and Noë combine to answer this point, however. Put simply, the point is this. One’s conception of the experienced object as physical is causally sustained by one’s previous experience of objects as having a richness and diversity not given in any one experience and it is on the basis of this association and the expectations that it gives rise to, that the phenomenology is modified. That is, our previous experience of tomatoes having backs, insides, tastes, feels, etc. gives rise to expectations and anticipations that actually influence the experience one has when one looks at one and simply registers its shape and colour. Ironically, perhaps, this theory has a lot in common with Alva Noë’s. What it has in common is the belief that factors which are external to the experience itself actually affect its phenomenology. So, in my way of putting it, what one anticipates as available actually affects the experience. Noë expresses this by saying: Phenomenologically, the world is given to perception as available.

(422)

The similarity is ironic because his approach is externalist and mine is internalist, in a very traditional way. The difference comes because he believes that the availability of other experiences is essentially connected with action. There seem to me to be the following questions. (i) Why should the availability of presently undisclosed aspects of objects necessarily involve reference to action on the part of the perceiver, rather than, for example, involuntary relative movement of the perceiver and object? (ii) If it does involve seeming action, why should this have to be understood in an externalist fashion? (iii) How can the availability of hidden features, whether understood in my way or in Noë’s, affect current phenomenology? The first two questions are meant in an anti-Noë sense, but the last is also a challenge to my approach. ⁷ See Susan Gelman (2003).

 ,   

139

Consider, first, (i). Why should Galen Strawson’s ‘weather watchers’ not be able to form a similar conception of the objects that float around them to the one that we form? (Strawson, 1994). This rhetorical question would have even more force if the passive entities in question themselves floated around on the breeze. No doubt, in our case our activity plays a role in causing us to have our object conceptions, but it is difficult to see how our activity enters into the conception of the object itself, in the way that the available contents of our experience of the objects do. To say that they do would seem to suggest a too immediate route to idealism, because it would seem to imply that, of their nature, objects are always objects-for-us: their objectivity is constructed, at least in part, out of our agency with respect to them. This seems to me to be an important point. What are presented in experience of objects are, as far as the nature of objects in our manifest image of the world is concerned, properties of the objects themselves, at least as modified for perspective and other features of ‘illusion’. In other words, these contents are appropriate to be included in our conceptions of the objects themselves. The fact that we managed to perceive these aspects as a result of actions of our own, as opposed to passively drifting around other objects, though perhaps causally relevant in our case, does not enter into the phenomenological presence of the object itself: it should not be part of the nature of the objects in general, as presented in experience, that we acted upon them. (ii) It is, I hope, illuminating to discuss Noë’s externalism, as found in (ii), because it throws light on a general implausibility of direct realism. Noe’s externalism is an application of Clark and Chalmers’s ‘active’ or ‘vehicle’ externalism to perception. Clark and Chalmers (1998) argue that our cognitive system (and, hence, our mind) often extends beyond the head, as when, for example, we use a library, a notebook or a computer as auxiliaries to our memories or our calculating capacity. It comes as a surprise that these banalities might somehow threaten the integrity of the Cartesian ego. The reason why they do not is as follows. We can divide an individual’s cognitive system into three parts. First there is the part that directly sustains conscious states. This part can also thereby contribute directly to behaviour. Second there is the part that does not sustain consciousness directly, but does provide information which directly (not via consciousness) influences how we behave. Much of the brain is like this. Third there is the part of the system that cannot directly influence either consciousness or behaviour except by having an adequately differential affect on one or both of the other two parts. The external auxiliaries are all in this category: my notebook cannot affect what I do except by modifying my brain. What is more, this affect must be adequately differentiated— it is not enough simply to ‘make contact’ with the brain. If I have the addresses 16 South Street, 81, Ashley Lane, and Nador utca 9 written in my book, then we assume that if these pieces of information are to produce properly differentiated behavioural responses, they must make equally differentiated impacts on my brain. This casts serious doubt on Noe’s claim that what is in the head may not

140

   

be sufficient for the mental state in question, for we do not accept that an external difference, not reflected in a neural difference, or difference in a relevant part of the nervous system could contribute to behaviour. It has often been pointed out against a standard version of the causal argument for sense-data that it begs the question against the direct realist to assume that some brain state would be sufficient to produce a mental state of the right kind: what is needed, according to the direct realist, is the external object, the whole causal process, and the brain state. (See, e.g., Noë, 2006, 411–12.) But this raises the question—which I do not think I have seen raised—of whether, and, if so, why, the brain state need be different for the perception of different things. Suppose I see a red square and then a green circle. If the sufficient conditions for these experiences include both brain state and object, why could not the brain state be the same in both cases, and the difference be made by the differences in the objects? On the direct realist conception, the brain contributes only the act of perceiving, which is transparent and hence no different from case to case: the difference is provided by the object. Why, then, need the brain states be different? There need be no difference with other kinds of act. If I take hold of an apple, then take hold of an orange of just the same size, the motions of my body can be exactly similar: it is only because of the external object that one is a taking hold of an apple and the other a taking hold of an orange. The direct realist theory of vision is sometimes called the torch beam theory, because it supposedly pictures perception as involving awareness reaching out from the subject to the world. Although this is only a metaphor, it seems to be well motivated in the case of non-reductive direct realism. But no-one would think that what the light beam falls on requires modification in the torch or the beam itself: the difference is entirely in the object. In general, with sets of causally sufficient conditions, a change in one of them is enough to affect the output: this change does not have to be accompanied alterations in all the other conditions. If I wish to make a copper sulphate solution and a saline solution, I do not need different kinds of water—copper-sulphateappropriate water and salt-appropriate water—the copper sulphate and the salt make the difference, the water is the same. So if, in the case of perception, the objects only make a difference to the final experience when in association with an appropriate difference in the brain, this is a radically different state of affairs from a normal case of when more than one condition is necessary. Indeed, the only plausible rationale for requiring appropriate differentiation in the brain would be the model adopted by the causal theorist, according to which the brain state possessed all the information necessary to be responsible for the final result, having appropriated the contribution of the object. This, of course, is exactly what the direct realist is rejecting. I say that this is the only plausible rationale, not that it is entailed by the need for a distinctive and differential brain state. Nevertheless, I think a direct realist, who wants the brain to be only a partial cause, whilst retaining the causal theorist’s requirement that the brain state be

  

141

adequately differentiated, wants to have his cake and eat it, lacks the courage of his convictions, and is on the verge of disingenuousness.

8.6 Availability and Phenomenology The most important challenge, from my perspective is (iii): how are we to make sense of the claim that features not directly present in experience affect the phenomenology of the experience? First, I want to claim that, quite generally, states that are not paradigmatically conscious can affect what it is like for one at a given moment. The most obvious case of this in my opinion is the following. Imagine two conscious subjects, A and B, who have exactly the same current explicit states of consciousness. What I mean by this is that their sense fields seem just the same way, their current conscious thoughts and desires and emotions are exactly the same. But A and B have entirely different histories, it is just a coincidence that their normal conscious states are exactly similar at that moment. The question is: is what it feels like to be A at that time just like what it feels like to be B? On the one hand, the way I have described the situation might seem to entail that it is, for what can ‘what it feels like to be X’ depend on other than his conscious states, for ‘feels like’ refers to consciousness. On the other hand, it seems to me that one’s memories condition what it feels like to be oneself, even when those memories are not part of current explicit consciousness. One might be tempted to unpack this latter thought conditionally. If asked about their past or their beliefs, A and B would give quite different answers. This is true and important, but it seems to me that there is something current and implicit in the subject’s sense of himself that underlies at least some of such responses. Adapting Berkeley’s jargon, one might say that part of the notion that we have of ourselves includes an implicit or tacit grasp on much information not currently being rehearsed in consciousness explicitly. It is still in some broader sense part of the phenomenology of our current existence. In other words, the baggage of one’s personality and history makes a more than purely behavioural/dispositional impact on one’s current consciousness. The general principle is that states which are not part of the current sensory foundation can be implicitly present and so can make a difference to the phenomenology. An analogy might be the way that a mood which you are not aware of being in, still colours one’s experience. In the case of ‘sense of self ’ the difference is not made to the perceptual phenomenology as such, but there are other cases when it does. If one is in the right mood, the water ‘looks inviting’, and this is how the water looks, because of some other mental state, which, in this case, may or may not be conscious, which ‘infects’ the experience. Similarly, what we have come to know about the world— what we have been habituated by experience to regard as available—‘infects’ our current experience. Nothing more is needed to explain this, however, beyond

142

   

experience of sensible qualities presented in the appropriate ordering, and the particular automatic or instinctive rationalizing tendencies of the human or higher animal mind.

8.7 Sense-Data, Direct Realism, and the Common-Sense Understanding of Perception We still need to complete the fourth task I set up at the beginning of this chapter, namely to show that the discourse of direct, or even naïve, realism is the appropriate discourse for describing our experience, even for a sense-datum theorist, and this not in any ‘inverted commas’ sense. We have already seen, through the argument about semantic direct realism in Chapter 7, that the judgemental or ‘conscious informational content’ component in perception has external things as its object. Our argument about the integration of the phenomenal and informational components argued that the phenomenal component is presented as being at the location in space which we think the object actually is, and so this is not an error theory. Nor is it a projection, if that means that careful scrutiny would show that what we are aware of is not an external thing. We are aware of the external world as it manifests itself to creatures of our kind; though how it manifests itself depends on our sensibilities, as well as its intrinsic nature. One might contrast Martin’s statement of naïve realism with my account of common-sense realism, incorporating the sense-datum theory, and see how similar they are. Martin says: According to naïve realism, the actual objects of perception, the external things such as trees, tables, and rainbows . . . and the properties which they can manifest to one when perceived, partly constitute one’s conscious experience, and hence determine the phenomenal character of one’s experience. (Martin 1997: 83–4)

My statement of common-sense realism, or direct but not naïve realism, is as follows. According to direct realism, incorporating the sense-datum theory, the actual objects of perception, the external things such as trees, tables, and rainbows, when perceived, manifest themselves to us in the ways appropriate to our sensibilities, and hence determine the phenomenal character of our experience.

Classic naïve realism is naïve in two ways. First it simply takes it for granted that the way objects manifest themselves must be, across the board, the way they are intrinsically. Martin’s statement does not make this explicit, but without it



143

relationalism, as normally understood, would not hold. Second, it embodies the assumption that, if the properties the objects manifest depend, in some instances, for their existence, on our sensibilities, it cannot be the object itself that we are aware of. This was assumed because it was assumed that the only possible form of direct perception was what I called in Chapter 7 ‘phenomenal direct realism’, accompanied by the first assumption above, namely the phenomenon in question must be of a quality the object possessed independently of our sensibilities. The position I am defending has some features of phenomenal indirect realism, because the phenomena depend, in part, on the effect produced on the subject: but it is also a kind of phenomenal direct realism because those phenomena present themselves as located where we take the object to be, and are actually the natural way for such objects to manifest themselves to creatures like us. This sophisticated or hybrid position became unavailable when the empiricists opted for an imagistic theory of thinking. Then there was no cognitive component that could reach beyond the sensory impact and use that impact as a vehicle for genuine cognition, because thoughts were just further images. If one has a genuinely non-reductive approach to the cognitive component one is not in the position of the imagist. Contrary to what I think is the received opinion, it seems clear to me that the standard scholastic view was a version of the sense-datum theory; the scholastic ‘percept’ is a sense-datum, but there was no sense that percepts constituted a ‘veil of perception’, for it never seemed to them that grasping form was a purely imagistic phenomenon. As we saw in the chapter on intentional theories, the standard line has been to treat percepts or ‘sensible content’ as in some way intentional—that is as representations of qualities, not instances. There seems to me to be no reason to do this. The intentionality comes in the intellectual or judgemental component in perception, and this is what constitutes the perceptual realism. But, as I argued in Section 8.2 above, that which embodies the perceptual judgement does not itself need to be a mere representation, rather than a genuine presence of a quality.

8.8 Conclusion Near the beginning of my earlier book on perception (1994: 4), I said that ‘one of the basic problems of perception’ was as follows. On the one hand, a little reflection – that is, thought that does not resort to any science that goes beyond common experience – shows that perception involves some sort of physical influence running from the external object to the senseorgan of the perceiver. On the other hand, the essential nature of experience seems to be that the subject reaches out to, and makes conscious contact with, the external object. How can a process in which the subject is a passive recipient of a

144

   

stimulus be the physical aspect or realisation of a process in which the subject reaches actively and consciously out into the world?

I think that the sense-datum direct realism that we have arrived at is the best way one can reconcile these apparent contradictions and save the appearances. The causal process from object to perceiver sets up a cognitive state in the perceptual consciousness of the perceiver which, given the organization of the phenomena, informs him of the world around him, making it immediately presented to him, in a way that is accessible to his sensibility. The phenomenal contents of these experiences are actual qualities, not mere representations of them, therefore they are in the family of sense-data or qualia. To put it even more succinctly: you can have everything you might have thought you were getting from direct—even naïve—realism, as follows. Objects perceptually seem to be ‘out there’; they are in fact ‘out there’ roughly where they seem to be, so perception is pretty accurate; objects manifest themselves the way they naturally do to creatures like us, with the kind of sense-organs etc we have. This is entirely consistent with the sense-datum theory, according to which the data are subjectively generated, because its ‘out there’ feature is a function of the cognitive component which is integral to sophisticated perception. This leaves open the question of how similar ‘in themselves’ objects are to the way they manifest themselves to us. Some might be worried by the idea that this is an open question, but anyone with any sense of the distinction between the ‘manifest image’ and the ‘scientific image’ will be aware of this issue anyway. It is the purpose of Part II to investigate this question.

PART II

W HA T T HE W OR L D I S , I N I T S E LF

9 The Problematic Nature of the Modern Conception of Matter 9.1 Introduction In Part I of this book, I argued that, in perception, we are in direct cognitive contact with the physical world, but with it in the manner that that world presents itself to creatures with our senses and capacities, and that this is entirely compatible with the sense-datum theory, at least as I interpret that theory in the Introduction. In this second part, I investigate what the world might be like in itself, as opposed to how it presents or manifests itself to us. I begin, in this chapter, by looking at the conceptions of the world that are presented by physical science, and by scientifically minded philosophers of a realist disposition. I find all these scientific realist conceptions very problematic. In the following chapters I look sympathetically into various idealist arguments in favour of a Berkelian account of the world. Some sober readers may find my conclusion hard to accept, but I hope they will find none of the arguments tendentious or unprovocative.

9.2 Sensible Qualities, the Nature of Matter and the Regress of Powers In looking at the scientific account of the nature of the physical world, we must begin from the changes that seventeenth-century science introduced. We naturally think of the physical world as possessing the kinds of properties that our senses perceive it as having. A serious problem arises, however, when we realize that, once those qualities that are sense-dependent (and, hence, consciousness-dependent) are discounted, the resultant conception of matter is attenuated to the point of emptiness. The argument proceeds as follows. The traditional list of sensible properties centres on what are usually known as primary qualities and secondary qualities. The secondary qualities—such properties as colours, sounds, smells, and tastes—have often been held to be perceiver-

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0010

148

     

dependent, and not part of the wholly mind-independent world.¹ But even if they are construed in a naive-realist manner, few philosophers would think them suitable to be taken as constituting the core of matter. The causal powers of matter, when explicated scientifically, are always put down to the primary qualities of objects. This leaves us with the primary qualities, as the materials from which to build up our concept of matter. From the seventeenth-century atomist tradition, these are the spatial properties, such as shape, size, motion through space, and solidity. In effect, this means that Newtonian atoms are volumes of solidity. Solidity is essentially the capacity or power to resist penetration by other bodies. Hume (2000: I iv 4, 150) objected to the idea that material bodies were volumes of impenetrability on the grounds that this left us with the concept of a body as something that resists penetration by other bodies, and that this is circular because the concept one is trying to define appears in the definition. It is not clear that Hume’s argument is a good one. To say that things are material bodies if they cannot interpenetrate each other seems to pick out the desired class of entities. Nevertheless the ontology of Newtonian atoms suffered from fundamental problems. They were supposedly absolutely rigid, but for such bodies to bounce off each other would require an infinite force at the point of change of direction. Conceptual criticism of the notion of inelastic, solid atoms, and the increasing scientific importance of notions such as energy and charge pointed away from an ontology of solid atoms to one of forces, energy, or fields; that is, to an ontology of powers, or entities defined in terms of what they are disposed to do.² Properties of these kinds are the primary qualities of modern physics. Rom Harré, drawing on the theories of the eighteenth-century Croatian priest and diplomat, Boscovitch, expresses the general form of the powers ontology as follows: Every fundamental theory must, as expressed in the language of physics, be a field theory. (1970: 313) The ultimate entities of the world, as we can understand it, must be point sources of mutual influence, that is centres of power distributed in space . . . (1970: 308)

Contemporary science seems to support this picture. Concepts such as mass, charge, spin all describe what particles will do, rather than any further intrinsic nature. According to quantum theory, the fundamental entity is the quantum field, and, indeed, according to some interpretations of the theory, such as the

¹ For a discussion of arguments against realist views of secondary qualities, see Robinson (1994), ch. 3; and for a more detailed statement of the arguments in the context of a discussion of idealism, Robinson (1982). ² For a lucid account of this development, see Harre and Madden (1975), ch. 9.

 ,    

149

‘many worlds’ version, this field is all that there is at the fundamental level. If these theories are to be taken as in principle complete, and realistic, then the fundamental ontology must be one of powers.³ Although the ontology of powers can be seen as an alternative to the Newtonian conception of body, for our purposes, the difference is not as significant as it may seem, because solidity, conceived of as impenetrability, is itself a power—the power to resist penetration. Not surprisingly, therefore, an accusation of vacuity similar to that deployed by Hume can be brought against any pure powers ontology. In outline the case is as follows. A power is a power to produce some effect, but if everything is a power, it is the power to produce another power (presumably by modification of a power-entity already present, not by creation of a power from nothing). Why this leads to a regress can be seen as follows. Let us call the first power A. We only know what A is if we know what kinds of thing the actualization of its potentiality give rise to. In other words, we only know what A is if we know what it is a power to do, what states would constitute its manifestation. Let us call the power which A is the power to produce, ‘B’. So what A is, is the power to produce B. But this is not informative unless we know the nature of B. B, being a power, is the power to produce some further power state, call it C. It seems that we are moving into a regress. You can understand what a power is only by reference to a further power, of which one has no specific conception unless you know the power state which would be its effect, and so on. And though I have stated this argument in terms of what we could know, the argument is not essentially epistemological. One could equally well say that what the nature of A is depends on what it is a potentiality for, for what a power is, is given by what it is a power to do. What it is a power to do is a function of what would constitute its manifestation, and if the nature of this latter can have no determinate expression, neither can the power which is defined in terms of it. If the physical realist is committed to the powers conception of matter, and if that conception is radically defective, then the physical realist is in trouble. Furthermore, his opponent will point out that, if the pure powers conception is vicious, then the powers that are supposed to constitute matter must produce something which is not itself a power, but a monadic quality. And such qualities are to be found as sensible qualities in the sense-fields of perceivers. So the physical world is a structured capacity to give rise to experience, and this is an idealist conception.⁴ We must look more closely at the regress argument to see whether it is really vicious.

³ Later in the chapter we will consider attempts to avoid a realism about powers. ⁴ This is slightly too swift. For a more detailed spelling out of this part of the argument, see Robinson (1982).

150

     

Disputes about whether a particular regress or circle is vicious tend to be between those who say that, so long as one can always name the next element along the line, the regress is not vicious, and those who say that, unless the process can be completed, nothing contentful has been proposed. There are regresses of both kinds and the problem is to decide into which camp a particular regress falls. The crucial issue seems to be whether the content of the early members of the series depends on the content of the later. Here is an example of a regress where the content of the earlier members do not depend on the later: if it is true that p, then it is true that it is true that p, and so on. No-one, so far as I know, thinks that this is a vicious regress. The reason is that neither the content of, nor our understanding of the content of, the assertion that p is true depends on the later elements in the list. No-one would suggest that you cannot understand what it is for p to be true unless you already understood what it is for it to be true that it is true that p. Another kind of non-vicious regress is where an infinite series approaches asymptotically to a limit, for in that case the earlier elements do not depend on the later. Pi is 3.14159 . . . , with no end. But, although the next number will make the figure more exact as an expression of pi, the number we have so far (that is, 3.14159 without the dots) is both a complete numerical expression in itself and accurate to a high degree of precision as a representation of pi. The significance or content of what has gone before—or that it should have any significance or content at all—does not depend on what comes next. The matter is otherwise in the case of powers. What it is to be a particular power does depend on what it is a power to produce, for to say that something is a power, without saying what it is a power to do, does not distinguish it from all other powers: which, if powers are all that there are, fails to distinguish it from anything else. So, for any given power, there must be a contentful nature to what it is a power to do, otherwise it is not differentiated from anything else. If ‘power to produce x’ is to be differentiated from any other power, there must be a specific content substitutable for x. If that content is ‘power to produce y’, this will have no specific content, differentiating it from any other power, unless there is a significant substitution for y. And so on. It is because the notion of ‘power to produce . . .’ is an incomplete expression that trying to complete it by putting an expression of the same form into the blank does not improve the situation. It might be productive to compare the regress of powers with one which, unlike pi or the regress of ‘is true’, is genuinely controversial. This is the regress of causes and the kin issue of whether the world could be infinitely old. There are those who think that the regress constituted by causes stretching infinitely back into the past would be vicious and those who think that it would not. Notice that here there is no question of the nature of any event or entity—say a present one—depending conceptually on the nature of the events or entities that went before. So, suppose

 ,    

151

that Aristotle were right and the world were infinitely old and humans had always existed. The essence of what I am—namely, a human being or a rational animal— would not involve essential reference in a regressive way to the previous members of the series. The definition can be given complete and well-formed for contemporary humans in their own right. The controversy about a regress of causes relates to the legitimacy of actual infinities, not to anything about the nature of the things that are supposed to participate in the infinite series. This case is, therefore, radically different from the regress of powers and defenders of the pure powers ontology consequently cannot seek any solace by appeal to a comparison with the dispute concerning an unending regress of causes.⁵ The argument has an interesting consequence that is not, I think, generally noticed. It is generally agreed that the candidates for being the non-power properties that arrest the regress must be sensible qualities. Sensible qualities can be thought of either as the properties of macroscopic objects, or as the phenomenal features of mental states. If one thinks of them in the former way, one might seem to have preserved the integrity of the physical world, for it would mean that the regress is halted within the physical realm itself: all that would be required is that one reconcile the existence of ‘emergent’ macroscopic features with the fundamental nature of microscopic reality. But this would not be a solution, for this approach has the unacceptable consequence that the microscopic entities are only possible in virtue of their ability to give rise to macroscopic objects with sensible qualities. Such a dependence might be acceptable if thought of as a matter of our ability to make sense of the microscopic—we must be able to model them in macroscopic terms—but does not make sense as a requirement on their very existence. The coherence of a world consisting of quarks or of the quantum field cannot be dependent upon their ability to produce properties found primarily in ‘middle sized dry goods’. One consequence that one might draw is that one must abandon the ‘pure powers’ ontology and concede that microscopic entities must possess more properties than those powers and dispositions that science attributes to them. This is seriously in tension with science, but we will return to it later. If, on the other hand, one thinks of sensible qualities as phenomenal features of mental states, then the attempt to cash the outcome of powers in terms of these qualities faces an even worse problem. The powers that physics attributes matter give rise to movement and distribution of matter, they do not give rise to phenomenal states. This is the classic ‘explanatory gap’ that makes reductionist accounts of mind impossible. In J. J. C. Smart’s phrase, mental states are connected to neural states by laws which are ‘nomological danglers’, that is laws which cannot be embedded into the main body of physical laws. The powers of the ⁵ For lucid discussion, from both sides, of whether actual infinities are coherent, see W. L. Craig and Q. Smith (1993).

152

     

micro-parts cannot, therefore, be given determinate sense by their tendency to give rise to phenomenal states, because that outcome is not built into the identity of those powers. How is one to deal with this? It seems, furthermore, that even if one could understand the physical powers in terms of their ability to produce conscious states, that would not capture the way science works, because physics develops to explain the manifest world, not initially or explicitly the operations of parts of the brain. Neither Newton nor Einstein, nor the founders of chemistry, were primarily concerned with laws governing the brain. So, if one regards the manifest world as a construct from phenomenal contents, as earlier chapters of this book argue, then it begins to look as if scientific theories are designed to explain the workings of something which is a construct from experience, though those theories cannot explain the intrinsic nature of experience. Furthermore, if we want to attribute an intrinsic nature to the creatures of scientific theory, we in fact do this by building models based on the manifest world, which, as we have argued, cannot itself be explained by science. These models become more and more abstract and seem to favour the view that science deals with abstract nomological structures, which are cashed by their relation to a phenomenal construct.

