Fabian Dorsch The Unity of Imagining

P h i l o s o p h i s c h e F o r s c h u n g P h i l o s o p h i c a l R e s e a r c h

Herausgegeben von / Edited by

Johannes Brandl • Andreas Kemmerling Wolfgang Künne • Mark Textor

Band 9 / Volume 9

Fabian Dorsch

The Unity of Imagining

Bibliographic information published by the Deutsche Nationalbibliothek The Deutsche Nationalbibliothek lists this publication in the Deutsche Nationalbibliographie;

detailed bibliographic data is available in the Internet at http://dnb.ddb.de

North and South America by Transaction Books Rutgers University

Piscataway, NJ 08854-8042 trans@transactionpub.com

United Kingdom, Ireland, Iceland, Turkey, Malta, Portugal by Gazelle Books Services Limited

White Cross Mills Hightown

LANCASTER, LA1 4XS sales@gazellebooks.co.uk

Livraison pour la France et la Belgique: Librairie Philosophique J.Vrin

6, place de la Sorbonne ; F-75005 PARIS Tel. +33 (0)1 43 54 03 47 ; Fax +33 (0)1 43 54 48 18

www.vrin.fr

2012 ontos verlag P.O. Box 15 41, D-63133 Heusenstamm

www.ontosverlag.com

ISBN: 978-3-86838-174-0

2012

No part of this book may be reproduced, stored in retrieval systems or transmitted in any form or by any means, electronic, mechanical, photocopying, microfilming, recording or otherwise

without written permission from the Publisher, with the exception of any material supplied specifically for the purpose of being entered and executed on a computer system, for exclusive use of the purchaser of the work

Printed on acid-free paper ISO-Norm 970-6

FSC-certified (Forest Stewardship Council)

This hardcover binding meets the International Library standard

Printed in Germany by CPI buch bücher.de GmbH

Table of Contents

Detailed Table of Contents ix

Acknowledgements xv

Introduction 1

Part One ‒ The Nature and Variety of Imagining

Introduction to Part One 23

1 Unified Accounts of Imagining 27

2 Key Features of Imaginative Episodes 59

3 Key Features of Sensory Imaginings 79

4 Imagination and Knowledge 115

5 Other Theories of Imagining 149

Part Two ‒ The Epistemological Account

Introduction to Part Two 191

6 O’Shaughnessy’s View 197

7 O’Shaughnessy’s Arguments 221

8 Critical Assessment 253

viii

Part Three ‒ The Dependency Account

Introduction to Part Three 275

9 The Representational Account 281

10 Visual Imagining As Experiential Imagining 313

11 Emotional Imagining As Experiential Imagining 337

12 Semantic Dependency, Simulation, and Pretense 365

Part Four ‒ The Agency Account

Introduction to Part Four 381

13 Imaginative Agency 385

14 Meeting the Desiderata 409

Conclusion 431

Bibliography 443

Index 455

Detailed Table of Contents

Acknowledgements xv

Introduction 1(i) The Need for a Unified Account 5(ii) Three Main Theories of Imagining 7(iii) Other Theories of Imagining 11(iv) The Structure of the Book 17

Part One ‒ The Nature and Variety of Imagining

Introduction to Part One 23

1 Unified Accounts of Imagining 271.1 Two Desiderata 281.2. Five Central Cases of Imagining 32

(i) Sensory Imagining 37(ii) Affective Imagining 41(iii) Intellectual Imagining 42(iv) Experiential Imagining 44(v) Imaginative Projects 46

1.3. Mental Episodes and Complexes Thereof 471.4. Non-Central Cases of Imagining 51

(i) Imaginative Dispositions 52(ii) Dream Representations 53(iii) Unusual or Pathological Cases 54(iv) Fanciful Expectation, Anticipation, Etc. 55

