Art, Representation, and Make-Believe: Essays on the Philosophy of Kendall L. Walton 9780367370169, 9781032013978, 9780367808662

Citation preview

Art, Representation, and Make-Believe

This is the first collection of essays focused on the many-faceted work of Kendall L. Walton. Walton has shaped debate about the arts for the last 50 years. He provides a comprehensive framework for understanding arts in terms of the human capacity of make-believe that shows how different arts – visual, photographic, musical, literary, or poetic – can be explained in terms of complex structures of pretence, perception, imagining, empathy, and emotion. His groundbreaking work extends beyond aesthetics to address foundational issues concerning linguistic and scientific representation – for example, about the nature of scientific modelling or to explain how much of what we say is quite different from the literal meanings of our words. Contributions from a diverse group of philosophers probe Walton’s detailed proposals and the themes for research they open. The essays provide an overview of important debates that have Walton’s work at their core. This book will be of interest to scholars and graduate students working on aesthetics across the humanities, as well as those interested in the topic of representation and its intersection with perception, language, science, and metaphysics. Sonia Sedivy is Professor of Philosophy at the University of Toronto Scarborough. Her research focuses on perception, aesthetics, and the later work of Wittgenstein. Beauty and the End of Art: Wittgenstein, Plurality and Perception (2016) offers a new approach to the diversity of art and beauty by bringing aesthetics together with the philosophy of perception and the later work of Wittgenstein. “Aesthetic Properties, History and Perception” shows how philosophy of perception and aesthetics inform one another in Art, History, Perception, a Special Issue of the British Journal of Aesthetics that she guest edited. She is currently writing a book on perception that draws on aesthetics.

Routledge Research in Aesthetics

The Aesthetics of Self-Becoming How Art Forms Empower Paul Crowther Philosophy and Film Bridging Divides Edited by Christina Rawls, Diana Neiva, and Steven S. Gouveia Paintings and the Past Philosophy, History, Art Ivan Gaskell Portraits and Philosophy Edited by Hans Maes Radically Rethinking Copyright in the Arts A Philosophical Approach James O. Young Philosophy of Sculpture Historical Problems, Contemporary Approaches Edited by Kristin Gjesdal, Fred Rush, and Ingvild Torsen Art, Representation, and Make-Believe Essays on the Philosophy of Kendall L. Walton Edited by Sonia Sedivy Philosophy of Improvisation Interdisciplinary Perspectives on Theory and Practice Edited by Susanne Ravn, Simon Høffding, and James McGuirk For more information about this series, please visit: www.routledge. com/Routledge-Research-in-Aesthetics/book-series/RRA

Art, Representation, and Make-Believe Essays on the Philosophy of Kendall L. Walton Edited by Sonia Sedivy

First published 2021 by Routledge 52 Vanderbilt Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2021 Taylor & Francis The right of Sonia Sedivy to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilised in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this book has been requested ISBN: 978-0-367-37016-9 (hbk) ISBN: 978-1-032-01397-8 (pbk) ISBN: 978-0-367-80866-2 (ebk) DOI: 10.4324/9780367808662 Typeset in Sabon by Apex CoVantage, LLC

Contents

List of Illustrations 1 Introduction: The Reach of Make-Believe

viii 1

SONIA SEDIVY

PART I

Fiction and the Verbal Arts 2 Fictionality in Imagined Worlds

23 25

S TAC I E F R I E ND

3 Walton and Fictional Characters

41

EILEEN JOHN

4 Walton on ‘the Paradox of Fiction’: Confusions and Misunderstandings

58

D E R E K M ATRAVE RS

5 Fear and Loathing in Fictional Worlds: Quasi-Emotion, Nonexistence, and the Slime Paradigm

74

E . M . DA D L EZ

6 Lyric Self-Expression

94

H A N N A H H . KIM AN D JO H N GIB SO N

7 Reading (With) Others

112

WO L F G A N G H UE ME R

8 The Puzzle of Fictional Morality S TUA RT B RO CK

127

vi

Contents

PART II

Visual Art, Photography, and Music 9 The Puzzle of Make-Believe About Pictures: Can One Imagine a Perception to Be Different?

145

147

SONIA SEDIVY

10 Holey Images and the Roles of Realism

164

J O H N V. K U LVICKI

11 Photography as a Category of Art

177

D I A R M U I D C O STE L L O

12 Transparency and Egocentrism

196

N I L S - H E N N E S STE A R

13 Photographs and Memories

214

C H R I S TO P H ER WIL L IA MS

14 Fiction, Fictionality, and Pictures

230

PA L O M A ATE N CIA - L IN ARE S

15 Understanding Humour, Understanding People, and Understanding Music

247

J U L I A N D O DD

PART III

Themes in Aesthetics: Agency, Appearances, and Norms

263

16 Style and the Agency in Art

265

G R E G O RY C U RRIE

17 Veridical Appearances of Production and Marxist Aesthetics

284

B RYA N J. PA RKH URST

18 Playing With the Rules of the Game: Imagination, Normativity, and Address in Aesthetics M O N I QU E RO E L O FS

302

Contents

vii

PART IV

Beyond Aesthetics: Meaning, Metaphysics, and Science

325

19 ‘Existence as Metaphor?’ Revisited

327

F R E D E R I C K K RO O N

20 Say Holmes Exists; Then What?

345

S TE P H E N YA B L O

21 Scientific Modelling and Make-Believe

367

RO M A N F R I GG

22 The Story of the Ghost in the Machine

384

A DA M TO O N

PART V

Walton in Conversation

401

23 Walton in Conversation

403

K E N DA L L L . WALTO N

List of Contributors Index

413 418

Illustrations

8.1 Percentage of the overall opinion across three stories of each story type 14.1 Kendall L. Walton Photograph of a Doll © Kendall L. Walton 2014 14.2 Jumonji, Bishin © Copyright. Untitled. 1973 14.3 Vicente Fernandez from Argentina hits from the sand to the second hole at Inverness Club in Toledo, Ohio, during the third round of the U.S. Senior Open, Saturday, June 28, 2003. Fernandez bogeyed the hole 16.1 Options for the canvas-context relation 17.1 Ron Mueck, Man in a Boat 2002 © Ron Mueck 18.1 Titian, Sacred Love and Profane Love (ca. 1514) 18.2 Raphael, Lady With a Unicorn (1505–06) 18.3 After Hans Holbein the Younger, Portrait of Henry VIII of England (ca. 1537)

134 236 239

240 273 286 304 306 308

1

Introduction The Reach of Make-Believe Sonia Sedivy

Kendall Walton’s work offers a comprehensive reorientation to the representational arts while reaching well beyond them. He proposes a novel perspective that focuses on our imagination and our capacity for makebelieve, and he highlights how make-believe involves props. Walton shows that focusing on make-believe explains paradigmatic representational arts such as paintings and novels, theatre and film as forms of make-believe with props. But he also shows how this novel perspective extends beyond the arts. His approach offers explanations of pictures and photographs in general not only artistic ones; stories in general as well as literary and performing arts; music; the nature of metaphor, and even the claims we make about fictional entities and existence. The cumulative effect is a framework that brings a variety of endeavours together that are representational in a new sense. Representations of this kind involve our capacity for imagination and overlap with the fictional. We will see that Walton’s framework emphasizes the socially or historically contextual nature of make-believe representation and many varieties of arts.

I. Walton proposes that we need to focus on things that have the function of props in make-believe rather than things we co-opt as props on the fly, as children do in their make-believe games. He eases us into his approach by discussing children’s games, such as imagining tree stumps to be bears. But this is because childhood make-believe is the human capacity that lies at the root of the fact that things have the function of props for makebelieve in social contexts. Walton shows how a complex structure comes into view if we highlight this fact. Firstly, things have the function of props for make-believe only in social contexts where there are norms or prescriptions for certain imaginings in response to features of certain objects – texts or pictures, for example. Secondly, such games involve us as participants. We are prompted by props to participate in make-believe games. We do so by engaging in prescribed imaginings about the props as well as ourselves. DOI: 10.4324/9780367808662-1

2

Sonia Sedivy

As participants in make-believe, we enter imaginary scenarios or ‘worlds.’ We imagine things about ourselves from the inside or experientially. Third, props have an independence from any one of us that gives them a kind of objectivity. What we are to imagine is prescribed by the prop. It follows that what is fictional is determined by props, it is not a matter of what anyone chooses to imagine. Evidently, representations that share this structure are fictional. As Walton writes “to be fictional is, at bottom, to possess the function of serving as a prop in games of make-believe” (1990, p. 102) and a prop mandates or prescribes certain imaginings rather than others. For example, any stump in the forest where the children are playing ‘stumps are bears’ is a bear. That there is a bear covered by leaves next to the creek is a prescribed imagining in the game whether anyone sees the stump or not; it is true in the game. Analogously, a text might evoke and prescribe imagining Lizzie Bennet poking fun at an oily suitor; a picture might evoke and prescribe imagining seeing ships on the high seas. But Walton’s approach also reconstrues the notion of the fictional. It offers a notion that “has little to do with contrasts between fiction and reality or truth and assertion” (1991, p. 380). Rather, what is fictional is what is to be imagined – as constrained and prescribed by props. Though it is sometimes said that there are fictional truths – such as the one about Lizzie Bennet mentioned earlier – Walton suggests we ‘resist’ this manner of speaking because it suggests that truth comes in varieties (1990, pp. 41−42). Instead of “fictional truth,” all we need is what his framework gives us: the notion of what a prop prescribes imagining or what is fictional or what is true in the fiction or make-believe. Yet, Walton is comfortable with continuing to say that fictionality is a property of propositions so that there are fictional propositions – as long as we don’t get hung up on the notion of a proposition and keep in mind that this too is a manner of speaking (1990, pp. 36−37). Again, all we need to countenance is that props mandate specific imaginings. This notion of fictionality as what we are to imagine is one of Walton’s distinctive contributions, and it plays a fundamental role in his framework. Though Walton aligns the notions of make-believe representation and fiction, he distinguishes them in the following respect. Representations are “things whose function is to be props” (1990, p. 52) in games of make-believe and he emphasizes that they need not be artefacts. In contrast, fictions are works, which is to say that they are props that are always human artefacts (1990, pp. 72, 103). This difference is due to Walton’s emphasis that naturally occurring pictures or designs – such as cloud formations or constellations of stars – have the function of props for make-believe games in our societies even though they are not produced or designed for this function. Similarly, he insists that we could read and enjoy a naturally occurring story if the surface texture of a boulder traces out letters that make up words and sentences, for example.

Introduction

3

By explaining a large range of representations in terms of uses of imagination and fictionality, Walton’s approach makes us at least pause and think about representation afresh. If he is right that many representations involve forms of make-believe, his work edges out more typical notions of representation from the core or central role they tend to occupy. These more typical notions include the idea that representations are of or about something; that there is a core notion of representation that divides into non-fictional and fictional varieties with the non-fictional being primary and the fictional derivative; or the idea that linguistic representation provides the model for explaining other kinds. His approach also makes us pause and think about the arts anew, challenging us to reconsider whether and how our responses are imaginative. What is remarkable is how much Walton explains in these terms, about both representational arts and other representations. Like variations on a theme, each account explains a different detailed structure of make-believe distinguished by the nature of the prop and the imaginative experience it mandates. Case by case, the specific explanations Walton provides in terms of make-believe have become leading contenders in each field. They set terms of debate about pictures, photographs, fictional texts, and beyond. To be sure, much of that debate is critical. Most of the chapters in this book examine Walton’s specific explanations critically. I will provide a preliminary outline shortly. But first, his work on the socially or historically contextual nature of the arts and make-believe needs to be brought to the forefront. Walton offers a landmark case for the historical or contextual nature of some artworks. The argument is made in “Categories of Art” (‘Categories’ henceforth) to challenge aesthetic formalism or what has come to be considered more broadly as aesthetic empiricism. But it stands as part of the turn towards contextual or historical explanation in the arts in the second half of the twentieth century. Walton’s stated aim is to delineate a group of aesthetic properties that fall outside of the formalist or empiricist view that aesthetic properties are restricted to what we can perceive in a work on an impoverished view of perception. He argues that there are aesthetic properties that are part of the ‘look’ or ‘sound’ or ‘felt quality’ of a work but that vary with and depend on historical or social context. Walton illustrates that properties such as the vividness of a painting – of Picasso’s Guernica for example – are both historical and perceptible in the following sense. Such aesthetic properties depend on the historical category to which the work belongs and they require trained perceptual skills whereby we perceive the work ‘in’ its historical category. To show that the vividness of Guernica depends on the historical category to which it belongs, Walton examines its aesthetic effects in two different social contexts where it belongs to different art categories. In our world, Guernica is a painting, whereas in the hypothetical scenario

4

Sonia Sedivy

it is a guernica. These are bas-relief type works whose raised surfaces have the colours and shapes of Guernica but in different mouldings so that different parts of the surfaces of each guernica “are molded to protrude from the wall like relief maps of different kinds of terrain” (Walton, 1970, p. 347). There are no paintings in this hypothetical context. The example illustrates how changing the social context and thereby the category to which the work belongs changes some of its aesthetic properties: We do not pay attention to or take note of Guernica’s flatness; this is a feature we take for granted for paintings, as it were. But for the other society this is Guernica’s most striking and noteworthy characteristic— what is expressive about it. Conversely, Guernica’s color patches, which we find noteworthy and expressive, are insignificant to them. It seems violent, dynamic, vital, disturbing to us. But I imagine it would strike them as cold, stark, lifeless, or serene and restful, or perhaps bland, dull, boring—but in any case not violent, dynamic, and vital. (Walton, 1970, p. 347, order reversed) The example shows that aesthetic properties such as vividness do not depend on the non-aesthetic properties simpliciter of a work but on properties that play a normative role for categories of art such as painting or guernica. Guernica has the non-aesthetic property of being flat in both contexts, but it is vivid or dynamic in our context and not in the other. Walton argues that what changes across the two contexts is that flatness is standard and colours and contours are variable for the category or comparison group of paintings, whereas three-dimensional moulding along with the one arrangement of colours and contours are standard for the category or comparison group of guernicas. In our context, the flatness that is standard for paintings makes the markings of the surface which are variable for painting – in this case Guernica’s sharp angles, edges, shapes, and black and white colours – stand out so that Guernica is dynamic or vital or violent. In the hypothetical context, bas-relief moulding of the markings is standard for guernicas, so that its flatness would stand out and Guernica would be bland or serene. As Walton puts the point, it is not (only) the work’s non-aesthetic properties such as colours or contours that determine aesthetic impact but also “which of its non-aesthetic properties are ‘standard,’ which ‘variable,’ and which ‘contra-standard’” (Walton, 1970, p. 338).1 In short, aesthetic properties like being dynamic or dull depend on the prescriptive or normative properties for categories such as painting or guernica to which a work belongs. But are categories such as painting or guernica both historical and perceptible? This is key for Walton’s argument. He needs to do two things: to show that such categories are historical and that they are perceptible, we can learn to perceive items ‘in’ them.

Introduction

5

First, Walton restricts the categories only to those where it can be plausibly argued that we could come to distinguish members of these categories through ‘trained’ perceptual skills: It is no use just immersing ourselves in a particular work, even with the knowledge of what categories it is correctly perceived in for that alone will not enable us to perceive it in those categories. . . . [P]erceiving a work in a certain category or set of categories is a skill that must be acquired by training, and exposure to a great many other works of the category or categories in question is ordinarily, I believe, an essential part of this training. (Walton, 1970, p. 366)2 Second, he argues for the historical nature of categories such as painting or guernica by proposing that category membership is determined by four conditions, two of which are historical. A work belongs in a category if it has “a relatively large number of features standard with respect to the category” and “[t]he fact, if it is one, that [the work] is better, or more interesting, or . . . when perceived in the category than it is when perceived in alternative ways” (Walton, 1970, pp. 357−358). In addition, there are two historical conditions. A work belongs in a category if (i) the category is “well established in and recognized by the society in which [the work] was produced” and (ii) the artist “intended or expected” their work “to be perceived” in the specific category or “thought of it” as being in that category (Walton, 1970, pp. 357−358). There may be numerous cases that are borderline or even undecidable on these criteria – such as innovative works that challenge existing categories and open new ones. But Walton’s point is that typically at least one of the two historical conditions applies so the categories he delineates are objective and historical. These considerations support his conclusion that aesthetic properties that depend on the normative or prescriptive properties for historical, perceptible categories of art can be perceived but only in a way that involves skilled perceptual understanding of the relevant historical category (or categories). These are properties that it is correct to perceive in a work. The significance of this view extends beyond its counter to formalism or aesthetic empiricism to our understanding that some artworks and some of their properties depend on historical facts and categories – that is why it stands as a landmark work in the historical turn in aesthetics in the second half of the twentieth century. But the contextual nature of art is an integral part of the make-believe framework, not just a strand in Walton’s early thought on aesthetic properties.3 Mimesis as Make-Believe (Mimesis henceforth) argues that things can have the function of props in games of make-believe only in social contexts, as we noted earlier. This does not suggest that props require

6

Sonia Sedivy

explicit conventions or that we follow explicit rules. Rather, Walton proposes the more subtle view that there are norms or prescriptions to imagine in that such prescriptions might be enforced – if questions arise. And as he puts it, “there must be social context to enforce the norms.” This entails that all forms of make-believe representation with functional props depend on specific historical context. Together, Categories and Mimesis hold that many works of art are multiply dependent on social context. Let’s continue with painting as our example. A painting is a picture and according to Mimesis, a picture is ‘society relative’ since the function of being a prop that prescribes specific imaginings depends on social context. According to Categories, painting is a perceptually distinguishable and historical category of art. Moreover, paintings belong to several specific categories that are both perceptible and historical such as Cubist or in the style of Picasso. The combined point is that artworks are multiply dependent on social contexts in that they are props for make-believe and they belong to historical media or styles. With this overview in place, let’s consider Walton’s approach in more detail and how the chapters in this book engage with it. In the next section, I will highlight (i) his approach to fiction and the verbal arts; (ii) his explanations of depictions or depictive arts – of pictures, photographs, and music – as forms of perceptual make-believe; and (iii) how his approach to fictional entities expands to broader issues about the arts and to topics outside of the arts such as scientific models and negative existential claims. The third section will briefy examine some important issues that are not covered by the chapters in the book. I will draw out Walton’s implicit view on the question whether art can be defned, and I will do so in relation to Arthur Danto’s work. I will also consider the ontological implications of Walton’s contextualism. Again, I will bring Danto’s view into the discussion. Walton and Danto are arguably the most infuential American philosophers of the second half of the twentieth and the early twenty-frst centuries, and they are both leading architects of historicism about art. Yet, their views do not tend to be considered in light of one another.4 The third section contextualizes Walton’s work in relation to Danto’s as well as to Frank Sibley’s. Walton briefy addresses the issues I raise here in some of his remarks at the conclusion of the book.

II. The volume begins with the verbal arts and Walton’s view of fictional characters and fictions in general. Walton proposes that verbal props such

Introduction

7

as novels or stories prescribe imagining what an explicit or implicit narrator conveys and he highlights the imaginative experiences such props evoke. We have seen that Mimesis explains that prescriptions to imagine are both necessary and sufficient for what is fictional or true in a work. But Walton has recently criticized his own view of fictionality. He now holds that prescriptions to imagine are only necessary but not sufficient (2015). Stacie Friend argues against Walton’s recent view that we cannot give necessary and sufficient conditions for fictionality by drawing on research on situation models and mental models. She argues that a work invites us to imagine a storyworld – to have an immersive de se experience of a complex multidimensional representation of a situation. She details how storyworlds can meet the objections Walton raises against his own earlier view. Her chapter shows how Walton’s notion of imagination might be developed in view of empirical research. Perhaps no part of Walton’s work has stirred as much debate as his account of our emotional engagement with fictions and fictional characters in particular. The key to Walton’s account of fictional characters or entities is the ‘switch of perspective’ that the make-believe framework affords. “The pretense construal has the appreciator pretending to describe the real world rather than actually describing a fictional one” (1990, p.  392). Walton explains metaphysical issues away by showing that we pretend to make assertions about characters and we do so within a make-believe world that we participate in. If assertions about fictional entities occur in make-believe games, the speaker is making assertions from within a fictional world about that world – which is the real world within the game – rather than making assertions from a perspective outside the game about a fictional world. Since the appreciator is a participant in the game, they make it fictional of themselves that they are making true claims about the real world – all the while of course, also knowing that they are engaged in make-believe. To support this approach, Walton analyzes diverse statements that seem to make reference to fictional entities to show that in each case the truth of these assertions can be explained without positing such entities. This is where the difficult detail and controversial argumentation lies. The desired outcome is that: What we should conclude is that it is our pretendings to assert, our games of make-believe, that are central to our conceptual scheme. It is this, not an ontological commitment to fictional entities, that plays an important role in our structuring of the world. (Walton, 1990, p. 404) Eileen John challenges Walton’s view with a realist approach to fictional characters that nevertheless embraces the importance of make-believe. John argues that Walton places too much emphasis on explaining fictional

8

Sonia Sedivy

truths about characters, which leads him to ‘confine’ fictional characters and any references to them to our games of pretence. She counters that to explain the ‘interest and importance’ of the pretence that artworks call for, we need to countenance the variety of what we do with fictional characters that lies beyond make-believe. We use fictional characters to imagine the lives of persons and we do this in a way that keeps the fact that they are representational devices in view. She argues that fictional characters are functional artefacts or representational devices and their reality is demonstrated through our extensive engagement with them. Her emphasis on function provides an alternative to the realist view that fictional entities are abstract created artefacts while showing that realism about fictional characters can be compatible with the make-believe framework. Derek Matravers and Eva Dadlez take up Walton’s much-debated view of our emotional engagement with fictions. Walton holds that the emotional experiences we have while engaged in pretence are not the same as the emotions we have when we are not engaged in make-believe. Many readers are troubled by his insistence that our feelings towards fictions are both ‘quasi’ and ‘real,’ which seems to suggest that they are emotions and yet not genuine emotions. Derek Matravers argues that much of this long-running debate misunderstands Walton’s view. He identifies six main mistakes that run through the critical literature – most notably, that there is a paradox of fiction to which Walton is replying and that Walton denies that we feel genuine emotions. Matravers offers a careful reconstruction that avoids all six ‘misinterpretations’ but shows a subtle remaining problem. He distinguishes between weak and strong imaginings to argue that weak imaginings pose a difficulty for Walton’s view. When we weakly imagine a narrative, we construct a representational model and such models might embed ‘ordinary’ emotional states. But this is contrary to Walton’s view. Matravers argues that there are no grounds within Walton’s approach for denying that it is possible to embed emotions in a mental model. Eva Dadlez challenges Walton’s view of the emotions we feel towards fictions while endorsing the make-believe framework. She offers a battery of considerations that highlight the role of thoughts in evoking feelings or emotions. She argues that our obligations, commitments, and motivations can induce genuine emotions, regardless of whether the object of those emotions exists. Fictions bring such thoughts to the forefront of our consciousness, thereby evoking emotions even though the targets of those feelings are fictional. Moreover, Dadlez points out that we can feel obligations to those who might occupy certain positions rather than to specific individuals, which also suggests that we can have feelings towards fictional characters. Walton’s innovative account of lyric poetry (and some music) is a prime example (Walton, 2015) of how he continues to expand his approach in

Introduction

9

surprising ways. His account of fiction in Mimesis leaves us with the implicit expectation that poems could be explained along the same lines as novels or stories – as prescribing we imagine what a narrator conveys. Instead, Walton recently explains much poetry (as well as some music) as akin to speeches, which are written to be used by others to express their own thoughts and feeling. He proposes that poems are similar in that they don’t use words but mention them, allowing or inviting us to use the words to express our thoughts or feelings (or to use musical motifs for expressing ourselves). This means that poems are not fictional in the usual Walton sense: insofar as the words are not used, the prop does not prescribe specific imaginings. But Walton argues that readers nevertheless engage in a game of make-believe. They pretend to use the words of the poem so that it is fictional in the reader’s game “that he asserts the declarative sentences in the poem” (Walton, 2015, p. 65). This is a wholly distinctive kind of imaginative experience, one in which we borrow someone else’s way of expressing themselves – a form of empathy not with someone else but with their way of expressing themselves. Hannah H. Kim and John Gibson highlight that Walton’s unique approach can explain lyric poetry as expressive without attributing a subject that speaks to us fictionally. But they raise the problem that many poems are voiced from a point of view that it would be inappropriate or unauthorized or even impossible to undertake. They argue that this issue can be addressed from within Walton’s approach. On their suggestion, the expressive subject can be seen as having “an implicitly plural grammatical function” – it is not so much a particular individual as a perspective or point of view that can give voice to ‘multitudes.’ Wolfgang Huemer highlights Walton’s view that acts of imagining evoked by works of fiction have a social dimension. Huemer expands on this social emphasis to argue that fictions have an important social or cultural role that derives from the fact that they offer “encounters with recognizable perspectives that can be attributed to concrete (fictitious) persons.” Such encounters depend on social norms and allow for greater ‘calibration’ and ‘fine-tuning’ with others. Stuart Brock takes an experimental approach to the puzzle of fictional morality to which Walton has drawn our attention. Walton (1990, pp. 154–156) highlights Hume’s ‘contention’ that we respond differently to moral and descriptive claims in fiction. We cannot ‘enter into’ moral claims and sentiments that deviate extremely from our own views whereas we readily entertain highly deviant descriptive claims.5 Walton (1994, 2006) goes on to distinguish three different puzzles: aesthetic, imaginative, and fictional. The first aesthetic puzzle is that deviant moral claims can be considered to be aesthetic flaws but deviant descriptive claims are not. The second puzzle is posed by imaginative resistance – readers resist imagining certain moral situations even though they recognize that these are fictional. The third puzzle of fictionality is perhaps

10

Sonia Sedivy

the most fundamental: readers ‘balk’ at interpreting deviant moral claims as being true in a story. Brock focuses on the third problem. He carries out an experimental study that carefully distinguishes readers’ responses to a variety of fictional stories with either deviant descriptive or moral claims. His findings support the core fictional problem against alternative interpretations and offer several interesting results. To explore Walton’s variations on the theme of make-believe further, the second part of this volume examines his distinctive view that pictures and depictions in general are props in games of perceptual game-believe. But what is perceptual make-believe? Walton points out that when we read a fictional narrative, we might imagine Lizzie poking fun or even imagine seeing her make fun of the oily suitor, but we don’t imagine seeing the words to be seeing Lizzie. As he puts it more technically, we don’t imagine of seeing the words on the page, that we are seeing Lizzie. In contrast, we do imagine of seeing a picture, that we are seeing ships on high seas, for example. Walton’s account of pictures exemplifies his view that perceptual make-believe involves imagining one experience to be a different experience. It yields a general notion of depiction that extends across different kinds of props that evoke experiences in different sensory modalities – or complex combinations thereof such as films and theatre. John V. Kulvicki and Sonia Sedivy probe Walton’s account of pictures. Sedivy examines Walton’s proposal in light of some current theories of perception. Her chapter asks whether the experience Walton posits can be explained by current theories and whether there is something to be learned from the fit or lack thereof between Walton’s account and current approaches. Specifically, she examines Walton’s view in light of theories that explain perception in terms of contents, in terms of relations to objects, and in terms of both contents and relations. Her aim is to get clearer about the specific kind of visual experience Walton posits, one that is both perceptual and imaginative. Kulvicki offers a new perspective on Walton’s approach to pictures and pictorial realism by focusing on the idea that picture-making involves norms and discussing one that has not been previously identified, namely that “pictures should be convex and hole free.” He argues that only Walton’s approach to pictures and pictorial realism predicts this norm. Walton’s view is that a picture is more realistic if it allows for greater perceptual engagement and thereby more imaginative engagement. Specifically, pictures need to support perceptual actions – such as scanning from left to right – that are similar to the perceptual actions one would perform of the imagined scene. Walton’s approach allows us to understand that pictures without holes or concavities are more realistic because they secure a key similarity in our perceptual actions. Scanning a picture from one point to another “corresponds to a represented path between the corresponding points in the represented space” (this book, Chapter 10). But holes disrupt how we scan a picture in a way that has nothing to do with what we are to imagine.

Introduction

11

Walton’s account of photographs is highly controversial because it contends that we see the photographed objects, albeit indirectly. He argues that photographs are transparent pictures, which means that we see through them to the world – much as we see through eyeglasses or telescopes to objects in the world – and we engage in perceptual make-believe with them. As always, Walton’s account turns on his explanation of the prop and the nature of the imaginative experience it evokes. He agrees with theorists who hold that photographs are distinctive in that they result from mechanical links to their object, even if the mechanical process is much manipulated. But he argues distinctively that it is because of this mechanical link to the object that one sees through the photograph to the object, one indirectly sees Aunt Mabel for example. And he adds the hallmark twist: one also imagines seeing Aunt Mabel directly. The distinctive nature of photographic props evokes one experience and evokes and prescribes imagining another one: one indirectly sees the object through the photograph and one imagines seeing the object directly. Four chapters engage Walton’s approach to photographs in this book. Diarmuid Costello argues that in keeping with Walton’s argumentation in Categories, art photographs fall into different historical categories rather than a single kind characterized by transparency. He discusses a variety of photographs where what we see is very different from the objects on which the works depend. In part his point is to drive home that artists can intervene in the mechanical process to an extent that disrupts any grip on the idea that we see the photographed object. But his point is also constructive: if we recognize that photographs like other works depend on standard, variable, and contra-standard perceptible properties, as Walton suggests, we can appreciate how photographs can differ in their kinds, their etiologies, and their aesthetic properties. Nils-Hennes Stear defends the transparency of photographs from the objection that we do not see the objects of photographs because photographs do not provide the egocentric information or connection that is necessary for visual perception. Stear examines different views of vision’s egocentric nature to argue that Walton’s approach can be defended from each one. Christopher Williams offers a way to keep the spirit of Walton’s approach while changing its detail completely. He proposes that photographs are aids to memory – like keepsakes or relics of the past – rather than aids for vision. The connection photographs provide with the past is through memory rather than perception and imagination. This means that photographs are not transparent, they do not allow us to see the past indirectly. But they give us a trace that connects our experience to the experience of the person who took the photograph. Williams’s argument builds on a Humean approach to memory and on the idea, familiar from work on personal identity, that recall of the past across persons could rest on transfer of memory traces from one person to another.

12

Sonia Sedivy

Paloma Atencia-Linares uses photographs to argue against Walton’s recent amendment to his account of fictionality. We have seen that Mimesis explains fictionality in terms of prescriptions to imagine – prescriptions are both necessary and sufficient for what is fictional or true in a work – yet that Walton (2015) now holds that prescriptions to imagine are only necessary but not sufficient. His argument focuses on a variety of photographs to illustrate that some of what we see through a photograph prescribes imagining things that are not fictional in the work. He holds out no hope of “a non-question-begging way of distinguishing” between prescribed imaginings that are fictional and those that are not. Atencia-Linares counters that we do imagine seeing just what is fictional in Walton’s examples but only insofar as we use an antecedent notion of fiction that guides what we imagine. Thus, her defense of the initial view casts doubts on whether Walton’s notion of make-believe can replace the pre-existing categories of fiction and non-fiction that we ordinarily use. Walton controversially explains music as a form of depiction. This means that like pictures, music evokes perceptual games of make-believe: “music still qualifies as representational in our sense: its function is to serve as a prop in listener’s games” (1990, p. 337). Yet, very little music seems to be depictive even in Walton’s sense, as he recognizes. For the most part, the listener does not imagine of hearing some music that they are hearing something else like the booming of a cannon or something more ‘abstract’ like ‘arrival’ or ‘conflict.’ Rather, Walton suggests that much expressive music evokes imaginative experiences where it is fictional “not that one sees or hears or otherwise perceives external things but that one experiences or is aware of (one’s own) feelings or emotions or sensations or sentiments or moods” (1990, pp. 335−336). His controversial thesis is that: In place of fictional perception of external objects we have fictional introspection or self-awareness. If I am right, this is likely to be true even of such stalwarts of musical purity as Bach’s Art of the Fugue; and to whatever extent introspection is analogous to the “external” senses, it will be reasonable to expand our understanding of “depiction” to include it. (Walton, 1990, p. 336) Julian Dodd challenges Walton’s approach as too inward turning or introspective. Dodd focuses on Walton’s argument (2015) that to understand a piece of music is to come to understand the complex intentional content of one’s own experience. Walton draws an analogy to understanding humour. He argues that we come to understand a joke as we recognize and acknowledge what we take to be its objects. This is different from just responding to the causes that make us laugh. But it is a process that remains focused on our own experience. Dodd criticizes Walton’s

Introduction

13

approach to understanding humour and counters with an account of musical understanding and appreciation that is not introspective. The chapters in Part III of this volume take up larger issues that Walton does not address directly. Gregory Currie examines the repercussions of taking a contextualist or ‘inclusive’ approach to the value of an artwork. He argues that a very ‘inclusive’ account of what determines a work’s value is compatible with a ‘conservative’ account of what an artwork is. Currie uses Walton’s arguments against the cobbler model of art as his point of departure. He agrees with Walton that appreciating art is unlike appreciating shoes. Our appreciation of works of art takes into account how they are made. Indeed, Currie argues that the agency of the artist and “facts about making play a central, organizing role in the identification of [a work’s] context” (this book, Chapter 16). But Currie argues that such contextual facts do not affect the ontological identity of a work of art. He argues for a conservative contextualism: works like paintings or sculptures are identical with objects, contextual facts are only relevant for appreciation. Such conservatism opposes expansionist contextualism, which argues that contextual facts are constitutive of the identity of works of art. Currie does not locate Walton’s approach among the positions he identifies. Bryan Parkhurst argues that Walton’s approach provides theoretical resources that can support Marxist art criticism. Parkhurst works with Walton’s view that how a work is made or contextual facts relevant to how a work comes about can be manifest in a work’s appearance. He focuses on new, unpublished work where Walton argues that if a work “appears” to have been made in a certain way, then it has appearance content that includes the proposition that it was created in that way. This recent proposal can explain the Marxist view that facts about the broader context of production may be part of a work’s veridical appearance. Monique Roelofs suggests that even though Walton identifies the ‘generative confluence’ of norms, imagination, and make-believe in art, his emphasis on rules doesn’t allow for transgressive imaginings that are both aesthetically and politically important. Emphasis on rules threatens to hollow out Walton’s orientation to historical context because it doesn’t capture the problematic nature of the ‘cultural sites’ where we engage with art. Roelofs details three iconoclastic interpretations of canonical paintings by Raphael, Titan, and Holbein from a short story by Julio Cortazar to illustrate the kinds of imaginative play that emphasis on rules obscures. She argues that instead of mandating or prescribing imaginings, works of art invite us to address them. Such invitations allow transgressive forms of address that are culturally and politically valuable. The fourth part of this volume reaches beyond Walton’s work on the arts. As Walton emphasizes “works of art are neither the sole nor the primary instances of representation in our sense” (1990, p. 7). Contributors focus on Walton’s account of negative existential claims and on extending his approach to explanatory models, especially scientific ones.