9.3 Contemporary Discussion of the Powers Regress I have expounded what I believe to be the strongest version of regress argument, and have spelled out what I think are its consequences. There are those, however, who either deny that there is a vicious regress, or seek to downplay its importance. I shall now consider some of those. (i) The most enigmatic is Simon Blackburn’s (1990). Blackburn’s initial response to the powers ontology is positive. He says of this picture of the world: ‘Is this the way it has to work? I believe so.’ (63) But he then considers a version of the ‘regress’ objection to this theory, accepting that the regress can only be ended because ‘[c]ategoricity in fact comes with the subjective view . . .’ (65). And he continues ‘[t]he trouble is that such events, conceived of as categorical, play no role in the scientific understanding of the world’ (65). This appears to be a complete surrender to the argument. Blackburn’s only response is to lapse into a sceptical detachment. After remarking that ‘I leave the issue in Hume’s hands rather than Berkeley’s’ (64, n.7), (which, I presume, is a way of refusing to admit that it might have idealist implications), he concludes:

    

153

It almost seems that carelessness and inattention alone can afford a remedy – the remedy of course of allowing ourselves to have any idea at all of what could fill in space. (65)

This sceptical hauteur does not seem to me to help the realist come to grips with the challenge that the argument presents. (ii) Stephen Mumford thinks that the argument rests on a view of dispositions which is akin to Ryle’s empiricist analysis where the ascription of a disposition is nothing more than ‘affirming the truth of a subjunctive conditional’ (33). He is attacking the argument specifically as it is stated in Robinson (1982), but if the criticism is to have any general application, it must be an objection to any attempt to define powers or dispositions in terms of the states that constitute their actualizations, meaning by that the state brought about when they are activated. Mumford spells out his objection: We need not look for the actualization of a disposition solely in its manifestation, however, for we understand dispositions to be actual whenever they are ascribed. This is the realist alternative that I will be defending. If we were to treat dispositions as actual properties that play a causal role in their manifestations, then we can understand why dispositions are actual even when not currently manifested. (33–4)

Mumford’s objection only works if essentially characterizing a power in terms of its actualization is equivalent to adopting the reductive account of powers and dispositions. Surely, however, the nature, or essence of anything which is a pure power must reside in that which it is a power to do whether one conceives powers realistically or reductively. I think that Mumford is misled by a possible equivocation in the sense of ‘actual’ in this context. The fact that a realist believes that a power is actual, in the sense of actually existing when it is not activated, does not mean that its nature is not expressed by reference to its actualization, in the sense of the manifestation of the power. Even when dormant, it is essentially the power or disposition to give rise to that actualization. (iii) George Molnar (2003: 173ff) titles the regress argument ‘always packing, never travelling’. Although not a defender of the ‘pure powers’ ontology, he does not believe that it can be refuted a priori, as the regress argument purports to do. He has two main responses to it. His first objection has some similarity to Mumford’s objection considered above, in that it concerns the connection between the real existence of powers and the truth of counterfactual conditionals. When Blackburn states his version of the powers ontology, he does so in a way that seems to imply that powers are either equivalent to, or at least imply, the truth of conditionals. (He says, e.g.: ‘An electrical field can abide, certainly, but that just means that there is a period of time over which various counterfactuals are true’

154

     

(63, my emphasis).) Molnar points out that there has been considerable literature in recent years devoted to casting doubt on whether dispositions can be analysed conditionally, or even whether they entail conditionals (176–7). But the argument as I have stated it above does not make explicit appeal to conditional statements, only to the essentiality to a power of what would constitute its actualization or manifestation. No one would deny that the nature of powers and dispositions essentially involve what would constitute their manifestations, and any problem derived from difficulties in formulating the exact connection between powers and conditionals that goes beyond this is not relevant. Molnar has a further objection directed against C. B. Martin’s statement of the regress argument against pure dispositionalism. Martin has also suggested that leaving out physical qualia at the level of the fundamental entities results in a Pythagorean ontology in which all is numbers, quantities, ratios and proportions, but there is no whatness, no quiddity, nothing that the numbers are numbers of and quantities of. Except of course further numbers, which returns us to the regress generated by pan-dispositionalism. I think Martin’s worry can be assuaged. If the property of exerting a certain force is a definite something that the numbers can measure, so is being the source of that force. That about the object that makes it a source of a force is a (quantitative) power property. It is open to the dispositionalist to say that this is where the quiddity lies, this is what the numbers are numbers of. (179)

The problem with this response is that it is hard to see what content one is supposed to attach to the italicized words in the claims that ‘the property of exerting a force is a definite something’ or that there is a ‘[t]hat about the object that makes it a source of the force’, if all the properties are pure powers. Molnar is talking as if that which lies behind the force is something other than the force itself, considered as potential or dormant, which on a pure powers ontology it is not. Assigning a numerical value to this potentiality clearly adds nothing relevant. Of course, it would make a difference if one could consider powers as Janus faced, by giving them some other nature additional to the capacity to act. I turn to considering various ways that the physical realist might do this in the next section. (iv) There is one influential attack on this argument that is not treated in Robinson (2009), and that is provided by Alexander Bird in his (2007). Bird’s response to the regress argument moves in two stages. First he claims that those who are prepared to accept vagueness in properties will not be disturbed by the argument. For example, ‘[t]hose who take an epistemicist approach to vagueness will reject this conclusion’ (2007: 136). He then acknowledges that, though the problem no longer exists based on the need for the determinacy of a property, there is still an issue about identity.

    

155

But even to those for whom indeterminate natures are permissible, the impermissibility of indeterminacy of identity should be clear, thanks to Gareth Evans’s (1978) argument that indeterminate identity violates Leibniz’s law. Bearing in mind that for the requirement the identity be determinate is more obvious than the requirement that nature be determinate it will be more effective to present the regress argument in terms of the determinacy of identity rather than of nature. (136)

The issue of identity, stripped its association with natures, is deemed soluble by appeal to graph theory. If I understand Bird’s argument, it is that the points on a graph can be adequately identified in purely relational terms, provided the graph is asymmetric. (If it is not asymmetric, then you have something like the indeterminate identification that would come with a mirror universe; you would not know which was which between two isomorphic points.) The argument by which Bird reduces the issue to what one might call one of bare identification thus leaving it seemingly open to a formal, geometrical solution, is, I think, entirely misconceived. Vagueness is irrelevant to the regress argument. The regress charge is that no identifiable content has been assigned to the powers, not that the borders of those powers are vague. Vague predicates standardly have clear paradigm cases—Yul Brynner’s baldness and Everest’s mountainousness are not contested and clearly identifiable. The powers regress’s claim is that no central case content has been given, not that the content is indeterminate in a sense like that of vagueness. Indefiniteness, in the sense of vagueness, is not the same as vacuity, and the latter is what the regress argument purports to prove. So the possibility (which I do not need to deny) of identifying a node on a graph or geometrical figure in a purely relational fashion is, it seems to me, completely irrelevant to the task of giving content physical properties in terms of the states they (paradigmatically) give rise to. Neil Williams (2019) uses a response which is, as he acknowledges, similar to Bird’s. He admits that the regress looks threatening to what he calls the identity of powers. But he thinks that the presence of ‘structure’ in the interrelation of the powers is enough to fix their identity. Talking of the dependence of the nature of a power on those that it interacts with, he says: Now the worry starts to emerge: the specific nature of the first power is given in terms of other powers, but those other powers must rely on yet more powers to determine their natures. And those further powers must do the same. The threat of a regress seems very real. (98)

He thinks that this is overcome by the complex interrelation of powers into a structure. The next sentence after the above quotation is:

156

     

And it would be, were we unable to fix the identity of all powers holistically.

I think that he has missed the core of the regress argument, and this rests on a certain ambiguity in the sense of ‘identity’. The regress argument is not claiming that one cannot distinguish between power P1 and power P2: in that sense of identity—‘which is which’—there need be no issue, the place in a structure could do the job, but only if the structure itself is a physically real structure. What the regress argument claims is that in a powers monism there is no content to what the powers are powers to do, and so they are vacuous as a possible model for physical reality. A purely formal, mathematical, or ‘syntactic’ structure is not merely, in some way ‘unsatisfying’, it is not an account of the full nature of anything physical, or otherwise non-abstract. But Williams still thinks that the pure powers ontology (which he calls ‘power monism’) is, in some sense, lacking in substance, and it is this secondary problem that quality is required to remedy. His misinterpretation of the ‘identity’ problem has repercussions for the adequacy of Williams’s own remedy for making the world more substantial by adding qualities. He argues very plausibly against any dualism of power and quality in the nature of the physical itself which gives the quality a nature clearly distinct from the power to which it belongs yet makes it a constituent of the power itself. His preferred solution is that the quality belonging to a given power is simply what it is like to be that power. (The sense of ‘what it is like’ does not carry any experiential commitment, merely a qualitative one.) This is not just a limitation to what we can know about that quality, it is all there is to it. It seems to me that this does not touch the regress, for there is still no more to what a power does than modify other powers: the fact that ‘there is something it is like to be’ those powers does not help to specify their manifestation, and it is the absence of this latter that empties them of content. It is interesting that Williams’s second favourite for being an account of the role of quality comes close to admitting this point. He is sympathetic to a dualism in which [qualitative] properties are included in manifestations. (112)

This is, of course, the idealist solution to the problem, and the one which Blackburn admits is the only one he can see, and to which he wishes to turn a blind eye! It is, in fact, the only solution which answers the regress.

9.4 Grounding Basic Powers The dispositions, capacities and powers of complex objects are grounded in and explained in terms of the properties of the more minute structures that make up

  

157

such complex objects. The problem that concerns us is with the powers of the ultimate elements of physical reality. If a pure powers ontology is not coherent, how should a physical realist think of basic powers as being grounded? Before considering this, we must remind ourselves of the locus of the problem. There are philosophers, such as Foster and Armstrong, who think that it is a conceptual requirement that powers or dispositions have an owner, but the regress argument does not depend on this. In fact, if one is a realist about powers, and not a Humean about causation in general, then I can see no objection to the idea that there could be, among other things, autonomous powers. The model would be a magnetic field with no solid chunk in the middle. The regress concerns, not the ownership of powers, but the determinate status of their final actualization. One might assume that, if powers are grounded in categorical properties, then their actualization could consist in the way they move or otherwise modify the categorical properties of other powerful objects. It may not be the case that the postulation of some mysterious quiddity to play the role of owner would help in giving content to the manifestation of the power, for the forces and fields postulated in science are defined by those effects that are ultimately observable: could effects on unknowable quiddities stand in place for these? We will have more to say about Lewisian theories later. For the moment we will consider attempts to ground powers in something not a power, bearing in mind the question of whether these theories, even if correct, solve the actualization regress. There are those who want to ground powers in qualities, and there are also the defenders of the ‘Janus faced’ option, which says that, in some sense, there is no difference between qualities and their powers: these are just two different ways of conceptualizing the same thing. Perhaps this would answer both the grounding and regress problems, because the qualitative ‘aspect’ could be thought of as the ground, and the effect of the power be the change—perhaps just of place—that it makes to the qualitative ‘aspect’ of other quality-powers. This would not solve the problem if the theory was as follows; powers have a qualitative nature, that is, not that they are associated with a quality, but that the powers themselves are qualitative. One might call this the theory of qualitative powers, or dynamic qualities. This suggestion, if coherent, would not touch the regress argument, because the identity of this entity would still consist in what it does, and that would still reside in the modification of another power. This theory would be a theory of what powers are, not a theory which ascribes some non-power feature to the powers. Putting this suggestion aside, the theory we are considering is that there is something that can be conceptualized either in terms of powers or dispositions, or in terms of qualities. The main problem I have with this theory is of making sense of the actual nature of this thing that can be seen in both ways. Harré illustrates what is wanted, taking the example of solidity.

158

     

Solidity is the alleged quality, the possession of which is responsible for the fact that two material things cannot occupy the same place at the same time and is logically connected with impenetrability, the power to resist penetration, in that the possession of the former is supposed to account for the manifestation of the latter. (1970: 305)

Martin at one time held a similar theory, with any intrinsic quality treated as ‘a two sided dispositional qualitative coin . . . The dispositional and the qualitative are equally basic and irreducible; there is no direction for one being basic in a property and the other “supervenient” ’ (1997: 216, citing his previous view). He moved, however, from this ‘Janus faced’ theory, to one according to which the dispositional and the qualitative are not different aspects of a property, but are identical. . . . the qualitative and dispositional are identical with one another and with the unitary intrinsic property itself . . . What is qualitative and what is dispositional for any property is less like a twosided coin or a Janus-faced figure than it is like an ambiguous drawing. A particular drawing, remaining unitary and unchanged, may be considered one way as a goblet-drawing and differently considered, it is a two-facesstaring-at-one-another-drawing. The goblet and the faces are not distinguishable parts or components or even aspects of the drawing, although we can easily consider the one without considering, or even knowing of, the other. The goblet-drawing is identical with the two-faces drawing. (216–17)

The analogy does not look helpful. What is interpreted as a goblet or as two faces, is neither of these things, it is a line or lines on a piece of paper. As there is clearly nothing accessible which is seen as a disposition and seen as a quality, what is really present must be some mysterious third unknown. Such a mysterious unknown plays a part in Harre’s attempt to cope with the regress. Rom Harré played a major role in popularizing the powers ontology in the philosophy of science, and in drawing attention to the importance of Boscovitch in the historical development of this conception of matter (Harré, 1970; Harré and Madden, 1975). It is interesting to see, therefore, how he copes with the problems that face it. Harré’s way of recognizing the problem is somewhat elusive. He refers to the dilemma that confronts anyone who tries to universalize the dispositionalist account of properties. It seems as if one must choose between the inelegant alternative of grounding science on ungrounded dispositions, and the alarming prospect of an indefinite regress of groundings. (1986: 296; my italics)

  

159

Given that one is dealing with the ultimate nature of reality, it is not surprising if it should turn out to be unfamiliar in a disturbing way, so the force of ‘inelegant’ and ‘alarming’ is unclear. Furthermore, it is hard to see what is inelegant about ungrounded powers, if the notion makes sense. Nevertheless, Harre does acknowledge it as a problem that he wants to solve. The immediate solution that he gives is as follows. Ideally the dispositions which theoretical micro-regresses require physicists to ascribe to unobservable beings, like quarks and gluons, would be grounded, at least in principle, in observable properties of the universe. These properties would be occurrent rather than dispositional, embracing such matters as the quantity and distribution of energy fields. (296)

This solution, which he derives from Mach, looks rather like a form of phenomenalism, because the subatomic world is being grounded, ontologically and not just epistemically, on the observable. (I put aside the oddity of calling energy fields ‘occurrent rather than dispositional’.) But later he develops the theory in a Kantian direction. The basic powers of matter are the product of the interaction between ‘the ur-stuff of the world’—which he prefers ‘to nickname . . . “glub”, to avoid any of the metaphysical temptations that arise from the connotations of the word “stuff ” ’—and our observation apparatus. Whatever this theory is, it is not straightforward physical realism. Harré compares his ‘glub’ to Kant’s noumenal world, having no characteristics accessible to us. The theory, therefore, seems to have more in common with Kantian idealism than with normal physical realism. Where does Harré’s theory leave us with regard to the problem mentioned at the start of this section—does it solve the actualization regress? It is difficult to see how making modifications in some Kantian ‘glub’ can be what gives differential content to the various forces and their values that are found in physics. John Heil has also developed a theory which tries to marry qualities and powers closely together, calling them ‘powerful qualities’ (Heil, 2010). He seems to me to have three separate arguments for this conception. First, he argues that if colours where always in fact associated with certain powers—for example, if blue spheres always attracted yellow ones—we would naturally think of them as necessarily connected. He then seems to back down, admitting that any such connection is stipulative and so can easily be thought contingent. Next, he considers the connection between shape and behaviour, suggesting that it seems appropriate that a sphere should role down a slope rather than ‘tumble’ down it. Certainly, the connection between geometrical properties and movement seems, in some cases at least, not arbitrary in the way that the connection between colour and any causal powers seems to be. It is not logically impossible that cubes should role down an incline but spheres stand still, but imagining a complete and consistent physical

160

     

dynamics based on entirely counterintuitive principles like this might prove very difficult, if not impossible. Geometrical properties, however, are not directly relevant, for the powers argument concerns the supposed fillers of space, not spatial properties themselves. These, it is argued, are all powers or dispositions, like solidity and mass. Heil’s third argument is that qualities must be essentially tied to powers, otherwise we could not perceive them. I cannot see the force of this argument. If secondary qualities are essentially subjective, then they constitute the contents of mental states, so we could detect the presence of a power by its effect on us, without it being associated with an external quality. This latter position is, in effect, the Berkelian position with regard to our awareness of God’s action in producing our experience. An idea that might be thought common to all the, roughly, ‘Janus faced’ approaches, is that the quality-power dichotomy is a function of our modes of conceptualization, rather than a straightforward feature of the world itself. This approach seems to make the world itself noumenal, and make it a complete mystery what makes these two forms of conceptualization catch on to the world. Any attempt to render this approach more intelligible always seems to lead one in a circle: there is only one kind of thing out there—not two distinct properties—but it has different ‘aspects’ or ‘features’ that quality talk and power talk latch on to: but how are these not the two different properties one started off trying to avoid? One is in a ‘same and different at the same time’ dilemma. Perhaps it is possible to stand back and say that it is useful sometimes to think of things as qualitative, and useful sometimes to conceive of them as dispositional. One tries to escape the dilemma by a kind of pragmatism. But why should it be ‘useful’ if these features are not ‘out there’? And in the more familiar cases, such as the solubility of salt in water or the fragility of glass, the qualitative and dispositional features of the objects seem to exist side by side and not as alternative conceptualizations. If, in the end, someone simply insists that these are the categories we think in, and how they latch on to the world is either a mystery, or that to even ask is a misconceived question, it is difficult to see how to pursue the argument.

9.5 Quiddities, and Similar Devices Lewis and Armstrong both add unknowable quiddities to the causal relations. Two things are worth noting about their doing this. First these quiddities do not, as you might imagine, do the job of a Lockean substratum, of binding an object together. Quiddities, unlike substrata, belong to properties, not to substances. So whereas an electron has a substratum (if one believes in such things) in which its properties of charge and mass inhere, for Lewis and Armstrong the quiddities go

,   

161

with the individual properties, so there are two quiddities involved in an electron. I don’t think this makes any relevant difference, though it seems to me strange. A rather different way of giving a categorical nature to fundamental matter, is Michael Esfeld’s notion of matter points, which figure in what he calls his ‘minimalist ontology’. It is all built on what he calls his ‘two axioms’. (1) There are distance relations that individuate simple objects, namely matter points. (2) The matter points are permanent, with the distances between them changing. (2020: 1893) He goes on to explain the nature of matter points, and their influence on each other as follows. This new version of Humeanism can be dubbed Super-Humeanism. While the standard Humean, following Lewis’s thesis of Humean supervenience (e.g. Lewis 1986a, pp. ix–x), holds that there are spatial or spatio-temporal relations connecting points and natural intrinsic properties instantiated at these points, the Super-Humean maintains that there are only sparse points that then are matter points with distance relations individuating these points; but neither is there an underlying space nor are there natural intrinsic properties. . . . Hence, on Super-Humeanism, parameters like mass and charge are no addition to being. The configuration of point particles and its change is all there is. Given that this change exhibits certain patterns, laws can be formulated, and given the laws, one can attribute parameters like mass and charge to the point particles. But these are not properties that the particles have per se, as something essential or intrinsic to them. They obtain them only through the regularities that the change in the distance relations among them exhibits. (1897)

There are two controversial claims here. One which he shares with Lewis and many others is his commitment to a Humean account of causation. The other, more peculiar to him, is the notion of a matter point. These entities have no intrinsic properties, but are purely punctiform, and can move, which they do according to the laws of nature, Humeanly construed. The only sense in which they can be called ‘matter’, as far as I can see, is that they follow the laws of physics, but they have none of the traditional properties of matter, for example, extension, solidity or mass. It is true that two of them cannot occupy the same point, but that cannot be because of their intrinsic nature, as they don’t have one; it is just another Humean fact about how they behave. Esfeld seems to take the nodes in a purely structuralist picture, call them ‘matter points’ and thereby think he has postulated

162

     

entities, but this ‘minimalist ontology’ seems to me to be a vacuous ontology, containing entities in name only. You cannot make something material simply be dubbing it ‘matter’ whilst denying it any recognizable material properties.⁶

9.6 The Humean Account of Causation: Against the Primitiveness of Regularity Esfeld argues against the postulation of laws or powers at the fundamental level in the following way: However, in subscribing to such commitments [that is, laws or powers], one does not provide an explanation of the change in the distance relations that is illuminating by contrast to accepting that change as a primitive as does the Humean. It is true that one traces that change back to modal properties of the matter points, modal structures instantiated by their configuration, or laws endorsed as primitive. But all these are defined in terms of the causal role that they exert for the motion of matter. The explanations hence are circular. For instance, one does not give a deeper explanation of attractive particle motion – i.e. answer the question why attraction happens in addition to subsuming it under unifying laws – in terms of mass or the gravitational force, because these are defined through the effect that they have (or can have or are the power to have) on the motion of the particles. . . . The same goes for laws admitted as primitive. All this is an instance of the scheme at which Molière pokes fun in Le malade imaginaire: one does not explain why people fall asleep after the consumption of opium by attributing a dormitive power to opium – although, of course, mass and charge, or quantum entanglement are sparse, fundamental properties or structures by contrast to the phenomenological properties of opium. Nonetheless, the Molière argument hits also these latter properties or structures: like the dormitive power of opium, they are defined in terms of the effects that they bring about under certain conditions – in other words, by the functional or causal role that they exercise for the evolution of the objects to which they are attributed. (Esfeld, 2020: 13)

In short, Esfeld’s argument is that one does not provide a real explanation of regularities by postulating something—a power or a law—the sole content of which is the claim that they make these regularities occur. Any seeming difference between ‘these regularities occur’ and ‘there is a regularities-making factor which makes them occur’ is just a verbal illusion. ⁶ See my comment in Section 9.7 below on the empty nature of the ‘particle(s)’ postulated in the Bohmian interpretation of Quantum theory.

    

163

This seems to me to miss the real point, which concern why the regularities occur at all and, particularly, why they persist. Foster (2004) and Robinson (1985) have very similar arguments against mere regularity theories on these grounds. The argument is essentially that if there were nothing enforcing or imposing the regularities in nature, chaos is what one could reasonably expect (or perhaps nothing at all). Foster takes the case of gravity and argues as follows (I have added the numbers). (i) First, given any body existing at a certain time, there is an infinite (indeed a non-denumerably infinite) range of logically possible forms of behaviour open to the body at that time, only one of which would be, in the prevailing conditions, gravitational. (61) (ii) Secondly, there have been a vast number of occasions when bodies have been observed, or scientifically monitored that would have revealed nongravitational behaviour had such behaviour occurred . . . (61) (iii) Thirdly, on none of these occasions has non-gravitational behaviour been detected . . . [occasional weird phenomena, e.g. levitation, excepted] (61–2) (iv) Now consider the hypothesis (H) that this consistency in the gravitational behaviour of bodies over the cases examined has occurred for no reason—that nothing has been responsible for ensuring or to any degree encouraging it . . . But if the situation postulated by H is one whose obtaining would be objectively hugely surprising then the rational expectation that this situation does not obtain, and that there is something which accounts for the past consistency of gravitational behaviour. (62) Notice that Foster talks of its being ‘hugely surprising’ and does not talk about it as improbable, though it is objectively surprising, and the rational expectation is that this would not happen. He also says that an outcome like H is ‘a very strange one’ and would ‘warrant huge surprise’ [italics added] No doubt Foster avoided appeal to probability at this stage because he was well aware of the Humean approach to probability, as found in Ayer’s book Probability and Evidence, which denies the legitimacy of a priori probabilities, beyond the facts of mathematical proportions. The only things that are allowed to be empirically improbable are those things that we know do not standardly occur—like flying pigs—and the regularities of nature do not fall into this camp. Foster regards the point that he illustrates by ‘the law of gravity’ as showing how implausible this radical empiricist position is. He argues (like Robinson 1985) that the fact that there are massively more disordered options than there are ordered ones is enough to rationally demand an explanation for why an ordered one happens.

164

     

If the regularity theorist persists, I think that there are at least two strong arguments which show that he must be a full-blown Humean sceptic and not a Lewisian realist. The first argument rests on Foster’s solution to the problem of induction. He summarizes this as follows: 1. There is a significant range of cases where the knowledge of a hitherto exemplified regularity provides rational grounds for an extrapolative inference. 2. In any such case—or, at least any case which is exclusively concerned with the workings of the physical world—the grounds can be made explicit by breaking the inference down into two further steps of inference, neither of which is, as such, extrapolative. 3. The first of these two steps is an inference to the best (the most plausible) explanation. What the explanation is advanced to explain is the occurrence of the hitherto exemplified regularity whose extrapolation is at issue. And this regularity calls for explanation because it is too extensive to be deemed coincidental—deemed to be something that occurred for no reason. 4. This explanation involves the postulation of some law or set of laws of nature, sometimes precisely specified, sometimes not. In the simplest cases, the postulation of these laws will form the whole explanation of the relevant regularity. In other cases, the explanation will additionally involve the recognition of certain standing conditions—sometimes precisely specified, sometimes not. Which of these forms of explanation is appropriate depends on other things we independently know or have reason to believe about the world. 5. The second step of inference is a deduction from the explanation—a deduction that the regularity (thus explained) will continue to hold for the unexamined cases, or will do so subject to the continued obtaining of the relevant standing conditions. I know of no other solution to the problem of induction, and rejection of the rationality of induction commits one to full-blown scepticism. The crucial point for current purposes is that without the postulation of laws, the ‘argument to best explanation’ will not get going; it would be circular to postulate regularities to explain regularities. The second point, I think, forces the regularity theorist into even deeper scepticism. He might say that we have to accept certain facts as basic or brute, and why not the regularities? The reply to this is that we could take the regularities of experience as brute, like the Humean phenomenalist, or even those of my own experience, like the solipsist. Hume has an explanation of why we do not do this, namely that it is a psychological impossibility—in modern terms, we are hardwired to be realists of some sort. But this is a Humean argument based on habit

   -

165

and conditioning—it does not show that there is anything rational—beyond the pragmatic—for doing it. A Lewisian might argue that the realist mosaic of quiddities makes a more complete and coherent picture than the fragmentary one of the radical phenomenalist or solipsist. This is correct, but why is its being a more satisfyingly complete picture a rational ground for thinking that it is true? I think that it is also worth drawing attention to the following point. The regularity theorist might say that he has only one brute fact, namely that the universe is organized in a regular, not an irregular or chaotic way. But it is not one brute fact. Each time a pattern repeats itself, it is a brute fact. Imposing regularity as a class is to make it rather like a Kantian category. Another objection that Foster considers might be described as ‘Goodmanian’— or ‘grue-ish’. Perhaps the laws only apply to the instances so far observed and will not fit future cases. His response is that such temporally or otherwise indexed laws would be analogous to that in which someone who purports to accept the principle of universalizability in ethics, but wants nonetheless to adopt a moral position which is advantageous to himself, insists that his basic moral principle is that everything should be done to further the interests of anyone whose fingerprints are thus and so – where he displays a photograph of his own prints as the specification of the type in question. (2004: 75)

So, to avoid arbitrariness, the defender of Foster’s theory . . . has the resources to justify his preference for explanations which, by postulating laws that impose uniform constraints across bodies, space, and time, are suited to the purposes of extrapolation. (75)

It seems that the Hume-Lewis regularity theory is totally implausible, and that we need laws of nature. We must wait until Chapter 12 to see what those laws must be, and, in particular, why they need a Divine lawmaker, as Berkeley believed.