x

2 Key Features of Imaginative Episodes 59

2.1. Imaginative Attitude 61

2.2. No Epistemic Function 68

2.3. Reference to Reality 73

3 Key Features of Sensory Imaginings 79

3.1. Diminished Vivacity 81

3.2. No Immediacy 83

3.3. Indeterminacy 85

3.4. Quasi-Observationality 87

3.5. Immunity to Error through Misidentification 94

3.6. Intellectuality 96

(i) Peacocke’s View 96

(ii) Sartre’s View 100

(iii) Imaginative Thought vs. Intention-in-Action 103

(iv) Explanatory Power 106

4 Imagination and Knowledge 115

4.1. Two Examples of Cognising Imaginings 116

4.2. The Absence of Alternative Grounds 122

4.3. Similarities to Perception (and Memory) 126

4.4. Differences from Perception (and Memory) 128

4.5. Imagination and Information 133

4.6. ‘Spontaneity’, Attention, and ‘Filling-In’ 138

4.7. Voluntariness and Cognitive Constraints 142

5 Other Theories of Imagining 149

5.1. The Quantitative Account 150

5.2. The Phenomenal Account 152

5.3. The Ontological Account 154

5.4. The Modal Account 161

5.5. The Attitude Account 168

5.6. The Spontaneity Account 176

xi

Part Two ‒ The Epistemological Account

Introduction to Part Two 191

6 O’Shaughnessy’s View 197

6.1. A Brief Synopsis 199

6.2. The Three Main Claims 202

6.3. The Argumentative Strategy 213

7 O’Shaughnessy’s Arguments 221

7.1. The Argument from Origin 222

(i) Standard Constraints on Cognitions 223

(ii) Wakefulness as a Further Constraint 232

(iii) The Origin of Imaginings 242

7.2. The Argument from Attention 247

8 Critical Assessment 253

8.1. The Explanatory Power of Negation Claims 254

(i) General Limitations 254

(ii) The Limitations of (NC*) 256

(iii) The Limitations of (NI) and Similar Claims 258

(iv) The Limitations of (NC) 264

8.2. The Extensional Adequacy of Negation Claims 266

Part Three ‒ The Dependency Account

Introduction to Part Three 275

9 The Representational Account 281

9.1. Hume’s Copy Principle 282

9.2. Objections to Hume’s View 290

9.3. Causal Echo Claims 292

9.4. Representational Echo Claims 294

9.5. Intellectual Imagining 302

9.6. Imaginative Projects 309

xii

10 Visual Imagining As Experiential Imagining 313

10.1. Martin’s Argument: Perspectivalness 314

10.2. A Second Argument: Egocentric Orientation 322

10.3. Replies to Objections 325

10.4. The Nature of Experiential Imagining 332

11 Emotional Imagining As Experiential Imagining 337

11.1. The Puzzle of Fiction 340

11.2. Walton’s Account of Emotional Imagining 346

11.3. Moran on the Various Forms of Imagining 351

11.4. Moran’s Account of Emotional Imagining 356

11.5. The Experiential Nature of Emotional Imagining 360

12 Semantic Dependency, Simulation, and Pretense 365

12.1. Semantic Dependency 366

12.2. Imagination as Simulation 368

12.3. Imagination as Pretense 373

12.4. Conclusion and Outlook 377

Part Four ‒ The Agency Account

Introduction to Part Four 381

13 Imaginative Agency 385

13.1. Mental vs. Bodily Agency 386

13.2. The Nature of Imaginative Agency 387

13.3. Qualifications 391

(i) Concreteness 391

(ii) Voluntariness 392

(iii) Directness 394

13.4. The Essentiality of Imaginative Agency 396

13.5. The Subjection to the Will Account 404

14 Meeting the Desiderata 409

14.1. Explanatory Power 409

(i) Informativity 410

(ii) Fundamentality 413

xiii

14.2. Extensional Adequacy 414

(i) Non-Imaginative Phenomena 414

(ii) Imaginative Phenomena 417

14.3. Non-Representationality: the Case of Imagined Pains 421

14.4. Passivity: the Case of Spontaneous Images & Thoughts 423

Conclusion 431

(i) The Agency Account 431

(ii) The Epistemological and the Dependency Account 434

(iii) Other Theories of Imagining 437

(iv) Some Loose Ends 439

Bibliography 443

Index 455

Acknowledgements

I am grateful to the following journals for permission to reproduce mater-ial: Enrahonar for parts of my paper ‘Emotional Imagining and Our Re-sponses to Fiction’ (Enrahonar, vol. 46, 2011; reprinted by permission ofthe publisher, Universitat Autònoma de Barcelona); Rivista di Estetica forparts of my paper ‘The Recreative Imagination’ (Rivista di Estetica, vol.54, 2013; reprinted by permission of the publisher, Rosenberg & Sellier);and Philosophical Explorations for parts of my paper ‘Transparency andImagining Seeing’ (Philosophical Explorations, vol. 13, 2010; reprinted bypermission of the publisher, Taylor & Francis).

My special thanks are due to Malcolm Budd, without whom I would nothave started and continued to work on the imagination; Mike Martin, whohas challenged and pushed me further on the topic than anyone else; LucyO’Brien, who made sure that I finished my PhD thesis on the imaginationin time; Sebastian Gardner and Gianfranco Soldati, who taught me a lotabout the phenomenological approach to the mind and the imagination; andRob Hopkins, who has always been most helpful in his criticism.

In addition, I am very grateful, both for their constructive criticisms andfor their sympathetic encouragements, to Jiri Benovsky, Davor Bodrozic,Johannes Brandl, Josep Corbi, Peter Goldie, Henning Hahn, David Harris,Frank Hofmann, Eduard Marbach, Aaron Meskin, Richard Moran, KevinMulligan, Matt Nudds, Martine Nida-Rümelin, Matt Soteriou, KathleenStock, Juan Suarez, Mark Textor, Gian-Andri Toendury, Cain Todd, Lam-bert Wiesing, Marcus Willaschek, Richard Wollheim, several anonymousreferees, as well as audiences at University College London, the University

xvi

of Valencia and the University of Fribourg.Finally, I would like to extend my gratitude to Johannes Brandl, Andreas

Kemmerling, Wolfgang Künne and Mark Textor, for their personal assist-ance and editorial guidance; the Graduate School of University CollegeLondon and the Swiss National Science Foundation, for their financial sup-port of my research; and to my family and friends, for their uncomplainingpatience during the last years.

For Evgenia

Introduction

Introduction

The principal aim of this monograph is to present and compare the maintheories of imagining in a systematic and opinionated way.

The presentation is systematic in so far as it is structured by the concernwith the unity and common nature of the many forms of imagining, such asvisualising, supposing or daydreaming. Accordingly, the book involvesboth a close study of the distinct kinds of imagining and of their main fea-tures, similarities and differences; and a critical discussion of the principalaccounts of imagining put forward in the literature. The book thereforeserves as an introduction to the main ideas and positions that have been ad-opted with respect to imaginative phenomena. But it does so in a strictlyproblem-oriented manner, namely from the perspective of the possibility ofproviding an account of imagining that unifies its different incarnations. Asa consequence, the discussion is organised around the possible sources ofthe unity of imagining and centres less on particular philosophers and theirspecific views.1

The comparison of the corresponding principal theories of imagining is1 Another reason for this is that ‒ with some notable exceptions, such as the works by

Husserl (2006), Sartre (1940), Casey (1976), White (1990), O’Shaughnessy (2000)and McGinn (2004) ‒ many philosophers have not presented fully developed and fo-cused theories of imagining, but instead embedded their claims in discussions ofother, often more general issues (see, e.g., Collingwood (1938), Ryle (1949/1963),Scruton (1974), Peacocke (1985), Walton (1990, Hopkins (1998) or Martin (2002a)).Even of the noted exceptions, who devote whole books or lecture series to the topic,some have not aimed at formulating a cohesive account of all forms of imagining,but rather focused on discussing particular kinds of imagining, or on developing alist of more or less loosely connected ideas and observations.

2 Introduction

opinionated, on the other hand, in so far as a more specific goal of the bookis to defend the Agency Account of imagining as the best unified account ofimagining available. The main claim of this theory is that imaginings are,essentially and fundamentally, mental actions of a certain kind. Accord-ingly, imagining is something that we actively and voluntarily do.

A unified account of imagining is characterised by the fact that it cap-tures the common nature of the central cases of imagining (e.g., visual-ising, supposing, or daydreaming) in fundamental terms and is able to dis-tinguish them from the central cases of non-imaginative mental phenom-ena, notably cognitive representations (e.g., perceptions, judgements, ormemories). The minimal goal of any theory intended as a unified accountof imagining should therefore be to achieve extensional as well as explan-atory adequacy: it should be valid for the paradigm cases of imaginativeand non-imaginative phenomena; and it should correctly describe and ex-plain the distinctive and fundamental nature of the former. A main theme ofthe discussion is therefore the elucidation of primary examples of imagin-ing, and of what distinguishes them from other, non-imaginative mentalphenomena. Another and closely related guiding issue is whether it is pos-sible to account for the specific character of these imaginings by identify-ing a set of features distinctive of them and responsible for their imaginat-ive status (i.e., their property of being instances of imagining). That is, thebook addresses the issue of whether it is possible to specify and elucidatethis status in terms of individually necessary and jointly sufficient condi-tions for something to be an imagining, at least with respect to theparadigm cases.

The issue of formulating a unified account of imagining has often beenneglected in the philosophical tradition. Many of the discussions of ima-gining in the past and the present have focused, not on the nature of ima-ginings, but on their role in our mental lives and our interactions with otherpeople and the world. It has been widely acknowledged that imagining isvery prominent in and significant for various parts of our lives, rangingfrom our emotional engagement with other people (e.g., Goldie (2000):194ff.) and our moral evaluation of actions (e.g., Johnson (1993)) to theaesthetic appreciation of artworks (e.g., Walton (1990) and its many fol-

Introduction 3

lowers) and even the acquisition of knowledge about the external world.2 Moreover, even when philosophers address the question of what it

means for a mental phenomenon to be imaginative (i.e., possess the prop-erty of being an imagining), they usually concentrate exclusively on spe-cific forms of imagining, notably on sensory or visual imaginings.3 Indeed,the neglect of the issue of what all imaginings have in common as imagin-ings may be closely linked to the neglect of non-sensory or complex kindsof imagining. Finally, philosophers writing about the imagination have notalways succeeded, or been particularly interested, in developing their ideasinto a proper theory of what is distinctive of imaginings.