14

Sonia Sedivy

To explain negative existential claims such as ‘Vulcan does not exist,’ Walton expands on the perspectival shift he proposes for dealing with talk about fictional characters. Assertions about fictional characters are made in pretence. In addition to engaging in pretence, Walton suggests that we can also allude to or betray or disavow assertions we make in pretence. A claim about a fictional character like “Gregor Samsa does not exist” disavows attempts at such referential pretence (Walton, 1990, p.  425). A claim like “Vulcan does not exist” does not disavow a particular pretence to refer; it claims that any attempt to refer in this way fails. Yet, pretence still comes into the analysis: “Vulcan does not exist” indicates unofficial games of pretence and asserts that such fictional uses would be unsuccessful. Walton’s approach is anti- or “irrealist.” It suggests that the predicate ‘exists’ does not express a property, it is used to characterize attempts at reference by means of official or unofficial games of pretence. Walton’s approach to existential claims is an especially fertile and controversial part of the make-believe framework. One principal objection is that the analysis does not capture the ‘phenomenology’ of what we seem to be saying when we deny that something exists – we seem to be saying something about the world and not about our attempts to refer. A related objection is that the analysis misfires because it turns existential statements into claims about claims. Instead of explaining the content of a claim that purports to be about Vulcan, Walton’s analysis suggests that we are really making a claim about claims that attempt to refer in this way. Frederick Kroon and Stephen Yablo modify Walton’s approach to address these objections. Kroon shows how Walton’s approach can explain negative existential claims without positing “unofficial games of make-believe” about referring expressions. If we countenance that there is some way that reference is fixed (for example through a causal chain of uses derived from an original baptism), then the same way of fixing reference would also hold in make-believe. “Gregor Samsa does not exist” asserts that the normal way of fixing reference does not pick out anyone in the make-believe. The statement expresses that the condition for fixing reference is not met in the fiction rather than asserting that an attempt at reference fails. Stephen Yablo offers an account “in the same spirit as Walton’s, but different in almost every detail” (this book, Chapter 20). He suggests that we preserve the spirit by modelling negative existential claims on per absurdum conditionals where the antecedent is followed by an absurdity. Consider “If Vulcan exists, then I am a monkey’s uncle.” Entertaining a per absurdum conditional is not very unlike Walton’s idea of pretending to refer in order to disavow or repudiate the possibility of doing so. “Both have us making as if to do something (refer with Vulcan, accept that Vulcan exists) as a prelude to critiquing the very act we have made as if to perform.” Yablo suggests that Walton’s approach has trouble providing content for existential claims because it denies that the singular terms in existential claims refer.

Introduction

15

But Yablo’s proposed route provides such content – content is entertained in the antecedent of per absurdum conditionals only to be denied by showing that it leads to absurdity. Roman Frigg and Adam Toon expand Walton’s make-believe approach to new territory by showing how it helps explain scientific models. Frigg suggests that Walton’s approach is important for explaining scientific models because it provides a detailed explanation of the representationality of make-believe props. Walton’s approach is especially suitable because it provides an account of fictionality in terms of prescriptions to imagine rather than through an opposition between truth and falsity or between truth and “fictional truth.” But Frigg argues that we need to supplement Walton’s approach with an account of how the features of the prop or model get us to the target domain and how exactly they figure in scientific representation. That is, Walton’s approach explains ‘what’ scientific models are, while Frigg provides further resources that explain how “the features of the model figure in scientific representation.” Adam Toon uses Walton’s work to help argue for mental fictionalism. Just as scientific models “represent the world by asking us to imagine it as other than it is,” ordinary talk about ourselves as having inner mental experiences and states is a fictional or imaginative way of capturing real complex patterns in our activities. Walton’s account of metaphor and prop-oriented make-believe provide the details for reconstruing ‘folk psychology’ as prop-oriented make-believe. We are the props; folk psychology is another term for the vast array of prescriptions to imagine about one another, and talk of ourselves as having hopes or fears is metaphoric, understood in terms of pretence within a game of make-believe governed by public prescriptions or rules.

III. With this overview in place, let’s briefly consider Walton’s implicit stand on two much-discussed issues: whether art can be defined and the ontological implications of contextualism. First, I will highlight how Walton’s work implicitly challenges more predominant ways of thinking that hold that art needs to be defined and demarcated from endeavours and objects that are not artistic. Second, I will examine Walton’s historicism or contextualism with the aim of clearing away some misinterpretations to examine its ontological implications. I will contrast Walton’s work on these issues with his contemporary Arthur Danto. Though both Walton and Danto argue for strong historicist positions – and they both often use indiscernible items in their arguments – their views provide independent and very different visions. A feature of Walton’s work that does not receive much attention is that it steers clear of traditional definitional or ontological questions in a way that offers the implicit suggestion that we should do so as well. Walton

16

Sonia Sedivy

spells out his doubts that art can be defined in “Aesthetics – What? Why? And Wherefore?” his 2006 presidential address to the American Society for Aesthetics. But this view can be gleaned from the make-believe framework from the outset. As we have seen earlier, rather than trying to explain or define a clear-cut endeavour such as art or representation, Walton delineates something else – make-believe representation – that cuts across the arts and non-arts. He is explicit about his aim: he is not trying to map exactly onto what we ordinarily say or to revise our categories. Rather, as he puts it, his theory brings into view what it is a theory of: a large swath of representation that centrally involves imagination and spans across many arts as well as non-arts (Walton, 1990, pp. 3, 7). Walton’s distinctive contribution on this score is a sharp counterpoint to Danto’s. Though both provide important arguments for historically contextualist approaches to art, they disagree over essentialism. Danto’s pivotal contribution is to argue that art is both historically contextual and definable. He argues that art’s nature is to embody meanings, and both meaning and embodiment depend on historical context. He shows that a relational definition can capture the nature or essence of art, “eternally the same” “regardless of time and place” (Danto, 1997, pp. 95, 165) while also establishing that art varies because it is related to contextual facts: embodiment of meaning is realized in its specific historical context and indexed to it. In contrast, Walton might seem to take up shop elsewhere – concerning make-believe or fiction – since he does not address the issue of definition. Most discussions contribute to this impression by addressing his specific accounts. But case by case, his explanations of make-believe representations put in place a framework that does not demarcate art. If the approach is correct, then given its cumulative scope, no explicit argument against definition is needed. This may be strategic. A philosophical claim that something can’t be done tends to serve as an invitation to try. The supposed neo-Wittgensteinian denial (in the 1950s) that art can be defined spurred Danto to produce a relational definition (in the 1960s) that opened the floodgates to more definitions across ensuing decades. Walton’s framework silently sidesteps dispute over definitions or theories with the battery of specific explanations just outlined. All but silently that is, except for the opening of Mimesis where he states that if we needed to explain the category of “representational art, we have the interminable and excruciatingly unedifying task of separating art from nonart” (Walton, 1990, p. 2). Walton also does not broach ontological issues. He does not explicitly address what kind of entity a work of art is. A univocal answer here might not be possible given the variety of arts that Walton discusses, including music, dance, theatre, novels, and poetry which perhaps raise distinctive ontological issues from those raised by visual arts such as pictures and photographs. Nevertheless, one might wonder whether his work holds

Introduction

17

some ontological implications. I will focus only on whether Walton’s contextualism holds ontological implications. And for ease of exposition, I will only consider visual art which might seem to be an “easier” case to think about because it can involve objects. In contrast, Danto explicitly focuses on the ontology of works of visual art, and his work distinguishes two options that historicism or contextualism about art yields. Danto argues for a strong view that hinges on the fact that embodiment of meaning depends on historical context in ways that the identity of an object does not. Danto argues that because the identity of an artwork is determined by the meaning it embodies, the work (i) determines which parts or properties of its counterpart object or ‘base’ belong to it, and the work (ii) can have properties that the counterpart object does not have and does not determine. If this is correct, if the work determines which parts and qualities of the bases belong to it, it might be possible to imagine cases in which no material parts and qualities are shared by works whose photographs exactly resemble one another, or which to all intents and purposes are totally similar under sensory scrutiny. (Danto, 1981, p. 102, my emphasis) Danto’s strong view challenges a weaker alternative about artworks. According to the weaker view, artworks are identifiable independently of historical or contextual facts. Works of visual art such as paintings, photographs, or drawings consist in an item that is identifiable independently of contextual factors. Historical factors play a role in the interpretation of the work or in the appreciation of its aesthetic properties, but they are not constitutive for the identity of the work. Unlike Danto’s view, Walton’s is implied not stated. I suggest that we can reconstruct the position without going beyond it in terms of extrinsic properties and coincident entities. But to do so, Walton’s view in Categories needs to be untangled from some prevailing misinterpretations. His view is often glossed in the following sorts of ways: (i) Walton argues that aesthetic properties depend on non-aesthetic properties; or (ii) Walton argues against formalism by showing that aesthetic properties depend on historical categories of art; or (iii) following Sibley, Walton holds that aesthetic properties supervene on non-aesthetic properties. These are not innocuous simplifications; each is incorrect in a way that obscures Walton’s contextualism and the ontological picture it suggests.6 Firstly, here is Walton’s key claim again about the group of aesthetic properties he is concerned with: “a work’s aesthetic properties depend not only on its nonaesthetic ones, but also on which of its non-aesthetic properties are ‘standard,’ which ‘variable,’ and which ‘contra-standard’” (Walton, 1970, p. 338). As I discussed earlier, this clearly states that these

18

Sonia Sedivy

aesthetic properties do not depend on non-aesthetic properties simpliciter, but on non-aesthetic properties that are standard, contra-standard, or variable for the category to which the work belongs. We examined that aesthetic properties like Guernica’s vividness depend on normative or prescriptive properties that are themselves context dependent or extrinsic. So, the first gloss is incorrect and misleads us that the aesthetic properties Walton is concerned with depend on non-aesthetic properties simpliciter. Secondly, Walton explicitly and quite narrowly circumscribes the categories of art that figure in his argument to perceptually distinguishable ones. In the second section of the paper, he specifies the sorts of categories he will be dealing with. “It is necessary to introduce first a distinction between standard, variable, and contra-standard properties relative to perceptually distinguishable categories of works of art” (Walton, 1970, p. 338, my italics).7 He maintains the qualification throughout the paper either in explicit terms or through his focus on perception. The paper in its entirety is concerned with the question of what aesthetic properties are perceivable insofar as a work is perceived in certain correct categories. The second gloss is incorrect because it leaves out Walton’s restriction to perceptually distinguishable categories. This restriction cannot be left out since it circumscribes the historical facts to those that one might perceive in a work with training. For example, to make this clear, Walton specifies that the argument applies to paintings and not to etchings – though it does apply to apparent etchings. The category of etchings as normally construed is not perceptually distinguishable in the requisite sense, for to be an etching is, I take it, simply to have been produced in a particular manner. But the category of apparent etchings, works which look like etchings from the quality of their lines, whether they are etchings or not, is perceptually distinguishable. (Walton, 1970, p. 339, my italics) As Robert Hopkins (2005) puts it, carefully circumscribed categories of this sort make up “Walton’s natural territory.” Perhaps, Walton’s approach could be expanded (Hopkins, 2005). But there is no question that Walton restricts the relevant historical facts and categories to ones that we might be trained to perceive, and his view is that we cannot perceive certain facts of causal origin in a work. His view is that just as we cannot perceive an etching, we cannot perceive a particular work by Renoir though we can perceive a particular work in the style of Renoir. The incorrectness of the second gloss is important. Given that Walton is only concerned with a subset of art categories – those that are historical and perceptually distinguishable – the aesthetic properties at issue are similarly carefully circumscribed. His argument shows that

Introduction

19

these restrictions go hand in hand. That is why the argument allows for myriad other aesthetic properties and is truly silent about them: there may be aesthetic properties that do not depend on the historical and perceptible category to which a work belongs, or that do not admit of right or wrong, or that are not open to perception even with training. It would be helpful to introduce a term for the group of aesthetic properties Walton delineates akin to other terms that have been suggested for his innovations – Waltonian mimicry (Kulvicki, 2006, and this book) or Walt fiction (Friend, 2008, and this book), or even Walton’s natural territory (Hopkins, 2005). But if the phrase Waltonian aesthetic properties feels too long, we need to keep qualifying that these properties are both historical and perceptible for lack of a handy term. Once the argument in Categories is untangled from the first two glosses, we can appreciate that it has the following ontological implications. (I will return to the third issue of supervenience shortly.) Firstly, some aesthetic properties are both perceptible and historical. Such aesthetic properties are extrinsic and they depend on extrinsic properties  – they do not depend on non-aesthetic properties such as colours or contours simpliciter but on normative or prescriptive non-aesthetic properties that themselves depend on social historical context. Secondly, works that belong to historical and perceptible categories of art are such that both their Waltonian aesthetic properties and their determining non-aesthetic properties are extrinsic. (Of course, a painting or a member of another category in “Walton’s natural territory” may have various other non-aesthetic properties.) The account of representational works in Mimesis is consistent in that it holds that artworks are props that have the function of prescribing certain imaginings. The identity of a prop is determined by prescriptions to imagine – and Walton is emphatic that prescriptions are dependent on specific social contexts. This means that the identity of the prop is determined by extrinsic properties and is ‘society-relative.’ If we put Mimesis and Categories together, we get a nuanced picture that emphasizes the prescriptive conditions for representational works of art and for certain aesthetic properties and their bearers. Mimesis tells us that a representational work is a functional prop that mandates or prescribes imaginings so that it is dependent on a social context. Such a prop may be coincident with an entity such as an object which has distinct identity conditions. Categories tells us that works – such as paintings or sonatas – that bear Waltonian aesthetic properties are dependent on extrinsic normative properties that are themselves dependent on social context. This suggests that such works are coincident with entities such as objects. In sum, representational works that belong to historical categories are coincident with things or entities whose identity conditions can be specified in terms of intrinsic properties or in terms of extrinsic properties different from those artworks depend on.

20

Sonia Sedivy

I have used the notion of coincident things or entities to capture Walton’s view. Here, it is important to untangle Walton’s position from the third gloss, which claims that Walton follows Sibley in arguing that aesthetic properties supervene on non-aesthetic properties (simpliciter) like colours or contours. But it is not correct to formulate Walton’s view in terms of supervenience for the same reason that it isn’t correct to apply the term to Sibley’s view. Sibley argued that aesthetic properties such as being serene or dynamic, trite or sentimental are ‘emergent.’ They depend on non-aesthetic properties in the following sense. “Any aesthetic character a thing has depends upon the character of the non-aesthetic qualities it has or appears to have, and changes in its aesthetic character result from changes in its nonaesthetic qualities” (Sibley, 1965, 2001, p. 35).8 Since Sibley states that “changes in its aesthetic character result from changes in its non-aesthetic qualities,” his phrasing seems to capture the core idea of supervenience that “there cannot be an A-difference without a B-difference.”9 But Sibley does not take a stand on any of the dimensions of supervenience that were subsequently differentiated as that notion came under intense scrutiny and debate: whether it provides entailment, reduction, grounding or is ontologically innocent; whether it is a purely metaphysical or an explanatory relation; and whether it is local or global.10 Given the additional fact that Sibley did not choose to use the term ‘supervenience,’ it is not appropriate to apply the notion to the dependence relation Sibley articulated. Similarly, Walton’s argument in Categories predates work that specifies the notion of supervenience and does not commit to that notion. Walton addresses Sibley’s view – as articulated in the early- to mid-1960s – with the primary aim of showing the historical nature of the sort of perceptible aesthetic properties Sibley identified. This is also Walton’s view: he would not characterize Sibley’s approach or his own in terms of supervenience.11 As that notion has been specified, it does not fit the dependence that either he or Sibley were attempting to capture. To detail the mismatch lies beyond the scope of this discussion. The notion of coincident entities – particular artworks and objects for example – captures Walton’s view without going beyond it.

IV. Empathy, imaginative resistance, metaphor, aesthetic values, sports as fiction – these are just some of the topics that Walton addresses but this book does not. Though the chapters written for this volume cover much ground, the fact that they leave much unaddressed is a testament to the scope of Walton’s framework. I hope that the collection will show some of the cumulative effect of his work. Part by part, his approach reconfigures how we think about a large array of human endeavours. Each piece brings

Introduction

21

a role of the imagination into view and turns on that role – accumulating to a stunningly extensive and deeply thought exploration of the human capacity to imagine. To be sure, this is not all. In the attempt to say more than a book can, we close with an informal ‘interview’ with Walton. The contributors joined together to suggest a range of questions, serious and fun, focused and broad ranging, which we hope will help us gain more insight into his thought.

Acknowledgements Many thanks to Andriy Bilenkyy, John Gibson, Derek Matravers, William Seager, and Christopher Williams for helpful comments on earlier drafts of this Introduction. Special thanks to Andriy Bilenkyy and Zachary Weinstein for their tireless editorial work and their helpful suggestions throughout the process of preparing the book for publication. I am deeply grateful to Kendall Walton for his participation and contribution to the volume.

Notes 1. Though I only reconstruct Walton’s Guernica example, his view does not rely on switching contexts and it does not suggest that aesthetic properties are primarily determined by the variable properties of categories of art such as the surface colours and contours of paintings as in this illustration. Walton offers a range of examples, most of which stay within one context and illustrate how certain aesthetic properties may be determined by standard or contra standard properties as well as variable ones. 2. Walton continues “it is no use just immersing ourselves in a particular work, even with the knowledge of what categories it is correctly perceived in, for that alone will not enable us to perceive it in those categories. We must become familiar with a considerable variety of works of similar sorts” (1970, p. 366). 3. There is an important change between these two key works but it concerns representationality. Categories includes representational or resemblance properties of visual art among the properties determined by the normative features of a historical category. Mimesis provides the new approach to representationality in terms of make-believe which applies to depictive visual art – as I outline in Section II. 4. But see Sedivy (2018). 5. Hume, “Of the Standard of Taste” (1757) in Paragraph 33. 6. One might counter that the first and second ways of glossing Walton’s views are just incomplete. Strictly speaking, this is correct since they could be added to. But my point is that they are typically stated as is without further qualification. Thanks to Derek Matravers for this point. 7. Walton continues: “A category will not count as ‘perceptually distinguishable’ in my sense if in order to determine perceptually whether something belongs to it, it is necessary (in some or all cases) to determine which categories it is correctly perceived in partly or wholly on the basis of non-perceptual considerations” (1970, p. 339). 8. Sibley’s other key point is that uses of aesthetic predicates for such aesthetic properties are not condition governed. The conditions for an aesthetic claim are only sufficient and defeasible (1959, 2001).

22

Sonia Sedivy

9. Yet, see Sedivy (2016, pp. 212–216). 10. McLaughlin and Bennett (2018). 11. See Walton’s reply to the first question in Walton in Conversation (this book, Chapter 23). He writes that the notion of grounding best fits the dependence relation Sibley proposed and that on his own view aesthetic properties do not supervene on non-aesthetic ones.

References Danto, A. C. (1981). The Transfiguration of the Commonplace. Harvard University Press. Danto, A. C. (1997). After the End of Art, Contemporary Art and the Pale of History. Princeton University Press. Friend, S. (2008). Imagining Fact and Fiction. In K. Stock & K. Thomson-Jones (Eds.), New Waves in Aesthetics (pp. 150–169). Palgrave Macmillan. Hopkins, R. (2005). Aesthetics, Experience and Discrimination. Journal of Aesthetics and Art Criticism, 63, 119−133. Hume, D. (1757/1985). On the Standard of Taste. Reprinted in E. Miller (Ed.), Essays: Moral, Political and Legal (pp. 227–249). Cosimo, Inc. Kulvicki, J. (2006). On Images: Their Structure and Content. Clarendon Press. McLaughlin, B., & Bennett, K. (2018). Supervenience. The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/archives/win2018/entries/supervenience/ Sedivy, S. (2016). Beauty and the End of Art Wittgenstein, Plurality and Perception. Bloomsbury Academic. Sedivy, S. (2018). Aesthetic Properties, History and Perception. British Journal of Aesthetics: Art, History and Perception, 58(4), 345−362. Sibley, F. (1959). Aesthetic Concepts. Philosophical Review, 68(4), 421–450. Sibley, F. (1965). Aesthetic and Non-Aesthetic. Philosophical Review, 74(2), 135–159. Sibley, F. (2001). Approach to Aesthetics, Collected Papers on Philosophical Aesthetics (J. Benson, B. Redfern, & J. R. Cox, Eds.). Clarendon. Walton, K. L. (1970). Categories of Art. Philosophical Review, 79(3), 334−367. Walton, K. L. (1990). Mimesis as Make-Believe, on the Foundations of the Representational Arts. Harvard University Press. Walton, K. L. (1991). Precis of Mimesis as Make-Believe: On the Foundations of the Representational Arts. Philosophy and Phenomenological Research, 51(2), 379−382. Walton, K. L. (2008). Marvelous Images. Oxford University Press. Walton, K. L. (2015). In Other Shoes, Music, Metaphor, Empathy, Existence. Oxford University Press. Walton, K. L. (1994). Morals in Fiction and Fictional Morality I. Proceedings of the Aristotelian Society Supplementary Volume, 68(1), 27–50. Walton, K. L. (2006). On the (So-Called) Puzzle of Imaginative Resistance. In S. Nichols (Ed.), The Architecture of the Imagination: New Essays on Pretence, Possibility, and Fiction (pp. 137–148). Oxford University Press.

Part I

Fiction and the Verbal Arts  

2

Fictionality in Imagined Worlds Stacie Friend

I. Introduction It is fair to say that Kendall Walton is responsible for bestowing on imagination the central place it holds today in various domains of aesthetics and the philosophy of art. Elaborated most fully in Mimesis as MakeBelieve (Walton, 1990), his proposal that we should conceptualize not just fiction, but representational art more generally, in terms of its role in prompting imaginings in games of make-believe is a position some may embrace and others reject, but no one – at least no one interested in these topics – can ignore. Among the many issues that Walton’s theory has influenced is the debate concerning how to understand “truth in fiction,” or fictionality in Walton’s terminology. It is in Walton’s sense fictional that Offred is a handmaid, but not that she is the Commander’s wife (in Margaret Atwood’s 1985 novel The Handmaid’s Tale). Since Offred and the Commander do not exist and the events of Atwood’s novel never actually took place, what justifies these conclusions? In virtue of what does a proposition count as fictional? One answer, derived from David Lewis (1983), is that “truth in fiction” is truth at a certain set of possible worlds. Walton rejects the claim that “fictional truth” is any kind of truth at all. Instead, he proposes that what is fictional is what a work prescribes imagining. In reading Atwood’s novel, we are supposed to imagine that Offred is a handmaid, not that she is the Commander’s wife; and this is why the former is fictional and the latter is not. Recently, Walton (2015) has raised objections to his own account. He now maintains that while a prescription to imagine is necessary for fictionality, it is not sufficient. He further argues that there are no systematically identifiable sufficient conditions. I disagree. To avoid Walton’s conclusion, what is needed is a more detailed account of the kind of imagining relevant to fictionality. I propose to develop such an account in this chapter. I proceed as follows. In the next section, I delineate the scope of fictionality by considering the range of representations to which the concept DOI: 10.4324/9780367808662-3

26

Stacie Friend

applies, which goes well beyond works of fiction in the ordinary sense. In Section III I argue that the kind of imagining that determines what is fictional is imagining a storyworld. I elaborate an account of the content of this kind of imagining by appeal to cognitive psychologists’ notion of constructing a mental model or situation model. After further clarifying the proposal in Section IV, in Section V, I address Walton’s objections to his own view. I argue that the framework of situation models has the resources to overcome the challenges he poses. However, the appeal to situation models does not explicitly concern imagining; situation models reflect only the content imagined, rather than the attitude taken toward it. In Section VI, I propose an account of the relationship between fictionality and imagining that does justice to the ambitions of Walton’s theory.

II. Fiction and Fictionality To grasp the scope of Walton’s notion of fictionality, it is essential to understand his conception of fiction. In Mimesis as Make-Believe (henceforth: Mimesis), Walton uses the terms fiction and representation (or representational art) interchangeably. His ambition is like that of Nelson Goodman (1976): to offer an account of representation across art forms, the core features of which for Walton are exemplified by paradigmatic works of fiction. Goodman defined representation in terms of denotation, making it difficult to accommodate non-denoting representations such as paintings of unicorns and novels about fictional characters. Walton instead defines fictions or representations by their function in serving as props in certain games of make-believe: specifically, games in which we are meant to imagine the contents of the works. The connection to children’s games is intentional. Just as the child pushing a toy truck across the floor while imagining that it is rumbling down the road plays the sort of game appropriate for the toy, the reader of Chinua Achebe’s Things Fall Apart (1958) who imagines that Okonkwo kills Ikemefuna plays the sort of game that is authorized for the novel. Not only does Walton’s approach avoid the problem Goodman faced with non-denoting representations, it also easily accommodates non-linguistic fictions (by contrast, for instance, with various definitions of fiction that rely on speech act theory).1 The category Walton carves out does not line up with ordinary classifications, however. For example, representations in Walton’s sense need not be intentionally created: rocks or clouds can serve as props in games of make-believe if they are accorded that role in a social practice (1990, pp. 87−88). Nonfiction texts such as histories and biographies are difficult to accommodate. On the one hand, it is counterintuitive to deny that they are representations; on the other, it is problematic to identify them as fictions. I have argued elsewhere that many works of nonfiction invite imagining in one form or another and should therefore be included in

Fictionality in Imagined Worlds

27

Walton’s category (Friend, 2008). For example, many histories ask us to imagine what it was like to live in another time and place. Easier (though less intuitive) to classify are pictures, which for Walton invite imagining by definition. To “see-in” a picture, in Richard Wollheim’s (1992) phrase, requires imagining, of looking at the picture, that it is seeing the represented scene. The same holds for photographs. Though Walton (1984) notoriously maintains that they are “transparent,” allowing us literally to see their subjects, this is indirect seeing; we still imagine that we are directly seeing the subjects, face-to-face. As a consequence, even journalistic photographs are classified as fiction for Walton. Because Walton’s concept of fiction or representation cuts across familiar distinctions – such as between artifacts and natural objects, fiction and nonfiction – it is useful to adopt a term of art, waltfiction, for the category (Friend, 2008). (I reserve the terms ‘fiction’ and ‘nonfiction’ for the ordinary, narrower categories, according to which novels and plays are fiction and histories and biographies are nonfiction.) I will focus on waltfictions that are uncontroversially works. For current purposes, then, waltfictions are defined as those works designed (or taken) to prescribe imaginings about their content, where imagining the content is participating in the game of make-believe authorized by the work. The invocation of games highlights two essential features of Walton’s account (Friend, 2014). The first is the significance of participation: rather than looking in at a fictional world from the outside, readers and audiences imaginatively participate in that world. In Walton’s theory, imaginings about the content of a work are generated by more basic de se games of make-believe, which differ according to the modality of the representation. When I look at Goya’s The Third of May 1808 (1814), I am supposed to imagine seeing (face-to-face) Napoleon’s troops about to massacre innocent civilians – or more accurately, to imagine of my looking at the painting that it (my looking) is seeing Napoleon’s troops ready to open fire. When I read Chinua Achebe’s Things Fall Apart, I am supposed to imagine of my reading the novel that it (my reading) is reading about people and events who share my world.2 Goya’s painting prescribes imagining that the troops are about to open fire in virtue of the fact that, within an authorized game of make-believe, we see that this is so. Similarly, Achebe’s novel prescribes imagining that Okonkwo kills Ikemefuna in virtue of the fact that, within an authorized game of make-believe, we read that this is so. These points draw attention to the second feature of Walton’s account associated with the appeal to games: the normativity of representation. Some moves in a game are licensed, while others are not. Just as nothing prevents a child from using the toy truck as a hammer, nothing prevents the reader from imagining that Okonkwo spares Ikemefuna’s life. Still, there is a clear sense in which this imagining is not authorized by the novel. Only some imaginings reflect what is “true in the fiction,” or in Walton’s terminology, fictional. What is fictional is what we are supposed to imagine.

28

Stacie Friend

Walton’s use of the term fictional is unfortunate in two ways, however. First, “fictionality” is ambiguous between a property of works (those classified as fiction) and propositions (those that the works prescribe imagining). Second, even when we restrict ourselves to the latter sense, describing the propositions as “fictional” is misleading given the breadth of the category of waltfiction. As noted previously, Walton’s aim is to avoid the implication that “fictional truth” is any kind of truth, such as truth at a set of possible worlds. But what he identifies is a broader notion of representational content for the full range of waltfictions and not merely for works of fiction in the ordinary sense. To avoid confusion, I will adopt a term I have introduced elsewhere (Friend, 2017): I will say that the content a waltfiction invites us to imagine is storified, and that the storified propositions are the story-truths.3 My aim is to give an account of storified content that avoids Walton’s objections.

III. Imagining a Storyworld Walton does not provide an account of the kind of imagining that is relevant to storified content. When a waltfiction invites us to imagine that P, we must, at a minimum, form a mental representation with the content P. When we imagine in response to a waltfiction, though, we do not merely imagine a series of propositions; rather, we imagine a world. The idea of “fictional worlds” is familiar, but the concept extends beyond novels and plays to other waltfictions. Looking at The Third of May 1808, I do not construe the scene I imaginatively witness to occur at a frozen moment in time, spatially cut off from anything else, populated by flat cutouts of people. To the contrary, I assume that there is a before and after, a surrounding landscape, “backsides” of the figures, and so on. In short, I imagine a world in which the events take place. I agree with Walton that we should reject an account of storified content that relies on the notion of truth at possible worlds. Instead, I take imagining a world to indicate a certain kind of mental activity, involving the construction of what psychologists call a situation model (van Dijk & Kintsch, 1983) or mental model (Johnson-Laird, 1983). This is a complex, multidimensional representation of a situation, what I will call the storyworld: people and things, their properties and relations, the unfolding events, settings, and so on. The notion of a situation model originates in the psychology of text comprehension, where it is typically contrasted with two other “levels” of memory representation associated with content: the surface code, which is the exact wording of a text, and the textbase, which is the semantic content of explicit sentences. Psychologists agree that understanding must go beyond these levels; to genuinely comprehend what we read, we must construct a “mental microworld of what the story is about” (Graesser et al., 2002).

Fictionality in Imagined Worlds

29

Though they may have propositional content, situation models are usually construed as having an analogical structure, for instance, representing relations between entities in an abstract spatial framework (see Tapiero, 2007). Many psychologists point to evidence that they include modality-specific representations, whether or not these are experienced as conscious imagery.4 This is compatible with our being able to describe the content of a work propositionally. Having constructed a complex representation of Okonkwo – his physique, his thoughts and feelings, his actions and relations to others, and so on – we will treat certain propositions about Okonkwo as storified: for instance, that Okonkwo is proud. (With pictures, the relevant propositions may include demonstratives: for instance, that the central figure’s arms are raised in that way.) The key point is that a situation model is a mental representation of a dynamic scenario rather than just a set of propositions. The details of the psychological literature on situation models and mental models will not concern me here (though I say more in Section 5).5 Instead, I want to indicate why the contents of situation models are good candidates for the contents of the imaginings invited by waltfictions. First, imagining a storyworld is plausibly objectual imagining: imagining O (the storyworld), not just imagining that P (Yablo, 1993).6 When we construct situation models, we form mental representations of individuals and their properties and relations – a “microworld” in exactly this sense. Second, situation models are postulated to explain how we comprehend texts and representations in other media. It is by forming mental representations of people, events, and situations that we are able to follow complicated stories, integrate information from different sources, and so on. Finally, situation models are constructed in response to different media. For example, we seem to form the same kinds of mental representations of narrative events when we watch films as when we read texts (Zacks et al., 2009). We are also able to integrate textual and visual information into the same mental representation of a situation (see, e.g., Glenberg & McDaniel, 1992). In other words, we construct representations that are, at some level, independent of source modality: representations of the situations depicted and not only how they are depicted. For Walton, the how is instead reflected in the format of the de se games of make-believe. Given that determining what is storified is essential to comprehending a story, the appeal to situation models has the potential to illuminate story-truth. In light of this connection, my proposal will be that what is storified by a waltfiction – the content we are meant to imagine – is (roughly) what the work invites us to include in our situation models. Notice that this proposal concerns only the content of imagining when we imagine a storyworld; it is not an account of what constitutes “imagining a storyworld.” There is no doubt that constructing a situation model is necessary for imagining a storyworld; one could not so imagine

30

Stacie Friend

without mentally representing the individuals and events portrayed. But does the construction of a situation model suffice in itself for a kind of imagining? A few theorists would say yes. For example, some psychologists take situation model construction to be an imaginative activity, allowing us to take the “mental leap into imagined worlds” (Zwaan, 1999). And the idea that mentally representing situations that are outside one’s current experience constitutes an exercise of imagination is not unprecedented.7 However, there are also reasons to treat imagination as narrower in scope. This is not only because we construct situation models in response to nonfictions whose contents we believe in their entirety (Friend, 2008; Matravers, 2014).8 It is also because describing all mental representations of situations as “imagining” potentially empties the concept of any theoretical interest. I will therefore not assume that constructing a situation model is sufficient for imagining a storyworld. However, this does not mean that my proposal fails as an account of fictionality in Walton’s sense. What is fictional or storified is the content to be imagined; specifying content is distinct from specifying the attitude taken toward it. If I imagine that ostriches can fly, I imagine the same proposition that some might believe, others desire, and so on. I return to the role of imagining in an account of story-truth in Section VI. For now, my proposal is restricted to the claim that what is storified in a waltfiction – the content to be imagined – is (roughly) what the work invites us to include in our situation models.