9.7 Scientific Realism about Quantum Theory, and Common-Sense I have discussed the ‘pure powers’ conception of the physical world, which is primarily a philosophical theory, based, in a general way, on science. I have discussed the theories of Esfeld and the Lewisians, which are essentially philosophical theories. But what about the theory of the physical world more directly derived from physical science, especially from the quantum-theoretical account of the world?

166

     

It is almost impossible to shake off the conviction that a scientific realist view must be nearer to common-sense than an idealist one. But contemporary scientific realism is a very strange beast. Roger Penrose starts a chapter on ‘the structure of the quantum world’ as follows. Quantum theory provides a superb description of physical reality on a small scale, yet it contains many mysteries. Without doubt, it is hard to come to terms with the workings of this theory, and it is particularly difficult to make sense of the kind of ‘physical reality’ – or lack of it – that it seems to imply for our world. Taken at its face value, the theory seems to lead to a philosophical standpoint that many (including myself) find deeply unsatisfying. At best, and taking its descriptions at their most literal, it provides us with a very strange view of the world indeed. At worst, and taking literally the proclamations of some of its most famous protagonists, it provides us with no view of the world at all. (1995: 237)

David Albert agrees that the theory has put us in a strange situation, but, as we shall see, he rejects the nihilistic response. A hundred years ago, physics aspired to produce a complete, and unified, and seamless, and philosophically realistic account of the entirety of nature. It aspired to tell us straightforwardly what the world is. It aspired, that is, to settle questions of metaphysics. And all of that came to look somehow quaint and childish under the spectacular assault of quantum mechanics. (2013: 52)

The weirdness of quantum theory centres on what is known as ‘the measurement problem’. The theory assigns objects a probabilistic location, but not in an epistemological sense, so that it is, say, 70 per cent likely to be in one state and 30 per cent likely to be in another, but that it is really is 70 per cent located in one and 30 per cent in the other. This is the problem of Schroedinger’s cat in the famous thought experiment. This is bizarre enough, but it is a fact that when one actually observes a situation, one finds that the object—in experimental fact, a particle—is simply in one of the places, as common sense would expect: the cat is observed as either alive or dead. Until the point of measurement, the object behaves as a wave, with a probabilistic distribution. This is what is described as ‘the collapse of the wave packet’, because the probabilistic distribution of the particle required by the theory resolves itself one way or another. The title ‘the measurement problem’ thus comes from the fact that, whenever one tries to determine, by some kind of measurement, where the particle is, you get an answer

   -

167

contrary to what the theory predicts. This is not taken simply to refute the theory, as one might expect, because the theory is, in all other circumstances, fantastically accurate, so the natural response is that there is something weird about the measurement situation. In fact ‘measurement’ usually means ‘observation’. The probabilistic distribution is supposed to hold for any physical device used to make the measurement, including the human brain. The only point at which the ‘collapse’ manifests itself is in the conscious state of the observer. This led the early theorists to attribute special powers to consciousness. This is dubbed the ‘Copenhagen interpretation’ of the theory because Niels Bohr, a Dane, is generally credited with originating it, though many others, such as Heisenberg, Born, and Wigner defended a version of it. The fact that it attributes such a fundamental role to consciousness in determining what happens in the physical world makes quantum theory seem to be a friend of a radical form of dualism, if not of idealism. Albert, being strongly committed to the realist camp, rejects this interpretation of quantum theory in strong terms. This [Copenhagen interpretation] strikes most of us nowadays as weird, glib, scary, oppressive, intolerant stuff. (52)

Albert’s fundamental objection is that the notion of measurement, though essential to scientific practice, is not exactly definable, and therefore not suitable to function in fundamental scientific theory. There is, also, no doubt a dislike of the dualist or idealist implications. Albert recalls hearing Wigner speculate that a dog could probably collapse the wave packet by observation, but a mouse probably could not, and regards this kind of speculation as a reductio of this approach. The objection that the notion of measurement is not exact enough to figure in fundamental science looks like a powerful objection, until one realizes that that is a misunderstanding of how the concept is being deployed. The Copenhagen interpretation is at least committed to dualism, and measurement, in the form of observation, represents a form of interaction between the physical and the mental. It is not, therefore, part of any, let alone fundamental, physical theory, any more than is any other theory about mind–brain interaction, from a dualist perspective. Wigner’s speculation is no different from the common worry, which even physicalists of certain sorts (perhaps not behaviourists) might share, of whether, for example, earthworms are conscious in the same sense (though not the same degree) as we are. Whether certain animals can feel pain is a common concern. It seems to be commonly held by philosophers of physics that it is part of the nature of physics to explain everything, but this is a pure dogma, and would not have been shared by Newton, for example, so the claim that measurement has no place within basic physics is irrelevant, for it only boils down to the claim that interactionist dualism cannot be integrated into physics.

168

     

I do not wish to defend the Copenhagen theory and can see that if another, less strange, theory could be defended, it would be preferable. Albert tells us what a realist understanding of quantum theory would have to look like. And this chapter is going to be taken up with what seems to me to be the simplest and, most straightforward, and most flat-footed way of thinking—in this new realistic spirit—about quantum mechanical wave functions, which is to think of them as concrete physical objects. The most striking and controversial feature of this approach is undoubtedly that the stage on which such objects must make their appearance, the stage (that is) on which any such understanding of quantum mechanics is going to have to depict the history of the world as playing itself out, is a mind-numbingly highdimensional space . . . , a space whose dimensionality is three times as large as the total number of elementary particles in the universe. . . . But it will be best, first, to sketch out the basic metaphysics. The sorts of physical objects that wave functions are, on this way of thinking, are (plainly) fields . . . On Bohm’s theory, for example, the world will consist of exactly two physical objects. One of those is the universal wave function, and the other is the universal particle. And the story of the world consists, in its entirety, of a continuous succession of changes of the shape of the former and a continuous succession of changes in the position of the latter. (53–4)

Albert is well aware that this picture of the world has little relation to common sense realism. The particularly urgent question (again) is where, in this picture, all the tables, and chairs, and buildings, and people are. The particularly urgent question is how it can possibly have come to pass, on a picture like this one, that there appear to us to be multiple particles moving around in a three-dimensional space . . . The business of actually filling in the details of these accounts is not an altogether trivial matter. (56)

A similar view is expressed by Barry Loewer. The idea (it comes from a paper by John Bell as a way of understanding Bohm’s account of quantum mechanics) is that the fundamental ontology of the world consists of a field described by the wave function and a single point particle (the ‘world particle’) both occupying a very high dimensional space (the number of dimensions determined by the wave function). The goal then is to provide an account of how the manifest world emerges from the motions of the particle

   -

169

(the motions of the particle are determined by the field). At first (and maybe second and third. . . . ) this is very contra to way we think of the manifest world and highly unintuitive. But the idea is that degrees of freedom of the particle (motions along various dimensions) can be collected into triplets each of which can be thought of as comprising the motions of an individual particle in 3-dimensional space. These ‘emergent particles’ move so as to implement ordinary physical objects. E.g. there are a large number of particles that move (and would move under certain counterfactual situations) just like the motions of particles that comprise e.g. my cat. In particular the wiggling around of these degrees of freedom of the world particle constitute the cat. Because the world particle wiggles in accordance with certain laws (describing the evolution of the field and the motions of the world particle) a kind of 3-D spatial structure emerges. The emergence is understood as a kind of functionalist implementation so one would have to be on board with this much functionalism . . . (Personal communication)

Albert acknowledges the difficulty of working out how the one omni-competent particle manages to be in all the right places to make up the normal macroscopic world, but there are two problems here. One is internal to the science, and concerns how one might derive the almost infinite particle locations from the laws governing the one thing. But, beyond that, there are the kinds of problems raised by Russell (1912). Is there any reason to think that macroscopic physical space—or time—is actually qualitatively the way it seems? If not, even if the scientific problem were to be solved, the most that the physical world could achieve would be a bare isomorphism with the world of experience: one would have constructed a macroscopic world, but not the manifest one of commonsense.⁷ Albert and Loewer do not, of course, represent the only realist interpretation of quantum theory available, but theirs is one of the most respected.⁸ In order to escape the non-realist interpretation of the quantum account of matter, one has to imagine that something called a field, which operates in a space of a dimensionality three times as large as the number of elementary particles in the universe, manipulates a single particle, through all these ‘many worlds’ so that it can play the role of all the myriads of particles that seem to be in them. It is interesting to see how Albert explicates the notion of a field: they are the sorts of objects whose states one specifies by specifying the values of some set of numbers at every point in the arena in which they live, the sorts of objects whose states one specifies (in this case) by specifying the values of two ⁷ See also Robinson (2009). ⁸ A very different view, for example, is presented in Maudlin (2013).

170

     

numbers (one of which is usually referred to as an amplitude, and the other as a phase) at every point in the configuration space of the universe.

It is not obvious that such a notion is realist at all, for it consists in assigning numerical values to points in space. Albert, Maudlin, and most philosophers of quantum physics reject the interpretation of that science which says ‘just calculate’—that is, don’t try to give a realist or metaphysical interpretation of it, but characterizing the basic entity—the field—as something just specified by assigning sets of numbers seems to be a way of admitting that ‘just calculating’ is all that one can do! What is certainly the case is that this conception of the field is open to the same objections as the pure structure and powers conception of physical reality that was discussed in the early sections of this chapter, for these numbers assigned to points measure only degrees and kinds of influence. It might seem that the ‘universal particle’ gives extra substance to this conception, but what the nature of such a particle is supposed to be, given that all the causal properties belong to the quantum field, is itself something of a mystery. In fact, in quantum discourse, whether something is a field or a particle seems to concern whether its behaviour needs to be represented probabilistically or determinately: something is a particle when a situation need not be represented as a superposition of possible outcomes. The use of the word ‘particle’ also suggests to the layman the traditional ‘billiard ball’ entity, and, if not that, something with some kind of causal powers of its own, but it is essential to the Bohmian scheme that all the causal properties belong to the quantum field. So the particle represents utterly passive physical existence. This is surely a notion that merits the same kind of mockery as Berkeley gave to the notion of matter as a bare substratum. This is essentially the same problem as we saw is faced by Esfeld’s ‘matter points’. Bohmian mechanics is not, of course, the only rival to the Copenhagen interpretation. There is the ‘Many Worlds’ (MW) theory, according to which the whole universe splits into two every time there is a superposition; one, for example, in which Schroedinger’s cat is dead and one in which it is alive. One of the points of this theory is to preserve the deterministic unfolding of the field according to Schroedinger’s equation, and to make this field the only constituent of the world: it does not need Bohm’s ‘particle’. But if the regress argument is sound, this ontology certainly falls victim to it, for it is a pure field, and, therefore a pure powers ontology. Furthermore, it has been plausibly argued that MW is inconsistent with the central role played by probability in quantum theory. The many worlds interpretation of quantum mechanics suggests that the world we live in is just one of many actual parallel worlds that emerge from the fundamental quantum state. Those parallel worlds are not spatio-temporally disconnected from us. In fact, they are in the same space-time as our world. It

   -

171

just happens that even though we may share our spatial-temporal locations, we (more or less) do not interact with the creatures in those branches. Moreover, my action right now may influence what branches obtain and what future branches look like. This has interesting (and perhaps terrifying) consequences. On singleworld interpretations of quantum mechanics, there is just one actual world and no multiplicity; Tom’s action to raise his arm has a tiny quantum-mechanical probability to turn him into a swarm of butterflies. But no one should worry about that because it is overwhelmingly unlikely to actually happen. However, it follows from the many-worlds interpretation (and the small but non-zero probability of the event on the single-worlds interpretation) that there will be an actual world among the parallel worlds in which Tom turns into a swarm of butterflies after raising his arm. Because many-worlds interpretation does not alter the underlying deterministic dynamics, this world will obtain as a matter of certainty. Similarly, on single-world interpretations of quantum mechanics. (Rubio and Chen)

Attempts by Wallace and others to answer this criticism, by saying that the probabilities reflect our preferences, are not convincing to many philosophers of physics. It seems that the ‘many worlds’ interpretation is, possibly incoherent, resting on a ‘pure powers’ ontology; counterintuitive in the extreme, and quite possibly in conflict with one of the basic ideas of quantum theory itself. There is also the Ghiradi, Rimini, Weber theory (GRW), according to which any given particle undergoes a collapse about once every hundred million years. Given the enormous number of particles in the universe, it would seem that this guarantees the stability of macroscopic objects. This avoids the eccentricities of the others, but seems both ad hoc and untestable. The first part of this chapter was concerned with the ontology of real powers, and followed the common-sense, lay idea that energy, fields, forces are things that really make things happen. In fact many philosophers who discuss the nature of fundamental physical reality either accept, or are open to, Lewis’s neo-Humean account of causation, which denies that there is any more to causation than how objects actually move, and how they would move under counterfactual conditions. There is no special ‘oomph’ to causes, forces etc. themselves. It is plain that someone in a broadly Humean position cannot hold that powers or causes are the only things that exist, because they don’t exist in their own right, but only characterize how things behave. This maybe explains why Lewis believes in mysterious quiddities; if dispositional properties like mass, charge and other physical forces were just those dispositions and forces, they would be nothing. So anyone in the Humean tradition about causes must have something over and above the forces. This has consequences for the versions of quantum theory discussed above. On the Bohmian scheme, the particle which speeds through three times as many

172

     

dimensions as there are apparent particles does so, not as impelled by some genuine force, but ‘in its own right’, so to speak, because the quantum field turns out to be nothing over and above the pattern that the particle describes. The many worlds theory and, I think, GRW, both want to restrict fundamental reality to the field.⁹ This would deliver a kind of instrumentalism, because the fundamental reality would become just a way in which something else— presumably non-fundamental reality—behaved. Such a view would rule out even the most attenuated form of reductionism of the non-fundamental to the fundamental. It seems pretty clear that any kind of realist about quantum theory should stay well clear of the Lewis–Hume approach to causality. It is also unclear what the role of quiddities could be in Quantum theory. The only real entity involved in this theory is the field, and its only properties seem to be its values at points in space-time, which are numerically assigned. Lewis’s talk concerns mass, charge, and the more normal sort of property, but these do not figure directly at the basic level. Does any variation of the value at different points in space time count as different properties, or are they all the same property, just in different degrees? Is there a different quiddity for each point? The only way to make sense of this, I think, is to follow Einstein’s suggestion, which we saw raised in Chapter 10, that space time is kind of ether, though lacking some of the causal powers ether was supposed to possess. Of course, I cannot, as an utter layman, judge any of these scientific theories, but I have two points in introducing them into the discussion. The first is that it is hard to see any of them as a version of common sense. Are they really less counterintuitive than Berkeley’s idealism? But the second is stronger than this. Berkeley’s idealism preserves both the right degree of realism and avoids the paradoxes of the theory. Philosophers of physics often complain that during their education as physicists they were told ‘just calculate’ and do not raise questions about the nature of the reality quantum mechanics supposedly characterizes. This suggests an instrumentalist or even phenomenalist understanding of the science. Maudlin (2013) and Allori (2013; forthcoming) want to take a similarly instrumentalist approach to the laws of quantum theory, whilst being realist about our ordinary three dimensional space. They emphasise the distinction between the mathematics and the physical ontology. The mathematics involves the 3N high dimensional configuration space, but that is not a feature of the underlying ⁹ Barry Loewer has insisted to me that the quantum field is not just a power, but an entity or kind of stuff. This might make it something like the view of space that we will see in Chapter 11 was adopted by Einstein, where space was compared to the ether. Again, it seems to me that this is much further away from common experience than the Berkelian perspective, which keeps the empirical world at the centre. And, of course, what its substantiality consists in remains a mystery; it is a gesture towards endowing nomological structure with a more intuitive reality.

   -

173

ontology, including real space. So Schroedinger’s equation tells us how objects in our normal space work, and the many dimensions of configuration space are just part of the complex structure of the law that governs those objects in that space. Ney rejects this on the grounds that it is arbitrary to be realist about Maxwell’s equations, but not about the theoretical framework that has replaced them. Berkeley’s idealism removes this inconsistency by having an instrumentalist understanding of all such laws. The defenders of three dimensional space talk of the ‘primitive ontology’ or ‘beables’ that constitute the intelligible, normal ontology, but a phenomenalist treats the empirical data—the sense-data—as the primitive ontology or ‘beables’, the rest are part of the understanding or interpretation we make of these. If we say that the phenomena are primitive, not mind-independent objects, and the rest is just nomological structure, this provokes the challenge that there must be some reality that sustains these laws. Berkeley says that reality is the Divine plan for human (or other forms of conscious) experience. This does not require a bizarrely dimensional space, nor a strangely characterless particle, nor many worlds, nor an ad hoc collapse of the wave packet. All you need is the laws, but they are sustained by a fundamental reality, just not one to which one needs to give a physical interpretation. One might wonder why would God allow for the indeterminacy that expresses itself in what is called ‘the collapse of the wave packet’? As a guess, I would hazard that it preserves both the determinism of Schroedinger’s equation, and a certain indeterminacy in the unfolding of the actual world. I cannot here discuss how this relates to free will, but, rather than leaving space in the indeterminism of the brain for our free choice, maybe it allows space for God to accommodate our free choices in how He allows the world to develop in response to them. So the idealist, of course, denies none of the science, either in the form of observational evidence or mathematical expression, but can at least claim that his account of how this nomological structure is realized is certainly no more counterintuitive than that which is forthcoming from current scientific realism, and maybe that it preserves the science without physical eccentricities. This is so even before one takes into account the more standard shortcomings of a physicalist account of the world. Quantum realism is no better at accommodating consciousness, thought or value than is a more traditional form of physicalism. The idealist picture already builds the first two of these into its world, and holds more hope of accommodating the latter than its more fashionable rivals. But these are issues I cannot pursue further here.

10 Two Suggestive Berkeleian Arguments 10.1 Introduction I tried to show in the last chapter that physical science cannot deliver an adequate realist account of the physical world. From now on, my arguments will tend to be based on more purely philosophical foundations, often inspired directly by Berkeley. In the present chapter I want to look at two suggestive intuitions that can be found in Berkeley’s writings, one that appeals to empiricists, the other has had rather more take up from objective idealists in the Hegelian tradition. Both lines of argument are, in my view, suggestive rather than demonstrative. The one with empiricist appeal is expressed in Berkeley’s claim that ‘an idea can be like nothing but an idea’ (Principles, Pt I, $8), which might be interpreted, in context, as ‘qualities are essentially features or modes of experience’, so that attributing monadic properties of a qualitative sort to mind-independent objects is a kind of category mistake, rather like saying they are thoughtful or bored or intelligent. The more rationalist one is what has been called ‘Berkeley’s Master Argument’. Berkeley’s statement of it is that we ‘cannot conceive of the unperceived’ (Pr. I, $22ff.), but I will try to give it a more sophisticated interpretation, namely that we cannot form a conception of the world that does not involve attributing to it features that are essentially projections of our modes of thought.

10.2 The Sense-Dependence of Qualities Robert Adams and Michael Dummett are good examples of contemporary philosophers who follow the first, empiricist route. The thought is that reality cannot be purely dispositional or formal-mathematical, so it must possess a qualitative core or base, and there are reasons to think that qualities are essentially experiential; so, therefore, is reality. Adams begins by putting forward the argument that we developed in Chapter 9, that powers can only be determinate if they result in things that are not merely powers, but qualities. The potentially resulting state of affairs is particularly important to defining a power or disposition, which is normally understood as a power or disposition to

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0011

 -  

175

produce a certain state of affairs (under certain conditions). Powers and dispositions will be defective in positive content if they do not derive enough qualitative content from the positively resulting state of affairs. (2007: 44)

Adams then cast doubt on whether the qualities that constitute the actualization of powers and dispositions could be realized in inanimate matter: In fact I do not see how to find a clear, known case of the requisite qualitative character that is not a quality of consciousness. (46)

He goes on to admit that he ‘can find no proof that it is true’ that they must be in consciousness, but provides four reasons for thinking that this is so. His intuition is best expressed in his fourth reason. An essential motivation for this discussion of intrinsic, non-formal qualities is the assumption that if there were things in themselves, there must be something that it is like, in itself, for them to exist. We may well suspect that this notion of what it is like, in itself, for something to be the case is borrowed from our knowing by experience what it is like, in itself, to see red and be in pain, to feel jubilant, and in general to be in one conscious state or another. Perhaps nothing could have the relevant kind of ‘inside’, or be anything ‘in itself ’, without having something like consciousness. To sum up the point in a slogan, perhaps nothing can be something in itself without being something for itself. (46–7)

Adams clearly thinks that the claim that ‘an idea can be like nothing but an idea’, when restated in terms of qualities, is intuitively powerful, but not demonstrable. Michael Dummett builds an argument for idealism on the assumption that it is obviously true that qualities are essentially mental, and uses this as a ground for thinking that our conception of the world beyond experience (the ‘transcendental world’ or TW) can only be purely formal. Dummett argues as follows. The conception of ‘the world as it is in itself ’ collapsed because, of our own resources, we can give no substance to the expression ‘like’ as it occurs in the question ‘What is the world like in itself?’ . . . But to express our goal in this search by means of a word such as ‘like’ that calls for an account of experience, asking in effect how we should experience the world if we experienced it as it really is, and not in any particular way, is unintelligible: the question needs to be replaced by ‘How is the world to be described as it is in itself?’ . . . That is why our search for a conception of the world as it is in itself ended with barren mathematical models of which it is senseless to think ‘This is what there

176

   

really is’, still less ‘That is all there really is’. We set out with a robust version of ‘scientific realism’ as our understanding of what science aims at: its task is, on this view, to uncover how things really are in themselves. We finished by lapsing into a purely instrumentalist interpretation . . . (98–9)

Dummett here first identifies qualitative nature (‘what it is “like” ’) with experienced quality, which is close to (7) below. Dummett then goes on to say that this leaves us with only the barest structural-cum-mathematical conception of the world of the sort attributed to science in the argument, and that this can be no better than an instrumentalist conception of reality. The unsatisfactory nature of this abstract instrumentalist conception of the world ‘in itself ’ forces Dummett to a Berkeleian theism. The world, in so far as we apprehend it and are capable of coming to apprehend it, is the world we inhabit; of what we are incapable of apprehending we cannot meaningfully speak. In asking after the character of reality, we are asking after that of the world we inhabit; to speak of a world transcending ours and, as it were, encasing it, is merely to employ a form of words devoid of any clear sense. (2006, 23) Since it makes no sense to speak of a world, or the world, independently of how it is apprehended, this one world must be the world as it is apprehended by some mind, yet not in any particular way, or from any one perspective rather than any other, but simply as it is: it constitutes the world as it is in itself. We saw that how God apprehends things as being must be how they are in themselves. But now we must say the converse: how things are in themselves consists in the way God apprehends them. . . . This does not imply that God understands what it is for the material world to exist independently of there being within it any sentient creature to perceive it . . . God’s knowledge of the material universe consists in the grasp of an immensely complex structure determining what will be observed by the various kinds of sentient creatures . . . and what will be discovered by the various rational creatures when they attempt to find out what things are in themselves . . . (101–3)

So we have here the following; (i) The notion of what the world is qualitatively like is equivalent to how the world would be experienced if it were experienced as it is. (ii) The notion of experiencing something ‘as it is’, as opposed to from a perspective (‘in a particular way’) is unintelligible. From this there follows:

 -  

177

(iii) The notion of what the world is qualitatively like is unintelligible. And, given that a mind-independent world needs a qualitative component to be concrete and not just formal or mathematical; (iv) The notion of mind-independent world is unintelligible. The problem with this argument is that (i) is more or less equivalent to the assumption that qualities are essentially mental, or, perhaps, more weakly, that they are necessarily experientable, because it makes the notion of quality experience-dependent. This is perhaps complicated by the use of the term ‘unintelligible’. It could either be equivalent to ‘makes no sense’ or ‘we cannot know its nature’. Dummett’s verificationist tendencies no doubt tend to equate these. But if the very idea of perceiving something as it is makes no sense, this latter point may not matter. But even if any grasp on the nature of quality depends on perceiving from a perspective, does that entail that one cannot form a conception of how it is in itself—some perspectives are very accurate—viewed directly from nearby at right-angles to a surface, for example. Dummett might want to argue that this is just a favoured perspective, not an approach to the idea of how it is ‘in itself ’, but that would be controversial. The form of the argument found in both Adams and Dummett might be put as follows. (1)

Science cannot uncover the intrinsic qualitative natures of physical reality.