As a result, most discussions of imaginings have been concerned either

2 Although imaginings are typically held not to constitute knowledge about the ex-ternal world or provide relevant epistemic justification by themselves (though seeChapter 4 for an opposing view), it seems beyond doubt that they are often involvedin other ways in the acquisition of knowledge. See, for instance, the discussions onthought experiments (e.g., Sorensen (1992) and Gendler (2000a)), the link betweenconceivability and possibility (e.g., Gendler & Hawthorne (2002)), the role of men-tal imagery in geometry (e.g., Giaquinto (1992)) and in counterfactual reasoning(e.g., Williamson (2008)), or the involvement of visualisation in memory-basedcognitive projects, such as the project of determining whether frogs have lips, orhow many windows are in one’s own house (e.g., Kosslyn (1980): 1, or Pylyshyn(2002): especially 164).

3 See Collingwood (1938): chs. 9f., Sartre (1940), Peacocke (1985), Hopkins (1998):ch. 7, and, it seems, Wittgenstein (1984c): vol. II, sec. 63-147, to name just a few ofthose who focus more or less exclusively on sensory or visual imaginings. Kind(2001) does the same: while she argues that all instances of ‘imagining’ involvesensory imagery, she distinguishes the resulting sensory ‘imagining’ from intellec-tual ‘supposition’. Scruton (1974), White (1990), Casey (2000), O’Shaughnessy(2002), Currie & Ravenscroft (2002) and McGinn (2004) address both sensory andintellectual imaginings, but no emotional imaginings (in contrast to emotional re-sponses to imagining), and imaginative projects only in passing and not in relation totheir commonality with imaginative episodes. Besides, although McGinn discussesvarious forms of imagining, he does not aim to provide a unified account of them.Instead, he argues merely that they form an ‘imagination spectrum’ which extendsfrom the most simple and temporally and conceptually prior imaginative phenomena(e.g., those involved in sensory representation) to the most complex and developedones (e.g., those involved in creativity; see McGinn (2004): 13). See also the discus-sion of the five main forms of imagining in Section 1.2 for further references to ac-counts which focus on particular kinds of imagining.

4 Introduction

with aspects of imaginings other than (though perhaps dependent on orotherwise related to) their imaginativeness, or with the imaginativeness ofonly certain kinds of imagining. Just a few philosophers have attempted toprovide a satisfactory account of imagining in its (more or less) full vari-ety.4 The less comprehensive philosophical discussions have certainly shedlight on important aspects of many different kinds of imagining, such assensory imaginings (e.g., visualising a face), intellectual imaginings (e.g.,supposing or imagining that it rains), affective imaginings (e.g., imaginingan itch), experiential imagining (e.g., imagining having the experiences ofsomeone else) or imaginative projects (e.g., daydreaming about beingrich).

But the imaginative nature common to all forms of imagining has re-mained largely uninvestigated. A good example for this and the underlyingway of thinking is Walton’s account of imagining (1990: ch. 1). He spendsconsiderable time on specifying ‘a number of dimensions along which ima-ginings can vary’, but also insists that we cannot ‘spell out what they havein common’ and have instead to be content with an ‘intuitive understandingof what it is to imagine’ (ibid., 19). Indeed, it is rather common in discus-sions about imaginings and their role in our mental lives to take for granted

4 See Scruton (1974): chs. 7f., Casey (1976), O’Shaughnessy (2000): chs. 11f., andMcGinn (2004) for clear examples. But none of them discusses imaginative projects(fantasies, daydreams, etc.) in any detail. Whether other proposals are intended orable to constitute a unified account of imagining is less clear. Hume’s account of allimaginings as forming a certain sub-class of ‘ideas’ suffers from its treatment of thedifferences between sensory and intellectual representations and between imaginat-ive and cognitive representations as quantitative (i.e., a matter of ‘vivacity’) ratherthan qualitative (see Chapter 9 for discussion and references). Although Ryle dis-cusses mainly sensory cases, his account of imagining as a form of ‘internal’ pre-tending or pretending ‘in one’s head’ might be applied to intellectual imagining aswell (see Section 12.3, and Ryle (1949/1963): chs. 7f., especially sec. 8.6). Whiteanalyses both visualising and intellectual imagining in terms of thinking of the pos-sible, but does not explicitly connect the two analyses (see Section 5.4, and White(1990): 122f; 184). Despite the fact that Currie and Ravenscroft treat both sensoryand intellectual imaginings as simulations of their respective cognitive counterparts(see Section 12.2, and Currie & Ravenscroft (2002): 11 and 49), their main concernis with the involvement of the imagination in adopting a perspective on the worlddifferent from one’s current one, rather than with the general nature of imagining(ibid: 8f. and 11).

Introduction 5

that it is sufficiently clear what it means to imagine something ‒ as it islikewise often assumed that we have a good grasp of, say, what it means toperceive or believe something, or to desire it.

One particular difficulty with this approach is, however, that imagining‒ perhaps in contrast to perception, belief or desire ‒ does not obviouslyconstitute a natural kind (more on this point in Section 1.1). In fact, thesuspicion that imagining ‒ understood as the class of visualising, suppos-ing, empathizing, daydreaming, and so on ‒ is not among the most basic orprimitive mental phenomena may explain why it has by no means been thesubject of philosophical investigation to the same extent, and with the sameexclusivity, as perception, belief and desire. Similarly, it is not evident thatthere is unity in imagining; nor, if so, what it consists in or comprises. Thismay lead to a situation in which a certain kind of imagining is postulated orappealed to in the context of explaining a given phenomenon; while thelack of any further elucidation of the nature of this type of imagining raisesin others considerable skepticism about the proposed role or even existenceof the kind of imagining in question.5

Such complications – as well as, more generally, the significance andprominent position of imagining in our lives and interactions with eachother and the world – provide sufficient motivation for the investigation ofthe possibility of a unified account of imagining. In particular, it will behelpful and illuminating to learn more about what it means for arepresentation to be imaginative, and how this relates to or influences thevarious forms of engagement or project involving imagining.

(i) The Need for a Unified AccountThe nature and unity of imagining is of theoretical interest both in itselfand in relation to many important aspects of our lives. But the general in-terest inherent in the question of whether we can provide a unified theoryof imagining and the significant role of imaginings in our mental lives are

5 See, for instance, the skepticism ‒ expressed in Budd (1992), Hopkins (1998): ch. 1,Wollheim (2003) and Dorsch (2012c) ‒ about the existence of the specific form ofimagining seeing the depicted, which Walton refers to in his account of pictorial ex-perience (see Walton (1990): ch. 8, and (2002)).

6 Introduction

not the only motivations for the search of a unified account. The expecta-tion that the central cases of imaginings share a common nature arises alsofrom the need to explain two facts about our actual treatment of such rep-resentations.