IV. Distinguishing Obligations In its rough form the proposal is subject to several objections. One worry is that some waltfictions, such as Escher paintings or Robbe-Grillet fictions, storify what we (arguably) cannot imagine, such as metaphysical impossibilities or logical contradictions.9 Certainly, we cannot construct a coherent situation model that includes contradictions. This is not a serious difficulty, though. As Walton points out, “There can be prescriptions to imagine a contradiction even if doing so is not possible” (1990, p. 64). In such cases we recognize that we are invited to imagine both P and notP, even if we do not manage the task.10 A more serious concern is that what is storified goes far beyond what we are required to imagine. Although it is storified in Goya’s painting and in Achebe’s novel that the people portrayed are prevented from floating into the air by gravity, we can appreciate either work without ever entertaining this story-truth (even implicitly). We should therefore reject the assumption that we are required to imagine everything storified. Some kinds of imaginings are plausibly mandated, insofar as they reflect a minimal requirement on understanding a work. A reader would fail to grasp the basic plot of Things Fall Apart if she failed to represent Okonkwo as killing Ikemefuna. A fuller appreciation requires representing more than

Fictionality in Imagined Worlds

31

this; for instance, readers should recognize that Okonkwo feels terrible guilt about the killing, even though this is not explicitly acknowledged. Still, even full appreciation does not demand entertaining every proposition that is storified. Walton adds a qualification to his original account to avoid the implication that we must imagine everything that is fictional. Here is his most recent articulation of the original view: “a proposition is fictional in (the world of) a particular work, W, just in case appreciators of the work are to imagine it, just in case full appreciation of W requires imagining it” (2015, p. 17). In a footnote to this formulation, Walton draws attention to a key but often overlooked passage in Mimesis: A proposition is fictional, let’s say, if it is to be imagined (in the relevant context) should the question arise, it being understood that often the question shouldn’t arise. In normal cases the qualification can be understood thus: If p is fictional, then should one be forced to choose between imagining p and imagining not-p, one is to do the former. (Walton, 1990, p. 40) Though Walton presents these claims as articulating the same position, there are actually a variety of different obligations identifed in these passages, which we would do well to distinguish. Using terminology I have introduced elsewhere (Friend, 2017), I draw the following distinctions among the obligations. A work mandates imagining that P (i.e., including P in our situation models), if failing to imagine that P would mean failing to comprehend the work at the most basic level. A work prescribes imagining that P if including P within our situation models contributes to full appreciation of the work (however full appreciation is to be understood). A work invites imagining on the following condition, adapted from Walton’s qualification: If the question arose and we had to choose between including P and including not-P in our situation models, we would be required to include the former. Although Walton often describes the condition for fictionality in terms of “full appreciation,” suggesting that what is at issue are prescriptions to imagine, I would argue that storified content is best captured by invitations to imagine in the aforementioned sense – that is, by Walton’s qualification rather than by his official position. For instance, I take it as storified that gravity operates in the worlds of The Third of May 1808 and Things Fall Apart. However, we can fully appreciate each of these works without ever representing this in our situation models. Still, if the question arose it would be absurd to deny that there is gravity in the relevant storyworlds. Notice that there is a great deal that we may include in our situation models that goes beyond the content of a waltfiction. Different readers

32

Stacie Friend

might imagine Okonkwo’s appearance in Achebe’s novel differently; different viewers might imagine different thoughts going through the minds of the soldiers in Goya’s painting. Neither the precise appearance nor the exact thoughts are storified, so these different imaginings are all compatible with the representational content. If the question arose, we would not be obligated to imagine that Okonkwo’s nose is of one shape rather than another, assuming that this is left indeterminate in the text. We are authorized or permitted to fill in such aspects of the storyworld as we wish, in our distinct games of make-believe.11 So we are invited to include a great deal more than we should include for full appreciation, or must include for minimal understanding, but less than we are permitted to include. With this in mind, I return to my proposal: that what is storified is (roughly) what a waltfiction invites us to include in our situation models, with the invitation to be understood in the aforementioned technical sense. Even when clarified in this way, however, the proposal faces further objections.

V. Sufficient Conditions One such challenge has led Walton to reject his original account of fictionality, concluding that “[p]rescriptions to imagine are necessary but not sufficient for fictionality” (2015, p. 17). The problem is that many works invite us to imagine – to include in our situation models – representations that are not storified. Walton discusses several kinds of case, which generally fall into two categories. The first sort of case is exemplified by plot twists in literature and film. For example, in Agatha Christie’s And Then There Were None (1940), readers are invited to imagine that Justice Wargrave is fatally shot, the sixth victim among the ten. It is only at the end that we discover from Wargrave’s confession that his death was faked and that he is the murderer. In response to Fight Club, readers of the novel by Chuck Palahniuk (1996) or audiences of the film by David Fincher (1999) are invited to imagine that the narrator organizes a “fight club” with the saboteur Tyler Durden; eventually we (along with the narrator) find out that Tyler Durden is the alternate personality of the narrator himself. In the second sort of case, the waltfiction contains an embedded representation. Walton offers examples of “iconic meta-representations,” such as Vermeer’s A Young Woman Standing at a Virginal (Walton, 2015, p. 20). The painting depicts a painting on the wall of the room in which the young woman is standing, of Cupid holding a card. For Walton, to understand the content of the painting-within-a-painting the viewer must imagine seeing Cupid, which entails imagining that there is a Cupid. Yet, there is no Cupid in the world of Vermeer’s painting. In Chaucer’s Canterbury Tales, each pilgrim tells a story, and readers are invited to imagine the content without necessarily taking it to reflect what is the case in

Fictionality in Imagined Worlds

33

the world of the pilgrimage. In response to the Wife of Bath’s Tale, for example, we are meant to imagine that an old woman magically becomes young once the Knight grants her power over him. Yet this is not storified in the world of The Canterbury Tales itself. With such examples in mind, Walton concludes that prescriptions to imagine are necessary but not sufficient for fictionality. In my terms: if P is storified, we are invited to include it in our situation models, but the converse does not hold. If this is right, what we are invited to include in our representations of the storyworld outstrips the story-truths. To address this problem, more detail about situation models is called for. Plot twists can be accommodated by considering the way in which situation models for narratives are constructed over time. The influential Event-Indexing Model is designed to offer an account of this process (Zwaan et al., 1995; Zwaan & Radvansky, 1998). I will explain the account with respect to reading, which is how it was originally proposed.12 At any given moment we construct a current model corresponding to a sentence or clause. For instance, the first sentence of And Then There Were None reads, “In the corner of a first-class smoking carriage, Mr Justice Wargrave, lately retired from the bench, puffed at a cigar and ran an interested eye through the political news in The Times” (Christie, 2007, p.  1). The sentence prompts readers to assume a spatiotemporal framework (a train carriage interior), within which are tokens of a person, a newspaper, and so on, with their properties and relations. In reading further, we continuously build up an integrated model, a global model that incorporates all the current models up to that point in the process of updating. We do this by making inferences from the explicit text that connect the various elements into a coherent representation of a full scenario. By the end of the novel, readers will have constructed a complete model, which is an integrated model resulting from processing the whole text. If we reflect on our reading later, amending the representation, we may arrive at a final model. Perhaps we think in more detail about why Wargrave did what he did, amending our mental representation. The solution to the challenge posed by plot twists emerges directly from this account: Rather than saying that what is storified is what we are invited to include in our situation models simpliciter, we say that it is what we are invited to include in the final model, after we have finished reading (or watching) and engaged in further reflection. Up until a certain point in Christie’s story, the reader’s integrated model may represent Wargrave as dead, but this will be updated by the time she constructs a final model. To address embedded representations, including Walton’s iconic metarepresentations, we must assume that situation models are internally structured. A reader who failed to keep apart the pilgrimage and prologues from the stories the pilgrims tell would fail to construct any coherent model of The Canterbury Tales. Similarly, a viewer who took

34

Stacie Friend

the painted Cupid to be present in the room with the young woman in Vermeer’s painting would fail to understand the content of the picture altogether. To address such cases, some theorists introduce distinct “mental spaces” or “sub-worlds.”13 I will borrow an idea from Leda Cosmides and John Tooby (2000), and say that the embedded representation must be associated with a “source tag” indicating its origin – for example, painting or Wife of Bath – which prevents or (in Tooby and Cosmides’s terminology) “buffers” it from interacting freely with other elements of the situation model, for instance, in supporting inferences. Thus, we include in our situation models for The Canterbury Tales, not a representation with the content that an old woman magically becomes young, but instead one with the content that in the Wife of Bath’s tale, an old woman magically becomes young. The source tag prevents us from drawing mistaken conclusions, such as that magic operates in the world inhabited by the pilgrims. A coherent final model for Vermeer’s painting will include an untagged, unbuffered representation of the painting on the wall and a buffered representation of Cupid, associated with the painting qua source. And a coherent final model for Chaucer’s tales will include an unbuffered representation of the Wife of Bath along with a buffered representation of the Knight and his deeds, associated with her telling qua source. If we include these refinements, my proposal can be formulated more carefully: What is storified in a waltfiction is what we are invited to include unbuffered in our final models for the work. So, it is storified in Vermeer’s painting that in the painting Cupid is holding a card, but not that Cupid is holding a card. And it is storified in Chaucer that in the Wife of Bath’s tale an old woman magically becomes young, but not that an old woman magically becomes young. Now, Walton (2015) considers something like my solution to the plot twist cases, but rejects it on the grounds that it does not by itself accommodate embedded representations. It is not clear to me why all the cases have to be handled in exactly the same way. As Richard Woodward (2014) points out, different problems may call for different solutions. In any case, my solutions are not unrelated; they are both independently plausible clarifications of the situation models picture. My proposal still does not mention imagining, however. In virtue of what does including P in our situation models constitute imagining that P? I consider this question in the next section.

VI. Kinds of Imagining I have suggested that the kind of imagining relevant to storified content is imagining a storyworld. I have also argued that we represent storyworlds by constructing situation models, and that P is storified so long as a work invites us to include it in our situation models – or more accurately, to

Fictionality in Imagined Worlds

35

include it unbuffered in our final models.14 It follows that what is essential is not the attitude we take toward P in isolation, but rather the attitude we take to the situation model as a whole. For this reason, I will say that we imagine that P in response to a waltfiction so long as we include P in a situation model that is imagined. In other words, imagining that P in response to a waltfiction means representing it as part of an imagined storyworld. The important question is, then, what it takes to imagine a storyworld, where this is not merely to construct a situation model, but to engage in a certain activity or take a certain attitude toward the model. What is this attitude or activity? Because imagining is heterogeneous (Dorsch, 2012; Kind, 2013), there are a variety of different possible answers. Here I consider two kinds of imagining that can plausibly be construed as “imagining storyworlds.” I will argue that only one of them makes sense for the scope of waltfiction. The first emerges from a familiar contrast between believing and imagining. Many forms of imagining are compatible with belief, such as experiencing mental imagery. However, the contrast concerns imagining in a specific sense, what I call mere-make-belief (Friend, 2008): that is, imagining that P without believing that P. Standardly construed as propositional attitudes, belief and mere-make-belief are distinguished according to their functional roles in cognition. Though we may make inferences and experience affect in response to both, only beliefs play a role in guiding ordinary actions, while the implications and effects of mere-makebeliefs are typically confined to a limited domain (see Liao & Gendler, 2019). For instance, if I believe that I have been bitten by a poisonous snake I am likely to rush to the hospital, whereas merely imagining the same, no matter how much “fear” I experience, need not move me off the sofa.15 Mere-make-beliefs are often described as quarantined from beliefs and therefore unable to motivate action. One possibility is that when we imagine a storyworld, we construct a situation model quarantined from our beliefs. However, it would be impossible to construct situation models for waltfictions this way, not least because we rely on background knowledge to fill in the gaps left open by a representation. For instance, when I assume that the soldiers and victims in Goya’s painting have backsides, I import my knowledge of ordinary human beings. On the other side of the coin, if I learn something about Napoleon’s troops from viewing the painting, I export this information to my beliefs.16 Situation models cannot be quarantined. Kathleen Stock (2017) proposes a way to include believed material within the content of what is imagined, though she puts the point in purely propositional terms. She maintains that we count as “imagining that P” even where P is believed, so long as it is inferentially connected to at least one other proposition that we do not believe. To borrow her example, suppose that I am in fact typing on my computer. I can still

36

Stacie Friend

imagine that I am typing on my computer if I also imagine that my office is on the moon (Stock, 2017, p.  146). In line with this suggestion, we could say that so long as a situation model includes at least some representations that we do not believe to be accurate concerning the real world, it represents an imagined storyworld, and our attitude toward the content constitutes imagining. Because most works of fiction include at least some invented material, it is plausible that (most) fiction invites imagining in this sense (Friend, n.d.). However, the same cannot be said for waltfiction. The category of waltfictions includes works whose storified contents we take to be entirely true, such as journalistic photographs and accurate historical narratives. The situation models we construct in response to such works are fully believed. If they prompt imagined storyworlds, this must be for some other reason. I suggest that a better approach will appeal to something like imaginative immersion, which occurs when we find ourselves transported into a storyworld.17 Psychologists of narrative comprehension measure transportation according to a scale that reflects levels of absorption, imagery, and affect (Green & Brock, 2000). In other words, there is something it is like to imagine in this sense, where the phenomenology involves sensorimotor imagery and/or experiential imagining – imagining doing or being – along with close attention and emotional response. This sort of imagining cuts across the fiction/nonfiction divide in a way that fits with the scope of waltfiction. Vivid historical narratives or memoirs can transport us into the lives of others, for example. Of course, not every waltfiction will invite a fully immersive experience, where we are “lost” in the storyworld. However, I suggest that given the role of de se imagining in Walton’s account, every waltfiction will by definition invite imagining that involves some degree of transportation: specifically, experiential imagining. For Walton, looking at pictures of any sort involves such a phenomenological or experiential dimension: not merely the experience of seeing the picture but also of imaginatively seeing what the picture portrays. Many texts prompt imagery, but even when they do not, narratives typically offer the experience of learning about events as they unfold. When I read Simon Schama’s Rough Crossings (2006), about slaves who fled the British to fight on the American side in the Revolutionary War, I respond to the events as if they are happening “now” rather than centuries ago. This here-and-now perspective plays an important role in emotional engagement, including anticipating what will happen next. Walton explains suspense in just these terms: Even when we know the outcome, we adopt the position of the ignorant within our games of make-believe, so that we experience anticipation, surprise, relief, and so on (1990, pp. 259−271). Let us suppose, with Walton, that all waltfictions prompt de se games of make-believe. At a minimum, this means that we imagine of our own

Fictionality in Imagined Worlds

37

experiences that they are experiences of learning – whether through reading, seeing, hearing, or what have you – about the individuals, events, and situations portrayed by the work. We construct situation models whose content is determined by what we take ourselves imaginatively to learn in the course of such games. Even if we believe the content in its entirety, then, the process of construction is guided by an imaginative project. It is reasonable to conclude that constructing situation models in response to waltfictions constitutes imagining a storyworld, and that we can therefore count representations included in the model as “imagined” (even if they are also believed). Now, there are numerous objections to this proposal. Although there is widespread agreement that fiction invites imagining, nonfiction is often defined in terms of the invitation to believe.18 There is further dispute concerning how much of our imagining in response to fiction or other works is self-involving in the way Walton suggests. Most importantly from the point of view of a unified account of waltfiction, many theorists deny that picture perception relies on imagining. Rather than engaging in the apparently complex activity of imagining, of our own visual experience, that it is an experience of the represented scene, these theorists typically maintain that seeing-in relies on more basic perceptual or recognitional capacities.19 I cannot offer a defense of the nature and range of waltfictions here. However, I will say that those who criticize Walton’s theory on one point or another very often lose sight of the overall picture. Whatever distinguishes representations in different media, such as novels and paintings, sculptures, and photographs, there is something in virtue of which they all count as representations. Walton’s contention that representations function as props in games of make-believe has substantial explanatory capacity. It offers an account of what unifies works in the category while at the same time – via differences in the de se games of make-believe invited by different media – explaining what distinguishes them. If this theory is wrong, it is wrong about representation in general and not just about specific forms of representation. Though I have not defended Walton’s overall theory of representation here, I have argued that there is an account of storified content, or fictionality in Walton’s terminology, that applies across the range of waltfictions. If this account is plausible, it provides a reason to think that Walton’s categorization makes a valuable contribution to our understanding of representations generally.20

Notes 1. Such definitions include those proposed by Currie (1990) and Stock (2017) among others. 2. It is controversial that imagined seeing or reading counts as “participation” in the fictional world. For example, though Wilson (2011) agrees that we imagine seeing the events depicted in film, he denies that this requires imagining seeing

38

3.

4. 5. 6.

7. 8.

9. 10. 11.

12. 13. 14. 15. 16. 17. 18. 19. 20.

Stacie Friend them face-to-face, from within the same space. The so-called “report model” moves us away from the fictional world even further: According to this view, with texts and films we imagine that we are reading or watching a nonfiction report of events (Currie, 1990, p. 73; Matravers, 1998). I set aside this controversy here, since my goal is to clarify Walton’s account, rather than to defend it against all objections. (Thanks to Gregory Currie for raising this issue.) Though this proposal avoids some of the confusion surrounding Walton’s terminology, it is obviously not ideal for capturing the content of nonnarrative representations. In the absence of a better option, the term should simply be taken to cover all representations within the scope of waltfiction. Evidence for this appears in Barsalou (1999), Zwaan (2003), and Kurby and Zacks (2013). See also Sanford and Emmott (2012). Matravers (2014) and Friend (n.d.) discuss the psychological literature on mental models in some detail. The contrast between propositional imagining and objectual imagining is a species of the more general distinction between propositional and objectual attitudes (Forbes, 2000). Objectual attitudes are unlikely to be reducible to propositional ones (see, e.g., Crane, 2013; Grzankowski, 2015). It is, for instance, central in Bühler (2011). Thanks to François Recanati for drawing my attention to this source. Matravers uses this observation to argue that because an imaginative response is not distinctive to fiction – a claim with which I agree – there is no interesting distinction between fiction and nonfiction – a claim with which I disagree (Friend, n.d.). Gendler (2000) argues that we can imagine contradictions. See Stock (2003) for a reply. See Friend (n.d.) for further discussion. In Friend (2014) I used the term authorization for what I here call an invitation. I now distinguish stronger and weaker conditional obligations from mere permissions. See García-Carpintero (2013) for a discussion of conditional obligations in fiction. The model has also been extended to films (see Zacks et al., 2009). On mental spaces, see Fauconnier (1994), Sanford and Emmott (2012). On subworlds, see Gavins (2007). I will drop the clarification for ease of exposition. Walton (1990) uses such observations to argue that what we experience is not genuine fear, or fear of the ordinary kind (though contrary to the usual interpretation, it may be a genuine emotion). See Friend (2016). The terms import and export are from Gendler (2000). The term transportation is Richard Gerrig’s (1993). See Ryan (2015) for a narratological discussion of immersion. For example, by Currie (1990), though he ultimately restricts this distinction between fiction and nonfiction to individual utterances rather than whole works. For criticisms of Walton’s account of picture perception, see, for example, Wollheim (1998), Nanay (2004), Budd (2008). I would like to thank Kendall Walton for fruitful discussions of these (and many other) issues over the years, and Gregory Currie for helpful comments on a previous draft, which have much improved the chapter.

References Barsalou, L. W. (1999). Perceptual Symbol Systems. The Behavioral and Brain Sciences, 22(4), 577–609.

Fictionality in Imagined Worlds

39

Budd, M. (2008). On Looking at a Picture. In Aesthetic Essays (pp. 185–215). Oxford University Press. Bühler, K. (2011). Theory of Language: The Representational Function of Language (D. F. Goodwin & A. Eschbach, Trans.). John Benjamins. Christie, A. (2007). And Then There Were None (Masterpiece Edition). Harper Collins (Original work published 1939). Cosmides, L., & Tooby, J. (2000). Consider the Source: The Evolution of Adaptations for Decoupling and Metarepresentation. In D. Sperber (Ed.), Metarepresentations: A Multidisciplinary Perspective (pp. 53–115). Oxford University Press. Crane, T. (2013). The Objects of Thought. Oxford University Press. Currie, G. (1990). The Nature of Fiction. Cambridge University Press. Dorsch, F. (2012). The Unity of Imagining. Gazelle Books. Fauconnier, G. (1994). Mental Spaces: Aspects of Meaning Construction in Natural Language. Cambridge University Press. Forbes, G. (2000). Objectual Attitudes. Linguistics and Philosophy, 23(2), 141–183. Friend, S. (2008). Imagining Fact and Fiction. In K. Stock & K. Thomson-Jones (Eds.), New Waves in Aesthetics (pp. 150–169). Palgrave Macmillan. Friend, S. (2014). Walton, Kendall. In M. Kelly (Ed.), Encyclopedia of Aesthetics (2nd ed.). Oxford University Press. Friend, S. (2016). Fiction and Emotion. In A. Kind (Ed.), The Routledge Handbook of Philosophy of Imagination (pp. 217–229). Routledge. Friend, S. (2017). The Real Foundation of Fictional Worlds. Australasian Journal of Philosophy, 95(1), 29–42. Friend, S. (n.d.). Matters of Fact and Fiction. Oxford University Press. García-Carpintero, M. (2013). Norms of Fiction-Making. The British Journal of Aesthetics, 53(3), 339–357. Gavins, J. (2007). Text World Theory: An Introduction. Edinburgh University Press. Gendler, T. S. (2000). The Puzzle of Imaginative Resistance. The Journal of Philosophy, 97(2), 55–81. Gerrig, R. J. (1993). Experiencing Narrative Worlds: On the Psychological Activities of Reading. Yale University Press. Glenberg, A. M., & McDaniel, M. A. (1992). Mental Models, Pictures, and Text: Integration of Spatial and Verbal Information. Memory & Cognition, 20(5), 458–460. Goodman, N. (1976). Languages of Art: An Approach to a Theory of Symbols (2nd ed.). Hackett. Graesser, A. C., Olde, B., & Klettke, B. (2002). How Does the Mind Construct and Represent Stories? In M. C. Green, J. Strange, & T. C. Brock (Eds.), Narrative Impact: Social and Cognitive Foundations (pp. 229–262). Lawrence Erlbaum Associates. Green, M. C., & Brock, T. C. (2000). The Role of Transportation in the Persuasiveness of Public Narratives. Journal of Personality and Social Psychology, 79(5), 701–721. Grzankowski, A. (2015). Not All Attitudes Are Propositional. European Journal of Philosophy, 23(3), 374–391. Johnson-Laird, P. N. (1983). Mental Models: Towards a Cognitive Science of Language, Inference, and Consciousness. Harvard University Press. Kind, A. (2013). The Heterogeneity of the Imagination. Erkenntnis, 78(1), 141–159.

40

Stacie Friend

Kurby, C. A., & Zacks, J. M. (2013). The Activation of Modality-Specific Representations During Discourse Processing. Brain and Language, 126(3), 338–349. Lewis, D. (1983). Truth in Fiction. In Philosophical Papers (Vol. 1, pp. 261–280). Oxford University Press. Liao, S., & Gendler, T. (2019). Imagination. The Stanford Encyclopedia of Philosophy. https://plato.stanford.edu/entries/imagination/ Matravers, D. (1998). Art and Emotion. Oxford University Press. Matravers, D. (2014). Fiction and Narrative. Oxford University Press. Nanay, B. (2004). Taking Twofoldness Seriously: Walton on Imagination and Depiction. The Journal of Aesthetics and Art Criticism, 62(3), 285–289. Ryan, M. (2015). Narrative as Virtual Reality 2: Revisiting Immersion and Interactivity in Literature and Electronic Media. Johns Hopkins University Press. Sanford, A. J., & Emmott, C. (2012). Mind, Brain and Narrative. Cambridge University Press. Stock, K. (2003). The Tower of Goldbach and Other Impossible Tales. In M. Kieran & D. McIver Lopes (Eds.), Imagination, Philosophy, and the Arts (pp. 107–124). Routledge. Stock, K. (2017). Only Imagine: Fiction, Interpretation and Imagination. Oxford University Press. Tapiero, I. (2007). Situation Models and Levels of Coherence: Toward a Definition of Comprehension. Routledge. Van Dijk, T., & Kintsch, W. (1983). Strategies of Discourse Comprehension. Academic Press. Walton, K. L. (1984). Transparent Pictures: On the Nature of Photographic Realism. Critical Inquiry, 11(2), 246–277. Walton, K. L. (1990). Mimesis as Make-Believe: On the Foundations of the Representational Arts. Harvard University Press. Walton, K. L. (2015). In Other Shoes: Music, Metaphor, Empathy, Existence. Oxford University Press. Wilson, G. M. (2011). Seeing Fictions in Film: The Epistemology of Movies. Oxford University Press. Wollheim, R. (1992). Art and Its Objects (2nd ed.). Cambridge University Press. Wollheim, R. (1998). On Pictorial Representation. The Journal of Aesthetics and Art Criticism, 56(3), 217–226. Woodward, R. (2014).Walton on Fictionality. Philosophy Compass, 9(12), 825–836. Yablo, S. (1993). Is Conceivability a Guide to Possibility? Philosophy and Phenomenological Research, 53(1), 1–42. Zacks, J. M., Speer, N. K., & Reynolds, J. R. (2009). Segmentation in Reading and Film Comprehension. Journal of Experimental Psychology: General, 138(2), 307–327. Zwaan, R. A. (1999). Situation Models: The Mental Leap into Imagined Worlds. Current Directions in Psychological Science, 8(1), 15–18. Zwaan, R. A. (2003). The Immersed Experiencer: Toward an Embodied Theory of Language Comprehension. In B. H. Ross (Ed.), Psychology of Learning and Motivation (Vol. 44, pp. 35–62). Academic Press. Zwaan, R. A., Langston, M. C., & Graesser, A. C. (1995). The Construction of Situation Models in Narrative Comprehension: An Event-Indexing Model. Psychological Science, 6(5), 292–297. Zwaan, R. A., & Radvansky, G. A. (1998). Situation Models in Language Comprehension and Memory. Psychological Bulletin, 123(2), 162–185.

3

Walton and Fictional Characters Eileen John

One after another she creates her fools, her prigs, her worldlings, her Mr. Collinses, her Sir Walter Elliotts, her Mrs. Bennets. She encircles them with the lash of a whip-like phrase which, as it runs round them, cuts out their silhouettes forever. . . . Sometimes it seems as if her creatures were born merely to give Jane Austen the supreme delight of slicing their heads off. (Woolf, 2012, p. 173)

It is hard, perhaps foolhardy, to argue with Ken Walton. His work is a powerful mixture of good sense, bold thinking, and meticulous reasoning. Why argue? One reason is that his work makes me want to give good arguments, to be more clear and decisive about tricky issues. I aim to do that here, on an issue that, as far as I am able to think it through, leads me to disagree with Walton. I want to defend the reality of fictional characters. I will try to benefit from Walton’s views on this issue. A realist about fictional characters needs to incorporate a role for make-believe, but that role does not settle the ontological status of characters. My alternative view is related to various kinds of realist positions, embodied in Margaret Macdonald’s claim that ‘it is perfectly ordinary and proper to say that an author has created certain characters and all that is required for them to function’ (Macdonald, 1954, p. 177). Broadly, I want to follow out the idea that fictional characters are things with distinctive functions and that understanding their existence in these terms allows us to make sense of the interest we take in them.1 Philosophical accounts of fictional characters tend to emphasize the interpretation of utterances containing apparent reference to fictional characters doing or undergoing things that a person would do or undergo. ‘Ishmael stayed at the Spouter-Inn.’ ‘Lily Bart eventually worked for a milliner.’ These utterances are a prime puzzle to be explained or paraphrased. What could they mean, if no such person does or undergoes these things? I think this starting point makes it more difficult to reach an understanding of what fictional characters are and why they engage DOI: 10.4324/9780367808662-4

42

Eileen John

us. Recounting the plot when reading Melville’s Moby Dick or Wharton’s House of Mirth is one moment in a larger activity that shows the existence of a character. Even if we move to the utterances that overtly identify characters as characters – ‘Jane Austen created Emma Woodhouse’ (Walton, 1990, p. 410), or ‘Zavalita is one of the most memorable fictional characters created by Vargas Llosa’ (García-Carpintero, 2010, p. 142) – we are focused only on the discourse that is supported by the larger activity. If we start rather with the larger set of things we do with fictional characters, we will have more to draw on as evidence and different things to do justice to. As García-Carpintero notes, for roughly opposed argumentative purposes, ‘the success of apparent references to fictional characters seems to be suspiciously easy to achieve’ (GarcíaCarpintero, 2010, p. 147). Picking up on the names may be easy, a matter of recognizing that conventions involved with fictional characters are in force (van Inwagen, 1977, p.  307), but knowing a fictional character adequately can be difficult. The easy use of the names should not obscure the sophisticated practices that fictional characters figure in and support. If we embrace the role of pretense, as part and parcel of engaging with a novel, that would only help to rule out the existence of fictional characters if their best hope for existence were as the people doing things like staying at the Spouter-Inn. Walton may think this is their best hope, a kind of holy grail for realism: ‘If we have a naïve, pretheoretical commitment to fictional people and things, it would seem to be a commitment to people and things that are in most respects perfectly ordinary’ (Walton, 1990, p. 387). Or, with verve, ‘If Donald Duck is anything he is a duck (a talking duck); not an invention or a cultural artifact’ (Walton, 2015, p. 103). But as soon as we abandon this commitment, as we must, does realism lose its allure? I do not think so, in part because I do not think this is our pretheoretical commitment. People are by and large competent, sophisticated, and undaunted when it comes to grasping fictional characters’ distinctive form of being. If there is another best or better hope, that is what needs to be ruled out. That a fictional character involves us in imagining the life of a person is intuitively a way of grasping what fictional characters are, rather than a way of showing there is no such thing. The burden of argument for realism is to show that fictional characters make a distinctive contribution to reality. If they did not exist, there would have to be a discernible loss. Speaking more metaphorically, they have to present some tension or resistance to thought and action, some evidence of them not being only and exactly what we imagine or wish. There must be something to learn about how to become acquainted with them and how to treat them. One kind of evidence I highlight concerns purposes that makers and users of fictional characters can have and that can be well or badly met. If there were no characters, we either could not pursue these purposes or could only do so in awkward – at best inefficient, at worst ineffectual – ways. Another kind of evidence

Walton and Fictional Characters

43

involves disputes or inquiries about the nature of a character. Understanding a character – what is most important to it, what it serves to do, what features enable it to do what it does – is sometimes an interesting, engrossing project, and one that people can disagree on.2 The examples I point to are only indicative, but I hope they show some possibilities for learning, resistance, and distinctiveness of contribution. Fictional characters are functional artifacts that we come to know through their functioning.3 This discussion engages only partially with debates about realism and anti-realism, specifically avoiding the hard work – hard for both realists and anti-realists – of showing how ‘fictional-character talk’ turns out to be true or false as desired. Realists have often drawn on the clarity of one kind of talk or another (‘In the novel, Lily became a milliner’ or ‘Ishmael is a fictional character’); a realist needs to link those kinds of talk and, broadly, needs to support characters functioning as they do. Characters are representational devices used for a host of purposes, including the imagining of a person. Perhaps, the anti-realist will not mind saying that there are such representational devices; it is fine with me if anti-realists want to join this party, as long as they are willing to call themselves realists.4

I. Walton on Doing Without Fictional Characters How does Walton approach the putative reality of fictional characters? In defending the good sense of nonbelief over belief in fictional entities, Walton draws on his overall account of fiction. The key to understanding assertive uses of sentences appearing to make reference to fictional entities is to take as primary their use in pretense. What is asserted by means of them is to be understood in terms of their role in make-believe. (Walton, 1990, p. 396) Appreciators of a work of fiction ‘are expected to play games of kinds authorized for the works they appreciate and, when they participate verbally, to make it fictional of themselves in such games that they speak truths rather than falsehoods’ (Walton, 1990, p. 398). When Sally, having read Mark Twain’s The Adventures of Tom Sawyer, says, ‘Tom Sawyer attended his own funeral,’ Sally’s claim is that the novel Tom Sawyer is such that to behave in a certain way, to engage in an act of pretense of a certain kind while participating in a game authorized for it, is fictionally to speak the truth. (ibid., p. 400)