(2)

Physical reality must have an intrinsic qualitative nature

therefore: (3) There must be an intrinsic qualitative nature to physical reality which falls outside the range of science. Then we can add the following: (4) Ordinary sense-experience cannot reveal the intrinsic qualitative nature of physical reality because those qualities that feature in experience are all sensedependent. (5) Science and ordinary experience are the only two possible sources by which we might come to know the intrinsic qualitative nature of physical reality. Therefore: (6) The intrinsic qualitative nature of physical reality must be such that we could never know it. Reflection on (4) and (6) leads these philosophers to conclude: (7) There are good reasons for thinking that qualities as such are sensedependent phenomena.

178

   

From which there follows: (8) A mind-independent physical reality cannot have an intrinsic qualitative nature. Given (2), we can conclude: (9) There cannot be a mind-independent physical reality: it must be essentially phenomenal or ideal. Given the framework within which we are arguing, the only contestable steps are (4) and (7). Once one has got that far, and conceded that everything we know as qualitative is essentially sense-dependent, then (7) becomes plausible—qualitativeness is essentially a feature of experience—but remains far from demonstrative. In Chapter 9 we examined in greater depth the arguments against a world of pure powers, and the difficulties in resolving the relations between intrinsic qualitative properties and dispositions or powers. Perhaps the same intuition can be expressed in a slightly different way, by seeing the realist view of sensible properties as a kind of category mistake, by adopting the following principle: no quality or property that can constitute the content of a subjective, mental state can univocally qualify something extra-mental.

This can seem plausible because we do not think that the ‘what it is like’ of pains, itches, feelings of love or thinking thoughts can be the sort of thing that it makes sense to ascribe to the extra-mental. But why should the ‘what it is like’ features of sense experience be any different, yet these are the sensible qualities. This would suggest that anything beyond the most attenuated form of realism—‘purely structural’—would rest on a category mistake.

10.3 The Physical World and the Nature of Thought 10.3.1 Berkeley’s Statement of the Problem and Some Initial Developments Arguments against physical realism based on the nature of thought are much more murky than those that follow an empiricist track, and, for historical reasons, are greeted with more suspicion by analytical and empiricist philosophers. One can trace more than one route to idealism via the nature of thought, but, for our purposes, it is best to start from Berkeley’s so-called master argument in Principles of Human Knowledge, section 23.

      

179

But say you, surely there is nothing easier than to imagine trees, for instance, in a park, or books existing in a closet, and no body by to perceive them. I answer, you may so, there is no difficulty in it: but what is all this, I beseech you, more than framing in your mind certain ideas which you call books and trees, and at the same time omitting to frame the idea of any one that may perceive them? But do not you yourself perceive or think of them all the while? This is therefore nothing to the purpose: it only shows that you have the power of imagining or forming ideas in your mind; but it does not shew that you can conceive it possible, the objects of your thought may exist without the mind: to make this out, it is necessary that you conceive them existing unconceived or unthought of, which is a manifest repugnancy.

There are many things that can be said about this passage. In sympathetic vein, Christopher Peacocke (1985) has taken it as making the essentially true point that we cannot imagine an object simpliciter, but only imagine it as it appears from a certain viewpoint. He does not, of course, claim that this is enough on its own to prove idealism. John Campbell (2002a) also thinks that Berkeley has identified a deep and serious problem, but one that forces one to direct realism, rather than to idealism. More critical responses, however, often begin by pointing out that Berkeley confuses perceiving something with conceiving or thinking about it: from the fact that you cannot think about an object without thinking about it, it does not follow that one cannot think about an object unperceived. The criticism of the argument from which I wish to start is connected to this one. It is that, in one way, Berkeley misses the intentionality of thought. He confuses the psychological content of the thought-episode with object that the thought is about. Because, platitudinously, one cannot think of an object without thinking of it, Berkeley concludes that what one thinks of must be internal to the act of thinking. No doubt it is because he follows Locke in having an imagist theory of thought, that he cannot properly grasp the idea that a mental episode essentially refers beyond itself. I raise this criticism, however, only to show its limitations. Although the intentionality of language shows how thought refers beyond itself to objects in the world, it does not so easily show how the kinds of facts that we impute to the world are separable from our ways of thought. I think that many of the most suggestive of the thought-based arguments for idealism can be seen as founded on a single idea. This is that our way of thinking about the world possesses certain logico-grammatical features which, on the one hand, could have no correlate in a mind-independent reality, but which, on the other hand, are ineliminable from the world as it is represented in our understanding. A good point to begin to appreciate this line of thought can be found in Strawson’s response to Austin on the correspondence theory of truth.

180

   

What ‘makes the statement’ that the cat has mange ‘true,’ is not the cat, but the condition of the cat, i.e., the fact that the cat has mange. The only plausible candidate for the position of what (in the world) makes the statement true is the fact it states; but the fact it states is not something in the world [my italics]. It is not an object; not even . . . a complex object consisting of one or more particular elements . . . and a universal element . . . I can (perhaps) hand you or draw a circle round, or time with a stop-watch the things or incidents that are referred to when a statement is made. Statements are about such objects; but they state facts. Mr Austin seems to ignore the complete difference of type between, e.g., ‘fact’ and ‘thing’ . . . Roughly: The thing, person, etc., referred to is the material correlate of the referring part of the statement: the quality or property the referent is said to ‘possess’ is the pseudomaterial correlate of its describing part; and the fact to which the statement ‘corresponds’ is the pseudomaterial correlate of the statement as a whole. (Pitcher, 1970: 37)

At first sight, it might seem that Strawson is simply being mislead by the double sense of ‘fact’, when it can mean either true proposition or state of affairs, the former being ‘on the side of language’ and the latter ‘on the side of the world’. Surely the condition of the cat—the property instances it possesses—must be just as real as the cat.¹ But even thinking in terms of states of affairs, there are many such states which seem fundamental to our thinking about the world and which do not seem open to a simple realist construal. There are many true and important kinds of facts about the world which seem to be a function not only of how the

¹ Someone who, I think, has defended the Strawsonian position is Thomas Hofweber in his (2016) and (2017). Unfortunately, I came across his work too late to give it the attention it deserves. I think his understanding of how objects alone can carry the full weight of ontology is expressed in the following quotation: Some things have to exist for certain propositions to be true. And the existence of some things guarantees the truth of some propositions. But all that does not mean that what is true is determined solely by what exists. What is true is determined instead by what exists and what it is like, but not by what exists alone. The same things can exist, but be different, either in their properties or in their relations among each other, and then different propositions will be true. What there is alone doesn’t determine what is true, but only what there is and how it is.¹⁷ Truthmaking so understood is quite unproblematic, but it doesn’t give one any argument against the picture of the world as one of objects and only objects, since no one is denying that these objects are a certain way, and thus have properties and relations among each other. Internalism makes clear how that can be while all there is are objects. (Hofweber: 2016, 288; italics added) Hofweber’s Strawsonian picture here is that objects and the way that they are is in the external world, but that the reification of the way that they are as properties, is the creation of language—something internal to the way we speak or think. This must represent Strawson’s view, for Strawson cannot believe that objects as ‘bare particulars’ are all that is out there. My problem with this, is that I cannot separate the notion of the way objects are from the properties they possess. The way things are must be thus and so, and we do not need linguistic predicates for being thus and so to constitute a property. Hofweber, I think, is close the Strawsonian synthesis of Kantian idealism, neutralized by a more or less ordinary language claim that ‘one cannot break out of language’. In this section I am tentatively defending a neoPlatonist view that reality depends on an objective Nous that is presupposed by the existence of a physical world.

      

181

world coldly and simply in its own right is, but to be a product of how the world is and of a judgement or assessment of how the world is. A simple illustration of this is as follows. The easiest kind of fact to take realistically is one that concerns the location of a particular object in space and time. Thus the thought or proposition ‘there is a pen on the table’, if true, simply reflects a situation in the world. It is made directly true by what there is in concrete reality. But negative propositions do not ‘picture reality’ in the straightforward way that some positive assertions do. The proposition ‘there is no pen on the table’ is not so simple as its positive counterpart. Absences are not part of concrete reality. A fact of an absence is more like an assessment of a situation with a certain thought or perspective in mind—in this case, a concern about a pen or pens. It is grounded in the concrete situation— what the situation positively is on the table’s surface—but it is not simply a representation of that alone. If one wishes to include negative facts in one’s ontology, a simple physical realism does not seem to be available. Even the category of states of affairs is more problematic than it may seem. David Armstrong believes that the world includes states of affairs and that everything in the world is spatio-temporal: hence that states of affairs are spatio-temporal. Indeed, anyone who wishes to include states of affairs within his physical ontology, realistically conceived, would seem to have to deem them spatio-temporal. But it is not clear that they have spatial location. Perhaps it is easy enough to assign location to the states of affairs of macroscopic objects possessing the traditional primary and secondary qualities. Thus a ball’s being round will be where the circumference of the ball is, and its being red, where its surface is. With more arcane properties, the issue is harder. A particular electron has location, but where is its having spin or its having mass? Are they everywhere the electron is? Relations also create further problems. One could draw a rough outline round the state of affairs of a’s being six inches away from b: it would include a and b, but what of the intervening space would it include? In Timothy Sprigge’s words: The point is that there is no distinguishable portion or piece of reality which is where the relation is exemplified as there is in the case of a property . . . There is not, so to speak, some sub-division of the totality of particular reality which actualises the relation to the exclusion of its contraries as in the case of properties. (1983: 164)

10.3.2 A General Statement of the Problem So Far Developed This pattern of argument can be captured in the following general structure: (1) Our conception of the world, as shown in our language about it, possesses logico-grammatical complexity, and certain features of this complexity cannot be treated straightforwardly as representations of aspects of

182

   

concrete, mind-independent reality, as has been shown by the examples cited in Section 10.3.1. Therefore: (2) If there is a realist conception of the world, it must be possible to separate those features of our conception of the world which can be treated in a straightforwardly realist way and those that cannot. It is the former that constitute the realist conception of the world, and the states of affairs they represent must be enough to constitute the realist’s world. (3) It is not possible to make the distinction specified in (2). Therefore: (4) A realist conception of the world is not possible. Accepting that there is a prima-facie case for (1), there are four ways of responding to the argument: (i) There are two ways in which one might deny (2). One of them is to deny that a realist is restricted to having only concrete elements in his reality. Perhaps a physicalist or a naturalist of a certain kind is committed to this, but a realist per se is not bound to deny that there is a realm of more or less abstract facts, grounded in but going beyond the concrete core of his reality. (ii) A second kind of objector to (2) regards the argument as wholly misconceived. This is a response that might come from someone with sympathy for the ‘ordinary language’ school of philosophy. Of course, this respondent will say, when one describes the world, the descriptions possess the logico-grammatical features of the language used. This is a trivial consequence of using language to describe the world, which is something one cannot avoid. It does not carry strange consequences for one’s conception of the world itself. (iii) The third response is to deny (3) and to take up the challenge to delineate clearly between those features of our conception of the world which are to be taken wholly realistically, and those which are not, and to explain how the former sustain the practices which give rise to the latter. This, I take it, is the point of ‘truthmaker theory’, which tries to explain how statements are made true by a world conceived as very much sparser than the way a literal reading of our statements would make it seem.² I share the truthmaker theorist’s intuition that his theory represents what the realist is actually committed to, but it is important to consider the other two responses. My reservations about the ‘abstract facts’ strategy are easily stated. The suggestion that a physical realist can accept abstract objects fails to distinguish between

² For a thorough working out of truthmaker theory, see Armstrong (2004).

      

183

different kinds of abstract objects. The realist can, at least if he is not a rigorous physicalist, accept universals, numbers, and propositions, but can he accept that the facts that make certain propositions about the physical world true are also abstract? It is one thing to allow the proposition that there is no cat on the mat to be abstract, another to treat the fact that makes it true as abstract. One possible response to this is to argue that the proposition ‘there is no cat on the mat’ exists when there are no minds, and its truthmaker is the actual positive state of affairs at the mat. At the very least, I think that this picture fits best with what I call below the distinction between descriptive facts and factual assessments, where the latter look to be what a mind could make of the situation rather than a simple reflection of what there is the world. I discuss this idea further below. The claim that all the phenomena that have been cited merely reflect the trivial fact that we talk about the world using language, and therefore do so in a way that reflects the properties of language, is difficult to make exact. It could be argued that the very objection brought against Berkeley’s argument works against this strategy. I said that Berkeley fails to give weight to the intentionality of thought and language. But if one does give this weight, then one has the resources to distinguish between thought and language on the one side, and what they are about on the other. Once having been allowed this distinction, it is difficult to see what could prevent one from inquiring further into the relation between the logico-grammatical features of language and the question of what there is ‘out there’ to be its object. If one denies the possibility of separating these two, it looks as if one is in danger of surrendering to my developed version of Berkeley’s argument as it stands. I suppose the natural response is to say that we can tell which features of our language reflect features of the world, but we cannot contrive a language to characterize the world that possesses only those features—and, of course syncategorematic features that are descriptively neutral. It is not easy to evaluate this argument until one has isolated those features important to the way we talk that cannot be realistically taken. The best way to investigate this is to consider the third option, namely to see how far truthmaker theory can go in capturing our articulation of reality.

10.3.3 Truthmaker Theory According to the truthmaker theorist, for every statement with a truth-value, there is something about fundamental reality that necessitates that that statement should have the truth-value that it does. For a true statement, this is its truthmaker. It is important that this is something about the fundamental reality. One cannot generally articulate the truthmaker of a sentence simply by repeating it, as one can when stating its truth conditions. So the truthmaker for a statement such

184

   

as ‘there is a hurricane approaching Florida’ will be some complex state of affairs involving elementary particles moving in a particularly rapid way over other particles that constitute a section of the Atlantic. The statement about the hurricane will have the same truthmaker as a statement describing the same event in greater detail: they will both be made true by the same hunk of reality. This still leaves open a choice between two conceptions of how this theory fits with realism. They can be stated as follows. (A) If realism is true, then if, for example, there is no cat on the mat, then it is true that there is no cat on the mat even if there are no minds. Therefore: If realism is true, then there must be both truthmakers and truthbearers for all true statements. The alternative is: (B) If realism is true, then if, for example, there is no cat on the mat, then there must be that in the world, even if there are no minds in it, which would make it true to say that there is no cat on the mat. The difference is that the latter does not commit one to the existence of truthbearers in the absence of minds. The point is the following. We are inclined to say that, from the realist perspective, the facts are what they are whether or not there are any minds to appreciate them. From the perspective of the second theory, there is no fact that there is no cat on the mat in a world without minds. In other words, when fact and truthmaker are different, the fact will not exist unless minds do. Does this matter? I want to distinguish between a pure description and a factual assessment. A pure description is a characterization of the world the content of which simply captures the nature of its truthmaker. In other words, the truthmaker can be identified with the state of affairs that is asserted to obtain in a pure description. A factual assessment is a characterization of the world whose content does not simply reflect the nature of its truthmaker. It seems to me that at least certain kinds of factual assessments are true and essential to our notion of a real world. If this is so, such a world could not exist without minds. A possible example of this is the following. Suppose a certain random event (for example, the decaying of a particle) occurred at t, but that it could have occurred at t’, and that had it occurred at t’ certain very different things stretching into the future would, or could, have happened. If you are a realist about possibilities, you will want to say that it is a fact that these things would, or could, have happened, whether or not there are any minds. The truthmakers for such statements will be certain immediate dispositions or powers of actual objects. The extrapolation to more remote future outcomes cannot be thought of as existing in these actual states of affairs. This is shown to be true by the case in which, had the event occurred at t’ not t, the future that ensued could itself have gone either of two ways depending on some further

      

185

random event. Let us express what actually occurred by p. What might have occurred is q. If q had occurred, then it might have been followed either by r or by s. The ‘might’ here is not epistemological, but is based on genuine indeterminacy. So if p had not been true, but q had, then either q followed by r could have been the case, or q followed by s. The truthmakers for either of those possibilities in the actual world would be the same. This is so because if there is a circumstance C in which there are a set of outcomes that are indeterministically possible, it is the same set of facts that underlie the possibility or probability of any of those outcomes, because they determine the range of probabilities as a whole. Nevertheless, it could be argued that the realist and common-sense intuition is that it is objectively true that there are the two possible outcomes, whether or not there are any ordinary minds to appreciate this fact. If this is the case then the existence of the various possibilities outruns what is present in concrete reality, though it concerns physical events. It seems to be the kind of thing that exists only in the light of a theoretical understanding of how the world works. Such a theoretical understanding is a form of factual assessment, and there is no assessment without minds. What was suggested above about the problematic nature of relations from a simple realist point of view also supports this conclusion. It is interesting that the distinction I introduced between pure description and factual assessment parallels the premodern treatment of relations. Most philosophers from Aristotle to Russell claimed that there was a subjective element in relations, because there is always comparative judgement, which presupposes mind, in the attribution of relations. (See Weinberg, 1965.) The normal verdict on subjective theories of relations is that they are rooted in the ontology of Aristotle’s metaphysics, as expressed in Categories. Aristotle divides the world into substances and their accidents, and accidents cannot be shared between substances. Relations, therefore, must have their truthmakers the monadic properties of objects, which are compared in judgements. But once one understands the treatment of relations in Russell’s predicate logic, one can see that the traditional view rests on a primitive logic. There is a slippage of argument here, however. It is true that Aristotle’s logic, being of subject-predicate form, reflects the ontology of Categories, but revision of logic does not necessarily remove all the philosophical reasons behind the ontology. Sprigge’s remarks about relations specifically, and the general considerations of this section suggest that the simple invention of the logical form aRb does not remove all grounds for thinking that certain fundamental states of affairs involve a comparative judgement about reality. The real foundation for this intuition may not have been a certain understanding of the logical form of language, but reflection of the ontology of states of affairs and facts. Neither the argument from possibilities nor from the nature of relations are conclusive against the kind of ‘hard minded’ and minimalist empiricist who is prepared to say, as A. J. Ayer was fond of saying, that the world is ‘just one damn

186

   

thing after another’. Such a person would concede that possibilities, and modalities in general are, in Blackburn’s terms, only ‘quasi real’, being projections of our mode of thinking. Similarly, in the case of relations, an austere realist might argue that, taking spatial relations as an example, all the reality need contain, concerning objects A and B, is that A is at a particular place-time, and B is at another and that these locations are all that are needed. This fact grounds the fact that there is a certain distance relation between them, but the relational fact can be thought of as comparative and, hence, a mind-dependent factual assessment. On the other hand, someone might reply that, though the relational fact is grounded on the fact of location, it is also part of fundamental reality, in that it must be as completely ‘out there’ as the locations, which it cannot be if assessment is an essential part of its constitution. I think we can say with some confidence that if one wants a reasonably rich and common-sensical form of realism, one will need some of what I have called factual assessments, and, hence the presence of minds. There is also the question of whether the minimalist conception can accommodate the kind of mind which has the kind of imagination and inventiveness to create the quasi-real world of modalities, negations, and relations. If these arguments have any force at all—and I think that this is a much more open question than it is about the other arguments presented in Part II of this book—then they show that we have no notion of a world as it is in itself that is wholly disembrangled from modes of conception that are definitely minddependent. But what sort of idealism does this give us? The natural upshot might appear to be that the world is dependent on our way of thinking about it, as Kant claimed. Out of this idea, by dint of various obscure moves, came the idealism of the Hegelian tradition.³ But there is an alternative interpretation of the arguments presented in this section. This is that the existence of a world depends on its relation to a mind, but not ours. This interpretation is Platonic or, more precisely, neo-Platonic, rather than Kantian. According to this Platonistic conception, there could not be an objective world if there were not also an objective Intellect understanding it. Our intellect merely participates in this Divine Intellect. How one might set about choosing between the neo-platonic and Kantian–Hegelian options is an issue we have no space here to discuss.

³ I have not been able directly to discuss the Kant-Hegel tradition in the way that might be expected of this book, but the arguments in this section are an analytic philosopher’s attempt to cast within his own terms what he can understand from that tradition. Ewing (1934) is a great, but largely forgotten work, by a philosopher with all the analytical skills and real sympathy with Hegelian idealism, which examines the arguments that have been brought for that tradition. Sprigge (1983, 1993) are also works wholly accessible to analytic philosophers, but which make a strong defence of British Hegelianism.

11 Bishop Berkeley and John Foster on Problems with Physical Realism about Space 11.1 Introduction: Two Berkelian Arguments about the Nature of Space I want to consider two arguments about the nature of space that occur in Berkeley which I think are not sufficiently discussed.¹ The first concerns the phenomenology of space, the second its physics. The first is the ‘mite’ argument and the second concerns Newton’s two thought experiments about absolute space, the ‘bucket’ thought experiment and the ‘balls’ thought experiment. I hope that they will strengthen support for an idealist approach to space.

11.2 Mites, Men, and Objective Space Berkeley states his argument as follows: PHILONOUS. A mite must, therefore, be supposed to see his own foot, and things equal or even less than it, as bodies of some considerable dimension; though at the same time they appear to you scarcely discernible, or at best as so many visible points. HYLAS.

I cannot deny it.

PHILONOUS. HYLAS.

And to creatures less than the mite they will seem yet larger.

They will.

PHILONOUS. Insomuch that what you can hardly discern, will to another extremely minute animal appear as some huge mountain. HYLAS.

All this I grant.

¹ I allude briefly to both these problems in the World’s Classics edition of the Principles and the Three Dialogues. The ‘mites’ argument I discuss on pp. xxvi–vii, and Newton’s thought experiments on pp. 221–2. I am very grateful to my Liverpool colleague, Barry Dainton, for help concerning the latter.

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0012

188

      

PHILONOUS. Can one and the same thing be at the same time in itself of different dimensions? HYLAS.

That were absurd to imagine.

PHILONOUS. But from what you have laid down it follows, that both the extension by you perceived, and that perceived by the mite itself, as likewise all those perceived by lesser animals, are each of them the true extension of the mite’s foot, that is to say, by your own principles you are led into an absurdity. Unfortunately, taken in its context, the force of the argument is ambiguous. At this point in the First Dialogue Philonous is arguing that the sensible qualities that we directly perceive exist in our minds, not in the external world. This hypothesis is ambiguous between the claim that the instances that we directly perceive of those sensible qualities are in our minds and those sensible qualities per se exist only in our minds. Because he confuses these claims, Philonous conflates arguments which refute naive or direct realism—like the argument from illusion—with arguments that refute realism in general, that is, including representative realism. This is a confusion which Hylas notices, and draws to Philonous’s attention, only later in the dialogue. The argument can be interpreted, therefore, as no more than a version of the argument from illusion and as establishing, if successful, no more than that the instances of spatiality of which we are directly aware exist only in our minds. But I think the argument can also be deployed for the more ambitious purpose of showing that the very idea of space itself cannot be separated from experience. In the case of this particular argument, that is, Berkeley’s failure to distinguish between representational and naive realism does not prevent the argument from having the full force intended. On this interpretation, the argument is meant to show the impossibility of mind-independent space by arguing that determinacy of spatial magnitude is mind-dependent. If one adds the premise that space, like anything else, must be determinate to be real, it would follow that space could not be real without the mind. The argument, therefore, on this interpretation, is that because things look bigger to a mite than to a human and there is no clear sense in which one of them perceives the ‘right’ size, then there is no fact about the size and, therefore, no such thing as real or objective size, and so no such thing as mind-independent spatial dimensions. The contrast with the normal ‘argument from illusion’ is that, in other cases of that argument, there are standard procedures or conventions which stipulate which appearance is, or is nearest to, the accurate representation of the object. So we know how to determine which appearances of the shape of an object should be deemed to present its real shape: it is the ‘head on’ view that presents it the shape that it is, and other apparent shapes are deemed to be ‘views from an angle’. Similarly, we can work out what are to be taken as the ‘standard conditions’ that show an object in its true colour, and what are the conditions in which we would say that it, for example, is

, ,   

189

white but looks yellow. The force of the ‘mite’ argument by contrast, is that, having fixed, by these sorts of means, what the ‘real shape’ of an object is, there are not, and could not be, any means of fixing what the ‘real size’ is. And that ‘real size’ is as essential as ‘real shape’ to the objective existence of spatial properties. The first objection to this argument that springs naturally to mind is that we can use measurement to pick out real size and that, irrespective of how it looks to mite or man, an object will have a specific measurable size; for example, it will be two inches long. But this will achieve nothing unless we can give objectivity to units of measurement, and an inch is a long distance to a mite and a small one to a human, so in agreeing that an object is two inches long the mite and the man have only an illusion of agreement. There is agreement on the relative sizes of the object and an inch on a ruler, but no agreement about the absolute size of either. It seems that subjects of radically different sizes can agree on the relative sizes of objects, but not on their intrinsic dimensions. One natural response to this is to argue that size is a relative matter and agreeing on the relative dimensions of things is agreeing on their actual dimensions. Man and mite agree about the relative proportions of all things and that is all that is involved in agreeing about their sizes. The matter cannot be solved quite so simply, however. It can be proved that a purely formal definition of space, in terms, say, of Cartesian co-ordinates and axes, is not a complete representation of the nature of space.² As well as such formal features, there must be something intuitive and qualitative, like the way that extension presents itself experientially in vision. But it is in respect of such a qualitative feature that things look many times larger to the mite: relative sizes are preserved, and the same geometry applies, but the qualitative interpretations of dimensions differ in intrinsic magnitude. If such an interpretation is essential to any real space, as I have claimed it is, then the problem remains. The following worry might arise at this point. If there can be two phenomenal spaces that differ only in their intrinsic magnitudes, and if the kind of qualitative element that makes this possible is essential also to any real space, then should it not be possible for there to be two physical worlds that differ only in their intrinsic magnitudes: more exactly, should not the suggestion that everything in the world might double in size, by dint of the doubling of the intrinsic qualitative element, without any alteration in the geometry, also make sense? But this suggestion is generally rejected as vacuous. One might reject the accusation of vacuousness on the grounds that it is verificationist, but it is interesting to note that an idealist is in a better position to accommodate the intuition that everything could not double in size than is the realist. For the idealist, the physical world in which things cannot

² John Foster shows this in The Case for Idealism (1982), 73–88.