The first is simply that, even after discounting the less obvious cases,we do group together a large variety of mental occurrences in the class ofimaginings, while excluding many others. Thus we accept visualising, sup-posing, daydreaming, being engaged with fictions, empathizing, and so on,as paradigm instances of imagining, or at least as essentially involvingsuch instances; but not seeing, judging, deliberating, or feeling an emotionor desire. If such imaginings had nothing in common with each other, butshared features with the non-imaginative phenomena, this tendency in howwe categorise mental phenomena would be rather mysterious: there has tobe something about the imaginative mental phenomena which causes us totreat them ‒ but not other mental phenomena ‒ as members of one and thesame class (see Scruton (1974): 91f.).

The second relevant observation is that our corresponding classifica-tions are stable, and that we usually have a good grasp of whether ‒ thoughnot necessarily of why ‒ a given mental state is imaginative or not. Threeobservations are particularly relevant here. First of all, we repeat the samecategorisations and do not locate certain representations today on one sideof a divide and tomorrow on the other. There is no change in opinion aboutwhether, say, visualising is an instance of imagining. Then, it normallydoes not take us much effort or thought to come to a classification of a giv-en mental phenomenon as imaginative or non-imaginative. Identifying aninstance of daydreaming as an example of imagining, say, is typically verystraightforward. Finally, we are seldom unsure about how to treat a certainmental phenomenon. This may happen with borderline cases (e.g., spontan-eously occurring images and thoughts), or when we are confronted with anunusual or rare kind of representation (e.g., Macbeth’s awareness of Ban-quo, or of the dagger). But, in most cases, we are not in doubt about wheth-er some mental episode is an instance of imagining. These three aspects ofthe stability of our relevant classification, too, strongly suggest the exist-ence of a certain kind of unity among imaginings, to which we furthermorehave easy epistemic access, but which is still in want of further elucidation.

Introduction 7

Otherwise, it would be very difficult to explain the firmness and ease withwhich we take the sometimes very different instances of imagining to be ofthe same kind, and why it is that they, but no other phenomena, count forus as imaginative. Any satisfactory analysis of imagining needs to providean elucidation of this unity of the paradigm instances of imaginings.

The idea pursued in this book is that the unity under consideration is dueto some features shared by and distinctive of imaginings. This is thesimplest and most straightforward explanation of their homogeneity; and itis adopted by all theories of imagining that have the ambition to provide aunified account. But it is not the only possible account of the two notedfacts and the resulting apparent unity of imagining. One could, for in-stance, maintain that our grouping together of the variety of phenomenadescribed as instances of one and the same type of phenomenon is merelyaccidental or conventional, without any substantial grounding in sharedfeatures of the mental phenomena concerned. But such a claim would bedifficult to support in view of the facility and assurance with which we cat-egorise mental states as either imaginative or non-imaginative. Until it hasbeen confirmed that such a strong form of skepticism is inevitable, therealistic hope for a more positive theory of the common nature of imagin-ings should outweigh any doubts about the possibility of a unified charac-terisation of imagining. It is thus reasonable to demand from a theory ofimagining that it account for the fact that we classify a large variety of phe-nomena as imaginative; and the fact that this classification is not a pure co-incidence or convention. A unified theory of imagining promises to providesuch an explanation by treating these facts as a consequence of the pres-ence of certain features shared exclusively by all imaginings.

(ii) Three Main Theories of ImaginingThis raises the question of which proposals for a unified account of ima-gining are on offer and should be considered. When looking at the theoriesof imagining put forward in the philosophical tradition, three major recur-ring themes can be identified: the relationship (or lack thereof) of imagin-ative representations to reality, their relationship to cognitive representat-ions, and their relationship to the will. These motives identify the three

8 Introduction

principal ways in which the common and distinctive nature of the core in-stances of imagining may be elucidated: in terms of their irrelevance forour epistemic interaction with the world, in terms of their dependency onperceptions and (occurrent) beliefs, or in terms of their origin in a specifickind of mental agency.6 The resulting theories of imagining may be la-belled the Epistemological, the Dependency and the Agency Account. Theprominence in the literature of these approaches and their central ideasabout the nature of imagining is indicative of the fact that they are the maincontenders for a unified theory of imagining. Accordingly, I concentratemy discussion primarily on these three proposals and devote to each awhole part of the book.

The Epistemological Account specifies the difference between imagin-ings and cognitions by reference to the idea that only the latter are con-cerned with the representation of reality ‒ at least in a particular sense stillto be specified. Our minds interact with the world by means of cognitionand action. Both forms of interaction are primarily the domain of cognitiverepresentations, such as perceptions, memories or beliefs. In accordancewith this observation, the versions of the Epistemological Account to befound in the literature identify a lack of cognitive concern with reality and,especially, the failure to play a substantial role in cognition as the distinct-ive feature of imaginings. Different versions of this approach may vary inhow they characterise the assumed lack of cognitive concern.7

It is conceivable that a unified account of imagining may also be formu-lated in terms of their specific insignificance for our active engagement

6 Although it might be possible to endorse a view that characterises imaginings interms of their relationship both to reality and to agency (without also tracing backone relationship to the other), such an approach to imagining would seem to be over-complex. As will become clear in the subsequent chapters, reference to the presenceand nature of one kind of relationship suffices to account for the distinctive nature ofimagining and, if necessary, for the presence and nature of the other kind of relation-ship. The general idea is that imaginings will turn out either to lack a cognitive con-cern with reality precisely because they are voluntarily formed by us; or instead to(be able to) be mental actions precisely because they are not cognitively constrainedby how reality is like.

7 The main proponent of this view is O’Shaughnessy (2000). But very similar ideascan be found in the writings of Hume (1739), Sartre (1940), Wittgenstein (1984c)and McGinn (2004), among others (see also the introduction to Part Two).

Introduction 9

with the world. The idea is that imaginings ‒ in contrast to, say, desires, in-tentions, beliefs or perceptions ‒ cannot motivate us to act or guide us inour actions (e.g., by providing us with information about our relevant en-vironment, or about adequate means to achieve our goals). However, theclaim that imaginings lack a guiding role in agency can be traced back tothe idea that they lack a cognitive concern with the world: they cannotguide us in action because they do not provide us with knowledge aboutthe relevant aspects of reality (i.e., the environment and the means). Theidea that imaginings cannot move us to act, on the other hand, is not onlycontroversial (see Note 16 in Chapter 1); it also does not distinguish themfrom many non-imaginative phenomena. Hence, it is not very promising toformulate a unified account of imagining specifically in terms of theirseeming unimportance for our actions; and not very surprising that thefocus in the literature on imagining has rather been on the apparent lack ofcognitive concern.