44

Eileen John

This, though complicated, seems right: the work of fiction asks us, among other things, to participate in a pretense of a certain kind, and that participation is manifest through thought and speech that enact or somehow acknowledge this pretense – the pretense that there are such people doing such things. The words ‘Tom Sawyer’ can be used appropriately in the course of this pretense. There really is a novel written by Mark Twain, with words that really prompt readers to engage in pretense. The readers’ participation is real, and they are likely to utter strings of words that are appropriate to participation. Walton’s account of what Sally says is careful not to use the phrase ‘Tom Sawyer’; he does not want to suggest that those words are used to refer to anything. ‘We need not suppose even that there is someone, or some character, about whom Sally pretends to speak’ (Walton, 1990, p.  396), because that would locate quantification outside of the pretense. It is the kind of pretense that is real and explanatorily basic. Walton classifies pretense activity involving apparent reference to individuals as ‘kind K’ (Walton, 1990, pp. 400–401), and he notes the difficulty of specifying a given pretense of that kind. There is a sense in which we do not have a ready paraphrase of Sally’s assertion that eliminates her apparent reference to Tom Sawyer. . . . The paraphrase I suggested . . . requires the introduction of a technical term, ‘K,’ . . . whose reference was fixed, by pointing to the use of a sentence containing the name ‘Tom Sawyer,’ to an instance of the kind K. Should we conclude that a commitment to fictional entities is deeply embedded in our language and conceptual scheme, even if there aren’t really any? No. For it is the use of names like ‘Tom Sawyer’ in pretense that enables us to fix the reference of ‘K.’ To pretend to refer to someone with the name ‘Tom Sawyer’ is not in any interesting sense to be committed to there being a referent of that name. What we should conclude is that it is our pretendings to assert, our games of make-believe, that are central to our conceptual scheme. (Walton, 1990, p. 404) Let me pause to note that this is a Walton kind of passage – kind W – that I especially appreciate. It is conversational, but bracingly clear. It steps back from the argument to consider a possible vulnerability – the apparent name ‘Tom Sawyer’ is hard to do without – and it gives this line of thought its properly big, challenging frame. Where does makebelieve sit in our conceptual scheme? Walton gives make-believe a central, fruitful, interesting explanatory role, and I am persuaded that we cannot understand fction without it. But the realists, in Walton’s view, ‘are overlooking or underemphasizing the element of make-believe that lies at the heart of the institution. They mistake the pretense of referring

Walton and Fictional Characters

45

to fctions, combined with a serious interest in this pretense, for genuine ontological commitment’ (Walton, 1990, p. 390). In that remark, Walton stresses that we can take a serious interest in the pretense that a work of fiction supports, and that an account of fiction must make sense of that interest. I think this requirement is hard to meet if there are no fictional characters. In relation to representational art, Walton grounds the interest of our apparent references to characters in the interest and importance of the make-believe characteristic of engagement with fiction: ‘Given that the (authorized) games appreciators play with works of art are important, people are bound to be interested in what sorts of pretendings-to-assert are, fictionally in these games, assertions of truths’ (Walton, 1990, p. 405). If talk ‘about’ Tom Sawyer has any point, it is because we are interested in the overall imaginative game that Twain’s novel allows us to play. In a way, I agree. The pretendingsto-assert, aiming to make it fictional that one speaks truly about Tom Sawyer’s doings, are not in themselves enough to sustain interest. But the safe confinement of Tom Sawyer and our apparent references to him within the pretense leaves us with too little to work with, to account for the interest and importance of the game of make-believe. If our primary concern as make-believers were to make it fictional that we speak truly, that would be a strange fixation on the merely fictional. Sometimes, even the great lover of fiction can pick up a novel and experience what might be called ‘fiction ennui’: ‘Do I really have to soak up all of these prescriptions to imagine people doing many quite ordinary things?’ Sometimes, the fictional truths concern inherently exciting, intriguing, or odd events (reversals of fortune, discovering a murderer, attending one’s own funeral), but we could not ground the interest of fiction solely in the interest of getting it right about those kinds of fictional truths. If our apparent attention to Tom Sawyer and the rest were to be explained by the project of imagining in a way that is accurate within the game, then it seems that our serious interest in pretense would importantly rest on that project. Getting the fictional truths right would be the contribution that the make-believe made to the interest of the game. Consider how Walton, discussing Bruegel’s Wedding Dance, rejects the idea of paraphrasing utterances about the painting in terms of features of the painted surface: But a statement detailing the relevant colored splotches would hardly seem to capture what is said in normal cases. . . . [O]rdinarily we are not much interested in the relevant combination of colors and shapes for its own sake, but are very interested in, and so are likely to talk about, what is fictional in the world of the painting and in the worlds of games authorized for it. We may not even notice the precise characteristics of color and line that make it fictional that peasants are making merry. (Walton, 1990, p. 415)

46

Eileen John

I suggest that we are not necessarily so interested in what is fctional in the world of the painting. Or at least our interest in truths within the fctional world needs to be carefully situated. I might have sociological, historical interests in sixteenth-century peasant life and be interested in the fctional truths the painting supports for that reason. But most of the time I expect that engaging in and adverting to the pretense, while it requires seeking accuracy about prescriptions to imagine, is a kind of baseline activity that becomes interesting to the extent that it is achieved via this painted record of choice, effort, and use of the medium. My sense is that we are interested in the colored splotches as they contribute to a possibility for imagining human life (it is kind of amazing that they do that), and we may be interested in ideas and feelings – ideas and feelings not confned in relevance to the fctional world – that the paint splotches stimulate as they serve to provoke imagining. Yes, I want to recognize whatever is to be recognized in the world of the painting, in this case by participating in or acknowledging a pretense in which I come to know that peasants are making merry. But that it is fctional that peasants did or still make merry makes an inconsequential claim on me. If I try to defend the interest of getting the fctional truths right in that world, I cannot simply say it is because I have an interest in accuracy within a game of make-believe. There must be more to it, or my serious interest in pretense will collapse into something like a fxation on fctional truth, which in turn is hard to defend as interesting.

II. A Serious Interest in Pretense Walton, of course, does not say that games of make-believe matter to us because we care about getting fictional truths right. I am trying to press him into that corner for argumentative purposes. He says things like this: But the pretense I speak of is serious business, even if it doesn’t involve seriously supposing that we actually refer to what we pretend to refer to. We engage in make-believe in order to think and talk about features of the real world—often ones that matter, and sometimes ones that are not easy to think or talk about in any other way. (Walton, 2015, p. 91) I want to endorse and follow up on this claim, especially the last clause: the distinctive contribution to reality of fctional characters lies in large part in their role in distinctive projects of thought and inquiry. One of Walton’s main ways of explaining this grip on the real is through the role of props in make-believe. Props, like tree stumps when playing that stumps are bears, or paintings of weddings, have real features that can prescribe imaginings: ‘Props are generators of fictional truths, things which, by virtue of their nature or existence, make propositions fictional’ (Walton, 1990, p. 37).

Walton and Fictional Characters

47

Children may play a game in which bicycles are horses, and a garage is a corral. The real world fact that a bicycle is in the garage makes it fictional, true in the make-believe, that a horse is in the corral. I call the bicycles and the garage props. . . . The colors and shapes on the surface of a painting and events occurring on stage in a theatrical production are props which generate fictional truths. (Walton, 2015, p. 92) Representational works of art are props also. What makes it fictional in La Grand Jatte that a couple is strolling in a park is the painting itself, the pattern of paint splotches on the surface of the canvas. It is because of the words constituting Gulliver’s Travels that fictionally there is a society of six-inch-tall people who go to war over how eggs are to be broken. (Walton, 1990, p. 38) Props are solidly real: stumps, bicycles, paint splotches, words. Maybe words are a little weird in how they ft into reality, but still, they have uncontroversial presence. The real features of props keep appreciators of fction anchored to reality. A very important kind of prop, with respect to explaining interest in pretense, is the appreciator herself. As participants in make-believe, our own reactions to the work can generate fctional truths within individual games of make-believe. My tears when reading can make it fctional that I am sorry for someone within the world of a novel. As props ourselves, We don’t just observe fictional worlds from without. We live in them (in the worlds of our games . . .), together with Anna Karenina and Emma Bovary and Robinson Crusoe .  .  . sharing their joys and sorrows, rejoicing and commiserating with them, admiring and detesting them. True, these worlds are merely fictional, and we are well aware that they are. But from inside they seem actual . . . and our presence in them .  .  . gives us a sense of intimacy with characters and their other contents. It is this experience that underlies much of the fascination representations have for us and their power over us. (Walton, 1990, p. 273) What we experience within the scope of a pretense is often affectively complex and intense, and that felt intimacy – often not easy to achieve with real people – is itself alluring and gripping. It is also plausible that appreciators can learn about themselves, while serving ‘as refexive props, generating by their actions and thoughts and feelings fctional truths about themselves’ (Walton, 1990, p. 274). We can take an interest in our own activity within a pretense.

48

Eileen John

A different explanation is driven by the fact that ‘Sometimes appreciators participate scarcely at all. Some representations positively discourage participation’ (Walton, 1990, p. 274). We can use props, and be supposed to use them, in ways that suppress or deprioritize the generation of fictional truths. As just noted, Props can be mere tools or vehicles for establishing fictional worlds which we find exciting or interesting or poignant or moving. But in some cases it is the props themselves that matter. Their role in generating fictional truths in a game of make-believe may help us to understand them, to think and talk about them, even if the fictional world itself is of no interest at all. (Walton, 2015, p. 92) Focus on the prop is appropriate with many works of art: Works that limit our involvement in fictional worlds include acknowledged masterpieces. .  .  . We may marvel at a work’s suitability for use in games of certain sorts; we may be fascinated by the combination of fictional truths it generates . . .; we may admire the artist’s skill and ingenuity in devising ways of generating fictional truths; we may delight in the devices by which participation is inhibited. Even in such ‘distanced’ appreciation, however, the thought of the work’s serving as a prop in a game of make-believe is central to our experience. (Walton, 1990, pp. 274−275) In appreciating such works, the make-believe is in view, but the proporiented appreciator is not immersed in it. The prop is exposed as a prop, and its ways of serving or inhibiting make-believe are real functions that can be the focus of our interest. Walton is right that we value affective intensity and immersion in pretense and are interested in how props do what they do. But I resist the idea that props tend to function in one or the other of these ways. They all function in both ways, or really the functions are not separable, and that combined experience is basic to engagement with fiction. Whether it is a comic strip, a formulaic romance, a suspense movie, or a George Eliot or a José Saramago novel, we experience props as things with a pretense function. When describing engagement with ‘content-oriented’, immersive fiction, Walton notes that, ‘We focus on what happens in the world of a story: on whether the hero will arrive in time to rescue the heroine, why Hamlet was so wishy washy, “who done it,” etc.’ (Walton, 2015, p. 93). However, if we are attending to heroes and heroines, we are thinking about props of a certain kind, characters with expected functions. Asking why Hamlet was so wishy-washy is likely to encompass

Walton and Fictional Characters

49

the significance of the character, not just a man’s history and psychology. Wondering about who committed the murder is not just curiosity that arises while imagining people caught up in a murder; it is what we do with a ‘whodunnit’. My claim is that even the most immersed appreciative pretense is permeated with its identity as a project of engaging with a prop.5 The consequences of this claim lead me to further agreement bound up with disagreement with Walton, though I expect Walton would want to part company just about immediately. One way of putting my view of fictional characters is that they are, partly, props in Walton’s sense. One function that identifies them is that they prescribe imagining the lives of persons and thereby generate fictional truths. However, they are not only props in Walton’s sense, and that is because of the permeation of makebelieve with characters’ further functions, of the kinds just hinted at. A character’s service as a prop, with the prop’s specific tie to fictional truth, is coordinated with other functions. If we coordinate pretense with those other functions, we have a better way of grounding serious interests in fiction.

III. Fictional Characters and How They Function I take fictional characters to exist as functional artifacts. We come to know them by using them, where that usage involves responsiveness to purposes we understand a character to serve. Such artifacts are made with representational resources such as written language, actors’ utterances and gestures, and drawn, painted, and sculpted forms. Those resources are intended to prompt experience that unfolds and is unified, at least to some degree, as experience of a person so represented (taking persons very broadly, as conscious agents, allowing for all kinds of non-human persons). The ‘so represented’ clause is needed to capture something important about these artifacts. They do not achieve independence from their use of representational resources, as, for instance, linguistic reference to a person commonly aims to do. ‘I saw Lucy Liu at the library!’ aims to share knowledge of Liu’s location, and probably of my amazing proximity to her, and the intention is that those facts, rather than the utterance itself, should guide future thought and feeling. Fictional characters exemplify the use of representational resources to prompt experience as of a person. Fictional characters always embed a claim to the effect that this is a way to represent a person, and they always raise the question of why one would represent a person in this way. Sometimes, the ‘why’ question will prompt us to challenge and critique the character – we can find the use of representational resources on offer in a given character to be problematic or unacceptable. Say, the character gives form to a fantasy that we want to reject or overemphasizes the role of some ideal. Such functions (giving form to fantasies and prioritizing the role of ideals in a

50

Eileen John

life) turn out to be projects well suited to the potential of fictional characters. Competence with fictional characters involves grasping the goals being pursued by using representational resources in this way. Those general aspirations and questions can explain why fictional characters are interesting and, sometimes, important. The ontological and epistemological status that I seek for fictional characters – that they exist as artifacts known through their use – means that what they are is shown best in the transactions that occur between the artifact and our imaginative and reflective experiences. I find it helpful to think of fictional characters as belonging to a very broad kind, comprising things that exist in virtue of some type of representational effort, and that reveal their properties in what happens as the result of people engaging with those efforts. Two other examples of this kind are political campaigns and classes one teaches. Campaigns and classes are real things whose existence requires that someone puts content-bearing materials into circulation, and the particular ways those materials are provided are essential to the campaign or class – the abstract content is not sufficient to identify the campaign or class. It is in the responsive experience of voters and students that the distinctive properties of a campaign or class, and their successes and failures, are explored and revealed. The fictionalcharacter project is distinguished among such examples by the centrality of ‘person-shaped imagining.’ A character, though an artifact made out of, say, chosen and arranged words, is not knowable by studying those words; to tell what a character is like, the representational materials have to be activated, putting them to some use, including person-shaped imagining. Understanding fictional characters as belonging to this (very loose) kind – artifacts that call for responsive engagement with representational efforts – helps to make sense of how we assess them, by being concerned with how they function. Criteria for success or achievement for fictional characters draw on our knowledge of persons, but not always because we try to map what a character involves us in imagining onto what we know about persons. Sometimes, a character’s function requires not being particularly familiar or convincing as a living person. Characters can call for forms of testing and appreciation that diverge radically from response to persons. That it is worthwhile to exemplify some possibility for representing persons could flow from the needs of a story, a pointed contrast with a character from another work, a desire to combat a prejudice, an experiment in nonpsychological explanation, the humor of a style of portraiture, and so on. And although what is true-in-the-fiction about the character matters to grasping how a character functions, imagining those facts is only one kind of response that is relevant to understanding a character. Let me cite a few examples of the kind of functioning of characters that I want to highlight.6 Here is critic Marcia Holly:

Walton and Fictional Characters

51

Virginia Woolf, for example, in creating Lily Briscoe in To the Lighthouse indicates that Mrs. Ramsay’s sensitivity and Mr. Ramsay’s abstract intellectuality are not givens of the sexes. Lily unifies in herself those qualities typically assigned as sex traits. In fact, the burden of the book seems to be that personal wholeness develops with and from the dialectical integration of qualities seen as opposing. (Holly, 1975, p. 43) Woolf’s character is here discussed as crucially serving to integrate qualities. This allows the character both to challenge what might seem like an apt contrast between women and men and to manifest a conception of ‘personal wholeness.’ Setting aside Holly’s specifc interpretation, we can see her account fully integrating what we are to imagine of a woman with aiming to understand the purposes served by that imagining. In Michael Bell’s discussion of J. M. Coetzee character Elizabeth Costello there is further such combined thinking and also assessment. Bell has qualms about the continuation of Elizabeth Costello from The Lives of Animals (which began as real lectures given by Coetzee, in part voicing lectures given by fictional Costello) into the longer, more novelistic work Elizabeth Costello. Bell says of the latter extension, the resulting work, even as it generates new interests of its own, nonetheless, in so far as it establishes Costello as a more conventionally fictional character, significantly flattens the immediate effect of the original lecture. By privileging her continuity as a character, it blunts the discomforting edge of her irruption into a real historical occasion. (Bell, 2007, p. 219) Bell’s larger argument is that Coetzee had a project of probing the nature and force of conviction, and that this project benefted from Costello’s odd way of not sinking into a fctional ‘reality’ in the earlier work (Bell, 2007, p. 221). Let me give Coetzee and Woolf a chance to speak as commentators and critics themselves. Coetzee links Turgenev and Dostoevsky around Turgenev’s character Bazarov in Fathers and Sons. ‘Turgenev put his finger on a new and ominous social actor, Bazarov the Nihilist’ (Coetzee, 2001, p. 141). Coetzee comments as follows. There was something puerile in Nihilism, as both Turgenev and Dostoevsky recognized . . . in its intellectual complacency . . . its mindless destructiveness, its hubris, and . . . its ill-disguised contempt for those in whose name it claimed to speak. . . . There is every reason to believe that Turgenev shared Dostoevsky’s reading of Bazarov. The

52

Eileen John Left, however, preferred not to recognize the critical dimension of the portrait; and Turgenev furthered this slanted reading by declaring mysteriously that Bazarov was himself. Dostoevsky was outraged by this move on Turgenev’s part. In his ongoing critique of Nihilism, we can imagine Dostoevsky as projecting the career of Turgenev’s hero into the 1860s. (Coetzee, 2001, pp. 141−142)

Coetzee, citing Joseph Frank’s work, sees Bazarov as then a forebear and critical impetus for Dostoevsky’s Raskolnikov in Crime and Punishment and Verkhovensky in The Devils (Coetzee, 2001, p. 142; Frank, 1997). Coetzee thus presents Turgenev and Dostoevsky as diverging around the critical force of Bazarov: was he revelatory of the damning problems of nihilism or not? Why a character can support such a disagreement is suggested by the terms Coetzee uses for the faws of nihilism: puerile, complacent, contemptuous, etc. These are more obviously the faws of people than of philosophical positions. Yet, we might want access to these terms if a position is fawed especially in its tendency to be wielded in such a spirit. Revealing that kind of faw might be especially well achieved through characters that get us to imagine lives affrming the position. Woolf in her criticism often writes about authors’ characters, as in her remark about Austen slicing off the heads of her fools and prigs. ‘Her fool is a fool, her snob is a snob, because he departs from the model of sanity and sense which she has in mind’ (Woolf, 2012, p. 174). The point of such characters is to convey, indirectly, a real model of sanity and sense. Woolf wrote in part critically of George Eliot’s characters: Those who fall foul of George Eliot do so, we incline to think, on account of her heroines: and with good reason; for there is no doubt they bring out the worst of her, lead her into difficult places, make her self-conscious, didactic and occasionally vulgar. (Woolf, 2019, p. 11) Woolf’s illustration of this problem is Maggie Tulliver in The Mill on the Floss, when Maggie is old enough to seek love. First Philip Wakem is produced and later Stephen Guest. The weakness of the one and the coarseness of the other have often been pointed out; but both . . . illustrate not so much George Eliot’s inability to draw the portrait of a man, as the uncertainty, the infirmity, and the fumbling which shook her hand when she had to conceive a fit mate for a heroine. (Woolf, 2019, p. 11)

Walton and Fictional Characters

53

I think we cannot locate the problem that Woolf identifes for Eliot without treating Maggie as a character, a heroine who serves to pose a question about what would count as a ft mate. That Eliot fails to portray a ft mate is due, according to Woolf, to the real challenge of meeting aspirations represented in these characters. Eliot’s heroines do not find what they seek, and we cannot wonder. The ancient consciousness of women . . . seems in them to have . . . uttered a demand for something—they scarcely know what—for something that is perhaps incompatible with the facts of human existence. . . . [T]he struggle ends, for her heroines, in tragedy, or in a compromise that is even more melancholy. (Woolf, 2019, p. 11) Woolf is interested in these heroines as showing a struggle with real conditions of human nature and society. Eliot, on Woolf’s reading, tries but fails to articulate and envisage a certain ideal in these characters; her characters serve as players within an inquiry into a form of good life.7

IV. Situating This View of Fictional Characters The idea that fictional characters are functional artifacts is variously related to other forms of realism about characters. Without sorting out the similarities and differences thoroughly, I want to highlight a few points concerning how this view relates to and diverges from views of van Inwagen on the ‘theoretical entities of literary criticism’ (van Inwagen, 1977, p.  302) and Thomasson on characters as abstract artifacts (Thomasson, 1998). For van Inwagen, fictional characters belong to a broader category of things I shall call theoretical entities of literary criticism, a category that also includes plots, sub-plots, . . . poems, meters, . . . influences, . . . recurrent patterns of imagery, and literary forms. (van Inwagen, 1977, pp. 302−303) The notion and vocabulary of characters are ineliminable from the ‘conceptual machinery’ and discourse of literary criticism (van Inwagen, 1977, p. 303). I turned to the discourse of literary critics for examples, though, like van Inwagen, I do not think fictional characters are essentially beholden to the work of professional critics (van Inwagen, 2014, p. 92). The notion of a character is woven into basic experience and understanding of fiction. van Inwagen further articulates a distinction between the properties a character has, for example, ‘being a satiric villainess’ or ‘having been created by

54

Eileen John

Dickens’ (van Inwagen, 1977, p. 305), and properties that are ascribed to it in a work of fiction, such as living in a shoe (van Inwagen, 1977, p. 305). The latter are the stuff of prescriptions to imagine. The questions I want to raise about van Inwagen’s approach concern his argument’s broad strategy, which takes fictional discourse to be the core evidence for realism: Since, therefore, I think there are true sentences of fictional discourse . . . that entail ‘There are fictional characters’ . . . and since I think one should accept the perceived logical implications of that which one believes, I conclude—tentatively, perhaps, but all philosophical conclusions should be tentative—that fictional characters exist. (van Inwagen, 2014, pp. 100−101) The reality that allows us to infer the existence of characters is the reality of true sentences. Their truth requires, on his view, that fctional characters exist. I think we have to push further to consider what explains the sentences. Why do we have this discourse? What kind of thing would make us need to distinguish predication of properties and ascription of properties in a work? That there are things made using words or other content-bearing forms that ask us to imagine the life of a person, explains why there is such discourse and why ascription is pertinent. If the discourse and its truth-aptness were the core data, the characters would indeed seem like theoretical entities, things whose reality we have reason to infer, but not things we come to know directly. I think fctional characters are not theoretical in this way; their existence is not inferred from the need to do justice to the discourse. The discourse aims to do justice to them and what we do with them. We use whatever vehicles of representation they are made of to imagine a person for a host of interesting and ordinary purposes. We encounter and respond to characters in experience, as we engage with a political campaign or philosophy class. Given these encounters, we try to generate discourse adequate to them. For this reason, Thomasson’s view of fictional characters as abstract artifacts is more congenial. Characters are ‘entities that can come into existence only through the mental and physical acts of an author – as essentially created entities’; they are artifacts ‘created by the purposeful activity of humans’ (Thomasson, 1998, pp.  6, 35). The artifactual approach makes sense of our interest in the purposes served by characters, even if the purpose behind imagining a life in certain terms is sheer entertainment – that can be an appropriate purpose for a character, though not for a person. For Thomasson, ‘fictional objects, in this conception, . . . are closely connected to ordinary entities by their dependencies on both concrete, spatiotemporal objects and intentionality’ (Thomasson, 1998, p. 12). Thomasson uses the relation of dependency to ground these abstract artifacts in non-abstract reality: a character’s existence may

Walton and Fictional Characters

55

depend on the existence of words or drawings, such that if the words or drawings went out of existence, the character could as well. But the character is not actually made out of words or drawings. Characters exist in the way that marriages, contracts, and promises do, as ‘cultural and institutional entities that . . . can be brought into existence merely by being represented as existing’ (Thomasson, 1998, pp. 12−13). So, fictional characters are created merely with words that posit them as being a certain way. For example, because characters are created by being written about by their authors, Jane Austen creates the fictional character Emma Woodhouse and brings her into existence . . . in writing the sentence: Emma Woodhouse, handsome, clever, and rich. (Thomasson, 1998, p. 12) I would say rather that Emma Woodhouse is a character made out of those words that Austen wrote, and that she does not come into being by being written about. Characters are not written about (in being made); they are the structured conglomerations of words or other representational materials that function for us in the ways discussed earlier. Austen’s novel is not written about Emma Woodhouse; Emma Woodhouse is an entity we become acquainted with and perhaps do interesting things with in reading the novel. In this respect fctional characters are not abstract. They have whatever concreteness the range of content-bearing forms such as words have, and they are used in relatively concrete, temporally located activities of perception, thought, and feeling. Precisely what each character can be used to do is of course not settled by the general way characters exist. We find out what we can do with Elizabeth Costello or Maggie Tulliver, but that a character exemplifies a way to represent the life of a person, and is to be used to imagine such a life for some purpose, is basic. Allowing things used in this way into our ontology does not strike me as getting ‘something for nothing’ – getting existence out of the bandying about of name-like phrases, with no requirement for non-trivial knowledge of a referent8 – because we do not just encounter an apparent name, but a carefully made structure of representational material. In these encounters we can face sophisticated knowledge questions. Exactly what do George Eliot’s heroines seek and why do they not find it? Is Bazarov to be rejected or does he represent a new freedom? That we reach such questions via these representations and exploration of their functions is a tremendous achievement on the part of artists and audiences. Fictional characters are tried and true devices that serve these and more simply pleasurable projects. Characters are not immediately what anyone wants or imagines them to be; we can be wrong or naïve about a character, disappointed by or critical of it, and artists making them may not achieve what they hoped. Walton says that make-believe allows us to think ‘about features of the real world—often

56

Eileen John

ones that matter, and sometimes ones that are not easy to think or talk about in any other way’ (Walton, 2015, p. 91). I think fictional characters help to bear out Walton’s claim.9

Notes 1. This discussion aims to clarify and improve on ideas presented in John (2016). 2. See, for example, E. M. Forster on changing his mind about a Jane Austen character, not having seen that all of her characters ‘are round, or capable of rotundity’ (Forster, 1927, pp. 112−117). 3. My focus on fictional characters is driven by the depth and complexity of what we do with them – the interest they have for us – and is not intended to dismiss comparable realist claims about such things as fictional events and settings. See Macdonald, for instance, on fictional locations and other elements (Macdonald, 1954, pp. 179−181). 4. See Stacie Friend’s analysis of these debates, especially pp. 143−144 on antirealism and the function of fictional discourse and pp. 147−148 on forms of realism (Friend, 2007). 5. Walton’s account of the inseparability of appreciation and criticism is relevant here (Walton, 1990, esp. pp. 393−396). See Friend on the problem of distinguishing internal and external perspectives (Friend, 2007, pp. 151−153). 6. Let me note, but not adequately address, Walton’s treatment of statements that look to be about characters. He invokes unofficial games of make-believe and betrayal of pretense. In saying, ‘Jane Austen created Emma Woodhouse,’ one plays or refers to an unofficial game of make-believe ‘in which to author a fiction about people and things of certain kinds is fictionally to create such’ – one makes it fictional that Austen created a woman (Walton, 1990, p. 410). To say, ‘Gregor Samsa is a (purely fictional) character’ is ‘to acknowledge, while betraying the pretense, only that there is a work in whose authorized games so pretending is fictionally to refer successfully’ (Walton, 1990, p. 422). While I think it is implausible that anyone imagines Austen creating living beings (à la Frankenstein’s creature?), my main strategy here is to show that characters cannot be betrayed as ‘purely fictional’ because they have relatively sturdy ways of existing and functioning. 7. See Michael Weston on characters engaging us with ‘a possibility of life perceived through a certain conception of that life’ (Weston, 1975, p. 86). Weston might not accept what I say characters are, but I take my discussion to be sympathetic to his claims about their functions. 8. See García-Carpintero (2010) on the ‘something-for-nothing’ and knowledge problems (esp. pp. 146 and 154−156). 9. With many thanks to Sonia Sedivy for making this project happen and to Stacie Friend for excellent comments, none of which have been adequately dealt with. My greatest thanks go to Ken Walton. I am still learning from him and feel tremendously lucky to have been able to engage with his ideas and way of doing philosophy all along the way.

References Bell, M. (2007). Open Secrets: Literature, Education, and Authority From J.-J. Rousseau to J. M. Coetzee. Oxford University Press. Coetzee, J. M. (2001). Stranger Shores: Essays 1986−99. Secker & Warburg. Forster, E. M. (1927). Aspects of the Novel. Harcourt, Brace & Co.

Walton and Fictional Characters

57

Frank, J. (1997). Dostoevsky: The Miraculous Years. Princeton University Press. Friend, S. (2007). Fictional Characters. Philosophy Compass, 2(2), 141−156. García-Carpintero, M. (2010). Fictional Entities, Theoretical Models and Figurative Truth. In R. Frigg & M. C. Hunter (Eds.), Beyond Mimesis and Convention, Boston Studies in the Philosophy of Science (Vol. 262, pp. 139−168). Springer. Holly, M. (1975). Consciousness and Authenticity: Toward a Feminist Aesthetic. In J. Donovan (Ed.), Feminist Literary Criticism (pp. 38−47). University Press of Kentucky. John, E. (2016). Caring about Characters. In G. Hagberg (Ed.), Fictional Characters, Real Problems (pp. 31–46). Oxford University Press. Macdonald, M. (1954). The Language of Fiction. Proceedings of the Aristotelian Society, Supplementary Volume, 28, 165−184. Thomasson, A. (1998). Fiction and Metaphysics. Cambridge University Press. van Inwagen, P. (1977). Creatures of Fiction. American Philosophical Quarterly, 14(4), 299−308. van Inwagen, P. (2014). Existence: Essays in Ontology. Cambridge University Press. Walton, K. (1990). Mimesis as Make-Believe. Harvard University Press. Walton, K. (2015). In Other Shoes: Music, Metaphor, Empathy, Existence. Oxford University Press. Weston, M. (1975). How Can We Be Moved by the Fate of Anna Karenina? Proceedings of the Aristotelian Society, Supplementary Volume, 49, 81−93. Woolf, V. (2012). The Common Reader (1st series). Read Books. Woolf, V. (2019). Pride and Paragon: Responding to the Life and Work of George Eliot. Times Literary Supplement, 15, 10−11 (Original work published 1919).

4

Walton on ‘the Paradox of Fiction’ Confusions and Misunderstandings Derek Matravers

Of all the various facets of Kendall Walton’s theory, it has been his work on our psychological interactions with fictional characters that has drawn the most critical fire.1 This debate is characterised by myriad confusions and misunderstandings (to which I have contributed my fair share). This chapter has six sections but falls into three broad parts. The first part runs quickly through some of the more egregious confusions and misunderstandings, hoping to set them to rights; the second presents what I take to be Walton’s positive view; and the third attempts to clarify a puzzle that arises from that view.

I. Existing Confusions The standard way of interpreting the (so-called) paradox of fiction is as an inconsistent triad (here in the words of Peter Lamarque and Stein Olsen, but examples are legion). 1. Readers or audiences often experience emotions such as fear, pity, desire and admiration towards objects they know to be fictional, e.g. fictional characters. 2. A necessary condition for experiencing emotions such as fear, pity, desire, etc., is that those experiencing them believe the objects of their emotions to exist. 3. Readers or audiences who know that the objects are fictional do not believe that these objects exist. (Lamarque & Olsen, 2004, p. 298) A view which is erroneously attributed to Walton is put forward as one of the ‘standard solutions’ to the paradox. Without going into detail, the following is, I hope, a fair representation of the kind of discussion one encounters. Walton denies (1); he holds that rather than experiencing actual emotions towards objects we know to be fctional, we experience ‘quasi-emotions’. However, there are good reasons to hold that this is wrong; that what we feel towards objects we know to be fctional are DOI: 10.4324/9780367808662-5

Walton on ‘the Paradox of Fiction’

59

actual emotions. Thus, we should hold onto (1) and instead reject (2). What prevents Walton from rejecting (2) is his holding a strong cognitive theory of emotions: that emotions necessarily involve the belief that the object of the emotion exists. However, the strong cognitive theory is untenable, which leaves the way open for the denial of (2). There are at least six separate mistakes in such a discussion. Firstly, there is no ‘paradox of fiction’. That is, it is a mistake to think there is a problem, introduced into the literature by Colin Radford, to which many have contributed, which Walton is attempting to solve (Radford, 1975).2 Indeed, Radford’s paper does not even appear in Walton’s (1990) extensive bibliography appended to Mimesis as Make-Believe. The term ‘paradox of fiction’ is itself misleading, as, arguably, there is neither paradox nor any connection to fiction (at least, as that term is traditionally construed) (Matravers, 2014, p. 102). In this chapter, I shall use ‘paradox of fiction’ to refer to the problem as discussed; that is, that which commentators generally take the problem to be. My defence of the claim that this is not what concerns Walton will emerge as I give an account of his positive view. The second mistake is simply textual. For Walton the term ‘quasiemotion’ does not name whatever emotion-like state we feel in the context of a make-believe but rather an emotion’s ‘physiological-psychological’ component (Walton, 1990, pp. 196, 251). The ‘quasi’ part of an emotion is a proper part of any emotion, whether felt inside or outside the context of make-believe. The misinterpretation is understandable given the term ‘quasi-emotion’ has connotations of something that is almost an emotion but not quite. Walton, however, gives several examples of actual emotions that have quasi-emotions as a component (Walton, 1990, pp. 244, 250−251). The third mistake is to construe Walton as holding that, when engaged with a fiction, we are ‘only pretending’ to feel an emotion. That objection is that we cannot be pretending, as what we feel in the context of make-believe has all the phenomenological vivacity of emotions felt in real life (Carroll, 1990, p. 74). It is part of the core structure of Walton’s theory that there is no systematic difference in phenomenology between emotions felt within and those felt without the context of make-believe; the ‘quasi-emotion’, the ‘physiological-psychological’ component, occurs in both.3 The fourth mistake is to hold that Walton’s view implies a falsehood: that the emotions we feel in the context of make-believe could be engaged or disengaged voluntarily (Carroll, 1990, p.  74). However, there is no reason to think that Walton’s view has this implication (see, for example, (Walton, 1990, p. 245)). The fifth mistake is to hold that Walton holds a strong cognitive theory of emotions. The issue is complicated, as we shall see; suffice to say, for the moment, that Walton explicitly disavows such a view. Citing the

60

Derek Matravers

phobic who ‘fears poor old Fido without judging him to be dangerous’ he concludes that ‘fear may not require a belief that one is in danger’ (Walton, 1990, p. 245).4 Finally, the sixth mistake is to hold that Walton denies that we feel genuine emotions within the context of make-believe. Walton explicitly repudiates this charge: ‘It goes without saying that we are genuinely moved by novels and films and plays, that we respond to works of fiction with real emotion’ (Walton, 1997, p. 275). I take it that what is meant is that our emotional responses to fiction are involuntary; have the structure of emotions (a cognitive component plus the quasi-emotion element); can be felt with as much or as little vivacity as our emotions felt outside the context of make-believe; and that they can be counted as part of our emotional lives (Williams, 2019). Such claims are compatible with his central claim, namely that these emotions are only felt ‘within imagination’. Confusion as to what is meant by being ‘genuinely moved’ and ‘real emotion’ has led most commentators simply to ignore this passage while at least one finds it an inconsistent part of Walton’s overall view (Sainsbury, 2010, pp. 217−218).