190

      

all double in size—as they can when one moves from a human world to a mite’s— is an abstraction created by measurement from the various worlds of experience, which are the ultimate constituents of reality. The qualitative element of space falls out as irrelevant in such an abstract projection of the formal features of experience. The argument can be summarized more formally as follows. 1. There is more to our conception of physical space than can be captured purely formally. 2. The only thing this non-formal element could be is something qualitative concerning the intrinsic nature of extension. Therefore: 3. There is a qualitative element concerning the intrinsic nature of extension in our conception of physical space. 4. The only source for acquiring anything of this sort is sense-experience, by which we get an intuitive grasp on certain qualities. Therefore: 5. There is an intuitive grasp on the intrinsic nature of extension, and, hence, of space, which is gained through sense-experience. 6. For any extension to be real it must have a determinate metric—that is, things in it must have determinate sizes. Therefore: 7. If the quality of extension grasped intuitively through sense experience is identical with the intrinsic qualitative nature of real physical space, objects possessing such extension must have determinate sizes. 8. Berkeley’s example of the mite shows that the quality of size grasped intuitively through sense-experience is observer-relative. 9. There could be no principled reason for thinking that the human or the mite had grasped the real size—the real degree of extension—of the physical objects in question. Therefore: 10. Berkeley’s example of the mite shows that there is no determinate size, and, therefore, no determinate degree of extension, attaching to the intuitive component in extension, considered as a property of physical bodies. Therefore: 11.

The intuitive quality of extension is not physically real.

Therefore: 12.

Physical space (which needs intuitive extension) is not real.

, ,   

191

This argument could perhaps be rejected by detecting an ambiguity in the notion of ‘our conception of physical space’, which occurs in line 3. This expression might have a more and a less psychological sense. Because space cannot be purely formal, when we imagine it there must be a non-formal, qualitative element in how we imagine it to be. Perhaps this imagined feature, however, bears no qualitative resemblance to the quality space possesses intrinsically. For some empiricists it was almost a common-place that the way space actually is, is neither like the way it looks, not the way it (that is, in this case, the extension of objects) feels. The argument would then only prove that no space qualitatively like our visual space— or, probably, by parity of argument, any other sensory space—can be physically real. So though step 11 is correct in its assertion that the quality of extension as we intuit it is not physically real, we cannot conclude that no space with a qualitative element could be physically real; we just cannot know what the qualitative element is. This reply is not convincing, however. Although the intrinsic quality of space may not be the same quality as that presented in any of our senses, it must be something directly analogous to such qualities, for they are the only model we have from which to understand monadic qualities. Any such quality must be the kind of thing that could, in principle, be the object of some possible sense or other. I can imagine no reason why it should not, therefore, be possible that something similar to what applies to the perception of mite and man should apply to it. This reading of the ‘mite’ argument enables one to connect it to two of Berkeley’s more discussed arguments. Berkeley argues that ‘an idea can be like nothing but an idea’. This can be interpreted as saying that monadic qualities are essentially mental, because they are essentially features of sensations. We have seen in the previous chapter that this is a plausible argument, and, if it is, then the attempt to give qualitative content to a mind-independent space must fail.³ The other argument with which it connects is Berkeley’s claim that we cannot conceive of primary qualities without secondary qualities. As the ‘new science’ claimed that this was how the world was really constructed, this was intended to undermine scientific realism. It is natural to read this argument as depending on Berkeley’s attack on abstract ideas. It then moves in two steps: (i) it is impossible to conceive that which one cannot image; (ii) it is impossible to image primary without secondary qualities. If one rejects the attack on abstract ideas, one will abandon (i). It is possible, however, to reconstruct the argument in a different way. Most, if not all, primary qualities have a spatial element. If space, for reasons that are nothing to do with abstract ideas, cannot be conceived without a qualitative element, and if such elements are essentially experiential, then, given that nothing experiential lacks some secondary quality, then no primary quality will be ³ I also argue this in the introduction to the World’s Classics edition of the Principles and Three Dialogues, pp. xiv–xvi.

192

      

conceivable as really existing without some secondary. I do not want to suggest that this is how Berkeley intended his claim that primary qualities cannot be conceived without primary be defended—he was, indeed, relying on the attack on abstract ideas—but I think it is useful to see how closely the ‘mite’ argument, the doctrine that nothing can be like an idea but an idea, and the interdependence of primary and secondary qualities, are connected. The argument that I would like to have made more plausible is, therefore, the following. Any real space requires a qualitative filling, and qualities of the appropriate kind have absolute extent, independent of relational features. But once abstracted from the phenomenal context, the attribution of absolute extent becomes arbitrary in a way that shows that such qualitative notions do not really belong outside the experiential. This shows that our notion of physical space is a hybrid, of the abstract-cum-relational and the qualitative, of a sort that is operationally sound, but which cannot be taken realistically.

11.3 Newton’s Thought Experiments and Absolute Space The ‘bucket experiment’ is one of two thought experiments that Newton employed to show that space was absolute not relative. Imagine a bucket containing water and suspended by a rope. Now imagine the bucket spinning—say turning it until the rope becomes tense and letting it go. At first the bucket will turn leaving the water stationary, but, as a result of friction, the water will start to turn with the bucket, until it is fully turning with the bucket and so stationary with respect to it. Centrifugal force will make the water higher at the edge, with a trough in the middle. Now suppose that the bucket is the only thing in the physical universe. It will not be spinning with respect to anything, but, as the state of the water shows, it is spinning. Hence it is moving through absolute, not relative, space. In the other thought experiment we imagine two metal spheres joined by a taught rope. If the spheres begin to spin round each other tension increases in the rope. If they are alone in the physical universe, they are not moving with respect to anything, yet the increased tension shows they are moving, hence they, too, are moving in absolute space. As far as I am aware, no-one has answered these arguments satisfactorily. Berkeley seems to miss the point. He summarizes his view of motion as follows. For to denominate a body moved, it is requisite, first, that it changes its distance or situation with regard to some other body: and secondly, that the force or action occasioning the change be applied to it. If either of these be wanting, I do not think that agreeably to the sense of mankind, or the propriety of language, a body can be said to be in motion. (Pr. 115)

 ’  -

193

On this basis he has concluded of Newton’s bucket: As to what is said of the centrifugal force, [namely] that it doth not at all belong to circular relative motion: I do not see how this follows from the experiment which is brought to prove it . . . For the water in the vessel, at that time at which it is said to have the greatest relative circular motion, hath, I think, no motion at all: . . . (Pr. 114)

Berkeley’s version of relativism is interesting in itself. He adds the second condition, that the moving object is the subject of a force, so that he can say that when he walks down the street, though he and the pavement are in relative motion, he is moving and the pavement is stationary (Pr. 113). He concludes that it follows that the water in the bucket is initially stationary, so we do not have here a case of relative motion. This is doubly mistaken. First, it is a case of relative motion if the one that is moving is the bucket, not the water. Second, and more important, the interesting case is where the water has ‘caught up’ with the bucket and is moving with it in a centrifugal way. Supposing the bucket and the water to be the only things in the universe, they are moving relative to nothing, but, as the efforts of the water to climb the sides of the bucket show, are in motion. The power of this example is that, at the initial stages, the water and the bucket are moving relative to each other. The relativist cannot, therefore, use the simple blocking move of saying that a lone bucket cannot spin, because there is no distinction between its orientation and the orientation of space. It originally spins in a way that is acceptable to the relativist. The water in the bucket continues to move relatively to the bucket, though increasingly slowly, until the process is complete and it is moving with the bucket, pressed up against its side. Then there is no relative motion, but there are the physical effects of motion. Mach showed a better grasp of the issue, but did not provide a convincing response. He claimed that we do not know that these phenomena would occur in an otherwise empty universe, as we cannot test such a claim. So, if there was nothing else in the universe, then perhaps the water would never rise up the sides of a lone bucket and we would never have grounds for saying that it was spinning. This strongly verificationist approach is not convincing because the centrifugal force is not deemed to be caused by the gravitational pull of other bodies, so their absence should not lead to the water’s remaining flat, according to the principles of physics. (The same consideration applies to the tension in the rope in the other example.) Berkeley concludes his discussion by saying: From what hath been said it follows that the philosophic consideration of motion doth not imply the being of an absolute space, distinct from that which is perceived by sense, and related to bodies: which that it cannot exist without the mind, is clear upon the same principles, that demonstrates the like of all other objects of sense. (Pr. 116)

194

      

Berkeley here seems to me to confuse two different things. He confuses denying that there is any absolute space outside the mind with denying that a single body could move in a way that was not relative to any other body. He conflates these two because he accepts Newton’s claim that if a body moved without there being any other body for it to be moving in relation to, then it must be moving in an absolute space. This is, indeed, a correct inference if one takes physical space as existing independently of mind, for then motion which is not relative to another body must be, if it exists at all, absolute, and if it occurs in a mind independent space, that space must possess absolute positions. But Berkeley does not accept such a mind-independent space, and he has the option of allowing that a lone object could rotate with respect to the viewpoint of a mind. In fact, Berkeley can turn Newton’s thought experiments to his own advantage by using them to show that there is no adequate account of mind-independent physical space. The experiments show that the relational view is defective; as I argued against Mach, physical theory seems to require that the lone bucket should spin and the balls rotate. Provided that there is something wrong with the absolutist notion, then it would seem that there is no adequate conception of mind-independent space, either relationist or absolute.⁴ The point which should have appealed to Berkeley is that the notion of the rotation of a single objects round its own axis makes clear sense in a visual or phenomenal field. So, without absolute space, only by reference to what it would be experientially like can the phenomena Newton cites make sense. If absolute space is a defective conception then space must be a construct from experience. This seems appropriate, for, if one asks oneself why, intuitively, it seems to make sense that a lone object in a universe might rotate, the answer is that one could imagine what it would look like. Similarly, one might argue that purely phenomenal visual space allows rotation of unique objects—a lone afterimage could rotate—and our intuitive notion of physical space is based mainly on our grasp of visual space. Notice that these things do not seem to be rotating with respect to one’s body but simply with respect to one’s viewpoint. This fact gives, I think, Berkeley the materials for replying that Newton’s thought experiments show that his ‘absolute motion’ requires not an absolute, mind-independent space, but a space possessing features essentially derived from its experiential roots, but it does so only if the conception of absolute space is independently defective. But is the notion of absolute, mind-independent space, demonstrably defective? Berkeley says ‘that [space] cannot exist without the mind, is clear upon the same principles, that demonstrate the like of all other objects of sense’ (Pr. 116). But appeal to these other considerations would undermine the point of the present argument, which is to extract an independent argument for an idealist approach to ⁴ Notice that the relativist cannot explain the motion of the lone objects by reference to a relationship to other possible objects, for to say that it would be moving relative to something at p presupposes an independent identity of that point, and so falls back into absolutism.

 ’  -

195

space. The traditional objections rest on the fact that absolutism means that the universe could have happened in any of an infinite number of different places without this making any empirical difference. Leibnitz thought this objectionable because it meant that God lacked a sufficient reason to put it in one place rather than another. Verificationists think that the absence of empirical difference means that the hypothesis must be vacuous. It may well be, however, that one subscribes neither to the principle of sufficient reason, nor to verificationism. What, then, is wrong with absolute space? Now one may be tempted to argue that though verificationism is not true across the board, there is something fishy in the idea that the very same universe could have happened in a different location: such an idea fails to do justice to the way in which physical space is bound up with the physical world, as part of the same system. This could be true of space, even if space were not simply the relation between its occupants. The argument which I have been suggesting so far runs like this: 1. If physical space is mind-independently real, then it is either absolute, as Newton thought, or it is relational, as Leibniz and Berkeley thought. 2. Newton’s thought experiments show that it is not relational, and the considerations just mentioned strongly suggest that it is not absolute. Therefore: 3.

Space is not mind-independently real.

4. A conception of space that ties it essentially to its perceivability from our viewpoint would accommodate the correct intuitions of both parties: that is, it would accommodate the rotatability of lone objects, whilst tying the identity of spatial points to their role in accommodation the empirical world. Therefore: 5. An idealist approach to space seems to be the best way of preserving our intuitions. This, I think, is an appealing argument provided that the first premise is correct in its claim that absolutism and relationism, as conceived in the debate over Newton’s bucket, represent the only realist options. Unfortunately for my idealist strategy, it is possible that Einsteinian space constitutes another option. The space-time of relativity endows space with physical properties, unlike the pure receptacle of Newton’s theory, but does not seem to collapse space into relational properties of its occupants.⁵ It is, indeed, comparable with the ether. In Einstein’s own words:

⁵ See the fascinating article ‘Absolute versus Relational Space-Time: an Outmoded Debate?’ by Robert Rynasiewicz, Journal of Philosophy, 1996, 279–306.

196

      

physical space and ether are only different terms for the same thing; fields are physical states of space.⁶

Thus space is non-reducible, but physically influenced by its occupants. . . . Einstein restricts the label ‘absolute’ to a specifically ‘causal’ sense, in that this ‘space’ conditions the behaviour of ponderable bodies without in turn being influenced by them. It is in this sense that Einstein emphasizes the nonabsolute character of the aether of general relativity . . .⁷

It seems that as well as the two options that I originally considered, there is also the possibility of conceiving of space as a medium for objects, itself endowed with physical properties: it is rather like the ether, except that, obviously, no further question of its location in space can be raised (as, in general, it was not actually raised in the case of the ether). Space is, thereby, essentially tied to the objects that occupy it, and there is no question of the universe as a whole being located in a different place. I seem to face a dilemma in my attempt to use Newton’s thought experiments to support idealism. If Einstein’s space can accept Newton’s conclusions, then there is a conception of physical space, other than the rejected form of absolutism, which can accommodate the phenomena. If, on the other hand, Einstein cannot accommodate Newton’s bucket, then I, in defending Newton, appear to be saying that Einstein’s theory of space is false, and this seems rather reckless. What follows is very tentative. How does Einstein’s theory stand on Newton’s argument? This, in turn, depends on how the special and the general theories stand with regard to the ‘bucket’ thought experiment. I think I am correct in saying that there is no agreement on this. Whether the special theory of relativity allows motion round an axis for an object in a single object universe must depend on whether such a universe allows the existence of an inertial frame by reference to which such motion could be deemed to occur. As far as I can see, it could not, for, without ascribing independent identity to points, there are insufficient references by which to set up all the co-ordinates required. A bucket is asymmetrical from top to bottom, and this establishes an axis, but nothing other than points distinguish the different parts on the symmetrical circumferences. (It might be thought that one could pick out individual atoms on the rim of the bucket, but this will not do, for they move as the bucket rotates—if it does—and so do not establish an

⁶ Quoted by Rynasiewicz (1996), 288. It is originally from Einstein’s ‘The Problem of Space, Ether, and the Field in Physics’, in his The World As I See It (New York: Covici-Friede, 1934), 274. ⁷ Rynasiewicz (1996), 297–8.

 ’  -

197

independent axis.) In comparing, in a concessionary spirit, relativistic space with the ether, Einstein said: As to the mechanical nature of the Lorentzian ether, it may be said of it, in a somewhat playful spirit, that immobility is the only mechanical property of which it has not been deprived by H. A. Lorentz. It may be added that the whole change in the conception of the ether which the special theory of relativity brought about, consisted in taking away from the ether its last mechanical quality, namely its immobility.⁸

In other words, Lorentz’s ether is rather like Newton’s space, in being a fixed frame of reference, and nothing else; and Einstein’s space is like that ether, except that, apart from the relations of objects in it, it determines no location, and so has lost the properties of a Newtonian space. The fact that, at the beginning of the rotation, the water moves with respect to the bucket might be taken to suggest that relativity can accommodate the later stages by allowing for the temporal dimension of space-time. This might suggest that the General Theory fairs better than the Special in accommodating the buckets, for the General Theory is a theory of space-time and not just of space. On the other hand, this might seem rather contrived, for, as the water catches up with the bucket, what happens is that the two inertial frames collapse into each other, and the original difference becomes irrelevant. Perhaps one could argue that Einsteinian space is devised for a world with a plurality of objects: it is, after all, about their relativities. If there were only one object, or, if you deem the water and the bucket to be at least two, if there were only objects operating insufficiently independently, space could not exhibit Einsteinian characteristics. But what would such a space be like; in particular, would the bucket spin and the water climb the sides? If space were physically real, then this question should have an answer, which it does not seem to have. It should have an answer because there would seem to be no reason why God, or bizarrely improbable accident, should not have created such a world. But if physical space were ideal, that is, existed as a form of organization for human (and perhaps other) experience, and if this thought experiment abstracts from any context which could be empirically genuine, then there need be no fact about the matter. We can imagine, say, visually, a world with a single rotating bucket: but it is not a physical world in which we could ever be physically located. It is not an option for which a theory of space designed to cope with a world which could ever be our physical environment need cope. I argued above that an idealist account of space was preferable to either an absolutist approach or a realist relationist approach. I then raised the question of

⁸ From Einstein (1922), 10–11: quoted by Rynasiewicz (1996), 296.

198

      

whether Einstein’s relativist account provided a third realist option. My conclusion is that it does not because it cannot say anything plausible about a ‘single bucket’ world, which, if realism were true, ought to be a possible world. A Berkeleian theory, on the other hand, can cope with the ‘single bucket’ world, because this is a world easily conceived experientially, but it can also allow the sophisticated worlds of the special and general theories of relativity, because they are empirically founded. ‘Physical’, as opposed to ‘naïve’ space is something which, from a realist point of view, has gaps which it ought not to have—there ought to be a fact about the ‘single bucket’ world, but there is not. Physical space, as an idealist construction, can tolerate such a gap. It seems to me, therefore, that Einstein supports Berkeley.

11.4 John Foster on Spatial Topology and Empirical Reality The ‘mite’ argument rests directly on Berkeley’s text, the ‘bucket’ argument leads somewhat further away, even into Einsteinian relativity. I want now to consider an argument about the ideality of space which, as far as I can see, owes nothing directly to Berkeley, but is entirely original to John Foster. Foster’s argument moves in two steps. The first involves the refutation of naïve or direct realism, leaving the realist with a traditional representative realism as his best option. Given Part I of this book, I take that objective as achieved, and the Part II concerns the consequences for our conception of the physical world that follows from this. Let us call the mind-independent world of the representative realist—the world that we cannot perceive directly, in a naïve realist fashion, the Transcendental World (TW). The second step of Foster’s argument consists in the claim that a Transcendental World (TW) need not have the same topology as the physical world (PW) and that, therefore, the space of the TW is not physical space. If the space of TW is not physical space, TW is not the physical world. This is, of course, given the intuitive premise that spaces which have different topologies cannot be the same space—a space cannot have two inconsistent topologies at once. This would seem to be a simple case of Leibniz’s law. Foster uses a thought experiment to show that TW and PW could have different topologies, and hence be different spaces. This rests on the possibility of what he calls nomological deviance. Foster says: Putting it intuitively, a space S (whether physical or non-physical) is nomologically deviant iff, if G is its geometrical structure, there is a different geometrical structure, G’ such that the laws that govern what takes place in S ensure that everything behaves as if S had the structure G’ rather than G. (1982: 130)

    

199

The possibility the TW has a different topology from the PW is demonstrated by the following thought experiment. Figure 11.1 is Foster’s diagram from Foster (1993: 299), with his explanation. The thought experiment illustrates nomological deviance in the movement of bodies in TW, which because it is universal—applies to all bodies, in the sense of all physical phenomena, cannot show up in any investigation, however scrupulous. So, in so far as ‘physical space’ is the space that manifests itself in ordinary experience and in physical science, physical space is the functional space—that in which Oxford and Cambridge are where they seem to be, not where they are located in the transcendental space that sustains the empirical world. As so far expressed, this argument only shows that the topology of physical space is not independent of the laws governing the behaviour of bodies in it. In order to be an argument for idealism, it would have to be the case that, if physical space were independent of experience, then it ought also to be independent of the laws that govern the behaviour of bodies.

Oxford

Cambridge t2

t1

t3

Oxfordshire

t4

Cambridgeshire

Journey in the empirically constructed world

R2 Corresponding to Cambridge t3

t4

R1 Corresponding to Oxford t3

t2

Corresponding to Cambridgeshire

Corresponding to Oxfordshire Process in TW

Figure 11.1 After Foster, 2008: 132.

t1

t2

200

      

Barry Dainton challenges the argument on the grounds that it is not true that a space can possess only one geometry at a time. His example is non-Euclidean space—space which is ‘curved’. If I understand the argument correctly, it is that such a space has two radically different features. Consider a portion of curved space. It is curved because anything travelling through that space, from A to B must follow a curved path, but for that path to counts as curved there must be a more direct line from A to B. So the space in question has the physical feature of being curved—physical because that it the only way bodies can move through it— but it must also have an inaccessible traditional geometry because without that contrast the term ‘curved’ means nothing (Dainton, 2001: 249–53). This is not, I think, a difference in topology, but it is also a case where the ‘deviance’ can be worked out from the behaviour of the objects, so it does not transcend science. Someone might also challenge Foster’s argument as follows. They might claim that it is not implausible to think of space (and time, but we need not bring that in) as, not an independent ‘container’ of objects and events, but as one of the interdependent parameters of the physical world, and of TW, in so far as it is the physical world. So the topology of any real, as opposed to abstract, purely formal, space is a function of—perhaps amongst other things—the way objects behave. Foster has to say that this cannot be the case for a mind-independent world. If space were characterizable in a purely formal, geometrical manner, then one could assign those Cartesian co-ordinates to points which preserved the smooth movement of the bodies, but if space in its own right possesses a qualitative nature, then this opens the way to the possibility of deviance. This argument has some intuitive appeal. The diagrams do indeed show how visual space can be imagined as detached from the laws governing the movement of bodies. But does it follow that this is possible for any kind of qualitative nature? Foster says it does. But of course, exactly the same considerations would apply whatever the nature of the external correlate, and irrespective of whether it was intuitively space-like in its own character. (1993: 306)

This might be challenged. It might seem obviously true for a space that is qualitatively similar to visual space—the diagrams show that. But is this a necessary feature of all possible qualitative fillings that physical space might have? The blind from birth have a conception of space which is dependent on touch and kinaesthetic sensation. It is, presumable, not purely structural but has a qualitative component. But can we imagine a tactile space that is independent of how objects, including our own bodies, seem to move? Could one represent a surface as a felt continuity, whilst, at the same time, the objects seeming to move discontinuously within it, especially if one includes the sense of continuous movement of one’s own body?

    

201

On reflection, however, the impossibility of such a thing derives not from the qualitative nature of tactile space, but from the ‘tunnel vision-like’ nature of fingertip tactile experience. Within the spatial extension involved in such an experience, there seems to be no space for the ‘jumping about’ of objects. But if one imagines an extended spatial tactile experience, such as might be suggested by a feeling with the whole flat of the hand, then phenomena like that in the visual case seem to make sense. It seems that any qualitative filler that gives extension ‘in one go’ (not punctiform, like fingertip) is thereby bound to allow for deviant movement within it. In so far as the qualitative component has a spatial nature in its own right, which it must do if it is to ‘flesh out’ the purely mathematical geometrical conception, it must allow a variety of movements within itself. I have not so far given Foster’s argument for the claim that the purely formal characterization of space cannot be adequate. [A] knowledge of the geometrical structure of space does not amount to a full knowledge of its intrinsic nature. For it does not tell us anything about the kind of thing that has this geometrical structure. In order for the space to be threedimensional network of distance relations that satisfies the relevant geometrical requirements, and these requirements can be specified in a purely formal and mathematical way. The simplest way of doing this is by saying that, relative to a suitable coordinate system, the distances confirm to the Pythagorean principle; in other words, for some one-to-one correlation between the points and ordered triples of real numbers . . . and for any pair of points p₁ and p₂ , if the ordered triple correlated with p₁ is (x₁, y₁, z₁), then the distance between p₁ and p₂ is the square root of {(x₁ – x₂)² + (y₁ – y₂)² + (z₁ – z₂)²}. But knowing that the distance relations meet these requirements does not tell us the nature of the spatial thing to which the points that stand in these relations belong. [This can be seen because] we can represent the set of all ordered triples of real numbers as itself forming a kind of abstract space with the triples serving as points, and with the distance relations by the Pythagorian formula itself ( . . . ). Trivially, this will have a three-dimensional Euclidean structure, but, as an abstract mathematical space, its nature is not, in any other way, remotely like that of the concrete space which forms an ontological ingredient of the physical world. Clearly, there must be something further about the intrinsic nature of physical space that its geometrical specification does not capture. It is also clear that this additional factor cannot merely consist in the concreteness of physical space.⁹ There must be something ⁹ Michael Martin, in discussion, in effect challenged this sentence, by arguing that it was possible that the extra feature physical space possessed was simply its spatiality, and that the fact that there could be abstract geometries that mirrored the geometry of space did not mean that physical space was such a geometry, plus some further qualitative feature. I do not find this point easy to unpack, but I think that physical space must possess the properties of extension and distance of the same kind as the properties of extension, distance, size, etc., which its occupants possess, and these are primary qualities.