The Dependency Account, by contrast, puts emphasis on the asymmetricdependency of imaginings and cognitions. It argues that imaginative phe-nomena are constitutively dependent on and, hence, have to be elucidatedby reference to their cognitive counterparts. While cognitions exist and canbe made sense of independently of imaginings, the opposite is false. Pro-ponents of the Dependency Account disagree about the nature of the kindof dependency concerned.8 Philosophers in the Humean tradition, for ex-ample, focus on the idea that imaginings depend on cognitions in so far asthey represent them (possibly due to a suitable causal link). Another prom-inent approach is to assume that, in some sense, imaginings imitate ormimic cognitions ‒ for instance, in the shape of off-line simulation or innerpretense. Besides, it is widely accepted that there is also some form of se-mantic dependency: that we acquire much of our ability to represent some-thing in an imaginative way by means of engaging in cognition (i.e., in per-

8 Hume (1739): 1.1.1.6ff. and 1.3.7.7), and O’Shaughnessy (2000): 363, endorse therepresentational version of the Dependency Account for all forms of episodic ima-gining; while Peacocke (1985): 22, Martin (2001): 273 and (2002): 404, and Dorsch(2010c) accept this view for sensory imaginings; and Dorsch (2011b) ‒ which ismore or less identical with Chapter 11 ‒ for emotional imaginings. Currie &Ravenscroft (2002): 49, and Ryle (1949/1963): 250ff., may be read as defending aversion of the Dependency Account formulated in terms of simulation or imitation.

10 Introduction

ceiving, sensing, judging, and so on) and developing the relevant recogni-tional and conceptual capacities.

Finally, the already introduced Agency Account proposes mental activityas the characteristic element of imagining. It maintains that imaginings are,in a particular way, essentially active; while cognitions are either passive,or active in a different way. The general idea is thus that imaginings aremental actions of a special kind: they are imaginative determinations or ex-pressions of the will. They differ in this respect from mental passions,among them: judgemental thoughts or occurrent beliefs that are formed inresponse to epistemic reasons; episodic memories that are determined bypast experiences; perceptions and sensations that are determined by theirexternal objects; and deviantly caused instances of the each of the preced-ing cognitive kinds of episode. Again, distinct versions of the Agency Ac-count vary in how exactly they individuate the active character distinctiveof imaginings.9

The particular version, that I intend to put forward in this book, main-tains that imaginings are mental actions, that aim at the active and directformation of mental representations with specific contents. The require-ment of representational specificity demands that the underlying motiva-tional states (e.g., tryings, desires or intentions) determine which particularfeatures are to be represented as being instantiated by which particular ob-jects; while the directness requirement is meant to ensure that the motiva-tional states end up determining the content of the formed representationswithout making use of epistemic or merely causal mechanisms of contentdetermination – and, notably, those mechanisms involved in the activelytriggered manifestation of mnemonic or other mental dispositions, or thoserelied on during the voluntary formation of beliefs on the basis of evidence– as means. My defense of this version of the Agency Account of imagin-ing is paired with a rejection of the Epistemological and the DependencyAccount.

9 The Agency Account (or something very much like it) has been endorsed by, amongothers, Wollheim (1973): 69, Scruton (1974): 95 (see in general 94-100), Levinson(1998): p. 232, n. 3, Kind (2001): 90ff., and McGinn (2004): 12ff.; 131f. Most ofthese endorsements have been inspired by Wittgenstein’s comments on sensory ima-ginings (see, for instance, his (1984b): vol. II, sec. 63 and 627; see also Budd(1989): 104ff.).

Introduction 11

However, since the present book is not only finite in its length, but alsoprimarily centred on the issue of the unity of imagining and aimed at giv-ing a balanced overview of the various approaches to this issue, the discus-sion of the Agency Account has to be confined to roughly the same numberof pages as the two other main theories (or, in the case of the DependencyAccount, as its most promising ‒ namely representational ‒ version). Thismeans, unfortunately, that there is not enough space for a comprehensiveelaboration and vindication of the Agency Account, which definitely meritsa monograph on its own. The chapters devoted to this theory therefore con-centrate just on the principal aspects of this theory, such as its characterisa-tion of imaginings as mental actions of a certain kind, the essentiality ofthe imaginative activity involved in voluntary imaginings, or the specialstatus of spontaneously occurring images and thoughts in relation to ima-ginative agency. Other necessary elements of a full defense of the AgencyAccount ‒ such as its application to daydreams and other imaginative pro-jects, or a discussion of pictorial experience, association, guessing, andother problematic cases ‒ have to be left for another occasion.

(iii) Other Theories of ImaginingIn addition to the three parts dealing with the main theories of imagining,there is a fourth and introductory part of the book in which, among otherthings, I briefly consider and argue against less important contenders for asatisfactory account of the unity of imagining. Some of these additionalviews start from the same ‒ or at least very similar ‒ basic ideas as the Epi-stemological, the Dependency or the Agency Account. This is part of theexplanation of why there are some considerable overlaps in convictions,arguments and motivations among the individual major and minor views.But these shared elements do not undermine the fact that the various ac-counts of imagining exclude each other as unified theories in so far as theyidentify different features as the fundamental constituents of the commonnature of the otherwise heterogeneous forms of imaginings.

One alternative proposal ‒ the Quantitative Account ‒ takes imaginativeand non-imaginative phenomena to differ solely in degree (e.g., a differ-ence in vivacity), rather than in kind. As a result, imaginings are assumed

12 Introduction

to possess the same essence as their non-imaginative counterparts.10

The Phenomenal Account assumes the most basic differences betweenimaginative and non-imaginative phenomena to be phenomenal in nature:that is, to be differences in what the respective episodes are subjectivelylike. Accordingly, theories of imagining should focus on the phenomenolo-gical description of imaginings and their non-imaginative counterparts,rather than on the formulation of more fundamental explanations of thenoted phenomenal differences. The Phenomenal Account therefore takes itto be distinctive of imaginings that they subjectively seem to have a certainkind of object, attitude or origin ‒ in contrast to some of the following the-ories which assume that it is distinctive of imaginings that they indeedhave a certain kind of object, attitude or origin.

The Ontological Account, for instance, elucidates the distinctive charac-ter of imaginings in terms of the specific nature of the imagined entities.The central idea is that there is a categorical ontological difference betweenobjects which are perceived, remembered or judged to be a certain way andobjects which are visualised or supposed to be a certain way. In particular,it may be said that imagined and cognised objects differ in their type of ex-istence, their spatio-temporality, their determinacy or (as also proposed byone version of the following account) their modal status.11

The central idea of the Modal Account is that, while it is characteristicof imaginings that they represent possibilities, it is characteristic of cogni-tions that they represent actualities.12 The most promising version of thistheory maintains that the difference is one in attitude, rather than content.This means that both types of representation are concerned with the samekind of objects and states of affairs. But while cognitions purport to repres-ent them as part of the actual world, imaginings lack such a commitmentand replace it with a stance of non-neutrality towards some possible world(or situation). That is, imaginings are said to make a claim, not about how

10 The idea that perceptions and imaginings differ merely in vivacity is central toHume’s view (see Section 9.1). Compare also Hopkins’ (2011a) approach to sensorymemories, according to which they belong to the class of sensory imaginings anddiffer merely contextually from its non-mnemonic members.