II. The Positive Account Walton evinces a certain impatience at what he calls ‘the undue emphasis that commentators have put upon this issue’ – that is, the issue of ‘the paradox of fiction’ (Walton, 1997, p.  275). It is easy to see why. In Mimesis as Make-Believe, Walton presents a comprehensive account which illuminates many aspects of the nature of representation. One of these aspects is our psychological interaction with fictional characters, and the paradox of fiction is only one aspect of that. Thus, a full grasp of Walton’s view can only be obtained by raising one’s eyes from the narrow discussion and engaging with the comprehensive theory. Failure to do so runs the risk of missing the justification for some of Walton’s claims. A good example of this failure is the proposed ‘solution’ to the aforementioned problem, namely the denial of (2). Clearly, we dissolve the paradox as stated by denying (2) yet, this still leaves everything unresolved (Stecker, 2011). The fact that we can feel actual emotions in the absence of belief (which, as we have seen, is something Walton does not deny) tells us little about the nature of emotions felt in the context of fiction. It tells us only that one reason for their not being classified as actual emotions does not apply, a claim that is compatible with their being actual emotions and with their not being actual emotions. Walton gives a summary of this general account of representation at the conclusion of Chapter 1 of Mimesis. Representations, I have said, are things possessing the social function of serving as props in games of make-believe, although they

Walton on ‘the Paradox of Fiction’

61

also prompt imaginings and are sometimes objects of them as well. A prop is something which, by virtue of conditional principles of generation, mandates imaginings. Propositions whose imaginings are mandated are fictional, and the fact that a given proposition is fictional is a fictional truth. (Walton, 1990, p. 69) That facts of the matter of any game can be described in a way that is external to the game or a way that is internal to the game. Consider children playing at mud pies. From the external perspective there are three globs of mud in a box. The globs of mud and the box are props which, in accord with the well-known principles of generation for the game, mandate the players to imagine there are three pies in the oven. This makes it the case that it is fctional that there are three pies in the oven. From the internal perspective, that is, the perspective of the players of the game, there are three pies in the oven and (assuming everything is working well epistemologically) they believe there are three pies in the oven. That is, their imagining (or ‘make-believing’, I shall use the terms interchangeably) that are three pies in the oven makes it fctional that they believe there are three pies in the oven. Now let us look at Walton’s best-known example of ‘fearing fictions’ to see how these general considerations transfer. Charles is watching a horror movie about a terrible green slime. He cringes in his seat as the slime oozes slowly but relentlessly over the earth, destroying everything in its path. Soon a greasy head emerges from the undulating mass, and two beady eyes fix on the camera. The slime, picking up speed, oozes on a new course straight towards the viewers. Charles emits a shriek and clutches desperately at his chair. Afterwards, still shaken, he confesses that he was ‘terrified’ of the slime. (Walton, 1990, p. 196) We can describe this from the external or the internal perspective. The flm is such that certain patches of light are props. That is, certain patches of light which are projected on the screen mandate Charles to imagine the slime is present. When the projected patches of light move in a certain way, Charles is mandated to imagine he is threatened by the slime. He does imagine this, and this causes quasi-fear. That he is caused to feel quasi-fear as a result of imagining he is threatened by the slime makes it fctional that he fears the slime. That is, it is fctional that Charles believes he is threatened by the slime, which causes him to feel fear. He experiences quasi fear as a result of realizing that fictionally the slime threatens him. This makes it fictional that his quasi fear is

62

Derek Matravers caused by the belief that the slime poses a danger, and hence that he fears the slime. (Walton, 1990, p. 245)

Let us mention only to put aside the issue of how a quasi-emotion can be caused by imagining. Walton briefy considers this question before concluding, rightly that it ‘does not matter for our purposes’; we can simply grant that it does, and that there is some (probably Darwinian) explanation. The key conclusion Walton draws from his discussion is that ‘it is not true but fictional that [Charles] fears the slime’ (Walton, 1990, p. 242) or, as he puts it in later discussion, ‘it is only in imagination that Charles fears the Slime, and that appreciators do not literally pity Willy Loman, grieve for Anna Karenina, and admire Superman’ (Walton, 1997, p. 281). It is important to note, as Walton does, that this is not some ad hoc move on his part, or some specially formulated solution to some problem (that is, to ‘the paradox of fiction’). As Walton says, it is implied by his general view; it only makes use of resources he has ‘on hand anyway’ (Walton, 1990, p. 249).5 Given Walton’s account, is there any reason to think otherwise? What motivation would anyone have to oppose the claim that it was fictional, rather than actual, that Charles feared the slime? As we have seen, one motivation is that the properties of the emotion, such as its phenomenology and its involuntariness, are thought to be incompatible with it being ‘merely’ fictional. We now know that this is simply an error; there is no such incompatibility (Walton, 1997, p. 281). Once this is put to one side, it is not clear what underpins opposition to the claim. Walton also provides two good reasons for accepting the distinction that his theory has delivered for us. The first is that the distinction is one we are familiar with, and need, in life. Consider another of Walton’s examples. ‘Let us say that stumps are bears,’ Eric proposes. Gregory agrees, and game of make-believe is begun, one in which stumps . . . ‘count as’ bears. Coming upon a stump in the forest, Eric and Gregory imagine a bear. Part of what they imagine is that there is a bear on a certain spot—the spot usually occupied by the stump. ‘Hey, there’s a bear over there!’ Gregory yells to Eric. Susan, who is not in on the game but overhears, is alarmed. So Eric reassures her that it is only ‘in the game’ that there is a bear in the place indicated. The proposition that there is a bear there is fictional in the game. (Walton, 1990, p. 37) If it were actual that there was a bear, Susan would be right to be alarmed. As it is fctional that there is a bear, she need not be alarmed. The example can be easily extended. Susan, arriving on the walk late, asks Gregory if Eric is enjoying himself. If Gregory were to reply ‘He is absolutely terrifed

Walton on ‘the Paradox of Fiction’

63

of being attacked by a bear’, that would (again) be likely to alarm Susan. However, Susan need not be alarmed; it is only ‘in the game’ that Eric is terrifed of being attacked by a bear. It is only fctional that he is so terrifed. Unless this distinction, the very distinction Walton is pressing on us, is observed there would be nothing to indicate to Susan whether she ought to be alarmed. However, that is wrong; there is something to indicate whether she ought to be alarmed. If it were actual that Eric is terrifed then she ought to be, but as it is fctional that he is terrifed she has no reason to be. The second reason for accepting Walton’s distinction stems from general reflections on the nature of the imagination. Walton cites the wellknown study from the literature on simulation. Subjects were asked to suppose that a Mr. Crane and a Mr. Tees are scheduled to take different flights, departing from the same time from the same airport. The limousine is caught in traffic on the way to the airport, and arrives 30 minutes after the departure time. Mr. Crane is told that his flight left on time. Mr. Tees is told that his was delayed and left just five minutes ago. The subjects were then asked which of the two is more upset? (Walton, 1997, p. 278) The point of the study is to show that subjects imagine being in the circumstances of Mr. Crane and Mr. Tees in turn. They ‘simulate’ the cognitive states of each of these characters, the various mental mechanisms turn, and generate some affective state of annoyance or being upset. An affective state of this sort is then attributed by the subjects to, respectively, Mr. Crane and Mr. Tees. By monitoring this process, subjects generally conclude that Mr. Tees will be more upset than Mr. Crane. As Walton says, some such story is fairly standard in the simulation literature, as is the view (which Walton quotes from Alvin Goldman) that ‘the output state should be viewed as a pretend or surrogate state, since presumably a simulator doesn’t feel the very same affect or emotion as a real agent would’ (Goldman, 1995, p. 189).6 Walton concludes that even if the person doing the simulating ‘may really be distressed or upset’, qualifying this with the parenthetical remark ‘(in one sense of those terms anyway)’, it is not true that the person doing the simulating ‘is annoyed at having missed her flight’ (Walton, 1997, p. 279). This is for the very good reason that, as everyone is fully aware, there is no such flight. The point is that the case of the simulator is relevantly like the case of Charles. Thus, if we are committed to saying this in the case of the simulator (which surely we are), then we are committed to saying it in the case of Charles. A non-actual scenario is being imagined. In the course of this imagined scenario, various affective states are generated. It is a misdescription to say, in either case, that it is a true

64

Derek Matravers

description of the imaginer that they feel this affective state towards the object; rather it is fictional (or imagined) that they feel this affective state towards the object.

III. An Objection There is, however, a further argument which aims to undermine Walton’s account that, although unsuccessful, casts an intriguing light on the project as a whole. It is structurally similar to the argument considered earlier. It aims to find scenarios which are relevantly like the case of Charles but in which the natural claim to make is that the object that is believed not to exist (or, at least, believed not to be present) is the object of an actual emotion. By parity of reasoning, this will support the conclusion that what Charles feels for the slime is an actual emotion. To understand this argument, we first need to investigate a persistent confusion in the literature on the paradox of fiction. Lamarque and Olsen’s version of (2), given earlier, is: 2. A necessary condition for experiencing emotions such as fear, pity, desire, etc. is that those experiencing them believe the objects of their emotions to exist. What is meant by ‘exist’? It could mean ‘currently exists’ or it could mean something like ‘of this world; it exists now, has existed in the past, or might exist in the future’.7 Neither option seems happy for the paradox of fction. If it means ‘currently exists’ it would follow that we could not feel emotions towards states of affairs in the past or states of affairs in the future. This would be a bizarre claim. It would entail denying that we could pity those who were operated on before modern medicine or dread an anticipated beating by a school bully – surely two paradigm instances of emotion.8 Moving to the weaker option, that ‘the objects of the emotions must exist now, have existed in the past, or might exist in the future’, brings in a raft of further implausibilities. It would entail, for example, that while Charles is watching Henry IV Part 1 and sees Prince Hal and Falstaff engaging in witty conversation, his response is bifurcated into actual admiration for Hal and fctional admiration for Falstaff. Furthermore, the paradox of fction seems in danger of vanishing altogether. If it is conceded that Charles can feel an actual emotion towards a state of affairs that ‘might exist in the future’ (that is, a hypothetical state of affairs), it is very diffcult to understand the prohibition on feeling an actual emotion towards a fctional state of affairs; after all, a fctional state of affairs in some ways just is a hypothetical state of affairs. Walton himself seems to opt for the first, stronger, option. To allow that mere fictions are objects of our psychological attitudes while disallowing the possibility of fictional interaction severs the

Walton on ‘the Paradox of Fiction’

65

normal links between the physical and the psychological. What is pity or anger that is never to be acted on? What is love that cannot be expressed to its object and is logically or metaphysically incapable of consummation? We cannot even try to rescue Robinson Crusoe from his island, no matter how deep our concern for him. (Walton, 1990, p. 196) Fear emasculated by subtracting its distinctive motivational force is not fear at all. (Walton, 1990, p. 69) That is, it appears that Walton is claiming that it is a necessary condition of feeling an actual emotion for some object that we, currently, can causally interact with that object. Now the argument comes into play. As we have seen, we can easily envisage a situation in which we feel an actual emotion towards some object (or state of affairs) with which, currently, we cannot causally interact: our pity for those who were carved up with no anaesthetic or our dread at some anticipated beating. Such situations lack the necessary condition for feeling an actual emotion that Walton specifes. However, we have no reason to think this undermines feeling an actual emotion in these cases; by parity of reasoning, there is no reason to think it undermines feeling an actual emotion in the case of Charles.9 The clearest way to understand Walton’s thought here is to look at his response to an argument of the aforementioned form from Richard Moran. Moran has described certain scenarios which he takes to be ‘paradigms’ or ‘central’ instances of the experience of actual emotion. He then claims that, as situations such as that of Charles are relevantly like these, there is no reason to deny that Charles feels an actual emotion (Moran, 1994). Consider the following, from Amanda Foreman’s biography of Georgiana, Duchess of Devonshire. Her right eye had swelled to the size of an apricot. . . . The children were dispatched to Chiswick so they would not hear their mother’s screams. . . . Georgiana’s illness and the experiments performed on her in the name of medicine were appalling even by eighteenth century standards. There was no anaesthetic except laudanum, no appreciation of cleanliness or even a basic understanding of the origins of infections. One of the doctors almost strangled her when he tried to force the blood up to her head in the belief that the eye needed to be ‘flushed’ through. Harriet told her lover, ‘After hearing what I did tonight I can bear anything’. (Foreman, 1999, p. 299) What we feel for Georgiana is surely actual pity and, as the situation of Charles and the slime is relevantly similar, what Charles feels for the slime is actual terror.

66

Derek Matravers

Walton replies to Moran’s argument as follows: Is the person who shudders on recalling her automobile accident terrified of the truck that she remembers careening into her car years previously? Moran thinks that, as a ‘paradigm,’ this example is not or should not be regarded as ‘paradoxical’. Whatever he means by this, it certainly does not follow that she is, literally, afraid of the truck when she recalls the accident. . . . Vivid memories like the ones Moran discusses involve imagining; one relives the remembered experience. The shudders result from vividly imagining the truck careening into one’s car, despite being fully aware that that is not now happening. One is terrified of the truck in imagination; there is no need to insist that one is (also) actually terrified of it. (Walton, 1997, p. 284) The crucial point to note, the point that has been missing so far, is that what concerns Walton is imaginative participation in games of makebelieve. The accident victim, let us call her Bella, believes that the truck exists (or, at least existed)10 and believes that, at some time in the past, it careened into her car. What interests Walton is an account of her reliving the experience; that is, of her make-believing that, right now, that there is a truck travelling towards her in her car. Of course, it is false that, right now, there is a truck travelling towards her. Does it make sense to say of Bella, when she is reliving the experience, that she is actually terrifed of that truck – that is, the truck travelling towards her? Recall that it is already built into Walton’s theory that the imagined emotion could be phenomenologically indistinguishable from the emotion felt on the actual occasion (although it would be odd in this case) and that it is a genuine emotion. With all this, it seems wilfully misleading to describe Bella’s state, when reliving the experience, as actual terror of the truck bearing down on her (for, as she is fully aware, there is no truck bearing down on her). The same considerations can answer the aforementioned argument. The reader, vividly reliving the doctor throttling Georgiana while she screams with pain, feels pity for Georgiana. However, as with Bella, there is no Duchess who is currently being throttled. The pity is part of the imaginative project; hence the reader does not feel actual pity for the throttled Duchess (they are fully aware that there is no throttled Duchess). What the reader feels is fictional pity for an imagined Duchess. By parity of reasoning, what Charles feels is fictional fear for an imagined slime. It looks as if the argument is defeated.

IV. The Broader Problem However, if this is the correct reading of Walton, it raises a broader question about Walton’s project. The argument Walton is running relies on the plausible claim that when we vividly relive some state of affairs we

Walton on ‘the Paradox of Fiction’

67

cannot causally interact with that state of affairs (as it is not taking place here and now). Hence, any emotion we feel towards the state of affairs that is vividly relived is an imagined emotion. Such an argument seems eminently sensible and I can think of no reason to dispute it. Note that there is no restriction on the nature of the state of affairs we choose to vividly relive in this way; it can be either fictional or non-fictional (as traditionally construed). Both my example (the Duchess of Devonshire) and Walton’s example (Bella) are non-fictional. Thus, the first conclusion to draw is that Walton’s discussion is completely unrelated to the distinction that underpins the paradox of fiction. Thus, on Walton’s view, the considerations that apply to our psychological interactions with fictions apply equally to our psychological reactions with non-fictions (once again, as those terms are traditionally construed). Is this a problem for Walton? It looks as if it ought to be. He sets out to give an account of fiction but the distinction he here relies upon is unrelated to the fiction/non-fiction distinction. However, to go over some familiar ground, there is, in Mimesis, a lack of clarity as to what Walton takes the extension of ‘fiction’ to be (Friend, 2008). On the one hand, he is happy to distinguish (at least initially) between ‘novels, stories, and tales’ and ‘biographies, histories, and textbooks’ (Walton, 1990, p.  3). Furthermore, he draws a clear contrast between fictions, which establish fictional worlds, and non-fictions, which describe the actual world (Walton, 1990, p. 70). On the other hand, Walton is no respecter of ordinary language; concepts need to be shaped to do the best theoretical work they can do (Walton, 2007). Thus, although the notion of fiction with which he works ‘is a natural descendant of the one used by booksellers and librarians. . . . [T]his is not to say that we should expect to draw the line just where they do’ (Walton, 1990, p. 72). Given that Walton would want ‘fiction’ to have the extension which would enable it to do the best theoretical work it can do, perhaps the reliving of Bella’s accident and the reliving of the Duchess’s operation would, for him, count as fiction. Whether this is granted or not, the concerns motivating the paradox of fiction have been left far behind. The lack of clarity over the extension of ‘fiction’ is an aspect of the well-known problem with the theory, namely that Walton does not give a detailed account of ‘the imagination’, leaving the term as ‘a placeholder for a notion yet to be fully clarified’ (Walton, 1990, p.  21). Thus, his account of what it is for a proposition to be fictional (those ‘whose imaginings are mandated’) leaves much still to be decided. There are certainly some places where he seems to be equating ‘imagine’ with something like ‘vividly relive’ – as is suggested by the earlier arguments. Furthermore, he appears to suggest that such an equation would net not all of what have been traditionally classed as fictions as well as some non-fictions: Some histories are written in such a vivid, novelistic style that they almost inevitably induce the reader to imagine what is said, regardless

68

Derek Matravers of whether or not he believes it. (Indeed this may be true of Prescott’s History of the Conquest of Peru). (Walton, 1990, p. 71)

However, leaning too much on the aspects of our concept of the imagination that link it to vividly reliving threatens to stretch our grasp of the imagination further than it can bear. This is what Walton says about representations featuring the written word. One can use words specifying a given proposition to ask other to imagine it. If a collection of words, a verbal or written text, is itself thought of as using such prescriptions, it is a prop; and if it is its function to be a prop, it is a representation in our sense. Some literary works might be thought of in just this way, as simply mandating the imagining of the propositions their sentences specify. (Walton, 1990, p. 353) If literary works are thought of ‘as simply mandating the imagining of the propositions their sentences specify’, this would fall short of vividly reliving their content. The mandate to imagine cannot be, on each and every occasion, the mandate to vividly relive whatever proposition the sentences specify; that would surely make engaging with representations impossibly time-consuming and exhausting. However, as we have seen, propositional imagining is built into the heart of the theory: that ‘propositions whose imagining is mandated are fctional’ (Walton, 1990, p. 69). What is mandated must be something psychologically less demanding; something such as the reader ‘representing the content to him or herself’ (or, as some people put it, thinking of the imagination as ‘the fctional analogue of belief’). Thus, the broader question raised by the discussion of ‘the paradox of fiction’ is the extent to which Walton preserves the unity of his theory only by relying on very different claims about the imagination. As we have seen, the theory relies on the imagination encompassing two different roles: a ‘strong role’ in which, in imagining a scenario, a reader vividly relives it and a ‘weak role’ in which imagination is the mental state by which a reader represents the content of a fiction to him or herself (or, as I shall sometimes put it, scenarios that are ‘strongly imagined’ or ‘weakly imagined’). It is not surprising that the discussion of Walton’s contribution has been confused. Unless one begins by specifying which aspect of the imagination one has in mind one is bound to go wrong. Certain things that are true of imagination in its strong role are not true of it in its weak role, and vice versa. Although propositions that are strongly imagined are also weakly imagined but not vice versa, it is worth exploring, briefly, whether some single conception of the imagination could be rescued that would serve Walton’s turn. Walton is keen, throughout Mimesis, to stress that ‘imagining

Walton on ‘the Paradox of Fiction’

69

(propositional imagining), like (propositional) believing or desiring, is doing something with a proposition one has in mind’ (Walton, 1990, p. 20). So perhaps the most consistent reading is simply to conjoin the two lines of thought: props mandate readers to imagine the propositions sentences express, where imagining is something stronger than simply giving the proposition a place in some represented content but (usually) weaker than vividly reliving whatever scenario is described (although it will sometimes be this). It is not clear that this irenic line of thought will work. The claim that, in order to count as an imagining, a mental state has to be at least some way along the road towards vividly reliving looks to be an empirical claim – and would seem to run counter to assumptions in the psychology of text processing.11 More pertinently, Walton himself claims that our imaginings are frequently non-occurrent. While it is true that non-occurrent imaginings are not necessarily unconscious (although surely they will often be), Walton also ‘suspects that even occurrent imaginings need not be conscious’ (Walton, 1990, p. 18). Thus, it looks as if Walton is committed to the view that, generally, readers are mandated to weakly imagine propositions and only sometimes strongly imagine them.

V. The ‘Paradox of Fiction’ and Weak Imagining We have explored strongly imagining and the role it has in emotions felt towards states of affairs that are (currently) not present to us. I shall devote the remainder of this chapter to exploring weak imagining. I do not propose to say anything concerning whether the weak role of the imagination can be the basis for a distinction between fiction and nonfiction (traditionally construed). Elsewhere I have argued that it cannot be.12 Instead, I shall focus on whether any issues remain concerning emotions felt towards states of affairs that are weakly imagined. It is generally agreed that the mental model readers use to represent the content of a narrative consists of both propositional and nonpropositional elements. For example, the model will represent its content in a way that preserves spatial information (Coplan, 2004). It also is generally accepted that the model will embody readers’ emotional responses to states of affairs.13 I shall assume this is true and also assume, to draw a contrast with strong imagining, that the emotional reactions are embodied in a way that is not immediately open to consciousness.14 That is, the existence of these embodied emotional responses is revealed through psychological tests – akin to the way that the embodied spatial information is revealed. Does the fact that emotions are embodied in readers’ mental models raise issues that have generally been dealt with under ‘the paradox of fiction’? Put another way, does weak imagining raise the same kinds of issues as strong imagining? Let us return to Walton’s reply to Moran, and the example of Bella who survived her car being hit by a truck. Earlier, we discussed Bella

70

Derek Matravers

strongly imagining the event – vividly reliving it. Bella must also weakly imagine the event; that is, represent it to herself in some way. Embodied in this representation will be certain emotional responses. Although this would seem the central case – weak imagining being universal in storing sequences of events in narrative structure and strong imagining the exception – it is difficult to know what to say about it. The key elements of Walton’s account (the ‘doing something’ with a proposition, the vivid reliving of the experience, and the subsequent state characterised by a quasi-emotion) are all absent. Nonetheless, as the central case, how (if at all) does it fall within ‘the paradox of fiction’? With weak imagining the form of argument Moran uses seems to have some bite. Inasmuch as Bella’s memory embodies emotional responses, surely those are actual emotional responses rather than imagined emotional responses. If so, then inasmuch as Bella’s weak imagining is relevantly similar to the weak imagining of a reader of fiction, we do not have that reason to deny that the emotional responses embodied in the weak imaginings of the reader of fiction are actual emotional responses. In the case of strong imagining Walton was able to appeal to the possibility of causal interaction: the distinction between the truck careening towards Bella in the here-and-now (actual emotion) and the truck careening towards Bella when she vividly relived the event (imagined emotion). As this distinction is absent in the case of weak imagining, to what can Walton appeal to defend himself against Moran? This question takes us closer to the kinds of consideration associated with the usual discussion of ‘the paradox of fiction’ and thus its attendant problems. The intuitively obvious way to draw the distinction is that Bella’s mental model represents a state of affairs that actually happened, while that of the reader of fiction represents a state of affairs that never happened. However, that takes us back to familiar problems concerning psychological attitudes to states of affairs that do not exist: future events and hypothetical events (which have also never happened) would fall on the side of imagined emotions when intuitively they ought to fall on the side of actual emotions. Walton could bite the bullet and run a similar argument to that which he ran in the case of strong imagination and claim that the mental models of both Bella and the reader of fiction feature imagined emotions. However, that does not seem right. Inasmuch as the question arises (i.e. are Bella’s weakly imagined emotions actual emotions or (merely) imagined emotions?) they are surely actual emotions. That is, inasmuch as there are emotions embodied in Bella’s non-occurrent memory of her being hit by a truck, they are actual emotions and not (merely) imagined emotions. A different response would be to argue that the question as to whether emotional responses in mental models are actual or (merely) imagined should not be asked. It is simply Quixotic to take a strong line. Part of the problem is that the term ‘the imagined’ is being made do too much work:

Walton on ‘the Paradox of Fiction’

71

to describe the representation to oneself of a state of affairs, to mark the contrast with the actual, and to serve as a synonym for ‘made up’, and possibly more besides. Hence, our linguistic intuitions, inasmuch as they are guiding us, will be all over the place. Even if that is true, however, it is not really the issue. Walton offers us an account of the place of emotional responses in fiction. Weak imagining is the standard case, and emotional responses are part of weak imagining. It would not really do to be able to say nothing about such cases.

VI. Conclusion Walton has occasionally complained of being misunderstood by his commentators (Walton, 1997). If the ruminations in this chapter are even vaguely correct, such complaints are amply justified (and I am as guilty as any). Although ‘the paradox of fiction’ provided work for a generation of aestheticians, the nature of the problem was never really clear and thus the ‘solutions’ generated more heat than light (Stecker, 2011). By contrast, Walton’s work is genuinely interesting as it is a facet of his overall account  – a grand theory that illuminates many aspects of representation and the arts. As aforementioned, the wrong way to think of it is as Walton coming up with some novel argument to solve some pre-existing problem. Rather, his concerns arise from general considerations of the account of representation in terms of make-believe and can be sorted out from resources ‘he has on hand anyway’. Thus, what interest there is in the problem lies in the light it throws on the overall theory. This is what I have tried to explore in this chapter. Walton has a good account of the emotions that arise in cases of strong imagination – although that account turns out to be unrelated to the traditional fiction/non-fiction distinction. With weak imagination the position is not so clear. Here, the argument that Moran and others have pushed (that there are cases that seem to be cases of actual emotions that are relevantly similar to cases that Walton would classify as imagined emotions) appears to have more bite. I should like to end on a personal note. I was in the final year of my PhD when Walton published Mimesis as Make-Believe. I have found the book, and the various papers that Walton has written that either are unrelated to the book or go beyond it, a source of great inspiration and enjoyment. Almost everything I have written on the philosophy of music and the philosophy of literature has either been influenced by, or been directly about, his work. In trying to puzzle this stuff out, I have frequently misunderstood and misrepresented Ken’s arguments. I have also committed the grave philosophical sins of oversimplification and overstatement. Despite this, Ken has been unfailingly patient and kind in print and in verbal responses and commentaries on my work and I am exceptionally grateful to him for that. I fear this chapter will simply be one more instance of his patience being tried.

72

Derek Matravers

Notes 1. I am grateful to Alex Barber, Dan Cavedon-Taylor, Sean Cordell, Eva Dadlez, Mark Pinder, and Joshua Thomas for comments on this chapter. 2. Stacie Friend has argued, in an unpublished work, that Radford’s paper itself has been widely misinterpreted. See also Teroni (2019). 3. This was pointed out as early as 1985 by Malcolm Budd (1985, p. 129). 4. Chris Williams rejects this argument on the grounds that (a) the phobic case is aberrant and (b) the phobic’s motivation to flee is ground enough to attribute a belief (2019). I have sympathy for (a) but not for (b). 5. This was pointed out to me, some years ago, by Emily Caddick-Bourne. 6. Although this interpretation is standard, there are other views. See Goldie (2000, p. 201), and perhaps even Walton himself (2015). Nonetheless, Walton’s point goes through even on these other views. 7. A laudable attempt to sort out these issues can be found in Tullmann & Buckwalter, (2014). However, as will become apparent, I do not think the attempt captures the whole story. 8. It is worth noting that Radford goes for the weaker reading, interpreting ‘exists’ as covering not only currently existing states of affairs, but ‘probable’ states of affairs (Radford, 1975, p. 73). 9. This was one of several poor arguments I ran against Walton in Matravers, (1998). In what follows, I will also be reconsidering several other poor arguments I ran against him in Matravers, (2014). 10. How we classify her emotion will not turn on whether she believes that the truck has been scrapped. 11. Psychologists seem not to consider the extent to which the construction of mental models is a conscious process, but their accounts suggest that, largely, it is not. See, for example, McKoon & Ratcliff, (1992). 12. In its weak role, imagination is the means by which a reader mentally represents the content of a narrative and this needs to be done whether the narrative is fiction or non-fiction. Hence, the content of all narratives, whether fiction or non-fiction, needs to be weakly imagined. My view is that the reader constructs a mental model (akin to that which narrative psychology refers to as ‘a situation model’) and that mental model consists of beliefs and propositions that are part of the content of the narrative that are not beliefs  – what Stacie Friend has called ‘mere make-beliefs’ – together with various sorts of non-propositional content (Friend, 2008). There might be a statistical difference in the mix of beliefs and mere make-beliefs in fiction and non-fiction, but, in essence, the story is the same in both cases. However, other construals of the representation, which allow a distinction between fiction and non-fiction, might be possible (Matravers, 2014). 13. See, for example, Oatley, (2002). For a recent paper that contains an overview of recent research in this area, see Nabi & Green, (2015). These papers seem to me to raise more questions than they answer; clearly, there is more research to be done. 14. In short, these are ‘participatory responses’ or ‘p-responses’ (Gerrig, 1993, pp. 65−96).

References Budd, M. (1985). Music and the Emotions. Routledge and Kegan Paul. Carroll, N. (1990). The Philosophy of Horror. Routledge. Coplan, A. (2004). Empathic Engagement with Narrative Fiction. The Journal of Aesthetics and Art Criticism, 62(2), 141−152.

Walton on ‘the Paradox of Fiction’

73

Foreman, A. (1999). Georgiana: Duchess of Devonshire. HarperCollins. Friend, S. (2008). Imagining Fact and Fiction. In K. Thomson-Jones & K. Stock (Eds.), New Waves in Aesthetics (pp. 150−169). Palgrave Macmillan. Gerrig, R. J. (1993). Experiencing Narrative Worlds: On the Psychological Activities of Reading. Westview Press. Goldie, P. (2000). The Emotions. Clarendon Press. Goldman, A. I. (1995). Empathy, Mind, and Morals. In M. Davies & T. Stone (Eds.), Folk Psychology: The Theory of Mind Debate (pp. 185−208). Blackwell. Lamarque, P., & Stein, H. O. (2004). Introduction to Part VI. In Aesthetics and the Philosophy of Art: The Analytic Tradition (pp. 183−184). Blackwell. Matravers, D. (1998). Art and Emotion. Oxford University Press. Matravers, D. (2014). Fiction and Narrative. Oxford University Press. McKoon, G., & Ratcliff, R. (1992). Inference during Reading. Psychological Review, 99(3), 440−466. Moran, R. (1994). The Expression of Feeling in Imagination. Philosophical Review, 103(1), 75−106. Nabi, R. L., & Green, M. C. (2015). The Role of Narrative’s Emotional Flow in Promoting Persuasive Outcomes. Media Psychology, 18(2), 137−162. Oatley, K. (2002). Emotions and the Story Worlds of Fiction. In M. C. Green, J. J. Strange, & T. C. Brock (Eds.), Narrative Impact: Social and Cognitive Foundations (pp. 39−70). Psychology Press. Radford, C. (1975). How Can We Be Moved by the Fate of Anna Karenina? In A. Neill & A. Ridley (Eds.), Arguing About Art (pp. 67−93). Routledge. Sainsbury, R. M. (2010). Fiction and Fictionalism. Routledge. Stecker, R. (2011). Should We Still Care About the Paradox of Fiction? The British Journal of Aesthetics, 51(3), 295−308. Teroni, F. (2019). Emotion, Fiction and Rationality. The British Journal of Aesthetics, 59(2), 113−128. Tullmann, K., & Buckwalter, W. (2014). Does the Paradox of Fiction Exist? Erkenntnis, 79, 779−796. Walton, K. (1990). Mimesis as Make-Believe. Harvard University Press. Walton, K. (1997). Spelunking, Simulation, and Slime: On Being Moved by Fiction. In In Other Shoes. Oxford University Press. Walton, K. (2007). Aesthetics—What? Why? and Wherefore? The Journal of Aesthetics and Art Criticism, 65(2), 147−161. Walton, K. (2015). Empathy, Imagination, and Phenomenal Concepts. In In Other Shoes: Music, Metaphor, Empathy, Existence. Oxford University Press. Williams, C. (2019). Why Quasi-Emotions Should Go Away: A Comment on Dos Santos. The Journal of Aesthetics and Art Criticism, 77(1), 79−82.