202

      

further about the nature of the concrete thing that has the geometrical structure – something that covers, as it were, the nature of the spatial material, or fabric, in which the geometrical structure is realized. (2008, 74–5)

So a purely formal characterization does not necessarily characterize anything spatial at all. Once the qualitative element has been shown to be necessary, the possibility of deviance seems to be unavoidable. Taking this as showing how Foster’s thought experiment works, we can set out the argument that rests on it in the following way: (1) Any non-abstract space must allow for the kind of discontinuous behaviour shown in the thought experiment. (2)

If TW is non-abstract, it must involve a non-abstract space.

Therefore: (3) Any non-abstract TW must allow for kind of discontinuous behaviour shown in the thought experiment. (4) Physics would not reveal this kind of nomological deviance if it occurred in a TW. (5) A completed physics does capture the nomological structure of the physical world (PW). Or, PW is what a completed physics captures. Therefore: (6)

If TW is non-abstract and contains nomological deviance it cannot be PW.

(7)

Nothing abstract is the PW.

Therefore: (8)

If TW contains nomological deviance it is not PW.

(9) Laws of nature are contingent, so, even if the actual world does not contain nomological deviance, it might have done so, or might come to do so. Therefore: (10)

The (or ‘any’) actual TW is not necessarily the PW.

(11)

Identities are necessary.

Therefore: (12)

The (any) actual TW is not PW.

Defence of the argument (i). (1) is what was defended just above. (2) seems to be correct—how could a TW—especially one which is thought to be PW, be concrete, yet its spatial features be purely abstract? (3) follows from (1) and (2) by hypothetical Syllogism.

    

203

(4) is ex hypothesis, from the nature of the thought experiment. (5) is a substantive claim we will have to consider. (6) follows from (4) and (5) because PW is something of which physics captures the nomological structure, and it does not capture the nomological structure of a TW with deviance. (7) asserts that PW, in the context of this discussion, is not an abstract structure. This enables us to eliminate the reference to abstraction in (8). (9) requires discussion. The rest of the argument follows by simple modal principles. So (5) and (9) might be disputed, as might the applicability of (11)—that is, someone might try to argue that the identity in question is not rigid. (9) and (11) concern the role of modality in the argument, (5) the definition—or scope or ambition—of physics. One could also challenge the significance of the conclusion, by saying that what strictly counts as ‘the physical world’ is not really central. This connects with challenging (11) and the general modal worries. In fact, all the objections concern the substantiality of the conclusion, rather than its validity or even soundness. I will look first at the modal issues, as raised by (9) and (11). One might challenge the argument by saying that though this deviant behaviour is logically possible, it is not metaphysically possible, and that that is what matters, so the logical contingency of natural law does not allow such nomological deviance. The rationale for this response would be as follows. It is reasonable—the best or simplest explanation of our experience—that bodies do not behave in the deviant way, and, given that this non-deviant behaviour is part of the nature of these objects, they are not actually capable of behaving otherwise, so (9) is not correct. Of course God could make them behave otherwise, or there could have been (God could have created) deviant objects, but it is not naturally possible that the objects there actually are should do this. The suggestion that they might is, therefore, merely a logically possible sceptical option. If the sceptical option obtained, then there would be a fissure between TW and physical space/world, but there is no more reason to opt for that sceptical possibility than there is for other sceptical options. Perhaps it is possible to respond to this objection by claiming that it is logical possibility that is relevant, not something called ‘metaphysical possibility’, which, after all, leads a rather indeterminate existence between the logically and the empirically possible. Thus one might argue as follows. 1. There is a conceptual (a priori) connection between experience and the PW. 2. There is no conceptual (a priori) connection between experience and any TW. Therefore:

3. No TW can be part of the conceptual or analytic identity of the PW.

204

      

Foster always admitted (in discussion) that his argument rested on a very fine line between what was probably the case, concerning the TW, and what was conceptually required for something to be the PW, so even if they marched strictly in parallel, they could not be identified. Someone might ask why we should accept 1, and agree that there is a conceptual, or a priori, connection between experience and the PW. It breaks down into two steps. (i) There is a conceptual (a priori) link between experience and physics. (ii) There is a conceptual (a priori) link between physics and what constitutes, or counts as, the physical world. (i) does not deny the role of mathematics in physics, but affirms that physics rests on and is confirmed by the phenomena, inside and outside of a laboratory. This should not be controversial. (ii) might be more controversial. It says that the physical world does not transcend what a completed physics would show: that is, it is not something noumenal which surpasses even the best possible theory based on the evidence that could be available to one. Suppose we accept (ii), what follows? There are two responses, a deflationary one and a Fosterian one, as one might call it. The deflationary one is ‘OK, so the TW is not strictly the PW, but it is the thing that actually causally sustains our experience and hence our empirical science, so the conceptual facts—if you can call them that—are insignificant.’ To give this theoretical back-up, one might deny (11) and say that it is contingently identical with the physical world, because it plays that role, as Olivier played the role of Richard III, but they are not picked out as identical by rigid designators. Foster rejects the idea that something not essentially the physical world might be contingently identical with it, or ‘play that role’, but maybe that thought is no more counterintuitive than many philosophical positions to which one is forced to resort. But this resorting to contingency shows that it is the conformity with physics that matters for being the physical world, not being mind-independent, for a mindindependent TW that failed to play the physical role would then not be the PW. Someone might argue that this then becomes an argument about how to use the term ‘physical world’, but it is more than that. In one way of using it, it can be treated as a rigid designator, in another, it is a name for a certain role that could be variously realized. Foster’s argument rests on the firm intuition that we mean it in the former way, but not as a hidden essence that transcends our scientific investigation. It is part of our notion of the physical world that it is accessible from the empirical world. Hume argued that, once our reasons for being naïve realists had been refuted, our reasons for thinking that there was a world out there more or less the way it



205

naively appeared had lost its foundation. Standard Lockean ‘representative realism’ was only a pretence to enable us imagine that we had saved our naïve realist— or, in David Chalmers’ word, ‘Edenic’—world. Hume’s scepticism about causation and explanation led him to claim that there was no explanation of the way experience went, and radical phenomenalists, such as Mill, followed him in this. I shall assume here, but will argue in Chapter 12, that some explanation of how experience happens and why it is as it is, is required. The question then becomes, what conception of the TW can we have, especially given that our conception of the PW puts no particular constraints on it, other than that it can sustain the world of experience that grounds science? Foster shows that, whatever the TW might be, it is not strictly the PW. A way of putting together Hume and Foster is that they both show that what it might be that lies behind common experience—the nature of the TW—is up for grabs, for it is not itself the PW in a strict sense: and, to challenge my figure of speech, we cannot grasp or grab it scientifically or in any other way. So this puts us in the position of trying to decide what this transcendental reality is, given that there is one. A realist might, at this point, appeal to Putnam’s claim that the brain in a vat is ‘really’ talking or thinking about the wiring in the computer that is controlling its experience (Putnam, 1981: 1–21). Perhaps our concept of the physical world has a two dimensional semantics, so what can be discovered by science is the primary intension, and the TW is the secondary. In the case of Dainton’s objection, discussed above, the secondary intension is the inaccessible Euclidean space into which the curved space is inserted as the primary intention. This seems to me to be a very dubious application of what is already, in my view, dubious externalist semantics. Natural kinds, as characterized initially by Putnam, have, as their real essences structures that science can discover. There is no suggestion that we are ‘really’ talking and thinking about things that transcend all investigation and discovery, such that we can never really have the faintest idea what we are talking and thinking about. So the appeal to real essences or secondary intentions does not add to the plausibility of the realist’s position.

11.5 Conclusion Foster’s argument is successful if one thinks that the physical world is something that necessarily is mapped by a completed physics. If you are prepared to abandon this and allow that in essence the physical world might be quite different from what scientific investigation and theory could reveal, even in its topology, then Foster’s argument will not disturb you. But, if you put together the non-Fosterian thought that what sustains the phenomena does not necessarily itself conform to physics, with the Hume-inspired thought that once naïve realism has been

206

      

abandoned we have no reason to think it at all like its manifest form, then it is up for grabs what the best account of the phenomena is. I argued in Chapter 9 that the theories of modern science, taken realistically, do not constitute the best account, and so we seem inevitably pushed in the direction of idealism. There is one form of idealism that straddles physicalism and genuine idealism, namely Kantian idealism. It is not my intention to enter into the business of Kantian interpretation. For my current purposes, what I mean by ‘Kantian idealism’ is the theory that there is something not essentially mental that lies behind our experience, but which is not subject to any of the forms of intelligibility that mark the world of experience or science, as we are capable of knowing it. It straddles the options because this ‘noumenal’ world is definitely not mental, but it transcends anything we can know as physical. On one interpretation of this ‘transcendence’, even the category of causation does not apply to the noumenal world, but then the whole point of postulating it as something that lies behind, and in some sense explains the world as we know it, seems to be undermined. I take it that it is this problem (together with Kant’s deployment of the antinomies) that led to Hegelian idealism. The very fact that our ordinary knowledge and our science enables us to cope, on the whole, with the world seems to me to show that, in broad structural, or ‘primary quality’ terms, whatever lies behind or world—if anything does—must mirror these structural features. So the only question is what, if any, other features beyond structural ones, the hidden world possesses. I believe I have discredited all the ‘mystery stuff ’ accounts that have been put forward, but it is impossible to refute a purely negative account. What needs examining is the viability of intelligible mentalist—and, in that sense, idealist—options. That is the next chapter.

12 Mentalist Alternatives to Berkeleian Theism, and Their Failure 12.1 Introduction Maybe I have persuaded some readers that some version of phenomenalism or idealism is worth considering—at least for curiosity value: the aporia I have raised concerning normal forms of physical realism merit a look at the non-realist options. But that does not simply leave us with Berkeleian idealism. First, there are two phenomenalist options. One is the Humean sceptical option, taken up by Mill and some positivists (though maybe neither Mill nor the positivists would have owned up to the sceptical label). The other is the Berkeleian, non-sceptical theistic, and, in a certain sense, realist option. Then there is the currently almost fashionable option of panpsychism. We need reasons for rejecting the Humean and panpsychist options if the field is to be taken by Berkeleian theism. In fact the non-Berkeleian or non-theist options have more in common than might at first be imagined. To see this, it will help to begin with Helen YetterChappell’s ‘Idealism without God’ (2017: 67) Setting this out she says: We want something outside of our finite minds to sustain objects when (finite) minds are not perceiving them, and to account for the regularity of our perceptions. But, even for the idealist, this does not need to be a god in any recognizable sense. There is no reason it must contain desires, intentions, or beliefs. It needn’t be an agent. On the view I offer, external reality – the tables, chairs, brains, stars, quarks around us – is constituted by a complex phenomenal unity, governed by laws of nature, structurally analogous to those materialists posit. This accounts for both the stability of the world external to us, and the regularity of our own experiences.

This is very similar, ontologically and nomologically, to Goff ’s cosmic version of panpsychism. According to the panpsychist, the intrinsic nature of the physical world is phenomenal down to its roots, and its nomological structure, at some level or other, is that of the physical sciences. Michael Pelczar’s version of Mill’s phenomenalism differs from this only in the fact that those parts of the physical world that are not in fact perceived are potentialities for the phenomenal realities

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0013

208

    

that both Yetter-Chappell and the panpsychist take to be actual. As Pelczar puts it, the physical world is the potentiality for what a Berkeleian God would perceive. So all three of these theories treat the physical world as some version of a phenomenal array, held together, at some level, by the laws of physics, but not requiring any further mental characteristics, of agency, intelligence or moral character, such as are normally attributed to God, to explain it. There are, of course, differences between these theories, and, indeed, within each general category—for example, Yetter-Chappell’s theory is more like cosmo-panpsychism, than it is like ‘smallist’ panpsychism—but the similarities are central to the metaphysical thrust of all these theories. It is, therefore my task in this chapter to try to show that it is not reasonable to dismiss the features of higher mentality, beyond the phenomenal, from the mental reality that is responsible for the physical world. It is not, however, my task here to perform a complete work of philosophical theology by trying to prove that the mind behind the phenomenal-physical world meets all the standards of orthodox theism. That would be a further work. I will be satisfied if I can get to the point at which Aquinas was able to announce, after each of the Five Ways, ‘and this we call “God” ’. Something similar also applies in the case of the other theories. A full discussion of phenomenalism and panpsychism would require another book, not a chapter. What I shall do is argue that Mackie’s objection to standard phenomenalism works, and that Pelczar’s attempt to take potentialities in a more robust fashion does not substantially alter the situation. Then I shall argue that the notion of mind involved in panpsychism is too crudely empiricist, and that this criticism also applies to Yetter-Chappell. Finally, I’ll argue that John Foster is right; the very notion of law of nature, which all these theories require, entails the existence of a governing intelligence—‘and this we call “God” ’!

12.2 Hume–Mill Phenomenalism The objection to the idea that our experience just occurs without any cause is well expressed by Mackie. Phenomenalism, without Berkeley’s God, is committed to the intermittent existence of things, and hence to interpreting successive observations of the same thing as the repeated springing into existence of complex groups of phenomena just like groups that passed out of existence some time before; and this sort of coincidence would occur not just a few times, but on innumerable occasions. Also, whatever observed phenomena we now explain in terms of unobserved causes would be left unexplained. In short, it would be an utterly improbable coincidence that our perceptions should lend themselves as well as they do to being interpreted as intermittent observations of a world of persisting things,

– 

209

unless there was some set of persisting entities related in some systematic way to what we perceive. The realist hypothesis, in a broad form which would include even Berkeley’s mind of God, is overwhelmingly supported by these considerations against its phenomenalist rival. (Mackie, 1969: 114)

Michael Pelczar is probably the most recent defender of the Hume–Mill kind of phenomenalism. His original response to this argument was essentially of the form ‘explanation must stop somewhere and phenomenal regularity is as good a place as any’. A phenomenalist does not believe that it is an improbable fluke that or statistical miracle that our experiences admit of interpretation as perceptions of an orderly physical universe. But nor need he believe that something explains why our experiences occur the way they do. For it is open to him to hold that it is a fundamental law of nature – or of experience that they occur that way. The point is missed often enough. ‘Caused versus random’ is a false dichotomy. Nothing (as far as we know) causes bodies to conform to the inverse square law of gravitation . . . Likewise, nothing causes physical phenomena to behave only in ways that conserve energy. That physical things behave in energy-conserving ways is not a consequence of some further fact about those things: it is a fundamental regularity in the things’ behaviour. . . . Well, according to a phenomenalist, it is a fundamental law of the phenomenal world that its constituent experiences are interpretable as perceptions of a physical world in which energy is always conserved. The main difference between the laws, according to the phenomenalist, is that the latter law is more fundamental. The relative coherence of conscious experience – coherence relative to what one would expect of randomly occurring experiences – is a reflection of the fact that experience is subject to certain fundamental laws, or exhibit certain fundamental regularities, rather than experiences being caused in certain ways. (Pelczar, 2015: 135–6)

Given that explanation must stop somewhere, why not at the level of phenomena? First, it is not clear that there are any laws at that level, and if there are, they must be abominably complicated. Imagine your visual field at the moment, then try to describe, in law-like terms, the flow of that experience without invoking the physical objects that we take it to be of. That is, one must refer only to phenomenal properties and invoke only laws that connect phenomenal properties. One might be tempted to argue that there must be some order at this level or it could not sustain the interpretation of it as a physical world, but how one could express that order without invoking the interpretation, I do not see. The radical phenomenalist must adopt the ‘philosophy of “as if ” ’; that is, he is bound to understand his basis

210

    

ontology—the phenomenal—in terms of another: the former only makes sense if seen as if it constituted something else. Can this be a proper place at which to end one’s explanations? The Berkeleian has no problem of this kind, because he attributes to God the intention to give phenomena of the sort that He does just so that it will constitute an empirical physical world for us. A teleology of this sort does not fit into a Hume–Mill scheme. Furthermore, it is essential to the Hume–Mill scheme that one has a Humean account of causation, because it makes no sense to have a realist account if there is no intelligible mechanism running on the purely phenomenal level, and if on the alternative, ‘as if ’ account, most of the things that are enter into causal relations— unperceived objects—are not themselves real. But I have already argued in Chapter 9, Section 9.6 that the Humean account of causation is radically implausible. Putting these things together, we could conclude that Hume–Mill phenomenalism is totally implausible. Later, however, Pelczar developed a less off-hand account of the potentiality that constitutes physical objects. I think it is fair to describe it as taking the potentialities for experience as being objective probabilities. They are not like Boscovitchian powers, that is, entities rather like magnetic fields, but neither are they just post facto rationalizations of what actually happens, or subjective tendencies to expect certain experiences, as Mill seems to think. On the more traditional phenomenalist system, the contribution that an unperceived, theoretical entity makes to experience is the difference it makes at the macroscopic or laboratory level, that is to what is deemed the evidence for its existence. This is not so in Pelczar’s version. A quark, for example, is the potentiality for the experiences that a possible observer might have of the quark itself: it is what, on one interpretation of Berkeley, Berkeley’s God would perceive when He perceived a quark. To put it slightly differently, a physical object is not a potentiality for experience, but a set of potential experiences. But that is not the only potentiality that a quark has: it also has the potentiality to affect other physical things—themselves also only potentialities—in the way that a correct physics says such physical things affect each other. It is not clear how these two sets of potentialities—one directly for experience, the other for physical interaction—are related to each other. The main problem with Pelczar’s theory, it seems to me, is that the ontological status of his potentialities is essentially ambiguous. For both Hume and Mill, explanation of the structuring of experience into a coherent world is provided by an associationist psychology—the tendency of the mind to pass from one experience to an anticipation of that which is normally associated with it. On most interpretations, this leaves Hume as a total sceptic about both the existence of an external world, and of the intelligibility of the notion of such a world. Mill is more ambiguous. He certainly accepts the psychological explanation of how we come about the idea of external objects that exist unperceived, and he affirms that the

– 

211

only idea we can have of such objects is that they are the permanent possibilities of sensation. It was not clear to his readers whether this account was really meant as a defence of the existence of an external world. In a footnote at the end of the chapter on the nature of the material world he says of ‘My able American critic, Dr. H. B. Smith’, Dr. Smith’ s criticisms continually go wide of the mark because he has somehow imagined that I am defending, instead of attacking, the belief in Matter as an entity per se. As when he says (pp. 157–8) that my reasoning assumes, contrary to my own opinion, ‘an a priori necessity and validity of the law of cause and effect, or invariable antecedence and consequence.’ This might fairly have been said if I were defending the belief in the supposed hidden cause of our sensations: but I am only accounting for it; and to do so I assume only the tendency, but not the legitimacy of the tendency, to extend all the laws of our own experience to a sphere beyond our experience.

The fact that he assumes ‘only the tendency, not the legitimacy’ of extending experience beyond its actual occurrence strongly suggests that the ‘permanent possibilities of experience’ are the product of a logical construction, sustained by the association of ideas, not of an inference to an external cause. I think that this is supported by Mill’s sophisticated Humean account of causation. He says in his Logic: The only notion of cause, which the theory of induction requires, is such a notion as can be gained from experience. The Law of Causation, the recognition of which is the main pillar of inductive science, is but the familiar truth, that invariability of succession is found by observation to obtain between every fact in nature and some other fact which has preceded it; independently of all consideration respecting the ultimate mode of production of phenomena, and of every other question regarding the nature of ‘Things in themselves’. (2020: 243)

As Mill clearly has no time for ‘things in themselves’, he has no time for the idea that they might produce phenomena. But the causation that he does believe in, namely ‘succession found by observation’ cannot extend to the ‘permanent possibilities’ themselves, so these are not causes of phenomena; rather they are what would later be called ‘logical constructions’. Whatever might be the case for Mill, Pelczar makes a realist interpretation or development of Mill’s ideas. Pelczar entirely rejects the subjectivist account of the nature of the potentiality that lies behind sensation. In his view, these potentialities are objective probabilities. The question then arises of what the reality of these probabilities or potentialities consists in. It is not in a psychological tendency

212

    

of the perceiving subjects, nor in some realist physical force or power. They seem to have to be self-standing truths, without any truthmaker. One way of giving this content might be to think of these potentialities as a complicated web of laws like those that a Berkeleian theistic phenomenalist would attribute to the mind of God, but treated in an abstract, Platonic way, and so dispensing with the theism. I shall be arguing later in this chapter that laws of nature require a God, but, even so, I don’t think Pelczar wants to adopt such a Platonism. Pelczar claims that his ontology is entirely mentalistic because there are only mental states and potentials for mental states, Phenomenalism is still an identity theory, since it identifies the world’s physical features with certain of its mental features. It’s just that the mental terms of the phenomenalist identities are potentials for experience, rather than actual experiences. (2019: 578)

But I can see no good reason for regarding the potentialities as mental. Potentialities for mental states are themselves mental only if the potentialities are themselves mental states, that is states of a mind, and, though they are that for a Berkeleian, because they are states of the Divine mind, they are not for ‘secular’ phenomenalists. This point, which might seem at first to be mainly verbal, is emphasized by the worry I drew attention to above, namely that the potentialities are Janus faced, because as well as being potentialities for producing experiences, they also function as laws of physics because they also govern or describe the relation of the physical atoms. My worry is this: as these potentialities are external to any mind, and follow the laws of physics in their own interactions, it seems strange to call them mental unless they and their interactions are logical constructions out of experience. This conclusion is, I think, strengthened by Pelczar’s discussion of what he calls the Sober Intuition. Sober Intuition: it’s possible for a world physically identical to ours to contain no conscious experience. (2019: 576)

Pelczar claims that phenomenalism can satisfy this intuition. Most potentials for experience go unrealized in our world and there is a world identical to ours in its potentials for experience, but in which no potential for experience gets realized. According to phenomenalism, such a world is physically identical to ours, despite containing no conscious experience. So phenomenalism is consistent with Sober Intuition. (2019: 577)



213

The reason that most potentials go unrealized in the world is that in most circumstances there is no conscious subject present and this normally means no brain in the appropriate place. The full physical duplication of our world, however, involves the duplication of brains. If this does not guarantee the presence of consciousness, it reinstates a kind of dualism and emphasizes the non-mental nature of the potentials; it seems to be part of their nature to interact with each other—nothing further is needed to achieve this—but, on their own they cannot give rise to consciousness. If they can exist without consciousness, and require a conscious principle to be present before they can exercise their potential for consciousness, they are not themselves mental. As much as I admire the austere but realist phenomenalism of Mill and Pelczar, I think that it can overcome Mackie’s criticism of phenomenalism only if it adopts either a straightforward realist theory of powers, as does Boscovitch, or a Platonistic account of laws. The former is a form of physical realism, the latter is strange in the following ways. First, Platonism normally concerns things that can be regarded as necessary existents. Second, laws of nature, which are contingent, normally connect actually existent phenomena, and only from that base are taken to apply to counterfactual situations. For the phenomenalist, however, the laws mainly cover the non-actual, and apply to the world of ‘as if ’ or ‘might have been’.

12.3 Panpsychism I have argued against panpsychism in Robinson (2016) and I do not intend simply to repeat all the arguments presented there, but to present further thoughts. The classic inspiration for modern Panpsychism is Russell’s neutral monism, as found in his Analysis of Matter (1927), and he was followed up, after a long dormant period for the thesis, by Maxwell (1978), Lockwood (1989), Galen Strawson (2006; 2017), Chalmers (2003), and Goff (2017). The motivation for Russell’s theory was that science uncovers only how things behave, not what they essentially are. This gap is plugged by postulating a mental core, which is also used to explain how mind can be present in living animals, namely because it is in matter from the start.¹ Chalmers states the issue as follows:

¹ One might ask the question why Russell, or Lewis, thought that there had to be anything more to the entities of physics than their powers, or existence as fields, in line with the Boscovitchean theories discussed in the previous chapter. It was not because of the fears of a ‘powers regress’ as developed there. It was because Russell and Lewis had a Humean theory of causation, and this rules out a realist theory of powers or fields. A Humean has to have something between which the nomic or constant conjunction relation holds. Similarly, Armstrong thinks that there must be non-dispositional properties in objects, but not because he is a Humean, but because he takes laws as fundamental, and laws need entities to relate—you cannot have something which is just a structure of laws—at least, in

214

    

Russell pointed out that physics characterizes physical entities and properties by their relations to one another and to us. For example, a quark is characterized by its relation to other physical entities, and a property such as mass is characterized by an associated dispositional role, such as the tendency to resist acceleration. At the same time, physics says nothing about the intrinsic nature of these entities and properties. Where we have relations and dispositions, we expect some underlying intrinsic properties that ground the dispositions, characterizing the entities that stand in these relations. But physics is silent about the intrinsic nature of a quark, or about the intrinsic properties that play the role associated with mass. So this is one metaphysical problem: what are the intrinsic properties of fundamental physical systems? At the same time, there is another metaphysical problem: how can phenomenal properties be integrated with the physical world? Phenomenal properties seem to be intrinsic properties that are hard to fit in with the structural/dynamic character of physical theory; and arguably, they are the only intrinsic properties of which we have direct knowledge. Russell’s insight was that we might solve both these problems at once. Perhaps the intrinsic properties of the physical world are themselves phenomenal properties. Or perhaps the intrinsic properties of the physical world are not phenomenal properties, but nevertheless constitute phenomenal properties: that is, they are protophenomenal properties. If so, then consciousness and physical reality are deeply intertwined. (2003: 130)

This has the interesting consequence that it produces a theory which is a kind of hybrid between idealism and materialism based on physics. In Russell’s words; While on the question of the stuff of the world, the theory . . . has certain affinities with idealism . . . the position advocated as regards scientific laws has more affinity with materialism than with idealism. (388)

In other words, the ontology is idealist, but the behaviour of these mental entities entirely follows the laws normally associated with matter. This is also the position adopted by the later followers, and it follows necessarily from the principle rationale for adopting the theory. This prompts the question of how one is to understand the relation between this idealist ontology and the totally physicalist—in the sense of conforming to fundamental physics—nature of the conduct of these mental entities. Given the supposedly mental nature of the things, one might expect their behaviour to follow recognizably psychological patterns, but this is not so. In fact there are Armstrong’s view. So those who take laws to be basic, whether Humean regularities, or laws more realistically construed, needs something intrinsic. On the other hand, a powers realist does not, at least at this point in the argument, or at the micro level of reality.