11 Some of these ideas can be found in Sartre (1940) and Casey (1976), for instance.12 The view presented in White (1990) comes close to the Modal Account, although

White presumably does not intend to restrict imaginings only to mere possibilities.

Introduction 13

things actually are, but about how things might possibly be.The resulting modal theory of imagining is, in fact, nothing but a spe-

cific version of the Attitude Account, according to which the differencebetween imaginative and cognitive episodes is fundamentally attitudinal.Minimally, the idea is that cognitions involve a cognitive attitude ‒ that is,the commitment that reality is just like they represent it to be ‒ while ima-ginings lack it. Versions of the Attitude Account may stay neutral on the is-sue of whether this means that imaginings possess instead a distinctivelyimaginative attitude. However, the need to distinguish imaginative epis-odes not only from cognitive episodes, but also from other kinds of episodehas moved proponents of the Attitude Account to assume that imaginingsare non-neutral as well, and to specify the commitment distinctive of ima-ginings in more positive terms, such as by reference to an act of positingentities as non-existent, absent or, indeed, non-actual.13

Next comes the Spontaneity Account which maintains that imaginingsdiffer from cognitions in their specific origin in the ‒ voluntary or involun-tary ‒ ‘spontaneity’ of the mind, rather than in the impression of the ex-ternal world onto the mind. In its most promising incarnation, the viewcomes close to the Agency Account ‒ especially in its insistence that theorigin of imaginings is not merely causal, but essentially involves somesubjective response or doing. But the two theories differ in that the Spon-taneity Account extends the realm of the imagination to the involuntary,thereby loosely drawing on a distinction between perceiving and imaginingthat is similar to the Kantian distinction between ‘receptivity’ and ‘spon-taneity’.

The Subjection to the Will Account is also similar to the Agency Ac-count. While the last-mentioned takes imagining to always involve volun-tary control, the first-mentioned requires merely that there is the possibilityof such control, without the need for its actual exertion. The Subjection ofthe Will Account has the advantage of being able to capture what seem tobe involuntary instances of imagining, while distinguishing them none theless strictly from, say, perceptions and impression, the passivity of which

13 The idea of a specifically imaginative attitude has been put forward notably in thewritings of the phenomenologists, such as Husserl (2006), Sartre (1940) and Casey(1976).

14 Introduction

resists any attempts to bring them under voluntary control.14

Finally, the Subpersonal Account ‒ the only one that I do not return to atone point or another in the subsequent chapters ‒ characterises imaginingsin terms of the specific nature or use of the subpersonal neurofunctionalprocesses or modules involved in imagining, as they are studied by thecognitive sciences. In other words, proponents of this approach maintainthat what is distinctive about imaginings is that they are subpersonallyformed in a special way, whether by means of the stimulation of certain‘imaginative’ areas in the brain, or by means of certain ‘imaginative’ waysin which standard cognitive mechanisms are employed. It is of course anempirical ‒ and, especially, experimental ‒ question whether it is possibleto identify neurofunctional areas or operations that are unique to imaginingand, if so, what their nature is. Accordingly, the formulation of the Subper-sonal Account is primarily a matter of the empirical sciences, and only sec-ondarily a matter of philosophy, which is by and large a non-experimentaldiscipline. In this respect, the Subpersonal Account differs sharply from themuch more philosophical theories of imagining listed above.15 This differ-ence is closely related to the three main reasons of why I do not discuss theSubpersonal Account any further.

The first is that the cognitive sciences are yet simply not advancedenough to formulate a satisfactory unified theory. Central to the explana-tion of this fact is that the unity of imagining has not been of much sci-entific interest, and that, hence, the relevant studies have typically focused

14 The noted similarities among the Agency, Spontaneity and Subjection to the WillAccounts may also be captured by reference to the fact that they are all instances ofthe more general Origin Account: that is, the view that imaginative and non-imagin-ative phenomena differ essentially in what does or can determine their occurrenceand representationality.

15 Of course, philosophy should take into account experimental findings; and philo-sophers may be engaged in experimental investigations. But philosophical methodo-logy is not itself experimental ‒ at least not in the sense of being concerned with thedesigning and carrying out of repeated observations or experiments with replicableresults. This is true even of the so-called ‘empirical’ or ‘experimental’ philosophy: itsimply incorporates some scientific discipline (e.g., cognitive psychology) and ap-plies its experimental methodology to empirical phenomena or problems, that are se-lected because of their relevance for certain philosophical issues. Hence, there is stilla division of labour between philosophy and the empirical sciences.

Introduction 15

on particular forms of imagining, notably visualising and belief-like ima-gining.16 What this may reflect is that the mental category of imagining isperhaps not a very interesting category with respect to the subpersonallevel; while, as noted above, it plays an important role on the personallevel. Indeed, at least to my knowledge, there has been no serious attemptso far to spell out, in terms of the neurofunctional entities and processesrecognised by the cognitive sciences, what the sensory, affective,intellectual, experiential and more complex forms of imagining havefundamentally in common.

Moreover, it is unclear whether it is reasonable to expect to be able toformulate such a theory. Perhaps there are several distinct and fairly inde-pendent neural or otherwise subpersonal correlates of the feature distinct-ive of all imaginings on the personal level ‒ for instance, one correlate foreach of the basics form of imagining. That is, it is unclear whether suchcorrelates need to share a scientifically interesting feature in virtue ofwhich they count as correlates of one and the same personal level phe-nomenon (i.e., imagining). For instance, green ‒ or, more generally, col-oured ‒ objects arguably do not have any physical property in common thatis significant from the perspective of physics.17 Perhaps the same is true of

16 Tye (1991), Thomas (1999) and (2010) provide an overview of recent empiricalstudies on sensory imagining, while Nichols (2006) and especially Gendler (2011)do the same with respect to different kinds of intellectual imagining and related phe-nomena, such as pretense, empathy or delusions. Harris (2000) offers a develop-mental perspective on intellectual imagining and pretense, while Currie &Ravenscroft (2002), Heal (2003) and Goldman (2006) look at it from a simulationistperspective. More specifically, Currie & Ravenscroft (2002) and Gendler (2011) dis-cuss evidence linking intellectual imagining to empathy, autism and delusion; whileWeinberg & Meskin (2006) develop an empirically-based account of intellectualimagining ‒ partly based on the work by Nichols & Stich (2003) ‒ and apply it tovarious philosophical problems ‒ namely the problems of emotional responses tofiction (see Chapter 11), of imaginative resistance, and of distinguishing imaginingfrom supposing (see Note 18 in Chapter 1). By contrast, the empirically informeddebate about the nature of sensory imagining has been largely dominated by thedispute between Kosslyn (see, for instance, his 1980) and Pylyshyn (see, forinstance, his 2002).

17 If at all, colour hues of surfaces should be identified with certain classes of reflect-ance properties, the members of which are individuated and grouped together interms of their dispositional power to bring about a certain pattern of stimulation in

16 Introduction

the neurofunctional correlates of imagining. Besides, the prospects of se-curing a satisfactory unified theory of imagining are likely to dwindle fur-ther, the more removed from the personal level an account becomes. Per-haps, generality is to be had only very close to ‒ or, indeed, only at ‒ thelevel of theorising about persons.