5

Fear and Loathing in Fictional Worlds Quasi-Emotion, Nonexistence, and the Slime Paradigm E. M. Dadlez

So much has been written about Kendall Walton’s account of makebelieve in general, and of his account of quasi-emotion in particular, that it is difficult to know how to launch into a discussion of it without unacceptable repetition. I will ask the reader to put up with an abbreviated rehearsal of the disagreements related to the quasi-emotion issue, a process that is sure to seem repetitious to some and unnecessarily telegraphic to others. I will then canvas some of the reasons to think that questions and problems raised in the course of these philosophical exchanges have not been resolved and conclude by making a few proposals of my own.1 I am largely on board with respect to Kendall Walton’s account of make-believe as an explanation of our encounters with fiction. However, I am not at all on board (indeed, I am over the side and swimming for shore) with respect to the part of it that introduces quasi-emotions. Let us imagine that homage has already been paid to Walton’s account of our imaginative immersion in fiction (one with which I almost entirely concur) as participation in games of make-believe. Walton presents us with a set of nested fictional worlds. The world of the work serves as a prop in the world of our game. What is true in the first is true in the second. Further, things are true of us in the world of our game that aren’t true of us in the actual world. In the world of the game I play with the beloved television series Stranger Things (which, coincidentally, also trades in nested worlds) it is true that young Will Byers is trapped in the Upside Down, a dark and terrifying mirror image of the ordinary world. But my game of make-believe also makes things fictional of me, insofar as my responses to the work are involved. It is true in the world of my game that I am aware of Will’s plight, that I believe he is in danger, that I am cognizant of and creeped out by the appearance of the Upside Down, that I fear for Will’s life. Certain conventions govern what turns out to be true in the world of my game. While Walton does not go into great detail respecting these, the limits of what such a game can make fictional of me are clear. For instance, while it is true in the world of my game (that is, imaginarily or make-believedly true) that I notice Will’s pallor and sickly appearance as residence in the Upside Down leeches his life force, it is not true in that DOI: 10.4324/9780367808662-6

Fear and Loathing in Fictional Worlds

75

world that I occupy the same space as Will, that light bounces off his pale visage and strikes my retina, that he is as aware of me as I am of him. And while it is true in my game (rather than that authorized for the work) that I shout hints to the estimable Dustin, it is never true in it that he hears them. Walton’s account of the nature of mimesis in the arts issues in a story that is so faithful to the reasons many of us love fiction as we do, so reflective of our fascination, and so accurate about the extent of our imaginative entrancement, that its popularity and wide-scale adoption are not at all surprising.

I. Of course, having acknowledged this, I am bound to move on to my one area of disagreement with Walton concerning his account of quasiemotion. Walton has it that, while what I describe as my fear for Will is both emotional and phenomenologically similar to or even indistinguishable from fear for a real person, I cannot really fear for what I know is a fiction. Of course, fictions can give rise to authentic fears concerning things we believe true of the actual world, as when the Demogorgon leads us to contemplate threats posed by political figures (whose resemblance to it may not be coincidental). But just as I do not really believe a boy named Will Byers is endangered but only imagine that this is the case (making it fictionally or make-believedly true that I believe it), so, Walton would say, I do not really fear for Will, but only imagine doing so (making it fictionally true that I fear for him). This affective response is called quasi-fear because it lacks the epistemic underpinnings and motivational and behavioral characteristics that fears typically involve. That is, it has some ingredients of a complete emotion, but not all of them. Questions have been raised repeatedly (by me and more celebrated others) about what Walton means when he maintains that we are genuinely moved by novels, films and plays, that we respond to works of fiction with real emotion, . . . our responses to works of fiction are, not uncommonly, more highly charged emotionally than our reactions to actual situations. . . . My make-believe theory was designed to help explain our emotional responses to fiction, not to call their very existence into question. My negative claim is only that our genuine emotional responses to works of fiction do not involve, literally, fearing, grieving for, admiring fictional characters. (Walton, 1997, p. 38) Christopher Williams has recently suggested that Walton means to include quasi-fear and quasi-grief under the umbrella of emotion, but insists on a distinction between these epistemically emancipated responses and the kind of fear and grief to which existential commitment is unproblematic

76

E. M. Dadlez

(Williams, 2019). That is, quasi-fear and quasi-pity belong to the genus emotion, but are of a different species than fear and pity. This is an ingenious explanation and may capture some of what Walton had in mind in the preceding passage. Indeed, I remember a lunch with Kendall Walton (large plates of squid salad seeming to commemorate the slime creature late of “Fearing Fictions”) during the course of which I peppered him with questions about quasi-emotion and received answers that were kinder than I deserved, and that were not incompatible with Williams’s analysis. Whatever may have been said over squid, Walton’s response to Catherine Wilson’s interesting challenge in her “Grief and the Poet” (Wilson, 2013) suggests Williams’s characterization is correct: That the aroused emotions are “no less than entirely real” is uncontentious and uncontested. Of course they are real, and are really emotional—“they” being, for instance, the responses we describe as Charles’ “fear of the slime”. . . . The only question is what sort of emotional experiences they are, whether these descriptions, taken straightforwardly and literally, accurately characterize them. . . . Wilson means to claim, I shall assume, that the emotions Charles and the reader of Anna Karenina (call her “Karen”) really experience are, literally, fear of the slime, pity for Anna, and sadness at her death. . . . There are good reasons to take it to be fictional that Charles fears the slime and that Karen pities Anna, quite apart from the question of whether these propositions are (literally) true.  .  .  . Charles’ and Karen’s experiences are emotional, affectively charged, in any case. And it is perfectly appropriate to describe them as “fearing the slime” and “pitying Anna” (respectively), since it is fictional that they do. (Walton, 2013) However, as an inveterate supporter of the thought theory – to be found in the work of Noel Carroll (1990, 1995), Richard Moran (1994), and early Peter Lamarque (1981) – I retain my objection to the quasi-emotion hypothesis. That is, I believe that full-blown emotions must have a cognitive component, but I believe that this component may be a thought entertained in imagination rather than a belief. For one thing, as Walton himself concedes, fears that he would not regard as quasi-fears may lack a belief component, as in the case of phobic reactions (Walton, 1990, p. 245). Walton follows this concession, however, by stressing that he wants to “think of fear in general as a state akin to that of a certain belief-desire complex, in its motivational force at least” (Walton, 1990, p. 245). But, in the same work, Walton also maintains that while “we cannot be said actually to pity Willy [Loman] or grieve for Anna [Karenina] or admire Superman,” (Walton, 1990, p. 204) “Emily Dickinson, being an actual person, can be an object of actual pity. One may really feel disgust for Ivan the Terrible or empathize with Julius Caesar”

Fear and Loathing in Fictional Worlds

77

(Walton, 1990, p. 252).2 Yet, if we can feel real pity and disgust for historical fgures, it must be without most kinds of motivating impulses, since these individuals are long dead and cannot now directly affect our lives any more than we can affect theirs. Pity for Emily Dickenson cannot depend on an impulse to alleviate her pain, nor can disapproving disgust toward Ivan the Terrible be part and parcel of some motive to bring him to justice or some wish that justice be meted out long after the fact. Granted, we can wish that Emily Dickenson had met a fellowINTP with whom she could happily consort, or that Ivan’s unfortunate daughter-in-law had managed to bat him over the head with some sturdy object. That is, we can talk about what we wish had been, but now cannot be the case. But this applies to fction as well – I can wish that Will had not been trapped in the Upside Down, or that Dustin’s bullies had received their comeuppance sooner than they did. In both historical and fctional contexts, we can wish for things which we are in no position to affect. Of course, my wishing that Will not be endangered doesn’t mean that I want the story to have gone differently. As Alex Neill once pointed out, something’s being a necessary condition of what I wish for does not make it the case that I wish for it in wishing as I do (Neill, 1995, p. 190). Studying hard is a necessary condition for receiving an A in my class, for example. Most of my students wish for that A, but a desire to study will often elude them. Further, empathetic reactions – since Walton says it is possible really to empathize with Julius Caesar – seem to lack both parts of the belief– desire/motivation complex to which Walton refers in Mimesis as MakeBelieve. While Walton himself has recently suggested “doing an end run around the mysteries of imagination” when it comes to empathy, maintaining that empathy need not involve imagining at all (Walton, 2015), many more familiar accounts of empathic emotion are such that Walton would have to characterize those responses as quasi-emotional. Let us consider these. I believe that Julius Caesar existed, of course, but empathizing with him emotionally involves imagining what it is like to be betrayed. That is, I undergo an experience of feeling betrayed when I am, in fact, not betrayed and do not believe I’ve been betrayed. I may believe that Caesar was betrayed, but I don’t believe that I have undergone the betrayal with which I empathize. The empathetic nature of the emotion situates me as the target of treachery. Neither of Walton’s conditions are met in this case – my motivations are not activated, and those I regard as betrayers and feel betrayed by haven’t betrayed me. Empathizing imaginatively places us in a situation that we don’t believe is our own. We adopt a first-person perspective on the other’s experience in order to empathize. According to Walton’s earlier writing, it “involves imagining oneself in the shoes of the person identified with” (Walton, 1990, p. 255). Thus, neither of Walton’s criteria applies if this characterization of empathy is true, since the person empathized with isn’t the object of our emotion. Instead,

78

E. M. Dadlez

his or her apprehension of danger or treachery is adopted by us. Since we do not believe we’re threatened or endangered, the epistemic condition cannot be met in cases of empathetic emotion and, as indicated, our motivation to act and our desires are similarly derailed. Now this could mean that all empathetic emotions are quasi-emotions for Walton, and much of what he writes in a later article (Walton, 1997) suggests as much. If empathizing involves simulating an experience one knows one isn’t actually undergoing, then perhaps, in a reversal from the era of the Julius Caesar claim, all empathetic fear is quasi-fear so far as Walton is concerned. Or, perhaps, empathetic reactions have a radically different structure from that depicted and need not involve imagination to begin with (Walton, 2015). While treating empathetic emotion as quasi-emotion (or as lacking imaginative components altogether) need not be regarded as a deal breaker, it should be acknowledged to appear counterintuitive in certain lights and to be at odds with Walton’s earlier positions.

II. It is worth mentioning in passing that there is some confusion about whether the belief in Walton’s belief-desire complex is thought to cause the emotion or to be a constituent of it. Most cognitivists about emotion such as Robert Solomon, who regard emotions as akin to judgments, take the latter course. The intentional object of the emotion is an epistemic object. Thought theorists (like Noel Carroll) who challenge the quasiemotion account merely allow that the intentional object could equally well be an imagined one – the content of a thought that is entertained in imagination and not believed, as is the case when we entertain hypotheticals. Whether it is supposed that the thought in question is a construal, as Robert C. Roberts (1988) maintains, or whether it is considered a state of object-directed affect with evaluative propositional content, as Patricia Greenspan (1988) suggests, both philosophers concede that such states can sometimes fall short of belief. In each case, whether of believing or of imagining, the object might not really exist except in our minds: in the first instance, because our belief about it could be false. So, if I fear for Sam because he boards a plane that I falsely believe has a history of malfunction, my fear is real, despite the fact that the danger my belief concerns is nonexistent. Similarly, if I fear for Will in the dangerous realm of the Upside Down, the thought that I entertain in imagination is about something that doesn’t exist. Thought theory would never propose (as has been mistakenly claimed) that thoughts were the intentional objects of our fear, any more than a theory that made beliefs the cognitive constituents of emotion would propose that beliefs themselves were its objects. In other words, as Peter Lamarque pointed out, we may be frightened by having a certain thought or belief, but we’re very seldom frightened of it (Lamarque, 1994, p. 105).

Fear and Loathing in Fictional Worlds

79

At the risk of arousing the ire of anti-Meinongians, it is worth noting that the objects toward which we take up the relevant emotional attitudes seem to be what those thoughts or beliefs are about, whether they reflect the truth or not, and whether we know that or not. So, the beliefs and thoughts that can figure in a causal story about our affective response are both real (though psychological) things, and it is in both cases their content that directs us toward the object of the emotion. But it is possible to offer an analysis of the preceding without resorting to a Meinongian approach. First, any affective response to fiction will necessarily depend on an array of broad conceptual beliefs that we already possess (Boruah, 1988, p. 114). Fearing for a character depends on our believing that the kind of situation he is shown to undergo is dangerous. This applies even in the case of Will’s sojourn in the Upside Down. The fear of those who say they fear for Will is buttressed by any number of general beliefs about what constitutes danger: the belief that being trapped in hostile territory is dangerous, the belief that it is perilous to be hunted by powerful predators, the belief that facing attack without the support of allies is risky. Our beliefs about what is or can be dangerous partly determine whether we will feel fear, in both fictional and real-world contexts. These broad beliefs – beliefs about danger in the case of fear, suffering in the case of pity, desirable abilities in the case of admiration – are usually dispositional and become occurrent when the emotion is triggered. They are necessary (not sufficient) for an emotional response. Of course, fictions or new situations in our lives can sometimes lead us to alter our doxastic repertoire by foregrounding exceptions to generalizations or offering new perspectives. The claim here is only that affective reactions that aren’t cases of simple contagion clearly build on conceptions of what constitutes danger when we respond fearfully or on conceptions of what constitutes suffering when we respond with pity. One wouldn’t pity someone for undergoing an experience one didn’t regard as unpleasant or fear for a person one thought perfectly safe. Such beliefs are tapped into in our response to fiction. They are perfectly authentic beliefs, albeit existentially uncommitted. So there are in fact underpinnings to our affective responses to fiction that are firmly grounded in our beliefs, just as Walton’s approach would require of fear and pity to which the ‘quasi-’ prefix did not apply. Next, consider the kinds of nonidentity cases discussed by Derek Parfit (1984, 2011), David Boonin (2008), and others. These address our moral obligations to not-yet-existing future people, as well as questions to which such considerations give rise. Some of the latter involve an important distinction between making existing people better or worse off than they might otherwise have been on the one hand and acting so as to impact whoever comes to occupy a certain position (or experience a certain kind of situation) on the other. So, for instance, imagine a scenario in which a woman could conceive an unhealthy child in the coming month

80 E. M. Dadlez or conceive a perfectly healthy child if she waits a year. If she declines to wait, she hasn’t wronged her existing child, which presumably has a less-than-optimal life that is nonetheless worth living. She wrongs a kind of office- or position-holder – an office that will be filled by whoever the child turns out to be (Dadlez & Haramia, 2015). Such offices generate genuine obligations on the part of office caretakers, independent of particulars concerning those who come to fill them. Parents have obligations toward future children who do not yet exist, pet owners toward not-yetadopted pets, teachers toward future students. Just as the relation between .  .  . office caretakers and the offices themselves . . . generates and justifies general obligations even without the existence of an actual person, the relation between a fiction’s audience and the beliefs that are tapped into by the fiction . . . can also generate and justify general obligations absent the belief that the person in question exists. (Dadlez & Haramia, 2015, p. 10) General beliefs about what can constitute danger in a situation are analogous to what has been described as a “position” or “office.” Indeed, Ronald de Sousa maintains that we come to know the vocabulary of emotion by association with paradigm scenarios. These are drawn first from our daily life as small children and later reinforced by the stories, art, and culture to which we are exposed. Later still . . . they are supplemented and refined by literature. Paradigm scenarios involve two aspects: first, a situation type providing the characteristic objects of the specific emotion type . . . and second a set of characteristic or normal responses to the situation. (de Sousa, 1990, p. 182) That is, our fear arises from what we have come to know or believe is dangerous. And fictions can act as just such paradigm scenarios vividly presenting situation types concerning which we have usually already made broad evaluative judgments – about what it is for a situation to be tragic or dangerous or unjust. Works of fiction that arouse affective response instantiate such situation types or possible experiences in their descriptions or depictions and thereby in the imagination. The fictional catalyst can be held to situate and direct a genuine emotional response arising out of our evaluative belief. We can have obligations to position holders, to experiencers of situations, as well as to existing people. Many evaluative beliefs that underlie affective response to fiction – about the kinds of situations we regard as unjust, or about behaviors whose exhibition we regard as deplorable – focus attention on positions and offices: on dreadful situations that could

Fear and Loathing in Fictional Worlds

81

be experienced or inappropriate conduct that could be exhibited This is certainly comparable to talk of office holders in other contexts. So, for instance, we are outraged by Dr. Brenner’s treatment of Eleven (in Stranger Things) as nothing more than a lab rat and of the citizens of Hawkins as mere cases of collateral damage. We pity and fear for Eleven and worry about the denizens of Hawkins, given the callous attitude taken toward innocent bystanders by Hawkins National Laboratory. Such reactions arise from a set of interrelated dispositional beliefs that the fiction has brought to the forefront of awareness. Our reactions stem from beliefs we may have about using people as means to ends, about when it is permissible to put communities at risk, about obligatory disclosure and informed consent. These are beliefs about actions that could be performed or experiences that could be undergone – they are about offices, the filling of which is illustrated in fictional descriptions or depictions. They reflect something about the obligations that exert a force on us both emotionally and morally: obligations not to use people as means to ends, not to abuse power, not to place innocent communities at risk without sufficient justification. So, it appears that emotional responses to fiction depend in a fundamental way on perfectly authentic evaluative beliefs and that they are also (especially in the case of emotional approval and disapproval) affiliated with obligations that can exert pressure on us in the broader arena of our lives. Necessary concomitants of our emotional response to fiction have both the epistemic respectability and the motivational force that Walton requires and can lay claim to both without resorting to any commitment whatsoever to the existence of the emotion’s object. Walton might well say that such broad beliefs must accompany any affective reaction, including that experienced in the course of a game of mud pies or monster (in which its being fictional I’m angry about a pie theft depends on my believing that stealing other people’s property is usually objectionable and its being fictional I fear the monster depends on the belief that pursuit by powerful predators is dangerous). These cognitions are not the ones whose absence makes the ascription of quasiemotion appropriate, for Walton. Why are they relevant? The claim here, however, is not only that such cognitions, plus the obligations and motivations that they generate, can ground an emotional response. The claim is that they are much more likely to do so in the case of fiction than in the case of other kinds of make-believe (and perhaps also in the case of response to immediate real-life events). So, first, if we can have obligations (as I think we can) to mere position holders and these can produce motivations to act, then it doesn’t seem much of a stretch to believe that we can have full-blown emotions about the experiences and situations that such positions involve, especially when these undergoings are represented to us with genius and artistry. Whether we characterize such responses as reactions to DeSousa’s paradigm scenarios

82 E. M. Dadlez or to Parfit-style positions or position-holders, the point here is to offer an analysis of fearing for or pitying Will Byers that situates Will as a placeholder for a specific set of experiences that we believe would be dreadful to undergo. This isn’t a case of pitying everyone in Will’s position even if we don’t know who they are, a suggestion famously railed against by Colin Radford. (I think, incidently, that it is perfectly possible to pity people even if I don’t know who they are. I pity anyone who had to brief President Trump on his poll numbers, for instance.) Rather, the object of our emotion is prospective and involves the constellation of (pitiable or tragic or fearful) experiences that comprise the situation type or position depicted in the fiction. What moves us (and often obligates us to behave in certain ways in prospective situations) is, in effect, our awareness of how it can be to undergo something like that – the plight of a prospective undergoer. I acknowledge that the target of the emotion is hypothetical, a positionholder rather than some particular person of whose existence we are currently aware. However, it is perfectly possible to be unproblematically moved by prospects for human experience. Fictional characters are not unique in providing, but are uniquely adept at illustrating and representing such prospects and at making them vivid and accessible. We can be moved by a situation type represented by a fiction, because the constellation of dispositional beliefs about how it might be to undergo that kind of situation has become occurrent. It is uncontroversial to contend that we often turn to fiction to explore how different kinds of experiences can be. At its best, fiction can make a fairly complex set of dispositional beliefs occurrent in such a way as to arouse an emotional response to a position represented by a character’s fictional situation. Fiction can make particular prospects for the world vivid and immediate. Since real life lacks the careful orchestration of fiction and provides readymade epistemic consorts for emotion, dispositional beliefs more often stay dispositional in the moment when we respond emotionally to actualities rather than fictions (though presumably they will more often become occurrent on reflection). That is, fiction is especially well adapted in eliciting such responses, to mustering the doxastic repertoire in such a way as to arouse an emotional response. I know some will not accept emotional responses to situation types or position-holders as an analysis of what it is to be genuinely moved by Will Byers’s plight, but simply regard the former as an additional source of genuine emotional response, along with real fear that the thriller will give one a heart attack or real admiration of a historical person depicted in the fiction. But in that case, it appears that instances of full-blown emotional response in the course of our interaction with fiction are far more frequent than Walton has been willing to allow. I propose further that few games of make-believe that do not involve fictions will give rise to affective reactions with situation types or position

Fear and Loathing in Fictional Worlds

83

holders as objects. My fictional outrage about the liberation of my pies is unlikely to involve outrage about the violation of property rights in general, or solidarity with the pie-deprived. My fear of the monster (imagine a roaring father who pursues a giggling child) is unlikely to batten on a sense of general susceptibility to attack. That is in part due, I will suggest in Section III, to a blurring of fictional and extra-fictional perspectives and to the role of personal agency in what is made believe.

III. Here is one further proposal: the discussion concerning quasi-emotion has been centered on the wrong cases as paradigms. This has led us to confuse affective responses that may legitimately be considered quasi-fears with real fears that arise as the result of contemplating fiction. To that end, I aim to draw a distinction among ways in which readers or viewers can function as props in games of make-believe. Participants in games that use fictions as props are, says Walton, reflexive props in these games, generating fictional truths about themselves and their actions and responses (Walton, 1990, p. 213). They imagine doing things and undergoing experiences. Their affective reactions make it true in the world of their game that they fear for or pity characters. Their beliefs that there are fictional dangers make it fictional of them that they believe there are dangers. All of the states of affairs described or depicted in the world of the work are true in the participant’s game. But game worlds are larger than work worlds, since they also include the fictional truths players (imaginatively) generate about themselves, using the work and their own reactions to it as props. Players make it true of themselves in the world of the game that they believe the events described in the work have transpired, that they are moved by them, that they are aware of them and reflect on them. As Walton points out, If to read a novel or contemplate a painting were merely to stand outside a fictional world pressing one’s nose against the glass and peer in, noticing what is fictional, but not fictionally noticing anything, our interests in novels and paintings would indeed be mysterious. . . . We don’t just observe fictional worlds from without. We live in them (in the world of our games, not work worlds) together with Anna Karenina, Emma Bovary and Robinson Crusoe, . . . sharing their joys and sorrows. (Walton, 1990, p. 273) I propose to ask about how we share those worlds. It is true in the world of the participant’s game, according to Walton, that s/he shares a world with the characters depicted or described in the fiction. What does this amount to? It won’t always on that account be

84

E. M. Dadlez

true in the world of the game that a participant occupies the same physical space as the characters, that she occupies the same room or car or stormtossed boat. A reader of Mary Shelley’s The Last Man is not playing a game in which both he and Shelley’s last existing human being each exist, physically, in the same world. And nothing seems to establish its being fictionally the case that participants are eavesdroppers and peeping toms, even though it is fictionally the case that they are privy to the content of private – sometimes very private – interactions. The unspoken rules or conventions of the games of make-believe we play with fiction may defuse such apparent paradoxes in any of several ways, says Walton. Problematic fictional truths may not be thought generated in the first place or may be de-emphasized for the sake of imaginative immersion. Two inconsistent truths may each be held true in the world of the game while their conjunction is not held to be fictionally true (Walton, 1990, p. 239). Walton concedes that: The appreciator is encouraged to concentrate on fictional truths about what the characters are up to, the predicaments they find themselves in, and what they think and feel. It is important also what fictionally he thinks and feels about them, but in abstraction, to a certain extent, from how they do or might affect him. (Charles’s confrontation with the slime, involving an aside as it does, is unusual in this respect). This allows the appreciator a kind of empathy with the characters, an ability to look at things more purely from their points of view, from a perspective relatively uncontaminated by his own personal concerns. The “objectivity” of this perspective contributes, no doubt, to the impression some may have that appreciators do not ordinarily belong to fictional worlds at all, that . . . they merely observe from without. This impression is mistaken, but we can think of the appreciator as having a rather “sketchy” or “ghostly” presence in the world of his game, in light of the restrictions on his role in it and the indeterminacy that often results from them. . . . Asides disturb this objectivity momentarily and reduce the indeterminacy. They . . . give him a slightly fuller presence in the world of his game. (Walton, 1990, p. 237) Here is my somewhat heretical fnal proposal, then, based on the distinction Walton draws between “ghostly” and fuller participant presence in the worlds of games of make-believe involving fction. I want to maintain that the “fuller” someone’s presence is (and sometimes the more it resembles participation in role-playing games) the more appropriate it will be to consider her affective responses to be quasi-emotions, on the ground that these import real-world considerations into the world of the game in a way that requires a kind of coordinated awareness of personal, sometimes extra-fctional, concerns together with imagining what

Fear and Loathing in Fictional Worlds

85

is fctional. That kind of dual attention, the so-called contamination that such cases bring into the picture, suggests something more closely resembling conscious pretense and an occurrent awareness of the fctionality of the object of the response. Explicit awareness of the fctionality of the events one contemplates undercuts the activation or reactivation of the broad evaluative beliefs discussed in the previous section. These are typically awakened by imaginative immersion in a perspective that renders beliefs about the fctionality of the narrative tacit. Consider the slime paradigm, which Walton acknowledges is unusual. In Walton’s groundbreaking “Fearing Fictions” (Walton, 1978), the slime monster in the film focuses on and then oozes toward the camera, making it fictional in the games of all viewers that they are the slime’s intended victims. Miguel Dos Santos’s recent foray into quasi-emotionality displays a parallel focus on a cinema-goer’s fear of the Grady sisters in The Shining (Dos Santos, 2017). But when we pause to consider the conventions that govern what is typically considered true in a game of make-believe, this kind of scenario is a real outlier because it involves fictional characters being regarded as personal threats by viewers or readers. Sometimes this is because the fiction contains characters which take up attitudes toward viewers. Sometimes this occurs just because viewers take up attitudes toward characters that aren’t prescribed by the work and that are unique to their own imaginative game. Consider the first of these. Any clever bit of cinematography having the camera pose as the eyes of viewers can make them the target of attention or danger in the fictional world, as was the case with Charles and the slime. If accompanied by a sudden shift of screen imagery, this approach almost invariably elicits a startle response, as when the shark leaps suddenly from the water toward the camera in Jaws. On stage, “Clap if you believe in fairies” makes it fictional that viewer actions can save Tinkerbell. But it is seldom the case that characters like the fearsome Demogorgon are as aware of us as we are of them in the world of the game. Roughly comparable to cases like those of the slime and Tinkerbell are games of make-believe in which we try to address or cajole characters (“No! Don’t go down that disgusting hole! It’s a gateway to the Upside Down!”) or in which we imaginatively assume that characters are taking or could take up attitudes toward us (even without any cinematographic or textual evidence of this). All of these seem to smack more of active pretense than imaginative immersion and to involve the kind of contamination of immersion with personal concerns to which Walton refers. First, then, any cinematic endeavor that deploys effects intended to elicit a startle response in concert with a point-of-view shot cannot count as eliciting an emotion either on thought theory or on Walton’s belief/ motivation-based approach. Famously, a startle response has no cognitive concomitant, whether thought or belief. So, a startle reaction, being a reaction to sudden changes in visual and sometimes auditory stimuli,

86

E. M. Dadlez

would be a reaction neither to what was imagined nor to what was believed. The startle occurs prior to the formation of a cognition. Nothing depends on the fictionality of the stimulus. We can happily relegate such reactions to quasi-territory without abandoning thought-theoretic approaches. However, Aristotelians who deplore spectacle might point out that attempts to elicit the startle response are often used as a shortcut to begin the process of building up a more cognitively robust fear for oneself or for characters without having to resort to plot development. Once the physiological component of fear is elicited, alertness to prospective dangers is guaranteed, and fear, whether quasi or actual, is likely to follow (though we should not forget that fear for oneself in such cases is not uncommonly fear of being startled yet again, which is an uncontroversial fear of having a genuine experience in the real world). I should note briefly that philosopher Lorraine Yeung, in challenging the kinds of approaches to emotion espoused both by cognitivists and thought theory proponents, points out that they neglect many kinds of affective responses, not giving formal devices and stylistic elements the proper credit for eliciting them, because of a near-exclusive focus on narrative content (Yeung, 2018). Sadly (since Yeung makes some excellent points), I must return to features of content, since these are crucial to the claim I’m presently trying to substantiate. Charles’s reaction to the slime may or may not involve a startle response. The slime’s greasy head is said to emerge from the undulating mass that is the rest of it and its beady eyes to fix themselves on the camera. Then, picking up speed, it is said to ooze toward the camera/ viewer point of vantage (Walton, 1978, p. 5). “Ooze” does not exactly convey the suddenness required to startle, though it is possible that the eye-fixing or acceleration are sudden. Even if Charles’s response cannot be considered a startle reaction, I will nevertheless maintain that it is one of quasi-fear, but on grounds that I will argue do not apply to Charles’s fear for the slime’s other (fictional) victims or to my fear for Will Byers. That is, I want to argue that when a work of fiction prescribes that characters take up attitudes toward readers or viewers (or prescribes that fictional events engulf viewers, as in the famous case of a train speeding toward the camera), this renders participant responses distinct from the “objective” or “ghostly” form of participation that Walton describes as typical. Consider first that there is a difference between having a fiction prescribe the role of slime victim or train victim cinematographically to the viewer by means of carefully calculated camera angles and having a fiction prescribe awareness of someone’s (i.e., a character’s) annihilation under the extrusions of the slime or the wheels of the train. Granted, POV shots are sometimes simply intended to make some character’s perspective on their situation more accessible and serve as an aid to empathetic imagination. But there are other instances in which this is not the case, and these are the ones I wish to consider first, because participation in

Fear and Loathing in Fictional Worlds

87

such games escalates from fictional awareness of events to the fictional undertaking of roles and the fictional performance of physical actions. As Walton puts it, our participation in the game of make-believe is fuller in such cases. There is a script for us as well as the characters (sometimes we will also enlarge on states of affairs depicted in the work). We are, in some respects, more like actors (or authors), and our attention is often focused on performing a prescribed role as well as the array of fictional incidents before us. The conscription of readers of novels into a story is not impossible either, although it is usually far more restrained than the following excerpt from a children’s book entitled The Name of This Book Is Secret by Pseudonymous Bosch: Good. Now I know I can trust you. You’re curious. You’re brave. And you’re not afraid to lead a life of crime. But let’s get something straight: if, despite my warning, you insist on reading this book, you can’t hold me responsible for the consequences. (Bosch, 2007, p. 1) Some works script a part for their audience. I submit that the conventions of games of make-believe that use such works as props must differ from conventions ascribable to other works. In the world of the latter, more familiar game, participants are aware of characters, inhabit the world of the game with characters, but characters are not aware of them and do not respond to them. Participants are emotionally and epistemically affected by characters, but characters are not affected by them. As indicated earlier, what is typically make-believedly true of me usually involves my attitudes and awarenesses and feelings. In the world of that game, participants react to fctional events and characters, but not the other way around. A world is shared in imagination, but the participant’s footprint in that world is faint. The former case, on the other hand, a case in which the fiction reacts to us, is quite different. Participants’ reactions here are sometimes due to their (atypical) incorporation in the fictional action. And where makingbelieve that fictional events and people are real things with which we share a world seems centrally to involve not attending to our belief that they are fictional, having fictional characters address us or ooze threateningly toward us prompts a whole different kind of two-way participation. The latter appears to require our splitting our attention between personal concerns necessitating a perspective outside the world of the game and the kind of imaginative immersion that lets awareness of fictional events roll over us as if it were an awareness of events in the world (underwriting a sustained perspective from within the world of the game). The split

88

E. M. Dadlez

focus resembles Walton’s game of mud pies more closely and is strongly reminiscent of role-playing games in which participants move rapidly between in-character and out-of-character perspectives, in and out of fictional and extra-fictional points of view. Consider Charles, gazing at the screen slime and warning the friend sitting beside him in the theater. If it is true in the world of his game that he is threatened by the slime, that it is coming his way with fell intent, then he should run or hide in addition to issuing warnings. Yet, the conventions of the movie-theater game preclude any such activity. What does the game require of him? The fate of Charles’s corporeal self has become part of the fictional world, and while it is only fictional that he is at risk, aspects of the actual, unfictionalized world begin to intrude on the story. Charles’s cry of “Yikes! Here it comes! Watch out!” (Walton, 1978, p. 19) raises the question of where “here” might be in the world of the game. The typical indeterminacy concerning our location in the worlds of the games that we share with characters is undercut in these kinds of cases. To emphasize again, it is Walton himself who distinguishes between a perspective “relatively . . . uncontaminated by personal concerns” (in the more typical response to fiction) where the appreciator’s presence in the world of their game is sketchy or ghostly and cases where asides like “Watch out!” disturb objectivity and reduce indeterminacy, enlarging one’s role in the game (Walton, 1990, p. 237). What of fearing fictional characters without there being fictional prompts stipulating that readers/viewers are threatened? If, in addition to fearing that Will Byers will fall into the Demogorgon’s clutches (fearing for Will) I also fear the Demogorgon on my own account, I am not just responding to the fiction but amending it by imagining new fictional facts. I am altering the prop to make it the case that the Demogorgon threatens me as well as Will in the world of my game. My role as Demogorgon victim is scripted by me, not prescribed by the work. As in the case of Charles, it involves a split between attention to personal concerns (plus the role I have crafted) on the one hand, and attention to ongoing fictional events on the other. Again, the continuity of imaginative immersion is disrupted. My tacit knowledge that what I contemplate is fictional is rendered explicit by aspects of those personal concerns. Just as what is true in the fictional world is true in the game of makebelieve, so some things that are true in the world of the game are true in the actual world. It is true in both the world of the game and the real world that my heart beats faster as I entertain the thought of Will’s plight. My much more controversial claim is that my fear for Will is true both in the world of the game and the real world. Of course, in the latter case, perhaps it is true in the real world only under the analysis that makes “fear for Will” a fear comprising or arising from the constellation of evaluations about danger made occurrent by the imaginative instantiation of the situation type to which those beliefs apply. That is, I’m characterizing