215

two very difficult issues contained in this problem. (i) The more obvious one is how one should understand the relation between the mental core and the physicalist causal properties associated with it, but the other is (ii) how to make sense of the mental core itself. (i) The relation between conscious core and ‘physical’ behaviour; (a) contingent or necessary, Humean causation or real powers?

Robert Adams in his invaluable article ‘Idealism vindicated’ (2007) is undecided between the merits of Berkeleian idealism and panpsychism. He sees the latter as a descendent from Leibniz’s monadism, and, on his version, the core mentality is responsible for the causal properties. He is puzzled, however, as to how the primitive consciousness of a monad—or an electron—could possibly explain the behaviour of such particles. What kind of primitive consciousness or sensational state could be thought of as being naturally responsible for the mass-behaviour or charge-behaviour of an electron, or the uniform and exact mathematical development of the quantum field? I believe that one should not lose grip of the intuition that consciousness is associated with a certain kind of responsiveness to stimuli. One considers the possibility that earth worms might be conscious because their responsiveness shows some of the flexibility we associate with being conscious—but not much. Consciousness completely disassociated from what seems to be its natural function is bizarre. Goff, in his most recent work tries to attribute ‘normal’ psychological properties to atomic entities by suggesting that individual atomic elements, such as electrons or quarks may have libertarian free will. The point of this move is to unify all causation in the world into one type, namely genuine agency. (Goff, 2020). Goff imagines these particles deciding, freely, to exercise their various powers of spin, charge, mass, etc., although there is nothing else that they can do. Apart from the strangeness of thinking of this as a libertarianly free kind of behaviour, there is also the problem of how an entity without structure, let alone an appropriate structure, be thought of as deciding or choosing, which seems to be a process with stages. Taking Lewis’s quiddities as the model, the relation between core and causal properties is entirely contingent. In Lewis’s case, this has to be so, for his account of causation is Humean, and mere constant conjunction cannot be a necessary consequence of the intrinsic nature of the things so conjoined: if it were an expression of that nature, then it would count as a real power of that nature, and hence not a Humean contingency. If this model is projected onto panpsychism, then we have a mental core, the nature of which has nothing to do with the way it behaves. It merely sits in reserve, so to speak, waiting for the time when it is to manifest itself as the minds we normally recognize. But I have already argued in Chapter 9, Section 9.6 that the Humean theory cannot account for the order in

216

    

nature, and this argument would apply equally in the case of panpsychism. Even putting this refutation aside, if we continue with the idea that all the efficacious laws are those of physics, even then the mental will be epiphenomenal, in the sense that its particular nature has nothing to do with how it behaves. Such a schizophrenic divide between the account of what there is and how it behaves is profoundly counterintuitive. It makes the ontology seem very ad hoc. On this view, there could be a world with non-mental quiddities which was just like ours in all its behaviour and laws, differing only in the fact that it was a zombie world at animal and human level. If one abandons the Humean account, replacing the view that the laws are just the pattern in the organization of the fundamental core with real powers, then one still faces the question of what the relation is between the core and the real causal powers. If they are taken to be contingently associated with the core, then as I argued in Robinson (1982: 122), this would seem to leave open the possibility that the core and the causal power might simply come apart, as if the field of magnetic influence and the metal magnet at its centre might float off in different directions. That is, if you a realist about powers, and if the core and the power are only contingently connected, one can raise the question of what binds them together. It will not do to answer that problem by simply saying that the power is a property of the core and that is why they cannot float apart. Compare this with other contingencies. A certain ball may be contingently red, but it is easy to say what binds the two together—it is the colour of its surface. It must have a surface and the surface has to have some colour or other, and, in this case, it is red. But why should some mental core have, for example, a certain charge or mass, or any charge or mass at all—these are not mental properties, in any normal understanding of the term. There is also a contrast with physical powers, such as, for example, magnetism, which is explained in terms of the arrangement of the electrons in a magnet. Goff suggests that we need ‘consciousness+’ to explain this relation: that is, that there is more to the mental core than what is the phenomenal component, and this extra ‘+’ explains the power, but this seems to be an entirely ad hoc move, and even raises the same problem; what attaches the phenomenal component to the ‘plus’ element? (i) The relation between conscious core and ‘physical’ behaviour; (b) the mental core and epiphenomenalism.

Epiphenomenalism is a problem which is recognized as facing anyone who believes in physical closure, but is not a reductionist about the mental. Philip Goff believes that panpsychism can avoid epiphenomenalism in the same way as physicalism can. I shall argue, however, that, though it would be strange to use the label ‘epiphenomenalism’ in the case of physicalist reduction, this theory is equally committed to the causal redundancy of all macroscopic states, including



217

psychological ones, given micro-closure. I shall also claim that panpsychism has extra problems. Epiphenomenalism follows from causal closure of the physical, if one does not want to allow overdetermination, that is, two causes, both of which are efficacious, though at least one of them would be sufficient on its own. Goff argues that the following is a case of ‘superficial overdetermination’, and, as such, does not entail ‘causal exclusion’ for the macroscopic level. The baseball smashed the window. Particles arranged baseball-wise acting in concert smashed the window.

It is ‘superficial’ because, though there are ‘competing’ causes, for the physicalist, one is grounded on, or identical with the other, so that, unlike the case where someone is killed by a bullet and a bolt of lightning, they are not really different, competing candidates for being the cause. Goff says the same is true for, for example, pains and the corresponding neural state, if the former is identical with, or grounded on, the latter. Goff then applies this to panpsychism, The constitutive Russellian monist [that is, one who thinks that complex psychological states are just made up from the conscious states of its parts], given the structural similarity between her view and physicalism, is able to give the same solution. The only difference would be that, as the constitutive Russellian monist tells the story, M is a micro-level event involving consciousness or protoconsciousness: The fact that I am in pain caused me to scream and run away. M (= the fact that certain micro-level entities with certain [proto]phenomenal states are related in such and such a way) caused my screaming and running away. According to a standard form of constitutive Russellian monism, the fact that I am in pain is grounded in – is nothing over and above – some fact concerning certain of my fundamental material parts and their (proto) phenomenal properties. If we identify that fundamental fact with the micro-level cause of my pain behaviour, then the over-determination of my pain behaviour turns out to be superficial and unproblematic. (2017, 155)

The first problem is that, if the world is closed under physics—that is, under the most minute micro-facts—then all more macro entities fail to make an extra difference to the location and motion of matter, for that is disposed entirely according to the forces of micro-closure. By the causal criterion of empirical reality, they are not real in addition to the micro-level, but just ways of conceptualizing the more fundamental level of reality into salient groupings which are of

218

    

relevance or interest. So your behaviour is what it is wholly because of the natural dispositions of your smallest parts, given their organizational relations, and does not originate from your psychology, except as a bye-product, and so is, in that sense, epiphenomenal. Looked at from another scientific perspective, your conscious states, behaviour, speech and thought are just the unfolding, according to Schroedinger’s equation, of that part of the quantum field that constitutes you. (See Robinson, 2016: part II, for arguments in detail for this view.) Is the situation any different for the panpsychist? It seems that it is not. The laws are the same as for the physicalist, and the macro, psychological level, makes no contribution in addition to the contribution of the micro parts, or the quantum field. So the fact that the macro psychological states of the ordinary subject are constituted by particles or forces that are themselves conscious does not allow the macro states to add anything to what is determined by the micro level. This has important consequences for Goff ’s theory. He believes that conscious macro-subjects, such as ourselves, are not reducible. He thinks that, if one is a ‘smallist’ and faces the combination problem, then they would be reducible (2017, 217–19). His response to this is that, if one abandons ‘smallism’ and adopts cosmopsychism—the view that the fundamental entity is the universe as a whole, and it is one unified consciousness, then ordinary subjects, such as ourselves, will not be reducible, because the system is not grounded bottom-up. But this is wrong as far as the problem of epiphenomenalism is concerned. Even if micro entities are, as Goff claims, just aspects of the macro whole, it is still true that the laws of physics operate over those aspects—for example, electrons and their properties, or properties of the quantum field at given points—not by reference to the macroscopic whole itself. In other words, even if one is an ontological holist, one’s physics still ranges over the constituent features, it is not top-down. Goff believes that, as long as one can claim both levels are real, ‘superficial overdetermination’ obtains, and the macro-level can be held to be causally efficacious. I have argued that that is not the case, because the relatively macro level adds nothing to the causal consequences of the micro, even if real. This is like the classic case of epiphenomenalism; the mental is regarded as real and irreducible, but adds nothing to the causal process. (ii) The nature of the panpsychist’s consciousness.

There is also the problem of how to understand the quality of consciousness in the minute elements of matter. The mentality of Leibniz’s monads consists, in Adams’ words, of ‘the subconscious perceptions of a substance so confused as to be totally devoid of consciousness’ (2007: 53). The ‘confusion’ arises because, far from being a simple sensation, monads reflect the whole universe. This is not a view adopted, as far as I know, by any contemporary panpsychists. But the idea that very primitive elements must, if conscious at all, have it in a very dim or subconscious



219

form, seems difficult to deny. Nevertheless Strawson seems committed to a form of Cartesian reflexiveness and transparency at all levels. He claims that ‘all awareness involves awareness of awareness’ and ‘all awareness comports awareness of itself ’ (2017, 198). This seems a pretty tough standard for the minutest of particles! Doesn’t ‘awareness of awareness’ involve something like recognition of the content of one’s state, with discrimination of its various features, and doesn’t this involve a complexity of mentality that seems to make no sense for an elementary particle that wholly lacks further structure—unlike a Leibnizian monad? Strawson’s Cartesian intuition is that without the ‘awareness of awareness’ you have lost the essence of consciousness, and, hence, the whole point of panpsychism. One might agree with this point about awareness of awareness, and think that it shows that a simple physical entity lacks the complexity necessary for possessing it. In response to the problems associated with attributing consciousness to the smallest elements (known in the literature as ‘smallism’) Philip Goff proposed ‘cosmopsychism’, which treats the whole universe as a single consciousness. This might be held to tune in better with modern physics if that holds that the quantum field is the most basis entity, and not the kinds of atoms that seem to be assumed by ‘smallism’. One must be careful not to confuse Goff ’s theory with one which credits the cosmos with a mind of its own; Goff ’s cosmic ‘mind’ consists only of the sum total of conscious states in the universe; it has no further mental states or mental acts of its own. This creates problems. Goff, like Strawson and, I think, most contemporary panpsychists, accepts that there is cognitive phenomenology, that is, that thought, and not just sensory experience, has a phenomenology of its own, where this includes the thought contents, or meanings, and not just any sensory experiences associated with thinking. The cosmic mind, therefore, is not just a compendium of all creation’s sensory states, but all the thoughts that occur, together with their propositional contents. There is, in this intellectual mélange, much contradictory material, which the cosmic mind is thinking. This has been treated by some critics as an objection to the theory. Goff ’s reply is as follows. What about the point that the cosmos will involves states of conscious cognition with profoundly contradictory content? My reply to that would be: so what? Cosmopsychism does not entail pantheism. We need not think of the universe as a supremely intelligent rational agent. Intelligence and agency are characteristics of highly evolved conscious creatures, which the universe is not. It is more plausible that the consciousness of the universe is simply a mess. It may be hard for us to imagine a single mental state involving such wildly conflicting contents, but I see no reason to think that such a thing is impossible. (2017, 243)

The picture here is of a mental state which is rather like a multi-dimensional sense-field, though it includes thoughts. In so far as there is a subject aware of this

220

    

field, it is entirely passive. It understands, in some sense the contradictory thoughts going on, but has no capacity to recognize the contradiction, for that would make it something more than the sum of the mental states embodied in the universe, and so reinstate some kind of autonomous, transcendent mind. There are various problematic features in this conception. First, what constitutes the unity of this field, if there is no capacity to recognize its features? It is as if the field or arena itself is an entity, independent of any subject. This is subHumean theory of the mind, because, though in Hume the mind is just a bundle of events, the bundle includes reflections on and thoughts about other elements in the bundle. In the cosmopsychic case, this would expand the range of that mind beyond the mental contents of the first order entities. The suggestion that the higher order thoughts belong to ‘highly evolved creatures, which the universe is not’ seems to suggest that we can think in a way that the universe cannot, even though our minds are just part of the cosmic consciousness. So ‘the universe’ thinks ‘I have just contradicted myself ’ when I think that thing, but it cannot recognize that my thought that God is good contradicts yours that God does not exist because it has no comparative capacity of its own. It entertains both thoughts without having the second order capacity to compare them. This is in contradiction to Strawson’s claim that to be aware is to be aware that one is aware. I think that both ‘smallism’ and cosmopsychism fail in similar ways, namely that they have a radically inadequate conception of what a mind is. Their model reflects the crudest kind of empiricism, according to which mentality is essentially just a form of sensory consciousness and thought is either imagistic-associationist, or behavioural. This was certainly the conception of Russell and of William James who also inspired neutral monism. This conception allows one to entertain the thought that consciousness is the kind of thing that ranges from the vividness of human consciousness, down to minimal levels which we cannot imagine. This brackets out the commitment that modern panpsychists have to cognitive phenomenology. What is the primitive form of thought a quark possesses, what protoconcepts, given its lack of internal complexity? Strawson modifies this crude empiricist model by his insistence that consciousness involves being aware that one is aware, without seeing that this involves a reflective sophistication that seems beyond the range of quarks because it involves an inner complexity. Goff ’s cosmic mind, too, lacks the complexity that makes it a unified consciousness, which surely involves the ability to notice and compare elements in one’s mind.

12.4 Idealism without God A similar failure to appreciate the importance of complexity in mind applies in the case of Yetter-Chappell’s ‘idealism without God’. Her view, quoted above, is that the physical world is

        

221

a complex phenomenal unity, governed by laws of nature, structurally analogous to those materialists posit. This accounts for both the stability of the world external to us, and the regularity of our own experiences.

Every other characteristic of mind—‘desires, intentions, or beliefs. It needn’t be an agent’—is missing. Her preferred view is a Humean ‘bundle’ theory of the self, which makes the physical world a co-conscious collection of all the possible perceptions of the physical world. Even a Humean theory includes beliefs, desires, intentions, and acts of willing in the bundle, but these are omitted in YetterChappell’s theory. It is like imagining your visual field as a mind in its own right, without any of the accompanying cognitive or affective states. It could be true that this is the only part of God’s mentality that is needed to constitute (as opposed to being needed to sustain) the physical world we experience, but the idea that it could be a mind in its own right is another matter. It is also appropriate to mention that this is only one way of taking a Berkeleian understanding of how God sustains or constitutes the physical world. Berkeley says that God’s ideas of the world are archetypal, and, one might think, therefore, not simply like ours but more complete: they are in a more fundamental form. I interpret this as meaning that they take the form of the nomic structure of the plan for the experience of creatures. It is not as if God has all the experiences we have and the ones we miss. Yetter-Chappell also admits that the phenomena in this mind are ‘governed by laws of nature, structurally analogous to those materialists posit’, so these laws must have a source external to the godless phenomenal field that constitutes our physical world—unless they are a happy accident within that phenomenal field. This leads naturally to the next section, which concerns the need for an active God to explain and make sense of these laws.

12.5 God as the Source of the Laws of Nature I have argued that the notion of standardly physicalist notion of matter is deeply problematic, and that, therefore some mentalistic ontology is to be preferred. I have argued that the bare phenomenalism of a Humean style is unworkable, both because it stops the process of explanation at a totally implausible point, and because it relies on a Humean theory of causation, which I have argued is inadequate. The option of panpsychism fails because it has a conception of consciousness and mind that is radically oversimple. So we are left with the Berkeleian option. But this involves God and for many philosophers this is a red line they will not cross. I have no intention here of trying to prove the existence of God in any of the standard ways—even if possible, that would require another

222

    

book. But I will argue that we must postulate laws of nature, and that such laws require the existence of God. John Foster in The Divine Lawmaker (2004) argues that we cannot do without laws of nature and the laws such as these require a law-maker, which can only be God. Nancy Cartwright agrees with the second part of this—that laws can be fundamental only if there is a God to enforce them—but thinks that an ontology that ascribes powers or capacities to objects in what she thinks of as an Aristotelian manner can avoid having laws in nature itself: we formulate laws, which are, if we are lucky, grounded in the powers of things. This way one can avoid the horror of having to postulate God. I think that Foster is correct; laws are unavoidable, and, therefore, so is God. To show that he is correct one needs to refute the other accounts of natural law that are available. These are: (i) The Humean—or Hume/Lewis—account of laws as mere regularities. (ii) The ‘secular’ theories of laws, such as Armstrong’s. (iii) Cartwright’s so-called Aristotelian option, which sees laws as generalizations based on the powers that objects naturally possess. The first of these has already been discussed and rejected. The second has not been touched on so far. The third seems to have been adequately covered by the discussion of powers in Chapter 9, where I discussed both a pure powers ontology and the problems that arise if you try to combine powers and qualities; what is the ‘essence’ of which the powers are the natural expressions?. But why not accept laws of nature as a secular, naturalistic option? First, there is the problem of understanding what sort of thing nomic necessity is. Foster says: We are left, then, with a fundamental problem of how to make sense of the notion of a law of nature. On the one hand, laws (if such there be) are not just regularities of behaviour, but forms of natural necessity . . . On the other hand, we have also to accept that the necessity in question is not a form of strict necessity. For we have to accept that, for each law, there are compositionally relevant possible worlds [that is, worlds with the same quiddities] in which it does not obtain, and in which, in its absence, the behaviour it prescribes does not always occur. The difficulty is in seeing how we can provide a satisfactory account of laws on this basis – an account which explains how there can be this special kind of necessity which is both real necessity but not strict. Unless we can provide such an account, then the notion of a law of nature will have to be abandoned as incoherent. (2004, 92)

In other words, how can things be contingently necessary? (This is the main problem with Armstrong’s answer.) Second, what kind of entity is a law? If it is abstract, how can an abstract entity regiment concrete reality? But if it is a strange kind of concrete entity, it must

        

223

possess some further intrinsic properties or inner structure: surely this concrete picture of laws is nonsense. . . . if laws are concrete entities, they would have to have intrinsic natures in addition to their world-governing role. But the idea of a law having such a nature – for example, of its being a certain type of space-occupying object, with shape and size and internal character – is just a manifest absurdity. It is clearly essential to our conception of a law that its whole nature consists in its world-governing role . . . (2004, 92)

How else can laws, if treated as a basic category work, except through a law-giver and enforcer? It seems generally to be accepted that the category ‘laws of nature’ was introduced in the seventeenth century specifically as a notion with explicitly religious implications. Cartwright suggests—but not as her favoured option—a semi-Berkeleian (van Fraassen-ish) option. Instrumentalism offers an alternative [to regularity theory and to Platonism]. God uses the ‘laws of science in making things happen in the world in much the same way as we do, as an instrument for calculating what is to happen. The laws are indeed writ in Nature but not in questionable regularities of the empiricist or Platonist account. They are written in the more complex, untidy particulars of everything that does in fact happen. (Cartwright, 2005: 20)

Her preferred theory is to treat the powers and capacities that objects possess as being the grounding for laws, which are themselves far from universally reliable generalizations about these powers and capacities. But this approach takes the notion of causal powers as essentially unproblematic, whereas what exactly it for objects to possess such features is one of the things at issue: simply appealing to the notion will not do, in the context of our debate.² If pure regularity theory is a bizarre fantasy, leading to a generalized scepticism; and the introduction of powers as a grounding for law cannot be coherently made out, and laws as natural phenomena are not really intelligible, then the Divine Lawmaker is the only rational hypothesis.

² It is also worth noting that the active powers invoked by philosophers who like to call themselves ‘Aristotelian’ are not Aristotelian powers at all. In Aristotle a power is, in the physical world, a potentiality to be affected in some way—it is a passivity. This ambiguity passes over into modern talk. A ‘passion’ is a dynamic impulse, but it is also a receptivity to being affected. Active powers in nature are in Locke, but not, or almost not, in Aristotle.

224

    

12.6 Conclusion Yetter-Chappel’s idealist, Pelczar’s phenomenalist, and the panpsychist all believe that the physical world is an actual, or potential, phenomenal array sustained by a mind or minds that lack any further character than their ability to passively experience the contents of their experiences, which, in some way, sustain the laws of physics. I have tried to show that, once one has reached some form of phenomenalism, a fuller and more intelligent sustaining mental background is required. There must be a law-maker for the laws that govern experience, and minds, thus, must be more sophisticated than passive screens for the phenomena, especially if the phenomena are complex and unified.

General Conclusion Part I was dedicated to defending what I deem to be a version of a sense-datum theory of perception. I claim that a theory is a sense-datum theory if it claims that qualities are actually present in, and not merely represented in, phenomenal content. Naïve realism (often now called ‘relationalism’), fails, I argue, because it is forced to have recourse to strange and counterintuitive denials of phenomenal content in ‘philosophers’ hallucinations (Chapter 1); and has nothing non-elusive to say about what philosophers call ‘illusions’ (Chapter 2). I claim that intentional theories, given the ‘blocking argument’, fail to be significantly different from sense-datum theories. In this, I agree with their naïve realist critics (Chapter 3). Their attempt to avoid this consequence rests entirely on the way they try to handle the difference between the content and the object of experience. None of the ways they try to deploy this distinction will accomplish the work they require (Chapter 4). In fact, the directness of perception can be understood in terms of three factors. First, there is the inevitable sense of objectivity that arises from what Hume called the constancy and coherence of experience (Chapter 6). Second it resides in the judgemental, or consciously informational, content implicit in perceptual experience (Chapter 7). Third it arises because the vehicles for this information—the coherently organized sensible phenomena, or sense-data—are themselves the accurate and appropriate ways that objects manifest themselves to creatures like us (Chapter 8). This overall conclusion is important because it combines the idea that how the world appears to us depends on the nature of our senses, with the idea that it is, indeed, the objects themselves that we perceive, through appearing in the way that they manifest themselves to creatures of our kind. It is further important because it leaves room for the question of what is the nature of physical objects themselves. The physical world must, indeed be such that how it appears to us is appropriate to it, otherwise we could not navigate our way through the world, but what the reality is that sustains these appearances is still, at this point, an open question. The view that the secondary qualities had no role to play in the physical world became established in the seventeenth century, and Locke, in his sympathetic consideration of Molyneux’s problem, indicated that visual and tactile spaces might be qualitatively different, clearing the way for later philosophers, for example, Russell, to claim that even primary qualities in the external world need

Perception and Idealism: An Essay on How the World Manifests Itself to Us, and How It (Probably) Is in Itself. Howard Robinson, Oxford University Press. © Howard Robinson 2022. DOI: 10.1093/oso/9780192845566.003.0014

226

 

have no more than a structural or mathematical similarity to the qualitative nature of their appearance. Berkeley’s idealist interpretation of the reality that lies behind experience seems like a radical development of this thought. My purpose in Part II is to suggest that the idealist solution is much less counterintuitive than it seems at first sight. Space, and its occupants, are the two basic features of the physical world. I have argued, with the help of Berkeley and Foster, that the nature of physical space cannot be severed from its empirical manifestation (Chapter 11). And that the modern conception of the matter that occupies space either leads directly to phenomenalism (as with the powers conception) or, in the case of quantum theory, is, at best, is at least as counterintuitive as Berkeley’s theory (Chapter 9). En Passant, I have directed sympathetic attention to the Berkeleian idea that the very notion of a quality—without which there can be no contentful world at all—is essentially experiential: and, with less conviction, I canvass his claim that our conception of the world has the lineaments of thought built into it (Chapter 10). I finally consider non-Berkeleian forms of mentalism, such as straight phenomenalism and panpsychism, and find them significantly less satisfactory than Berkeleian theistic idealism (Chapter 12). Part I is independent of the conclusions of Part II, but it is a necessary prolegomenon. I have some hope of convincing people of my thoughts on perception, and hope, at least, to raise worries about the metaphysics of physical reality, if not wholly convince the reader, by the arguments in Part II.