My second reason for not further investigating the prospects of the Sub-personal Account is that it is not really in competition with the unified the-ories aiming to capture the personal level of imagining. Rather, the twokinds of account should be seen as complementing each other. While thepreviously introduced theories aim to capture the nature of imagining in re-lation to subjectively accessible features (e.g., representationality, rationalrole, origin, passivity, and so on) and in contrast to other conscious mentalphenomena (i.e., perception, feeling, judgement, action, and so on), theSubpersonal Account intends to identify the neurofunctional foundation ofwhichever aspects of our conscious mental lives are essential to imagining.That is, the Subpersonal Account is concerned with how imagining is real-ised in the brain, and not what it means to imagine something from thepoint of view of persons. Accordingly, personal and subpersonal theories ‒that is, philosophy and the cognitive sciences ‒ have different goals andemploy distinct notions of fundamentality. The same is true of physics orbiochemistry and their concern with happenings in the mind or brain.

My third and final motivation for not spending more time on discussingthe Subpersonal Account is that the more philosophically minded theoriesshow a certain priority over it ‒ most notably with respect to issues of epi-stemic authority. If philosophy comes up with a satisfactory account ofimagining that makes good sense of the unity of the various forms of ima-gining on the personal level, it would be no reason to revise or give up thisphilosophical view that the cognitive sciences were to reveal that there are

the relevant opponent processing channels ‒ and, hence, to bring about certain phe-nomenally individuated colour experiences. However, from the perspective of phys-ics, these classes of reflectance properties ‒ in contrast to the reflectance propertiesthemselves ‒ are of no scientific interest. Physicists have been concerned withdeveloping and experimentally testing theories of how much of the incoming lightsurfaces reflect, but not of the extent to which surfaces stimulate the opponent pro-cessing system or cause phenomenally different colour experiences (see Dorsch(2009a) and (2010b)).

Introduction 17

no significant or interesting commonalities among all the different in-stances of imagining. Instead, and in the absence of further reasons toquestion the philosophical view, it would be more reasonable to concludethat imagining, as a personal-level phenomenon, has no single correlate onthe subpersonal level. Similarly, the fact that physics is not really con-cerned with colours does not undermine the attempt to formulate an ad-equate philosophical theory of colours (see Notes 17f. above). Of course,philosophical theories of imagining need to be given up if they contradictempirically well-supported versions of the Subpersonal Account (if thereare any). But, on the whole, the best philosophical theories of imaginingtell the cognitive sciences what to look for (e.g., for the neurofunctionalcorrelate of a particular kind of mental agency), and not the other wayround.18 Indeed, whether a given version of the Subpersonal Account issuccessful should in part be evaluated in respect of whether it captures thefeature, or set of features, identified as being distinctive of all instances ofimagining by the best philosophical unified theory.

(iv) The Structure of the BookAs already touched upon, the book consists of four parts. Part One has thefunction to set the stage for the subsequent critical discussions of the Epi-stemological, the Dependency and the Agency Account. It does so in threesteps.

First, it specifies the two main desiderata for any unified account ofimagining: extensional adequacy and explanatory power (see Chapter 1).As part of this characterisation, Part One involves a characterisation of thefive central forms of imagining to be unified: sensory imagining, affectiveimagining, intellectual imagining, experiential imagining, and imaginativeprojects (see Chapter 1). It also includes a description of the main featuresof imaginative episodes to be explained ‒ most notably their lack of a cog-nitive attitude and of an epistemic function ‒ and, indeed, provides a firsttentative reason for the endorsement of the Agency Account, namely that it

18 See Dorsch (2010b) and Dorsch (2010a) for a similar division of labour betweenphilosophy and the empirical sciences in the case of colours and of hallucinations,respectively.

18 Introduction

promises the best explanation of the quasi-observationality of sensory ima-ginings (see Chapters 2f.).

Second, Part One is concerned with the relationship between imaginingand knowledge. In particular, it argues that certain instances of sensoryimagining can constitute knowledge in a way very similar to that of per-ceptual experiences and, in response to this observation, asks in whichsense (if any) imaginings can still be said to be uninformative with respectto the external world (see Chapter 4).

Third, Part One finishes with a discussion of the less important unifiedaccounts of imagining listed above (see Chapter 5). The only exceptionsare the Causal Account and the Subjection to the Will Account, which arebriefly discussed in the Chapters 9 and 13, respectively; as well as the Sub-personal Account, which is, as mentioned above, not further discussed atall.

Part Two is concerned with the characterisation (see Chapters 6f.) andrejection (see Chapter 8) of the various versions of the Epistemological Ac-count of imagining, formulated in terms of the alleged inability of imagin-ings to be reliable or rationally integrated with our picture of reality, to in-form us or provide us with knowledge about the external world, or to pos-sess the commitment, justificatory power or epistemic function essential tocognitions. The central component of this discussion is a detailed andscholarly examination of O’Shaughnessy’s version of the EpistemologicalAccount.

Part Three deals with the various versions of the Dependency Account.Chapter 9 highlights the relevant causal and representational elements inHume’s conception of imaginings and reveals its continuity with more con-temporary ‘neo-Humean’ views that understand the assumed dependencyof imaginings on cognitions in purely representational terms. Together withthe following two chapters, it is also devoted to the applicability of the res-ulting Representational Account ‒ that is, the representational version ofthe Dependency Account ‒ to intellectual imaginings (see Section 9.5),imaginative projects (see Section 9.6), sensory imaginings (see Chapter 10)and emotional imaginings (see Chapter 11). The conclusion to be put for-ward is that, while the Representational Account is probably true of sens-ory and emotional imaginings, it fails to pay justice to the nature of the

Introduction 19

other two forms of imagining. Part Three concludes with a negative assess-ment of the prospects of devising a unified account of imagining by refer-ence to the idea that imaginings simulate or otherwise imitate cognitions(see Chapter 12).

The final Part Four develops and defends the Agency Account of ima-gining. It does so, first of all, by spelling out the main thesis of the theory,distinguishing it from the Subjection to the Will Account and presenting anargument in favour of the endorsement of the Agency Account (seeChapter 13). The central claim of this account is that imaginings are to beidentified with mental actions that have the purpose of producing mentalmental representations and of applying direct control over which states ofaffairs are thereby represented. And it is crucial for the truth of this view isthat the kind of mental agency involved is essential to the formedimaginative representations, thus ruling out the possibility of passiveimaginings. In addition, Part Four illustrates how the Agency Account canmeet the two desiderata for unified theories of imagining establishedearlier and deal with potential counterexamples, notably the imaginativecounterparts to non-representational phenomena, as well as spontaneouslyoccurring images and thoughts (see Chapter 14).