Fear and Loathing in Fictional Worlds

89

my fear for will as a fear arising from an awareness of a way it could be to be so placed. My fear of the Demogorgon, however, is true only in the world of the game. Under consideration are cases in which we amend fictions to include us as characters, or in which the fiction actually includes features which make temporary characters of participants, cases that genuinely make it fictional of us that we are threatened by fictional characters or events, or that we’re addressed by them, or that we’re given an opportunity to save them by clapping (or the opportunity to hasten their flittery demise by refraining from doing so). Any feelings we have on account of such participation – fear of the slime or Demogorgon or speeding train, excitement about being invited to lead a life of crime, enthusiasm about being able to contribute to Tink’s salvation (or her death) – are affects I am happy to characterize as quasi-emotions, insofar as the involvement of personal concerns tends to make our belief in the fictionality of events explicit rather than tacit. It is also worth noting that fear presents an especially difficult case in this context. Fear has been by far the most popular emotion to use in examples of response to fiction, but it is also a strangely complicated emotion. We can fear something, fear for another, or fear empathetically with another. So, in giving examples, it makes a difference whether we are responding on behalf of some fictional entity, as when I fear for Will Byers, or whether we respond on our own behalf, as when I fear the Demogorgon or Charles fears the slime. To fear something on our own account is acknowledge threatening possibilities for ourselves and our lives. (Hence, the “personal concerns” Walton acknowledges would appear to disrupt the fictional narrative and our concentration on it with extra-world considerations.) To fear something is acknowledged by some philosophers of emotion to be shorthand for fear of some future possible event involving that thing. We fear prospects. Like hope, fear is forwardlooking. So, when I fear the charging hippo, what I really fear is the possibility of being trampled by it. When I fear the slime, I fear being engulfed by it. When I fear the oncoming train, I fear being run over. But then fear of fictional characters seems to be a fear of prospects for our own lives – becoming Demogorgon chow or slime fodder. It seems to involve either fictionalizing those lives in ways that run-of-the-mill games of make-believe with fiction don’t accommodate (though games of mud pies clearly do) or exporting fictional threats into the world we actually inhabit. Either way, imaginative immersion would seem to be rendered less continuous and more sporadic. Roughly, I think that an excellent case can be made for quasi-fearing fictions in those instances where imaginative continuity is disrupted by an importation of the participant into fictional action formerly exclusive to the work, such that s/he has what Walton calls a “fuller presence” in the world of the game. The more fictional facts about the participant

90 E. M. Dadlez overshadow the narrative, the more likely there is to be a disruption in his or her focus on events in the work’s world. This seems in many cases to involve something resembling active pretense rather than imaginative immersion – warnings are shouted to one’s theater-mates and to characters, pillows are used as shields, feet are tucked up on the furniture out of the way of the scuttling, fanged creatures depicted in a film. This kind of make-believe clearly seems to involve quasi-fear – phenomenologically similar to ordinary fear for oneself but without any actual threat, more or less like a roller-coaster ride. There is, in other words, more awareness that it’s a game one is playing, that things are fictionally rather than actually the case, just because personal considerations become a focus of attention. However, fear for fictional others doesn’t involve anything like this apart from similar affective reactions. We don’t get reminded that things are only fictionally true by imaginatively immersing ourselves in the work (i.e., playing the kind of game in which our presence is thought “ghostly”). David Novitz describes a typical immersion in fiction – the kind in the course of which one loses all track of one’s physical surroundings and focuses only on the story, as one in which any explicit beliefs about the fictionality of the events we contemplate are rendered tacit (Novitz, 1987, p. 83). An epistemologist might say that our belief that we contemplate a fiction is not occurrent but merely, at least while immersion lasts, dispositional. We believe it, but we don’t attend to it. It is my contention that reactions to fictions that are (imaginatively) understood to take up attitudes toward us are much more likely make beliefs about the fictionality of what one contemplates occurrent. We are used to such reactions in other contexts, of course. Some tendencies to focus on form rather than content, as when we are struck by authorial or thespian virtuosity, are necessarily tied to our imaginative departure from the world of a game in which fictional incidents are real. But immersion can be curtailed by content-related considerations as well. I want to say that the more a game of make-believe resembles an RPG or Walton’s game of mud pies, the likelier it is that our belief about the fictionality of the events we contemplate will come to the fore, in part because our attention is more likely to be split between a world in which we’re endangered and a world in which we’re not. In fact, I am inclined to argue that games of make-believe that do not involve fiction, whether RPGs, games of mud pies, or reenactments, involve a greater investment in player agency (player impact on action and events in the game) than games involving fictions do. Because such investment necessitates concerns about skill and scoring and winning, and because these concerns are external to the world of the game, imaginative immersion is less continuous. I do not think that cases of first-person narration, descriptions of character thoughts in novels, or POV shots intended to elicit empathetic

Fear and Loathing in Fictional Worlds

91

responses involve the same kind of immersion-disrupting personal foregrounding. These make us aware of what it is like to be placed as the character is, or how it feels to be threatened as the character is, but will not create a make-believedly causal two-way interaction between ourselves and characters any more than empathy with real people does. We have no (fictional) agency in such cases, but are more or less going along for the ride by occupying some character’s point of vantage and observing what their experiences are like from the inside. In an effort to argue that the majority of our affective responses to fiction ought not to be characterized as quasi-fear or quasi-pity, I have tried to establish three things. First, it appears that responses Walton is willing to regard as genuinely and literally emotional do not fit the belief/ desire account of emotion that he proposes in Mimesis as Make-Believe, an account intended to buttress the claim that we cannot literally fear for or pity fictional characters. Belief in the object of one’s affective response need not be considered necessary for robust fear or pity, given Walton’s acknowledgment of phobic reactions. Nor can motivational or behaviorinciting effects be considered necessary, since Walton acknowledges that we can feel genuine pity and distress for historical persons and on account of historical incidents (see also Dadlez, 1997, 2002). Empathetic emotional responses to real people fail to meet either one of these conditions (just as fictions are said not to meet them), at least on standard accounts of empathy (if not Walton’s own), so my angry empathy with some real person who has been unjustly treated might be thought to stand or fall with my fear for Will Byers in the dread domain of the Upside Down. Either these are forms of genuine anger and fear, or they are quasi-anger and quasi-fear because emotions (as Walton says) cannot literally take such objects. Second, I have tried to show that our more typical responses to fiction involve broad evaluative beliefs that are existentially uncommitted, but no less epistemically respectable for that. Constellations of such beliefs appear to infuse every affective reaction to the fictional and can account both for participant emotions and related obligations and motivations when some or all are made occurrent. Such beliefs and the obligations to which they give rise can (this is a necessary, not a sufficient condition) arouse an awareness of how it may be to undergo a given experience. This awareness is something that I’ve identified with emotional response to fictional characters. Whether this can be treated as an analysis of what it may be to be genuinely moved by fiction, or whether it should be regarded as frequent concomitant of affective reactions to fictional entities, it still appears to suggest that more reactions to fiction are fully emotional than has been hitherto acknowledged. Finally, I have conceded that there may be a continuum along which affective reactions experienced in the course of games of make-believe may fall, some being more easily relegated to quasi- or make-believe territory than others. However, I think that a designation of quasi-fear or quasi-anger

92

E. M. Dadlez

would more likely involve the degree to which the world of one’s game licenses or prescribes interactions that make dispositional beliefs about the fictionality of events being contemplated occurrent, something that is often related to games which allow for a two-way interaction between fictional entities and the participant or places a significant emphasis on player agency and intervention in the narrative. This two-way interaction can elicit enough self-consciousness and self-reflection to render beliefs about fictionality that were hitherto tacit, quite explicit. Where it does, designations like quasi-fear may be appropriate.

Notes 1. I would like to thank my colleagues Mark Silcox and Derek Matravers for their comments and criticisms. 2. It isn’t clear that Walton still holds this view in later work (e.g., Walton, 1997). Perhaps, Walton would now maintain that imagining one witnesses a historical event means that the affective response itself is in some respects imaginary, but that believing Ivan the Terrible was a villain would result in actual anger. Thanks to Derek Matravers for holding my feet to the fire on this.

References Boonin, D. (2008). How to Solve the Non-Identity Problem. Public Affairs Quarterly, 22(2), 31. Boruah, B. H. (1988). Fiction and Emotion: A Study in Aesthetics and the Philosophy of Mind. Oxford University Press. Bosch, P. (2007). The Name of This Book Is Secret. Little, Brown & Company. Carroll, N. (1990). The Philosophy of Horror, or Paradoxes of the Heart. Routledge, Chapman & Hall. Carroll, N. (1995). Mimesis as Make-Believe [Review of the Book Mimesis as Make-Believe: On the Foundations of the Representational Arts, by K. Walton]. The Philosophical Quarterly, 45(178), 93−99. Dadlez, E. M. (1997). What’s Hecuba to Him? Fictional Events and Actual Emotions. Penn State Press. Dadlez, E. M. (2002). Quasi-Fearing Fictions. Film and Philosophy, 5−6, 1−13. Dadlez, E. M., & Haramia, C. M. (2015). Fictional Objects, Future Objectives: Why Existence Matters Less Than You Think. Philosophy and Literature, 39(1A), 1−15. Dos Santos, M. F. (2017). Walton’s Quasi-Emotions Do Not Go Away. Journal of Aesthetics and Art Criticism, 75(3), 265−274. Greenspan, P. S. (1988). Emotions and Reasons: An Inquiry Into Emotional Justification. Routledge. Lamarque, P. (1981). How Can We Fear and Pity Fictions? British Journal of Aesthetics, 21(4), 291−304. Lamarque, P., & Stein, H. O. (1994). Truth, Fiction and Literature: A Philosophical Perspective. Clarendon Press. Moran, R. (1994). The Expression of Feeling in Imagination. Philosophical Review, 103(1), 75−106.

Fear and Loathing in Fictional Worlds

93

Neill, A. (1995). Fiction and the Emotions. In A. Neill & A. Ridley (Eds.), Arguing About Art: Contemporary Philosophical Debates (pp. 272–290). McGraw-Hill. Novitz, D. (1987). Knowledge, Fiction and Imagination. Temple. Parfit, D. (1984). Reasons and Persons. Oxford University Press. Parfit, D. (2011). On What Matters. Oxford University Press. Roberts, R. C. (1988). What an Emotion Is: A Sketch. Philosophical Review, 97(2), 183−209. Sousa, R. D. (1990). The Rationality of Emotion. MIT Press. Walton, K. (1978). Fearing Fictions. Journal of Philosophy, 75(1), 5−27. Walton, K. (1990). Mimesis as Make-Believe: On the Foundations of the Representational Arts. Harvard University Press. Walton, K. (1997). Spelunking, Simulation and Slime: On Being Moved by Fiction. In M. Hjort & S. Laver (Eds.), Emotion and the Arts (pp. 37–49). Oxford University Press. Walton, K. (2013). Comment on Catherine Wilson, “Grief and the Poet”. British Journal of Aesthetics, 53(1), 113−115. Walton, K. (2015). In Other Shoes: Music, Metaphor, Empathy, Existence. Oxford University Press. Williams, C. (2019). Why Quasi-Emotions Should Go Away: A Comment on Dos Santos. Journal of Aesthetics and Art Criticism, 77(1), 79−81. Wilson, C. (2013). Grief and the Poet. British Journal of Aesthetics, 53(1), 77−91. Yeung, L. (2018). The Nature of Horror Reconsidered. International Philosophical Quarterly, 58(2), 125−138.

6

Lyric Self-Expression Hannah H. Kim and John Gibson1

Poems aren’t trees. But they do contain phrases, sentences, paragraphs, verses which readers can, if they wish, use themselves. The words are there ripe for picking, no matter what the poet was doing in writing them down, and no matter what the reader takes her to have been doing. (Walton, 2015, p. 62)

I. It is part of the very idea of modern lyric poetry that it foregrounds a poetic speaker – the “lyric I” – whose formally articulated thoughts, feelings, and experiences are presented as the primary objects of artistic interest and aesthetic regard. Lyric poetry is not merely an expressive artform but, evidently, an intensely and intentionally self-expressive one, concerned with conjuring up a sense of a subject which is attempting to make itself known through poetic means. Yet, ever since the zenith of anti-intentionalist and anti-authorial accounts of artistic meaning, the philosopher of poetry has worked against a backdrop of theory that makes it appear at best naïve and at worst a logical error to endorse even a modest literalism regarding lyric self-expression, where “literal” indicates that the self is actual and is thus fit to express the psychological states of a real subject. Hence, the default view in poetics is the same as it is nearly anywhere philosophers write about artistic expression: the lyric I speak on behalf of a fictional persona and the “poetic discourse” it generates is therefore a variety of fictional discourse. Kendall Walton has recently shown something remarkable. We can be entirely literal about lyric self-expression, and we can do so by locating the relevant self not in the body of the poet but of the reader. Put crudely at first mention, by identifying in a particular manner with the content and expressive form of a lyric – with what the poem says and how it says it – the reader, if you will, puts herself on loan to a poem and thereby makes its language expressive, in fact literally and genuinely self-expressive. Walton’s claim is surprising. He has given us the most DOI: 10.4324/9780367808662-7

Lyric Self-Expression

95

influential theory of fiction in contemporary philosophy, and it is striking that he is unwilling to apply it to a phenomenon that it would seem so naturally suited to explain. That is: why doesn’t Walton treat a poem as a prop that prompts games of make-believe? Walton has applied his theory of fiction to many fundamental questions in aesthetics and the theory of representation, and understanding his reluctance in the case of poetry reveals a complexity to his work that is often overlooked. Our interest is in exploring and expanding the Waltonian model of lyric self-expression. The critical dimension of our chapter concerns the limitations of Walton’s model given the nature of voice in poetry and common features of the semantic and expressive behavior of the lyric I. We argue that after a few revisions of the small to medium sort, Walton’s model can accommodate the worries we raise. It can do so by enlisting a more expansive notion of the subject of lyric poetry than Walton – along with the vast majority of contemporary theories of the lyric – will countenance. What we will say about the subject of lyric self-expression will be every bit as surprising as Walton’s own views are: the “I” of lyric poetry often has an implicitly plural grammatical function. Poets often intend the lyric I to “contain multitudes,”2 as Walt Whitman had it, and thus to stand for and speak on behalf of the experiences of more than one mere subject, fictional or otherwise. Seeing how this works, we argue, reveals something important about the nature of readers’ affective and cognitive identification with the lyric I, the kind of imagination central to lyric poetry, and what, if anything, hangs on ontological questions regarding the source of poetic expression.

II. Walton’s model aligns with the work of a small but influential group of philosophers and critics who break with the widespread habit of producing theories that do to poems what we are much happier to do with novels: cast them as intuitively, basically, and perhaps necessarily in the business of projecting fictions.3 Lyric poems commonly elaborate a firstpersonal exploration of the nature of experience, selfhood, memory, and similar aspects of the ways in which the I encounters both itself and the world in which it is situated. Yet, very often these poetic investigations are not presented in narrative form4 and, absent the telling of a story, we have nothing amounting to a “story world”5 and, to this extent, nothing that obliges us to treat a poem as generating fictional characters and an imagined world in which they reside. And even if a lyric poem does provide a narrative, the anti-fictionalist about lyric poetry will argue that that dimension of the poem is unlikely to be the site of proper poetic activity or the focus of our attention as readers of poetry. Aesthetic immersion in lyric contexts makes central the rhythmic, imagistic, figurative, formal, and expressive quality of poetic language itself.6 This, in short, is what it

96 Hannah H. Kim and John Gibson means to appreciate a lyric poem as a lyric poem. And while the imagination might be harnessed in both the creation and appreciation of these dimensions of a poem, the point anti-fictionalists will insist on is that it is not of the sort that requires us to see a poem as projecting fictions, even in the Waltonian sense of regarding propositions in poetic contexts as functioning as props in a game of make-believe.7 Yet, the more serious reasons for refusing to model poetic discourse on fictional discourse are essentially cognitive and ethical. To abandon this model would seem to be to permit the poet to speak on the worldly side of the divide between the fictional and the real, and thus to allow the philosopher of poetry to assert that there is a categorical difference between fictional utterances and poetic ones. As the eminent literary theorist Jonathan Culler has it, the lyric poem is not to be modeled on a “world-projecting fiction,” since the lyric poet is typically in the business of making “real statements about this world” (Culler, 2015, pp.  107, 119).8 Culler calls the language of lyric poetry epideictic discourse: “public poetic discourse about values in this world rather than a fictional world.” Anti-fictionalism in respect to lyric poetry is best seen as an attempt to get the lyric I and the reader to inhabit the same ontological space: to show that poets typically speak from a position within the world so that they can address our experiences and values directly. Poetic discourse may be oblique in any number of respects, but, according to theories like Culler’s, not in the sense that it is constitutionally bound to speak indirectly and across worlds – from the imagined world a poem projects over to ours – when it wishes to make features of actual human predicament the focus of poetic activity. To see what the philosophical problem amounts to, consider one way of thinking about a term that is clearly relevant to this debate (though sadly largely ignored in the philosophical literature): voice. As Kukla and Lance have it, voiced expressions are inherently “personal” (Kukla & Lance, 2009, p. 59). Idiomatically, when one “gives voice” to a set of concerns, feelings, or beliefs, one calls on the hearer to entertain a conception of a subject who is the bearer of these psychological states (perhaps in a particular historical and social context). In this respect, utterances in which voice plays a vital role have essential reference to a self who is represented as hanging together in a particular way. When we experience a use of language as voiced, we take it to be an attempt to speak on behalf of a self and its thoughts, passions, and experiences. We can give voice to the states of others – I might speak on behalf of the plight of my neighbor or the hypocrisy of my generation – in which case our utterances are expressive but not, or not exclusively, self-expressive. But in properly self-expressive utterances, the concerns, interests, and beliefs of the utterer are made the objects of attention. This is why Kukla and Lance claim that our grasp of a voiced expression typically culminates in an act of recognition or acknowledgment

Lyric Self-Expression

97

(Kukla & Lance, 2009, pp.  134–152). The immediate point is that our responses to voiced expressions indicate that we grasp that the speaker has revealed something about herself, which is to say, has engaged in an act of self-expression, properly so-called. To motivate our presentation of Walton’s view, note that we undoubtedly experience the presence of voice, understood in this light, in lyric poetry; this is inseparable from our experience of a poem as selfexpressive. Thus, one way of putting the philosophical issue is to ask: whose voice is it? This is not merely to wonder about the identity, or the ontological status, of the speaker. Our experience of lyric utterances as voiced, poetic content as expressed, obliges us to give a philosophical account of what so much as entitles us to apply the concept of voice and the category of expressive language use to a poem. The particular minefield the anti-fictionalist must traverse is that nearly all of the standard responses to questions of this sort reintroduce fictionalizing tendencies to our understanding of poetic discourse, just now through expressive instead of narratological avenues. For if one wishes to embrace positions such as Culler’s – and on the whole we do – a story is owed as to how the possessor of this voice, if it is not the actual poet, can speak to us so directly and epideictically about our worldly values, experiences, and circumstances. If we say that this voice belongs to an imagined speaker, we appear to invoke a non-actual speaker and so ultimately a fictional speaker (we return to this idea later). And if we are tempted by the Heideggerian move, as Culler and his allies are, of arguing that the poetic text itself – or the language of a culture, or Dasein, etc. – “speaks,”9 we have to explain how we can regard an inanimate object or abstraction as capable of such loquaciousness if it is not because we are imagining it to be so. This hardly seems to make good on the promise of literalism regarding the subject of poetic self-expression. So how can we explain how a poem gives voice to the concerns, anxieties, and predicaments of an actual or otherwise real subject? At first blush, these anti-fictionalist arguments sound remarkably unWaltonian, certainly to anyone who has read Mimesis as Make-Believe. Yet, Walton embraces much of this. In fact, Walton offers a theory of poetic expression that echoes the central features of Culler’s epideictic model, though with one crucial difference. Whereas Culler wants to establish that the lyric I makes “statements about the real world,” Walton claims that it does not make genuine statements at all. Walton argues that poets can often be seen as “thoughtwriters,” and that thoughtwriters, like speechwriters, compose a body of text not for the sake of expressing themselves but “for others to use in expressing their thoughts (feelings, attitudes)” (Walton, 2015, p. 54). Just as a speechwriter may not “believe in” anything she writes, a thoughtwriter may not “seriously mean” or assert anything she writes because she doesn’t use the words she inscribes; she only mentions them. Walton is thus happy to say that a poet’s only

98

Hannah H. Kim and John Gibson

contribution “might consist just in making words available for readers to use” (Walton, 2015, p. 54). This line of thought is condensed nicely in the following passage: Rather than understanding a poem on the model of an ordinary assertive or expressive utterance, addressed to or overheard by a listener, we might understand it on the model of a speech written by a speech-writer, for use by another person. Speechwriters don’t use the words they inscribe (not insofar as they are simply speechwriters); they mention them. They produce a text for use by someone else. The speechwriter doesn’t assert the declarative sentences he comes up with; his client asserts them when she delivers the speech. (Walton, 2015, p. 61) And it culminates in the following claim about the lyric poem: Is it nonfiction? Yes, if that just means that it is not fiction. But it isn’t a typical work of nonfiction, an ordinary instance of nonfictional literature. The poet, the author of the work, didn’t use the words in the usual manner, but only mentioned them. She didn’t assert its declarative sentences. The reader alone uses the words (the word types)—if he chooses to do so. The poem doesn’t serve as an actual vehicle of communication, not the usual kind at least—not even a pretended or attempted one. (Walton, 2015, p. 63) One way of understanding Walton’s gambit is in the following terms. The broader debate in poetics concerns how we should provide a principle of animation for the lyric I: how can we conceptualize the poetic speaker as endowed with suffcient subjective life such that we can justify our ascription of psychological and expressive predicates to it? Walton is not interested in solving this puzzle so much as showing that the thoughtwriting model allows us to demystify it, certainly in many cases. When readers treat poems in the manner he describes, we have an elegant and earthbound way of explaining how poetic content gets expressed and so poetic language itself becomes properly expressive, since the reader provides us with the fgure of a speaker who is using this language to reveal or otherwise manifest features of a self. Through an act of frst-personal lyric identifcation, the content of a lyric poem now becomes about a subject. It is given a subject of whom its declarations and descriptions can be seen as true. Walton finds support for his argument in the fact that language acquisition is itself a social endeavor and works largely on the inheritance of expressive phrases that went well before us. For Walton, poets supply readers with nuanced and rich tools for engaging in successful acts of

Lyric Self-Expression

99

self-expression; they are great instructors in how to give voice to our experiences and inner lives, much as an artist might aid an art student by providing the right color shade of paint. According to the thoughtwriting model, a “successful” or “good” poem would, in Keats’s words, “strike the reader as a wording of his own highest thoughts, and appear almost a remembrance.”10 The point is, when we experience a poem as properly self-expressive, it is because we, the readers, give voice to the thoughts and sentiments elaborated within it. Indeed, they become asserted thoughts and expressed feelings through us: it is by virtue of an act of readerly identification that they are made expressive of the mental states and experiences of an actual subject. True, the actual poet may think and feel them, too. But when this happens, it is because the poet, like the reader, puts a lyric to a particular kind of expressive use. On a Waltonian model, then, to experience a poem as voiced and so literally self-expressive is to experience it as expressing features of my personhood, though presumably with much more elegance and aesthetic success than I can if left to my own devices. In this respect, we, the readers of a lyric, furnish the self to whose thought, feelings, and experiences the poem “gives voice.” Walton is careful not to overgeneralize his claim, and his arguments are full of responsible qualifications to the effect that “sometimes,” “often,” and “it is possible that” poems can function this way, all of which makes it challenging for the critic to find the right level at which to pitch complaints. Nonetheless, his argument is the only one of which we are aware that purports to show us how to be both literalist about lyric self-expression and anti-fictionalist about poetic discourse, and, for the moment at least, we will treat the Waltonian model as issuing a bold claim about lyric self-expression in general, even if Walton himself would take umbrage. Walton is surely right to acknowledge that the vast history of lyric poetry will readily provide counterexamples to most sweeping claims we make about its nature. What is at stake, however, is showing that the default model of lyric self-expression shouldn’t be thoroughly fictionalist, and this is a point on which Walton, Culler, and we agree. The disagreement concerns just what a literalist model should look like and whose self it should make central to lyric self-expression.

III. If read in a perversely verbatim spirit, Walton’s model raises obvious concerns. Our arguments in this section will perhaps just provide reasons for not taking his model in this spirit, and in the next section we will offer suggestions for how the model might be revised to avoid the worries raised here. The problem is the particular manner in which Walton’s model has readers identify with the content of a poem. It is no doubt true that readers can use poems for the declaration of personal thought and feeling. But we fear that, on its face, Walton’s model demands entirely too

100

Hannah H. Kim and John Gibson

much of the reader’s person and in a manner that will obscure basic features of poetic experience. Even if our arguments are an exercise in philosophical pickiness, they will point up cases to which Walton’s model does not comfortably apply and thus motivate the philosopher of poetry to try to account for them. The overarching lesson will be that the thoughtwriting model applies well to some poems, and often to parts of poems, but not to entire poems themselves. The class of poems that especially resist a wholesale thoughtwriting treatment will include those with particular social or political settings, those which require a strong self-other distinction, or those with expressions that should not be disassociated from the subjects whose experience they draw from. Take a notion that is central to poetics: poetic address, and think of it as the mode of imaginative speech that grants the lyric I its bewildering liberty in respect to the range of things momentous and ordinary, abstract and concrete, to which it can address itself (eternity, a red wheelbarrow, one’s unborn child, and so on).11 As J.S. Mill famously wrote, we experience poetry as “overheard. . . . The peculiarity of poetry appears to us to lie in the poet’s utter unconsciousness of a listener.” That is, readers typically experience themselves as indirect auditors, listening in on an intimate exchange between the poetic speaker and the state of affairs its speech targets. It is immensely difficult to see how we can respect the integrity of our experience of poetic address if we do not treat the lyric I as an independent expressive subject. As Monique Roelofs has it, “we direct modes of address at people, nonhuman animals, and things,” and thus if we experience address in a poem, it would seem that the lyric I is quite capable of directing and animating its own speech, quite autonomous of the reader and the expressive use to which she puts its language (Roelofs, 2020, p.  2). Put simply, if the words of a lyric poem are, as Walton argues, “mentioned but not used” prior to readerly identification, how could poetic address be experienced as a form of address at all? Moreover, a great many of the emotions readers experience in respect to lyric poetry are other-directed: pity, sympathy, curiosity, admiration, and scorn are all at times perfectly legitimate responses to the “discourse” the poetic speaker generates. These emotions imply clear self-other differentiation and call on the listener to feel something about the state of another expressive subject, which would strongly suggest that in poetic contexts there is at least one subject acknowledged to be expressive who is not identical to the reader. Or consider empathy, which is arguably the most common and basic form of emotional bond between readers’ experience with the lyric I. True, empathy is a first-person emotion, but it is another’s first-person perspective that one tries, always with varying degrees of success, to assume. In Rae Langton’s words, empathy is marked by a “shift in self-location,” and this requires a conception of the other as already subjectively animated, lest there be no other self whose thoughts, feelings, and subject position I imagine myself to be experiencing.

Lyric Self-Expression

101

Sometimes, much of the cultural, critical, and political work of a poem will be undone when it is the reader who gives voice to the thoughts, feelings, and attitudes the poem elaborates. More specifically, the meaning of some claims that look like assertions in a lyric poem are obscured when the reader is seen as providing the grounds for their expressivity. As an example, take Langston Hughes’s “Dinner Guest: Me” (1965), which begins with an “I”-statement: I know I am The Negro Problem Being wined and dined, Answering the usual questions When the “I” here is being identifed with “The Negro Problem,” surely not just any reader can supply the self-substantiating “I.” In many instances, this would entail making the poem issue a false claim. In one sense, the claim is trivially false since an individual cannot literally be “The Negro Problem,” an abstract phenomenon that manifests differently across time, place, and cultural context. But the opening line is not a simple metaphor; individuals in certain social standings – usually the non-dominant ones – often experience the phenomenon of answering for or “standing in” for their social group. The claim “I know I am/ The Negro Problem/Being wined and dined,/Answering the usual questions,” then, invokes a lived experience, and, without the requisite life experience supplying the truth conditions of the “I know I am/The ____ Problem” (where ____ might be flled in with “Negro,” “Genderqueer,” “Immigrant,” take your pick), the expressive “I” is empty, or worse still, presumptuously and oppressively usurped. Is it possible to think that the reader is being asked to try on the African American identity so as to be able to embody the “I” here? This is the selfprojecting fictionalizing move that we might have expected Walton to make with poems. But, as we noted, he does not take this stance with poems, and we agree that one should not. Regarding a lyric poem such as Hughes’s as a work of fiction that invites games of make-believe risks trivializing the poem and the forms of experience it strives to give voice to. It does so for the obvious reason that it makes its utterances about, and articulated from a position within, a fictional world, to that extent diminishing the poem’s ability to speak directly about our world and the forms of alienation and oppression in effect there (Culler, 2017, p. 33). In this respect, we should be strongly inclined to ensure that our theories of lyric poetry explain how we can see works such as “Dinner Guest: Me” as making genuine claims and as creating a reading experience that readers must take up and consider in relation to the shared actual world in which we live. Though Walton does not endorse treating lyric poetry as fiction – “the poem isn’t a work of fiction, any more than a speech written by a

102

Hannah H. Kim and John Gibson

speechwriter is” – the thoughtwriting model itself still might be seen to be ill at ease with lyric poetry’s ability to speak in an epideictic register about the real world (Walton, 2015, p. 63). This is because it risks rendering the subjective seat of the lyric I a kind of mask that virtually anyone can try on and adapt. As noted earlier, Walton is careful not to overstate the thoughtwriting model’s capacity to account for lyric poetry, but he does think the experience of reading a poem often involves treating the poem as a piece of thoughtwriting (Walton, 2015, p. 57). Walton thinks that lyric poems, though nonfictional, often invite the usual activities prescribed by a work of fiction, especially when a poem’s claims are not easily adoptable: If the ideas or attitudes expressed in a poem are ones the reader doesn’t accept, she may be unwilling to think or utter them assertively, and unable to do so sincerely . . . Does the reader, in these cases, have no choice but to read the poem either as another person’s serious utterance, or as the unusable and perhaps inappropriate handiwork of a thoughtwriter? Enter the imagination, pretense, role-playing. (Walton, 2015, p. 64) If I disagree with the sentiments expressed in a poem, I may try them on, imagining uttering the words “seriously” to see what it feels like to express such thoughts or attitudes—and probably what it feels like to endorse or accept or adopt them. (Walton, 2015, p. 65) However, the kinds of poems where the thoughtwriting model runs up against its limits concerns not those where the reader fnds herself confronting ideas or attitudes with which she disagrees. They are the ones where the fexible transferring of the “I” is unauthorized, inappropriate, or impossible. It is not a matter of differing opinions, desires, and attitudes, nor even cases of imaginative resistance. It is rather that the reader sometimes, in fact often, lacks the requisite authority or experience to stand behind a poem’s voice. Certain points of views are simply not the kind of thing that can be, nor should be, taken up and inhabited; there are limitations to the kinds of perspectives we can responsibly “pick up and try out,” and “Dinner Guest: Me” provides one good example of such a perspective. The politicization of much art in the twentieth century – and the explicit attention to issues of subject-position and identity politics in contemporary lyric poetry in particular – ensures that the question we raise of Hughes will apply widely. Perhaps all we are saying is that sometimes subjective expressions created for others’ use – whether in a poem or on a bumper sticker – come with limits or suggestions for who may and may not use them as modes of self-expression. “Dinner Guest: Me” shows that sometimes role-playing, even first-person identification itself, is decidedly problematic and risks

Lyric Self-Expression

103

undercutting a poem’s point, which, for white American readers, is presumably to get them to acknowledge the experience of others, not to self-identify with them. In this particular instance, if the non-black reader locates herself in the I of the poem, this will lead her to miss what Hughes in fact wishes such a reader to identify with, a contrasting “white mind” also present in the poem: Murmuring gently/Over fraises du bois,/“I’m so ashamed of being white.” The second “I” here appears in a sentence in quotations; distance is created, and it is clear that the speaker merely mentions the indexical (as well as the sentence) without using it. There’s also the genteel demeanor and the irony of only admitting one’s shame in the context of delicacies and comfortable black company (a dinner party guest capable of answering the usual questions, and an unmistakable nod to W.E.B. Dubois), all pointing to the limited circumstances in which this other “I” is capable of expressing herself. Note that it would be a mistake to regard Hughes’s poem as a special case because it gives the reader two “I”s yet invites the reader, depending on features of her personhood, to take up and try only one. And this serves as a foil for other lyric poems which include only one instance of the “I” that do not straightforwardly invite the reader to take up and try on. In fact, “I, too” is another poem from Hughes that features a poetic speaker that resists being used by any old reader, beginning with the lines “I, too, sing America./I am the darker brother” and ending with the line “I, too, am America.” This is in part a poetic response to Walt Whitman’s “I Hear America Singing,” and its indirect point is that the “I” of Whitman’s lyric poetry, which often conducts itself in a universal voice, simply cannot speak on behalf of such a copious range of subjects. For these reasons, we maintain that much lyric poetry does not function as speeches do: most are not produced for others’ use, and treating the semantic grounds of the lyric “I” as shifting is often problematic. This finding is not surprising once we revisit the poet-speechwriter analogy. After all, most speeches are hardly written for everyone; if a political speechwriter produces a text for the State of the Union, a professor would be misusing the artifact were she to read it as a class lecture. Both speechwriters and poets have intended speakers in mind (even if the poet aims to speak to most humanity), and veering beyond the realm of appropriate usage would amount to misusing the artifact. Again, Walton did not intend his thoughtwriting model to explain all of our engagement with all poetry. But then the question becomes, what kinds of poems are best captured by his model? We might answer that some poems, especially poems that wish to speak on behalf of “general human experience,” would be appropriate for the thoughtwriting model. But here we need to consider