Bibliography Ackrill, J. L. (ed.) (1987) A New Aristotle Reader, Oxford: Oxford University Press. Adams, R. (2007) ‘Idealism vindicated’, in D. Zimmerman and P. van Inwagen (eds), Persons; Human and Divine, Oxford, Clarendon Press, 35–54. Albert, D. (2013) ‘Wave function realism’, in A. Ney and D. Albert (eds), The Wave Function: Essays on the Metaphysics of Quantum Mechanics, Oxford: Oxford University Press, 52–7. Allen, K. (2015) ‘Hallucination and imagination’, Australasian Journal of Philosophy, 93, 287–302. Allori, V. (2013) ‘Primitive ontology and the structure of fundamental physical theories’, in A. Ney and D. Albert (eds), The Wave Function: Essays on the Metaphysics of Quantum Mechanics, Oxford: Oxford University Press, 58–75. Allori, V. (forthcoming) ‘On the metaphysics of quantum mechanics’, in S. le Bihan (ed.), La philosophie dela physique: d’aujourd’hui a demain, Editions Vuibert. Aquinas, T. (1948) Summa Theologiae, trans L. Shapcote, O.P., London: Burnes and Oates. Armstrong, D. (1968) A Materialist Theory of the Mind, London, Routledge and Kegan Paul. Armstrong, D. (2004) Truth and Truthmakers, Cambridge: Cambridge University Press. Austin, J. L. (1962) Sense and Sensibilia, Oxford: Clarendon Press. Ayer, A. J. (1963) ‘Privacy’, in Concept of a person, London: Macmillan, 52–81. Previously in Proceedings of the British Academy, 1959. Barnes, J. (1971) The Ontological Argument, London, Palgrave Macmillan. Bennett, J. (1971) Locke, Berkeley, Hume, Oxford: Clarendon Press. Berkeley, G. (1710/34) The Principles of Human Knowledge. In many modern editions, including Principles and Three Dialogues for Oxford University Press’s World Classics, 1996, ed H. Robinson. Latest reprint, 2009. Blackburn, S. (1990) ‘Filling in space’, Analysis (special half-centenary volume), April 1990. Bradley, F. H. (1897) Appearance and Reality, Oxford: Clarendon Press. Bradley, J. ed. (1996) Philosophy after F. H. Bradley, Bristol: Theommes Press. Brentano, F. (1977) The Psychology of Aristotle. Berkeley: University of California Press. Brentano, F. (1995) Psychology from an Empirical Standpoint, London: Routledge. Brewer, B. (2011) Perception and its Objects, Oxford: Oxford University Press. Burge, T. (1986) ‘Cartesian error and the objectivity of perception’, in P. Petit and J. McDowell (eds), Subject, Thought and Content, Oxford: Clarendon Press, 117–36. Burge, T. (2003) ‘Tyler Burge replies’, in M. Hahn and B. Ramberg (eds), Reflections and Replies: Essays on the Philosophy of Tyler Burge, Cambridge: MIT Press, 289–470. Burge, T. (2005) ‘Disjunctivism and perceptual psychology’, Philosophical Topics 33, 1–78. Burge, T. (2007) Foundations of Mind, Oxford: Oxford University Press. Burge, T. (2010) The Origins of Objectivity, Oxford: Oxford University Press. Campbell, J. (2002a) ‘Berkeley’s Puzzle’, in T. S. Gendler and J. Hawthorne (eds), Conceivability and Possibility, Oxford: Oxford University Press, 127–43. Campbell, J. (2002b) Reference and Consciousness, Oxford: Oxford University Press.

228



Cartwright, N. (2005) ‘No God; no laws’, in E. Sindoni and S. Moriggi (eds), Dio, la Natura e la Legge. God and the Laws of Nature. Milan: Angelicum-Mondo X. Cassam, Q. (2014) Berkeley’s Puzzle: What Does Experience Teach Us? Co-authored with John Campbell, Oxford: Oxford University Press. Chalmers, D. J. (2003) ‘Consciousness and its place in nature,’ in Stephen P. Stich and Ted A. Warfield (eds.), Blackwell Guide to the Philosophy of Mind, Blackwell, 102–142. Child, W. (1994) Causality, Interpretation, and the Mind, Oxford: Oxford University Press. Chisholm, R. (1957) Perceiving: a Philosophical Study, Ithaca, NY: Cornell University Press. Clark, A. and Chalmers, D. (1998) ‘The extended mind’, Analysis vol. 58, no. 1, 7–19. Coates, P. (2007) The Metaphysics of Perception: Wilfrid Sellars, Perceptual Consciousness and Critical Realism, London: Routledge. Craig, F. and Phillips, I. (2020) ‘Austerity and illusion’, Philosophers’ Imprint, vol. 20, 1–19. Craig, W. L. and Smith, Q. (1993) Theism, Atheism and Big Bang Cosmology, Oxford: Clarendon Press. Crane, T. (1992) The Contents of Experience, Cambridge, Cambridge University Press. Crane, T. (2001) Elements of Mind, Oxford: Oxford University Press. Crane, T. (2005) ‘What is the problem of perception?’, Synthesis Philosophica, 20, 237–64. Crane, T. (2006) ‘Is there a perceptual relation?’, in T. Gendler and J. Hawthorne (eds), Perceptual Experience, Oxford: Oxford University Press, 126–46. Dainton, B. (2001) Time and Space, Chesham: Acumen. Dennett, D. (1991) Consciousness Explained, London: Penguin Press. Dretske, F. (2000) Perception, Knowledge and Belief, Cambridge: Cambridge University Press. Dummett, M. (2006) Thought and Reality, Oxford: Clarendon Press. Einstein, A. (1922) ‘Ether and the theory of relativity’, in A. Einstein, Sidelights on Relativity, London: Methuen, 1–24. Einstein, A. (1934) ‘The problem of space, ether, and the field in physics’, in A. Einstein, The World as I See It, New York: Covici-Friede. Esfeld, M. (2020) ‘A proposal for a minimalist ontology’, Synthese, Vol. 197; 1889–1905. Evans, G. (1982) The Varieties of Reference, Oxford: Clarendon Press. Ewing, A. C. (1934) Idealism: a Critical Survey, London: Methuen. Farkas, K. (2013) ‘Constructing a world for the senses’, in U. Kriegel (ed.), Phenomenal Intentionality, Oxford: Oxford University Press, 99–115. ffytche, D. (2013) ‘The hallucinating brain: neurobiological insights into the nature of hallucinations’, in F. Macpherson and D. Platchias (eds), Hallucination: Philosophy and Psychology, Cambridge, MA: MIT Press, 45–63. Fish, W. (2009) Perception, Hallucination, and Illusion. Oxford: Oxford University Press. Foster, J. (1982) The Case for Idealism, London: Routledge and Kegan Paul. Foster, J. (1993) ‘The succinct case for idealism’, in H. Robinson (ed.), Objections to Physicalism, Oxford: Clarendon Press, 293–313. Foster, J. (2000) The Nature of Perception, Oxford: Oxford University Press. Foster, J. (2004) The Divine Lawmaker, Oxford: Oxford University Press. Foster, J. (2008) A World for Us, Oxford: Oxford University Press. Foster, J. and Robinson, H. (eds) (1985) Essays on Berkeley, Oxford: Clarendon Press. Gelman, S. (2003) The Essential Child: Origins of Essentialism in Everyday Thought, Oxford: Oxford University Press. Genone, J. (2014) ‘Appearance and illusion’, Mind 123: 339–76. Goff, P. (2007) Consciousness and Fundamental Reality, Oxford: Oxford University Press.



229

Goff, P. (2017) Consciousness and Fundamental Reality, Oxford: Oxford University Press. Goff, P. (2020) ‘Panpsychism and free will’, Proceedings of the Aristotelian Society. Goldschmidt, T. and Pearce, K. (eds) (2017) Idealism: New Essays in Metaphysics, Oxford: Oxford University Press. Hardin, C. L. (1988) Color for Philosophers, Indianapolis: Hackett. Harre, R. (1970) Principles of Scientific Thinking, London: Macmillan. Harre, R. (1986) Varieties of Realism, Oxford: Blackwell. Harre, R. and Madden, E. (1975) Causal Powers, Oxford: Blackwell. Harman, G. (1990) ‘The intrinsic quality of experience’, Philosophical Perspectives 4, Action Theory and the Philosophy of Mind, 31–52, Atascadero, Ridgeview Publishing Company. Heil, J. (2010) ‘Powerful qualities’, in A. Marmodoro (ed.), The Metaphysics of Powers: Their Grounding and their Manifestations, London: Routledge, 58–72. Hinton, J. M. (1973) Experiences, Oxford: Clarendon Press. Hofweber, T. (2016) Ontology and the Ambitions of Metaphysics, Oxford University Press. Hofweber, T. (2017) ‘Conceptual idealism without ontological idealism: why idealism is true after all’, in T. Goldschmidt and K. Pearce (eds) (2017) Idealism: New Essays in Metaphysics, Oxford: Oxford University Press, 124–41. Hume, D. (1739/2000) A Treatise of Human Nature, eds D. F. Norton and M. J. Norton, Oxford, Oxford University Press. Husserl, E. (2001) Logical Investigations, vol. 2, London: Routledge. Irwin, T. H. (1990) Aristotle’s First Principles. Oxford: Clarendon Press. Jackson, F. (2004) ‘Mind and illusion’, in P. Ludlow, Y. Nagasawa, and D. Stoljar (eds), There Is Something about Mary: Essays on Phenomenal Consciousness and Frank Jackson’s Knowledge Argument, Cambridge, MA: MIT Press. Johnston, M. (2002) ‘The Obscure Object of Hallucination’, Philosophical Studies 120: 113–183. Johnston, M. (2004) ‘The obscure object of hallucination’, Philosophical Studies, 120: 113–183. Kalderon, M. (2011) ‘Color illusion’, Nous 45: 751–75. Lewis, C. I. (1929) Mind and the World Order, London: Charles Scribner’s Sons. Lewis, D. (2009) ‘Ramseyan humility’, in D. Braddon-Mitchell and R. Nola (eds), Conceptual Analysis and Philosophical Naturalism, Cambridge, MA: MIT Press, 203–22. Lockwood, M. (1971) ‘Identity and reference’, in M. Munitz (ed.), Identity and Individuation, New York: New York University Press. Lockwood, M. (1989) Mind, Brain and Quantum: The Compound ‘I’, Blackwell: Oxford. Ludlow, P., Nagasawa, Y., and Stoljar, D. (eds) (2004) There is Something about Mary: Essays on Phenomenal Consciousness and Frank Jackson’s Knowledge Argument, Cambridge MA, MIT Press. Lycan, W. (1987) Consciousness, Cambridge MA: MIT Press. Mackie, J. L. (1969) ‘What’s really wrong with phenomenalism?’ Proceedings of the British Academy, 55: 113–127. Mackie, J. (1976) Problems from Locke, Oxford: Clarendon Press. Martin, C. B. (1997) ‘On the need for properties: the road to Pythagoreanism and back’, Synthese 112: 193–231. Martin, M. G. F. (1997) ‘The reality of appearances’, in M. Sainsbury (ed.), Thought and Ontology, Milan: Franco Angeli, 77–96. Reprinted in A. Byrne and H. Logue (eds) (2008) Disjunctivism; Contemporary Readings, Cambridge MA: MIT Press. Martin, M. G. F. (2002) ‘The transparency of experience’, Mind & Language 376–425. Martin, M. G. F. (2004) ‘The limits of self-awareness’, Philosophical Studies 120: 37–89.

230



Martin, M. G. F. (2006) ‘On being alienated’, in T. S. Gendler and J. Hawthorne (eds), Perceptual Experience, Oxford: Oxford University Press, 354–410. Martin, M. G. F. (2010) ‘What’s in a look’, in B. Nanay (ed.), Perceiving the World, New York: Oxford University Press, 160–225. Maudlin, T. (2013) ‘The nature of the quantum state’, in A. Ney and D. Albert (eds), The Wave Function: Essays on the Metaphysics of Quantum Mechanics, Oxford: Oxford University Press, 126–53. Maxwell, G. (1978) ‘Rigid designators and mind-brain identity’, in C. W. Savage (ed.), Perception and Cognition, Minneapolis: University of Minnesota Press, 365–403. McDowell, J. (1986) ‘Singular thought and the extent of inner space’, in P. Petit and J. McDowell (eds), Subject, Thought and Context, Oxford: Clarendon Press. [Reprinted in J. McDowell, Meaning, Knowledge and Reality, Cambridge, MA: Harvard University Press, 1998. Page references in this chapter are to the reprint.] McDowell, J. (1994) Mind and World, Cambridge, Mass.: Harvard University Press. McDowell, J. (2008) ‘The disjunctive conception of experience as material for a transcendental argument’, in A. Haddock and F. Macpherson (eds), Disjunctivism, Oxford: Oxford University Press, 376–89. Mill, J. S. (1949) A System of Logic, London: Longmans. Mill, J. S. (2020) A System of Logic, Ratiocinative and Inductive, Outlook Verlag. Molnar, G. (2003) Powers: a Study in Metaphysics, Oxford: Oxford University Press. Morris, T. V. (1984) Understanding Identity Statements, Aberdeen: Aberdeen University Press. Mumford, S. (1998) Dispositions, Oxford: Oxford University Press. Ney, A. and Albert, D. (eds) (2013) The Wave Function: Essays on the Metaphysics of Quantum Mechanics, Oxford: Oxford University Press. Noë, A. (2006) ‘Experience without the head’, in Perceptual Experience, eds Gendler and Hawthorne, Oxford University Press, 411–33. Nudds, M. (2009) ‘Recent work in perception: Naïve realism and its opponents’, Analysis, 69(2), 334–46. O’Connor, T. and Robb, D. (2003) Routledge Companion to the Philosophy of Mind, London: Routledge; Oxford: Oxford University Press. Co-authored with John Campbell. Pautz, A. (2021) Perception, London: Routledge. Peacocke, C. (1983) Sense and Content, Oxford: Oxford University Press. Peacocke, C. (1985) ‘Imagination, experience and possibility: A Berkeley view defended’, in Foster and Robinson (1985). Peacocke, C. (2001) ‘Does perception have a non conceptual content?’, Journal of Philosophy, 98: 239–64. Pelczar, M. (2015) Sensorama: A Phenomenalist Analysis of Spacetime and Its Contents, Oxford: Oxford University Press. Pelczar, M. (2019) ‘Defending phenomenalism’, Philosophical Quarterly 69: 574–97. Pitcher, G. (1970) A Theory of Perception, Princeton: Princeton University Press. Plantinga, A. (1974) The Nature of Necessity. Oxford: Clarendon Press. Price, H. H. (1932) Perception, London: Methuen. Putnam, H. (1981) ‘Brains in a vat’, in his Reason, Truth and History, Cambridge, Cambridge University Press, 1–21. Recanati, F. (2012) Mental Files, Oxford: Oxford University Press. Robinson, H. (1982) ‘Matter: turning the tables’, in H. Robinson, Matter and Sense, Cambridge: Cambridge University Press, ch. 7. Reprinted in T. Crane and K. Farkas (2004), Oxford: Oxford University Press, 107–19.



231

Robinson, H. (1985) ‘The general form of the argument for Berkelian idealism’, in J. Foster and H. Robinson (eds), Essays on Berkeley, Oxford: Clarendon Press, 163–86. Reprinted in T. O’Connor and D. Robb (eds) (2003) Routledge Companion to the Philosophy of Mind, London: Routledge, 81–102. Robinson, H. (ed.) (1993) Objections to Physicalism, Oxford: Clarendon Press. Robinson, H. (1994) Perception, London: Routledge. Robinson, H. (ed.) (1996) Berkeley’s Principles and Three Dialogues. Oxford University Press’s World Classics. Robinson, H. (2003) ‘The ontology of the mental’, in M. Loux and D. Zimmerman (eds), The Oxford Handbook of Metaphysics, Oxford: Oxford University Press. Robinson, H. (2008) ‘Why Frank should not have jilted Mary’, in E. Wright (ed.), The Case for Qualia, Cambridge, MA: MIT Press. Robinson, H. (2009) ‘Idealism’, The Oxford Handbook to the Philosophy of Mind, ed. B. McLaughlin, A. Beckermann, S. Walter, Oxford University Press, 2009, 186–205. Robinson, H. (2011) ‘Two Berkelian arguments about the nature of space’, in T. Airaksinen and B. Belfrage (eds), Berkeley’s Lasting Legacy: Three Hundred Years Later, Cambridge: Cambridge Scholars Press, 79–90. Robinson, H. (2013) ‘The failure of disjunctivism to deal with “philosophers’ hallucinations” ’, in Fiona McPherson and Dimitris Platchias (eds), Hallucinations, Cambridge MA: MIT Press, 313–30. Robinson, H. (2016) From Knowledge Argument to Mental Substance, Cambridge, Cambridge University Press. Rubio, D and Chen, E. (unpublished) Evil in a Quantum Multiverse. Russell, B. (1912) Problems of Philosophy, Oxford University Press, 1967: originally published 1912. Russell, B. (1927) Analysis of Matter, London: George Allen and Unwin. Rynasiewicz, R. (1996) ‘Absolute versus relational space-time: an outmoded debate?’, Journal of Philosophy, 93: 279–306. Sainsbury, M. (2005) Reference without Referents, Oxford: Clarendon Press. Schellenberg, S. (2011) ‘Ontological minimalism about phenomenology’, Philosophy and Phenomenological Research 83: 1–40. Schellenberg, S. (2013) ‘Experience and evidence’, Mind 122 (487): 699–747. Schellenberg, S. (2018) The Unity of Perception: Content, Consciousness, Evidence, Oxford: Oxford University Press. Searle, J. (1983) Intentionality: An Essay in the Philosophy of Mind, Cambridge: Cambridge University Press. Searle, J. (2015) Seeing Things as They Are: A Theory of Perception, Oxford: Oxford University Press. Siegel, S. (2010) ‘Do visual experiences have content?’, in in B. Nanay (ed.), Perceiving the World, Oxford: Oxford University Press. Smart, J. J. C. (1963) Philosophy and Scientific Realism, London: Routledge and Megan Paul. Smith, A. D. (2002) The Problem of Perception, Cambridge MA: Harvard University Press. Snowdon, P. (1992) ‘How to interpret “direct perception” ’, in T. Crane (ed.), The Contents of Experience, Cambridge: Cambridge University Press, 48–78. Snowdon, P. (2005) ‘Some reflections on an argument from hallucination’, Philosophical Topics 33: 285–305. Soteriou, M. (2016) Disjunctivism, London and New York: Routledge. Sprigge, T. (1983) The Vindication of Absolute Idealism, Edinburgh: Edinburgh University Press.

232



Sprigge, T. (1993) James and Bradley: American Truth and British Reality, Illinois: Open Court. Strawson, P. (1950a) ‘On referring’, Mind 59: 320–44. Strawson, P. (1950b) ‘Truth’, Proceedings of the Aristotelian Society, reprinted in P. Strawson (1971), Logico-Linguistic Papers, London, Methuen. Strawson, P. (1974) Subject and Predicate in Logic and Grammar. London: Methuen. Strawson, G. (1994) Mental Reality, Cambridge, MA: MIT Press. Strawson, G. (2006) ‘Realistic monism: why physicalism entails panpsychism’, in A. Freeman (ed.), Consciousness and Its Place in Nature: Does Physicalism Entail Panpsychism?, 3–31. Strawson, G. (2017) The Subject of Experience, Oxford University Press. Travis, C. (2004) ‘The silence of the senses’, Mind 113: 57–94. Weinberg, J. (1965) ‘Relation’, in J. Weinberg, Abstraction, Relation and Induction, Madison: University of Wisconsin Press. Williams, N. (2019) The Powers Metaphysic. Oxford: Oxford University Press. Wittgenstein, L. (1951) Philosophical investigations, trans. G. E. M. Anscombe, Oxford: Blackwell. Wright, C. (2008) ‘Comment on John McDowell’s “The Disjunctive Conception of Experience as Material for a Transcendental Argument” ’, in A. Haddock and F. Macpherson (eds), Disjunctivism, Oxford: Oxford University Press. Yetter-Chappel, H. (2017) ‘Idealism without God’, in T. Goldschmidt and K. Pearce (eds), Idealism: New Essays in Metaphysics, Oxford: Oxford University Press, 66–81.

Index For the benefit of digital users, indexed terms that span two pages (e.g., 52–53) may, on occasion, appear on only one of those pages. Adams, R. 174–8, 215, 218–19 Albert, D. 166–9 Alford-Duguid, D. 15n.4 Allen, K. 30n.11 Allori, V. 172–3 Armstrong, D. 24n.9, 45, 156–7 On ‘States of affairs’ 181 On ‘laws of nature’ 222 Aquinas, St. T. 82 Aristotle. 81–2, 129, 185 “austere’ naïve realism 40–2 Austin, J. 122 Ayer, A. 92, 110, 185–6 Bird, A. 154–5 Blackburn, S. 152–3 Blocking function 58–61 Bohm, D. 166–9 Bohr, N. 167 Boscovich, Fr R., (S. J.) 3, 148 Brentano, F. 51, 63–6, 81–2 Brewer, B. 42–5, 91, 122–3 Burge, T. 91, 97–105, 111–12 Campbell, J. 17, 22, 42–5, 56–7, 59, 91, 95–7, 124, 179 Cartwright, N. 222–3 Cassam, Q. 92, 109, 121 Chalmers, D. 6–7, 213–14 Charles Bonnet Syndrome 30n.11 Chen, E. 170–1 Child, W. 22, 112–13 Chisholm, R. 121n.5 Clark, A. 139–40 Coates, P. 123–4 Common factor theory 15, 22 Crane, T. 51–63, 70–1, 79 Critical realism 123–5 Dainton, B. 200 Dennett, D. 8 Direct realism; varieties of 116–17 Generic Direct Realism (GDR) 116

Phenomenal Direct Realism (PDR) 116 Semantic direct realism 116 Also Chapter 7 in toto. Disjunctivism. 11, 35–40 Drake, D. 123 Dretske, D. 69, 73 Dummett, M. 175–8 Einstein, A. 195–7 Esfeld, M. 161–2 Matter points 161–2 Ewing, A. 186n.3 Farkas, K. 92, 106, 109 ffytche, D. 30 Fish, W. 16, 20–2, 26–8, 37, 41, 44–7 Foster, J. 18, 156–7 On topology of physical space 198–205 Against Humean causation and defence of induction 163–4 On Divine lawmaker 221–3 French, C. 40–2 Frasier, R. 127n.1 Genone, J. 47–8 Goff, P. 213, 215–20 Grice, P. 12 Hallucinations – non-arbitrariness 18–21 Hardin, C. 5–6 Harré, R. 148, 157–9 Heil, J. 159–60 Hinton, J. 22 Hofweber, T. 180n.1 Hume, D. 1, 6–7, 92, 102–5, 107–14, 138, 204–5 Husserl, E. 63–6 James, W. 220 Jaworski, W. 129n.2 Johnston, M. 18, 69, 74–6, 78 Kalderon, M. 41, 43, 45–7 Kantian idealism 186, 206

234



Leibniz, G. 194–5 Lewis, C. I. 4–5 Lewis, D. 160–1 Locke, J. 1–2 Lockwood, M. 84, 213 Loewer, B. 168–9 Lorentz, H. 197 Lycan, W. 69, 73 Mach, E. 193 Mackie, J. 208–9 Marr, D. 136 Martin, C. 154, 158 Martin, M. 16, 18, 20–6, 35–9, 43, 50n.1, 113–14, 142, 201n.9 Maudlin, T. 172–3 Maxwell, G. 213 McDowell, J. 22, 91–6, 124, 131–2 Mental files 84–7 Molyneux’s Problem 134n.5 Moran, A. 13n.3 Mill, J. S. 1, 208–13 And direct reference 82–4 Molnar, G. 153–4 Mumford, S. 153 Neo-platonism 186 Newton, I. and absolute space 192–8 Ney, A. 172–3 Noë, A. 136, 138–40 Nudds, M. 66–7 Pappineau, D. 105–7, 109 Pautz, A. 7n.1, 110–11 Peacocke, C. 43, 60–1, 179 Pelczar, M. 1, 207–13 Penrose, R. 166 Perspectivalism 42–7 Phenomenal principle 33–4 Phenomenalism (Hume-Mill style) 208–13 Philips, I. 40–2 Pitcher, G. 22 Plantinga, A. 85–6 Price, H. 7–8 Putnam, H. 205 Qualia 4–5, 70–1, 103 Quantum theory 165–73 Bohmian interpretation 166–9 Copenhagen interpretation 167

Ghiradi, Rimini, Weber (GRW) theory 171 Many Worlds 170–1 Schroedinger’s cat 166–7 And idealism 172–3 Quiddities 160–2 Radar – not like perception 126–7 Recanati, P. 84 Russell, B. 1, 213–14 Rynasiewicz, R. 195–6 Schellenberg, S. 59–60, 69–70, 72, 76–9, 117–22 Searle, J. 51–8, 70, 79 Sellars, R. 123 Sellars, W. Manifest world 102–3, 105–7 Scientific image 102–3 Sense-data 4–8, 33n.1, 115–16 Sense-data and SDR 125 Sense-data not an. Error theory 132–4 Sense-data and depth perception 132–3 Siegel, S. 130–1, 136–7 Smart, J. 23–5 On nomological danglers 151–2 Smith, A. D. 69–72, 124, 128 Snowdon, P. 12n.2, 18, 34–5, 37 Solidity 148 Soteriou, M. 20–2, 28–30 Sprigge, T. 181, 186n.3 Strong, C. 123 Strawson, G. 213, 218–20 Weather watchers 139 Strawson, P. 84, 179–81 Transparency 66–8 Travis, C. 128 Truthmaker theory 178–86 Tye, M. 44, 66, 69 Wallace, D. 171 Weinberg, J. 185 Wigner, E. 167 Williams, N. 155–6 Wittgenstein, L. 21 World maps 84–6, 90 Wright, C. 93–4 Yetter-Chappell, H. 207–8, 220–1