Many of the parts, chapters and sections of the book are more or lessself-contained and can be read fairly independently of the others. Engage-ment with Part Two and the Epistemological Account requires mainlyknowledge of Chapter 1 and Sections 2.1f., although reading at least thefirst half of Chapter 4 will be helpful as well. In addition, the reader maydecide to skip most of Chapter 6 and all of Chapter 7 if (s)he is not inter-ested in the details of O’Shaughnessy’s theory of imagining and his moregeneral approach to conscious awareness. For Chapter 6 begins with a briefsynopsis of the main claims and arguments of O’Shaughnessy’s, whichprovides all the information necessary to proceed directly with Chapter 8.Part Three (the Dependency Account) and Part Four (the Agency Account)presuppose even less than Part Two, namely really not much more than ac-quaintance with the content of Chapter 1, as well as perhaps Sections 2.1f..Both Chapter 10 and Chapter 11 ‒ on sensory and on emotional imagining,respectively ‒ can be read largely on their own, which is in part due to thefact that they are based on previously published material. Similarly, the dis-

20 Introduction

cussion of the (neo-)Humean approach to imagining and its rejection as aunified account in Chapter 9 is relatively self-contained; as is the discus-sion of the idea that imaginings imitate cognitions in Chapter 12. Chapter13 constitutes a reasonably free-standing vindication of the Agency Ac-count, with some additional support by Chapter 14. Besides, the Chapters1, 2, 3 and 5 in Part One may be used as introductory readings to the topicof imagining, given that they highlight the main forms and features of ima-gining and portray many of the theories of imagining to be found in theliterature.

Although the outlook and structure of the book is primarily systematicin its focus on the issue of the unity of imagining, it also includes substan-tial critical studies of the views of particular philosophers. These views donot always constitute proper theories or provide comprehensive accountsof imagining, or of some of its specific forms (the views of Hume and Wit-tgenstein are good examples). But even if not, they are informative and in-structive with respect to the ambition to provide a unified theory of ima-gining and, more generally, to understand what it means to imagine some-thing. Discussions of the views of the following philosophers, among oth-ers, are scattered across the book: Hume (Chapter 9); Husserl (Sections2.1, 3.1f., 5.2f. and 5.5); Sartre (Sections 3.4ff., 4.5f., 5.2f. and 5.4f.); Wit-tgenstein (Sections 3.4f., 4.7 and 5.3); Ryle (section: 12.3); Casey (Sec-tions 3.3, 5.2, 5.5f. and 13.5); White (Section 5.4); Peacocke (Sections 3.2,3.5f. and 9.4); Walton (Chapter 11 and Section 13.5); Moran (Chapter 11);O’Shaughnessy (Chapters 6ff. and Sections 9.4ff.); Martin (Sections 2.3,3.2, 9.4 and Chapter 10); and Currie and Ravenscroft (Sections 12.1f.). Es-pecially Husserl’s theory of imagining would deserve more attention. But itis some compensation that most of his relevant and important ideas still re-ceive due attention in so far as they can be found again in the Husserlianaccount defended by Sartre (1940).

Finally, due to the partially modular structure of the book, many pas-sages depend on or allude to discussions in other parts, chapters, sectionsor notes of the book. Throughout the text, I use capital letters (and noabbreviations) to distinguish these internal cross-references from externalreferences to particular elements or portions of cited works.

Part One

The Nature and Variety of Imagining

Introduction to Part One

This opening part of the book has two main functions. First, it is intendedto provide an overview of three different things: (i) the different forms ofimagining; (ii) their main features; and (iii) various unified accounts ofimagining (i.e., those that differ from the three main theories to be dis-cussed later on in being less well-developed, less plausible, or less promin-ent in the literature). In short, it is meant to serve as an introduction to thephilosophy of imagining. Second, this part also aims to set the stage for therest of the book. In particular, its purpose is to specify the main desideratafor a unified account of imagining and to highlight some features of ima-ginative episodes that are to become crucial during the upcoming discus-sions.

In the first chapter, I spell out the two main desiderata and list the fivecentral forms of imagining. The next two chapters are devoted to a descrip-tion of the main differences between imaginative and cognitive episodesand, in particular, their sensory kinds. In Chapter 2, I highlight the differ-ences in attitude and epistemic function between imaginings and cogni-tions and argue that both may refer to aspects of reality, though in virtue ofdifferent referential mechanisms. Chapter 3, by contrast, focuses on im-portant characteristics of sensory ‒ and, especially, visual ‒ imaginings,which distinguish them from perceptual experiences. The followingchapter addresses the relationship between imagining and the acquisitionof knowledge. More specifically, I argue that sensory imaginings can, un-der the right circumstances, lead to knowledge in a way that is very similarto the way in which perceptual experiences may result in knowledge. And I

24 The Nature and Variety of Imagining

question Sartre’s and Wittgenstein’s claim that imaginings strictly differfrom cognitions in being uninformative. The final chapter critically dis-cusses six unified accounts of imagining that are ‒ both because of their in-herent weaknesses and because of their lesser popularity in the recent de-bates about the nature of imagining ‒ of lesser importance than the Epi-stemological, the Dependency and the Agency Account to be considered inthe subsequent parts. The six accounts in question are the Quantitative, thePhenomenal, the Ontological, the Modal, the Attitude and the SpontaneityAccount. The discussion of a seventh minor view ‒ the Subjection to theWill Account ‒ has to wait until Section 13.5.

Before I move on to the main desiderata for a unified account of imagin-ing, three notes on my terminology are in order. First, when I speak ofknowledge, I intend to refer both to intellectual (or belief-constituted)knowledge and to successful perceptions, episodic memories and othergrounds for such intellectual knowledge. There is some debate aboutwhether, say, perception constitutes a form of knowledge, or rather just aground for knowledge (see, e.g., Williamson (2000) on the equation ofevidence with knowledge). Without intending to prejudice the rightful con-clusion to this debate, I assume ‒ mainly in order to simplify the expositionof respective positions and arguments ‒ that knowledge can be realised notonly by intellectual episodes, but also by non-intellectual episodes. In linewith this stipulation, I will say, for instance, that perceptual experiences orepisodic memories may constitute knowledge.1 Second, instead of alwayswriting that episodes constitute or realise knowledge of reality, I will alsosay that they cognise reality. This is not meant to imply that the respectiveepisodes are cognitive episodes (i.e., perceptions, memories, judgements,and so on). As will become clear, there may very well be non-cognitive ‒and, more specifically, imaginative ‒ episodes that may realise knowledgein a way very similar to, say, successful perceptions or episodic memories.Third, I normally use the term imaginative to denote the property of beingan instance of imagining, and not the property of being inventive or ori-

1 I also intend this not to imply disjunctivism about episodes or states of knowledge.In particular, I stay neutral on whether belief is a proper constituent of knowledge(see Williamson (2000) for further discussion), or whether perceptions andhallucinations share some or all aspects of their nature (see Martin (2006) andDorsch (2012a) for further discussion).

Introduction to Part One 25

ginal. The imaginativeness of a mental episode is therefore identical withits general status as an instance of imagining, rather than with a specificquality of its representationality: all ‒ and only ‒ imaginings are imaginat-ive in this sense.