104

Hannah H. Kim and John Gibson

the distinction between first reading versus repeated recitation in poetry engagement. Walton insightfully points to repeatability as an important feature that the thoughtwriting model accounts for. Invoking an example of someone lost in the desert and reciting a Psalm to soothe himself, Walton writes: The words are there in the adventurer’s memory. He deems them appropriate in his situation, and uses them. Notice that the Psalm is in the first person. (“Yea though I walk through the valley of the shadow of death, I will fear no evil. . . . ”) But the adventurer needn’t take its first-person pronouns to refer to David or to a fictional narrator. “I,” in his recitation of the Psalm, refers to himself. (Walton, 2015, p. 67) Of course, the example works because the adventurer has the requisite circumstances to embody the voice. But also notice that this is not the frst time that the reciter is encountering the poem; this is what he does long after he has read the poem because the poem is familiar to him. So thoughtwriting, it seems, works as a model for explaining how poems are engaged with after the reader is well acquainted with them.12 Poems, especially of the lyric variety, encourage repetition. After reading a poem, we might add it to a repertoire of future expressions, but when we frst read a poem, we do not “try out” the words. We have to frst accept it as is, perhaps positing a speaker and/or a voice who is making use of the words involved. Without such a frst-pass engagement, we could not even understand what is being expressed. This suggests a further point. We can grant that the thoughtwriting model captures what readers do with expressively apt phrases from poems. But insofar as most readers don’t use the entire poem as a thoughtwriting piece – that is, we don’t recite the whole work to express our own thoughts and feelings but usually just a handful of particularly apt lines – thoughtwriting looks to yield just a partial account of our experience of a poem. Indeed, it appears to offer an account not of our experience of a poem but rather of just a few of its lines. As a final concern, consider another idea that is widespread in poetics: that lyric poetry often intends to produce, and to be experienced, as an “event” and that this matters to our understanding of the kind of immediacy it seeks with the reader. Culler defends the idea that poetry is not a representation of an event – a mimesis – because it “attempt[s] to be itself an event” (Culler, 2015, p. 35). Again, as Mill put it, we take ourselves to overhear a poem, as though the lyrical exchange occurs here and now, enacted at the moment of audition. Many of poetry’s formal and rhetorical devices serve this purpose, not the least the manipulation of the present tense to produce impressions of contemporaneity. This is taken to contrast strongly with how we encounter the content of literary

Lyric Self-Expression

105

narratives. A representation of a narrative event in a novel, for instance, is experienced as reported to us;13 the lyric poem typically presents an event as happening before us or at least within earshot. There is a cognitive element to this as well. It suggests that we should not regard a poem as a mere depository of tools to be used for thinking and communicating but a more dynamic event in which we bear witness to a mind actively consuming, working through, and producing thoughts and feelings. Our more memorable reading experiences are built on experiencing poems as an event, a happening and not a mere object or a cache of already-there thoughts. This modification helps us recover an important aspect of poetry that is underemphasized in the thoughtwriting model: somatic pleasure. Much of what produces the event-like quality in a poem is its rhythm and rhyme: the poetic present that is felt and experienced. Sometimes, or most of the time, reading a poem is not simply, or even primarily, an experience of “thought” but an exploration of sensory experience that foregrounds the material, imagistic, and prosodic features poetic language. Take a well-known example of Romantic poetry: William Blake’s “Tyger.” What makes the poem so mesmerizing is not its content but its trochaic unfolding. The rhythm of the trochaic heptameter and the repetition of the opening stanza create the sense that there is a specific happening that we participate in, a distinct event that we experience. There is a playful sense of temporality, and this itself is a source of our pleasure, even of affective identification. Regarding the poem as a piece of thoughtwriting risks collapsing the moving and patterned temporality and thereby turning the poem into a timeless receptacle containing nuggets of adoptable thoughts and expression.14 Even the “iterability” of a poem is made mysterious if we approach a lyric just for content, representation, or cognitive activity; if I came to a poem looking for a thought, I would lack reasons to reread it once I grasped and “tried on” the thought. But we find ourselves returning to our favorite poems over and over again, and this only makes sense once we conceive of the reading experience as an iterable event that we find value in. We have argued that Walton’s thoughtwriting model for poetry is an elegant explanation of a poem’s expressive power that does not rely on a fictional persona nor require a subject whose expression we are hearing. However, in some instances, it does violence to a poem by disassociating the expression from the subject whose expression the poem contains. Walton’s presentation of the thoughtwriting model leaves room to discuss which poems would be appropriate and inappropriate candidates for such a treatment. We’ve shown one example, “Dinner Guest: Me,” which shows the limits of thoughtwriting as a poetic mode of engagement. We’ve also shown that thoughtwriting may be more appropriate for capturing what the reader tends to do after she has become familiar with a poem, and most of the time, just with parts of the poem.

106

Hannah H. Kim and John Gibson

IV. We will conclude by outlining a strategy for addressing these concerns that makes use of many of the core elements of Walton’s thoughtwriting model. The arguments of the previous section functioned to keep alive the question of who expresses the “I” in a lyric poem. As we saw, much philosophy of poetry gives us two options for answering this, both singular and decidedly individual: the owner of this I is the author or a fictional persona. Our suggestion, put crudely at first mention, is that this ownership is often, in crucial respects, collective and so plural. The lyric I often “gives voice” to experiences that are manifestly not those of one mere subject. They are often shared forms of experience, and the implied subject of the lyric I is thus a creature not of fiction but of culture: it organizes and gives expression to thoughts, desires, anxieties, and feelings of a constellation of selves. If this idea sounds extravagant, note immediately that it is implicit in Walton’s theory. The thoughtwriting model offers one way of thinking about how the lyric I can “contain multitudes,” to quote again Whitman’s apt phrase, since its central idea is that readers find themselves expressed in a poem. For Walton, the range of subjects that can provide the self for poetic self-expression is thoroughly plural, indeed potentially limitless. What we find especially promising about his theory is how effectively this pluralizing move demystifies literalism about lyric self-expression. It provides an altogether ordinary sense of how this self can be seen as real and so capable of generating “epideictic discourse,” that is, of speaking on this side of the line that separates the fictional and the real. The various concerns we raised are ultimately variations on a single theme: in certain cases, lyric poetry demands from the reader clear forms of self-other differentiation in a manner that obliges us to see the lyric I as expressive of a subjecthood that is very much in the business of using, and not merely mentioning, words. Our suggestion is really an application of ideas already present in Walton’s model, tweaked so that these cases can be accommodated. We need some recourse to the imagination in order to account for how we encounter this potentially plural subject of lyric self-expression. And we need to think of the imagination in a way that does not risk reintroducing the fictionalizing tendencies discussed earlier (surely certain poems welcome these tendencies; the idea is that theories of the lyric should not introduce them, or even require them, from the get-go). Consider a simple example that suggests an intuitive distinction between two ways we can employ the imagination when thinking about others. Say that you have just heard a report on the news about a tragic event that has happened very much elsewhere in the world. You find yourself captivated and feeling deeply for the afflicted. You know nothing about the people, apart from the description of the tragedy provided by the news

Lyric Self-Expression

107

report. As you are thinking about their plight, you begin to imagine, in an ordinary sense of the term, what their experience was like. Depending on your interests and mood, you might shape an image of someone undergoing the harrowing events as they were described to you. And you might do so in a manner that this someone plays the role of a character in how you think about the event. You imagine a person who functions as a fiction in your employment of the imagination. But you might also – we suspect more naturally – not imagine a particular person at all and certainly nothing like a character. The image you form might just involve a broad-stroke picturing of the class of people, very much real, denoted by the description, say an image of those of whom the tragic descriptions are true, where this image pictures subjects generally, not individually. In this case, our imaginative activity does not culminate in an image of a particular suffering person but of a kind of experience that certain kinds of people had, a general and impersonal “they,” if you will. This employment of the imagination is typical not of fantasy but of our attempts to enrich our understanding of large segments of cultural life. In its most pedestrian and potentially risky form, it is a matter of relying on “received ideas” and the varieties of conceptual typecasting they invite. In more nuanced employments, it is an imaginative “filling out” of the view from there, where “there” indicates a shared form of social and cultural experience that is not my own. We often use the imagination this way when we try to put a human face on abstract understandings of collective identity and shared experience. The point here is that it need not be a particular human we imagine. If so, the imagination can’t be said to generate a fictional persona or anything character-like. Its goal is to track features of real experiences and to put us imaginatively in touch with the real people, general and plural, who are the subjects of these experiences. So our suggestion: while the lyric I surely represents a first-person mode of poetic expression, what it articulates calls us to engage in an act of imagining roughly of the sort just described. This, in turn, often explains how we animate the figure of the poetic speaker. What gives credibility to this view is a feature of the function of voice in lyric-poetic contexts. As we noted earlier, and as the Hughes example makes very clear, the lyric I often does not just “speak”; it speaks on behalf of. It represents not just a voiced subject but a poetic speaker who is attempting to give voice to forms of experience broader than her own. Our sense of both what the lyric I is speaking about and whose experiences it is giving voice to are thus often plural in nature. The business of lyric I is to “contain multitudes” in this sense,15 and it is a move that preserves literalism about lyric self-expression roughly along the lines Walton’s model does. The feature of poetry that makes the lyric I so fit to travel on Walton’s model also supports our suggestion. In realist novels, selves are standardly presented as subjects of experiences in a biographic sense, certainly when the self is playing the role of protagonist or villain. Their

108

Hannah H. Kim and John Gibson

inner lives, relationships, and basic events of their life-stories are often presented with great determinacy. They are experienced as more or less fully articulated individuals, and the reader is granted intimate insight into who they are, where they come from, and how their lives unfold. Lyric poetry, however, tends to present selves in a radically and intentionally diminished manner, and rarely is the I presented in a biographic sense. In fact, the I of lyric poetry is often nothing more than a center of perceptual, cognitive, and affective attention: the subject of an experience. It is a self effectively reduced to a perspective. The economy of expression, condensation of thought and feeling, and “semantic finegrainedness”16 characteristic of lyric poetry serve this ability to elaborate highly particularized scenes of thought and feeling that often do not appear to belong to any particular person at all. Apart from the elaborated perspective, little else is given about this person. This explains why Walton’s model can cast the lyric I as so transferable and such an easy object of readerly self-identification. The suggestion we are offering is that for the kinds of cases Walton’s model cannot accommodate, this same feature explains why we so easily and readily experience the lyric I as a generalized subject with an effectively plural expressive function. There are two ways to interpret our suggestion, one modest and the other less so, and what matters here is that each leads to the same destination: an affirmation of an epideictic and literalist account of lyric selfexpression. The less modest suggestion would be that we can see much lyric poetry as distilling into a solitary I – the poetic speaker – a collective subject. That is, the claim concerns the ownership of the I and casts it belonging to many. The more modest suggestion would be that our argument has shown that regardless of how we conceive of the subject lyric I, what it functions to give voice to is plural in nature and so that the content, if not the vehicle, of lyric utterances betokens “multitudes.” Little hangs on the ontological questions since our conception of poetic voice furnishes real, existent, and often plural subjects whose experiences are picked out in poetic self-expression. The kind of imaginative engagement a lyric poem makes central to appreciation constructs the link to actual subjects and very real regions of cultural life and in a manner that allows us to affirm, with Culler, that it is a form of “public poetic discourse about values in this world rather than a fictional world” (Culler, 2015, p. 115). So, constructing a fictional persona might be fine, too, as long as it does not replace or obscure the real and plural this-worldly subjects whose experiences are voiced. Fictions, after all, can be mouthpieces of the real, a claim no one involved in these debates would deny.

V. Our work in this chapter has been expansionary and exploratory. Though Walton has made great contributions to nearly every area in

Lyric Self-Expression

109

the philosophy of art, it is perhaps unsurprising that his contribution to the philosophy of poetry has not received as much attention: philosophers of art themselves are on the whole happy to ignore poetry. This is a shame for many reasons. One reason we have tried to make visible here is that the debate between literalists and fictionalists regarding lyric self-expression is especially deserving of more philosophical attention, and that Walton’s thoughtwriting model has given it its most novel and provocative position.

Notes 1. The coauthors shared equal responsibility for the research and writing of this chapter. They would like to thank David Hills, Andreas Elpidorou, Sonia Sedivy, and, especially, Monique Roelofs for helpful discussion of the ideas presented here. 2. The line occurs in Song of Myself, 51. 3. For an excellent overview, see Culler (2015, ch. 3). For recent work in a broadly analytic register, see Claudia Hillebrandt (2015) and Anna Cristina Ribeiro (unpublished) for arguments in favor of seeing the language (of much, all, or some) lyric poetry as nonfictional. Jenefer Robinson (2005, ch. 9) is perhaps the foremost proponent of positing (fictional) persona to account for works’ expressivity. 4. Lyric poetry is of course standardly contrasted with narrative poetry (epic poems, ballads, etc.). See Simecek (2015) for an excellent discussion of the anti-narrative conception of lyric poetry. 5. Even if some importation of worldly facts is typically required to fill out all the implicit truths that obtain in the story. See Currie (2010), Matravers (2014), Voltolini (2016), Stock (2017), and Friend (2017) for discussion of contemporary theories of fiction and their stance on such issues. 6. Of course, some prose works highlight their rhythmic, imagistic, figurative, formal, and expressive qualities as well – but we would say much more readily that a lyric poem loses something essential or distinctive to it when it fails to highlight those qualities than to say the same of a prose work that fails to highlight those qualities. 7. If the poem contains a metaphor, Walton might partly analyze a poem in terms of make-believe where the focus is on the “prop” – the metaphor – that serves as an invitation to a make-believe certain things about the target (see Walton, 1993). However, this is an analysis of a particular figurative language in a poem, and insofar as the focus is on the metaphor-constituting word choices, it is not a wholesale fictionalizing treatment of the poem that posits a fictional world from which the poem (ostensibly) speaks. 8. Of course, fiction can make statements about the real world (e.g., thought experiments, models). And some philosophers, notably Derek Matravers and Stacie Friend, question the fundamentality of the fiction/non-fiction divide in these contexts. But note that such statements are not “real” at least in the sense that they are cast as spoken from a different world (such that we would import some content into our world by, for instance, recognizing suitable parallels between the fictional world and the real world). Our aim in this chapter is to explore the possibility of an anti-fictionalist and literalist model, that is, whether the speaker can be seen as originating from, and directly addressing features of, this world.

110

Hannah H. Kim and John Gibson

9. This tradition originates in the work of Käte Hamburger (1993, originally published in German in 1957). Her manner of distinguishing fictional and lyric discourse is crucial here. For discussion, see Culler (2017). 10. See Ribeiro (2014) for an especially detailed account of the value of this in respect to our experience of poetry that traffics in negative emotions such as sadness, despair, and grief. 11. See Waters (2003) for a discussion of address to which we are indebted. 12. Kivy writes that a notable kind of contribution literature is capable of making has to do with what the reader does after reading – the analysis that occurs in the reader’s head during the postpartum period of contemplating – which is in itself an integral part of the literary experience. “Serious literary works,” he writes, “have a sloppy outer boundary” (Kivy, 1998, p. 23). 13. David Lewis (1978) and Gregory Currie (1990) argue that fictional truth is what the storyteller pretends to share as known fact. Free indirect discourse in novels might challenge this quick gloss that fictional narratives report some content since it blurs the boundaries between omniscient and character perspectives, undermining a sense of an objective storyworld or “what’s known as fact.” However, we maintain that a subjective presentation of the storyworld is still a representation (akin to a “report”) and that lyric poetry, in that vein, is often not in the business of reporting anything, subjective or otherwise. 14. Eileen John (2013) challenges the idea that poems are in the business of offering a thought and argues that poems should not be seen as depositories that encourage simple uptake of nuggets of thought. Her argument is important, and we unfortunately do not have the space to engage with it here. We trust that arguments in the next section keeps us from running afoul of the skeptical arguments she raises. 15. Note that “multitude” can refer to a plurality of individuals and also a plurality of groups (i.e., the lyric “I” may also speak on behalf of different groups of people). 16. See Lamarque (2015).

References Culler, J. (2015). Theory of the Lyric. Harvard University Press. Culler, J. (2017). Lyric Words, Not Worlds. Journal of Literary Theory, 11(1), 32–39. https://doi.org/10.1515/jlt-2017-0004 Currie, G. (1990). The Nature of Fiction. Cambridge University Press. Currie, G. (2010). Narratives and Narrators: A Philosophy of Stories. Oxford University Press. Friend, S. (2017). The Real Foundation of Fictional Worlds. Australasian Journal of Philosophy, 95(1), 29–42. https://doi.org/10.1080/00048402.2016.1149736 Hamburger, K. (1993). The Logic of Literature. Indiana University Press. Hillebrandt, C. (2015). Author and Narrator in Lyric Poetry. In D. Birke & T. Köppe (Eds.), Author and Narrator: Transdisciplinary Contributions to a Narratological Debate (pp. 213–234). De Gruyter. John, E. (2013). Poetry and Directions for Thought. Philosophy and Literature, 37(2), 451–471. https://doi.org/10.1353/phl.2013.0029 Kivy, P. (1998). On the Banality of Literary Truths. Philosophic Exchange, 28(1), 17–27. Kukla, R., & Lance, M. N. (2009). “Yo!” and “Lo!”: The Pragmatic Topography of the Space of Reasons. Harvard University Press.

Lyric Self-Expression

111

Lamarque, P. (2015). Semantic Finegrainedness and Poetic Value. In The Philosophy of Poetry (pp. 18–36). Oxford University Press. Lewis, D. (1978). Truth in Fiction. American Philosophical Quarterly, 15(1), 37–46. Matravers, D. (2014). Fiction and Narrative. Oxford University Press. Ribeiro, A. C. (2014). Heavenly Hurt: The Joy and Value of Sad Poetry. In P. Destrée & J. Levinson (Eds.), Suffering Art Gladly: The Paradox of Negative Emotions in Art (pp. 186–206). Palgrave Macmillan. Ribeiro, A. C. (unpublished). Semantic Powers: Fiction. In Beautiful Speech: The Nature, Origins and Powers of Poetry (in progress manuscript). Robinson, J. (2005). Deeper Than Reason: Emotion and Its Role in Literature, Music, and Art. Clarendon Press. Roelofs, M. (2020). Arts of Address: Being Alive to Language and the World. Columbia University Press. Simecek, K. (2015). Beyond Narrative: Poetry, Emotion and the Perspectival View. British Journal of Aesthetics, 55(4), 497–513. https://doi.org/10.1093/ aesthj/ayv041 Stock, K. (2017). Only Imagine: Fiction, Interpretation and Imagination. Oxford University Press. Voltolini, A. (2016). The Nature of Fiction/al Utterances. Kairos, 17(1), 28–55. https://doi.org/10.1515/kjps-2016-0016 Walton, K. L. (1993). Metaphor and Prop Oriented Make-Believe. European Journal of Philosophy, 1(1), 39–57. https://doi.org/10.1111/ejop.1993.1.issue-1 Walton, K. L. (2015). In Other Shoes: Music, Metaphor, Empathy, Existence. Oxford University Press. Waters, W. (2003). Poetry’s Touch: On Lyric Address. Cornell University Press.

7

Reading (With) Others Wolfgang Huemer

I. The Pragmatic Dimension of Fiction Works of fiction are omnipresent in our everyday life. If a foreign anthropologist came to describe the contemporary European or Anglo-NorthAmerican cultures1 from an “external” point of view, she might even diagnose an obsession with fiction, as just about all of us consume works of fiction virtually every day – be it a TV series, a movie, a novel or a short story, a painting, a cartoon, or simply the horoscope in the daily newspaper. We might be particularly obsessed, but we are definitely not alone. The practice of storytelling and the engagement in fictional scenarios are recurrent traits in all cultures and periods of history, though they do, of course, differ substantially in the concrete forms in which they are realized. Very often, works of fiction make use of techniques that otherwise serve to describe or depict real persons and events, but they often do so in a playful manner and portray persons who have never lived and events that have never taken place (at least not in exactly the way they are presented). When they do so skilfully and at a high level of accomplishment – as it is the case with the great works of art – they draw attention not only to the content, the scenarios, or the characters that are presented but also to the way in which they are presented, to the techniques of representation.2 As a result, many of the theoretical reflections on works of fiction – in literary theory, aesthetics, and art criticism, but also in philosophy of literature – focus on “formal” aspects of the works.3 These questions are most interesting and highly relevant for our understanding of art. They do run the risk, however, to eclipse a more basic question: What is it that motivates persons to engage with works of fiction in the first place? Where does the interest in fictional scenarios come from? Creating works of fiction is a laborious and often painful process and also reading or watching them, especially when one does so in a critical and engaged manner, requires time and energy, so the question of why persons invest their resources in works of fiction seems most relevant. To address this point, it can be useful to take not only the formal aspects of the concrete realization of an artwork into account or to restrict one’s DOI: 10.4324/9780367808662-8

Reading (With) Others

113

attention to the question of how levels of meaning can emerge from the material qualities of the work. In addition to these queries that are situated at the syntactic or semantic levels, as it were, I want to suggest to also take the pragmatic dimension into consideration and raise the question of what people do with works of fiction, how they use them and what role they play in social communities. This allows, as I will argue in more detail below, to construe works of fiction as encounters of concrete persons. Works of fiction are particularly powerful when it comes to express perspectives; they can, thus, be used by authors to present concrete points of view and by readers to better grasp what familiar phenomena might look like from a different point of view. The perspectives expressed in the works are, of course, merely displayed; they are not necessarily endorsed and should not be confused with that of the real author – nor, for that reason, do readers automatically absorb them and make them their own. By putting concrete perspectives on display, works of fiction allow readers to get a fuller understanding of other perspectives and contrast them with their own. This can lead them, if they think it necessary, to adjust their own ways of seeing things – in ways that are similar to that in which we fine-tune our own perspectives to that of members of our social community with who we directly engage and interact in everyday life. An enquiry that focuses on the pragmatic dimension does not need to collapse into a sociological study that describes, on the basis of empirical data, the ways in which actual persons do, in a given place or a given period, create or appreciate works of fiction. Rather, it will aim at gaining a deeper understanding of delineating what we might call the “logical space” in which these practices take place. The goal, thus, is not to document the moves that are actually made, but to understand which moves are possible within the practice. An enquiry at the pragmatic level will aim to reveal, in other words, the rules and conventions that are constitutive for the practice.

II. The Social Dimension of Imagination The guiding question of a study of fiction along these lines should, thus, take the form: “why do people care?” Why do people care to spend their time and energy to come up with fictional scenarios and make the products of their fantasies accessible to others? Why do they care to engage in works that depict unreal scenarios that seem to have no friction with the real world and cannot be put up to a “reality-check”? The observation that works of fiction are artefacts, created by persons4 for other persons, that carry meaning or express a point of view might be a promising starting point for addressing these questions. It invites us to conceive of fiction as a form of communication or a contribution to a conversation between members of a community.5 In this light, a work of fiction would be conceived as a unidirectional act of conversation

114 Wolfgang Huemer that allows a sender (author/artist) to transmit a message to the receivers (readers/spectators). Put in these terms, however, the limits of the analogy become quickly manifest: it risks to reduce the role of fiction to the transmission of information and assigns a purely passive role of receiving messages to the reader. Both points seem inappropriate. For one, works of fiction notoriously do not transmit (true) information,6 nor do authors support the claims that might emerge from the text with evidence or arguments. Two, many (though not all) of the motives one can have to consume a work of fiction require active participation from the reader or spectator. Sometimes, one might read a book or watch a film just to relax or to kill time, but typically we do not just passively absorb, but critically receive a work, it arouses aesthetic appreciation in us, triggers our imagination, and encourages us to explore counterfactual scenarios. All these reactions require active participation from the reader or spectator. Kendall Walton’s theory of fiction (cf. 1990) and especially his conception of games of make-believe have provided for the required extension of this perspective that allows to take the aspects I have mentioned into consideration. Walton, as is well known, emphasizes the role that works of fiction play for our imaginative activities. “Imagination” is notoriously an umbrella term (cf. Kind, 2013) that is used for a whole range of phenomena and which, thus, can be studied from very different perspectives that highlight very different aspects. While some philosophers prefer to focus on the subject, the person who is imagining, and conceive of imagination as a private mental act, a lived experience, Walton underlines the social dimension. He reminds us that imagination is not essentially private; rather, it is an activity that can always be shared with others. One way to do so is to create works of fiction that prescribe the audience to imagine something. The work establishes fictional truths and so presents a fictional world to the reader for her to explore. Fictional worlds are insulated from the real world, but they are not like “distant planets”; rather, by inviting them to actively participate in games of make-believe, the readers or spectators become part of this fictional world. This highlights an important aspect of Walton’s theory: imaginings in games of make-believe always have a self-referential or de se component; a participant of a game of make-believe is in one way or another always part of the world of the game. While an onlooker might, like a critic, just analyze or “stare at” the work,7 an appreciator who uses the work as a prop in a game of make-believe becomes more directly involved: “it is in a first-person manner that appreciators are to, and do, imagine about themselves; they imagine, from the inside, doing things and undergoing experiences” (Walton, 1990, p. 213f).

Reading (With) Others

115

III. Make-Believe and Experience Our knowledge is – at least in great part – based on experience, which often is less comfortable as one would hope. To make experiences one has to go out into the real world and risk to be exposed to difficulties or to encounter situations that are potentially dangerous. Moreover, to gain experiences that substantially widen our cognitive horizons, one has to leave the terrain of the familiar behind and make adventurous explorations of hitherto unknown phenomena or territories. In contrast, readers of a novel sit in their cozy armchairs; they move but their eyes and occasionally their hands, just to turn the pages. What relevant kind of experience could they possibly have? It is a commonplace that also imagination plays a crucial role in the advancement of understanding, but it is less obvious exactly how imagining fictional scenarios can add to our knowledge or widen our cognitive horizons. Walton’s approach provides an interesting framework for addressing these problems. In games of make-believe, we explore fictional worlds. Where these fictional worlds share relevant features with the real world, these explorations can allow for valuable conclusions.8 For my purposes, the first-person involvement in games of make-believe is of particular interest. When participating in a game of make-believe, according to Walton, we immerse into a fictional world of which we become a part. And though we might just be sitting in an armchair, we imaginatively explore the fictional scenarios and the protagonists’ fictitious experiences. As we are drawn into the fictional world in first person, as it were, these experiences gain a new quality and allow us to draw relevant conclusions. Yet, we can do so in a protected environment, without exposing ourselves to any risk. Is this a safeguard way to make relevant experiences that add to our knowledge? Walton does suggest so when he states: There is a price to pay in real life when the bad guys win, even if we learn from the experience. Make-believe provides the experience— something like it anyway—for free. Catastrophes don’t really occur (usually) when it is fictional that they do. The divergence between fictionality and truth spares us pain and suffering we would have to expect in the real world. We realize some of the benefits of hard experience without having to undergo it. (Walton, 1990, p. 68) It is important not to confuse this kind of cognitive beneft with the acquisition of phenomenal knowledge or knowledge what-it-is-like, that is, the kind of knowledge that Mary gains when leaving her black-andwhite room in Frank Jackson’s famous thought experiment (cf. Jackson,

116 Wolfgang Huemer 1982). Jackson famously argues that there are aspects of experience that a physicalist theory cannot shed light on: the qualitative aspects that one can know only if one has made the relevant kind of experience oneself. As examples for this qualitative aspect, Jackson mentions “the hurtfulness of pains, the itchiness of itches, pangs of jealousy, . . . the characteristic experience of tasting a lemon, smelling a rose, hearing a loud noise or seeing the sky” (Jackson, 1982, p. 127). The characteristic feature of this kind of knowledge is that it is about aspects of experience that are ineffable; it is, in other words, impossible in principle to fully express them with words. This entails, however, that also literary works of fiction have to fall short in imparting this kind of knowledge to the reader: even a most skilfully drafted poem cannot teach the colorblind what red looks like; nor can a most detailed description in a novel provide an acquaintance with the sensation of pain to a reader who has never experienced pain. To fully understand (literary) descriptions of the relevant states, it seems, one needs to draw on one’s own, personal reservoir of experiences, which in empathy one uses as a sample for the experiences of others (cf. Walton, 2015a). Thus, the participant’s first-person involvement brings colour into the fictional world, as it were, and enriches it with the relevant experiential qualities that she draws from her past experiences. “Our real selves make themselves felt in what we imagine, as well as in what we feel and the manner in which we imagine what we do” (Walton, 2015d, p. 283). In consequence, Walton suggests that rather than imparting new knowledge, participation in games of make-believe reveals facts about oneself that were (or easily could have been) familiar to me beforehand. In this way it can help me, for example, to “clarify my interests and desires” (Walton, 2015d, p.  279) or to get a more profound understanding of myself. Should we conclude that works of fiction cannot impart experiential knowledge to the reader, but merely mirrors the experiences she has made in the past? Are they but Freudian tools that allow us to gain access to those aspects of our “real character and personality” (Walton, 2015d, p.  277) that we normally repress or, for some reason or other, tend to overlook? If so, experiences described in works of fiction would not add to our knowledge about others, but merely reveal facts of our inner life to which we should have had access and which we could have known independently of the work. Walton seems to acknowledge this point, but hints that deepening our understanding of ourselves goes hand in hand with getting a better understanding of the others and the situations in which they live: These self-imaginings are important even when our main objective is to gain insight into others. In order to understand how minorities feel about being discriminated against, one should imagine not just instances of discrimination but instances of discrimination against

Reading (With) Others

117

oneself; one should imagine experiencing discrimination. It is when I imagine myself in another’s shoes (whether or not I imagine being him) that my imagination helps me to understand him. (Such imaginative understanding may be what has been called Verstehen.) And when I imagine this I also learn about myself. (Walton, 1990, p. 34) Literary works of fction, especially those that describe the imagined scenarios in a colourful and detailed manner, draw the reader into their fctional world and make her become a participant in a game of makebelieve. Thus, they make it easy to imagine oneself being part of a fctional scenario, be it in the shoes of the characters or in that of a bystander who merely observes the situation. It should, thus, be easy to empathize with the characters and to get a better understanding of the situation in which they live. By switching perspective in imagination, one might gain a better understanding of what it means to bring up a child as a single mother, to be target of racial attacks, or to live in poverty – and so, according to Walton, learn about myself. Things are not that easy, though. As I continue reading, it will come natural to confront the way I think I would move or feel in this situation with the characters’ course of action and the feelings ascribed to them. When reading some works, I might conclude that my (hypothetical) ways to act or react to the situation more or less match with that of the characters; when reading other works, I might realize that they differ substantially. In the first case, it is difficult to see how the work of fiction can reveal new insights about my own personality or character. Rather than learning about myself, it would merely confirm my way of seeing things and consolidate my habitual patterns of behaviour. In the latter case, however, I would have difficulties to put myself in the shoes of another. The course of action described in the work might be inacceptable for me – but make sense for a person who holds convictions or principles that are at odds with all I stand for. In this case, parallel imagining (cf. Walton, 2015a) would be blocked. Rather, my imagining to act or react this way in the given situation would require me to imagine being someone else who sees things and feels about them very differently from how I do; but then, how could this reveal something about my personality and character?

IV. Coherence and Credibility A more profound analysis of phenomenal knowledge, I fear, does not help us to overcome this impasse. It seems more promising to focus on the holistic character of games of make-believe. As Walton states, “[f]ictional truths come in clusters, and so do one’s imaginings of the propositions that are fictional” (Walton, 2015c, p. 18). When participating in games

118 Wolfgang Huemer of make-believe, we do not imagine isolated propositions one by one, we rather imagine a whole battery of propositions that add up to a fictional world – which is insulated from the real world; the cluster of imagined propositions is quarantined (cf. Salis & Frigg, 2020, p. 31), as it were. This allows us to take the propositions into consideration independently of their truth value and without having to worry about whether they are compatible with the rest of our beliefs; it makes, in other words, the free game of imagination possible. But why should we accept an author’s invitation and participate in the game in the first place? The way in which Walton develops his account of make-believe, starting with an analysis of children’s games, shows nicely that we have a natural inclination or curiosity to participate in games of this kind. Moreover, “[e]ngagement in make-believe tends to be infectious” (Walton, 2015b, p.  100). Works of fiction, thus, have a genuine fascination for us and we are even more willing to accept an author’s invitation to participate in the game when we come to note that the work in question has already caught the attention of others. However, as all young authors who are despaired over the failure of their debut novel know too well, an invitation alone does not suffice: a work has to fulfil minimal requirements to merit the attention of a broader audience. I shall mention two of them that seem to be particularly relevant from my argument: coherence and credibility. It seems obvious that the cluster of propositions a reader is prescribed to imagine should be coherent. A set of propositions that contains obvious contradictions9 would not add up to a self-contained fictional world and participation in the game of make-believe would become impossible. Coherence is merely a formal criterion, though, that can easily be met by an attentive author (at least if assisted by a meticulous editor). For our discussion, credibility seems to be a more central requirement. It is a fact that readers get easily irritated when they come across descriptions that are too fanciful or far-fetched, when the characters’ behaviour is completely unrealistic, when their “inner life” is incomprehensible, or when the plot is unconvincing. Yet, it is more difficult to be defined in a clear-cut manner, it seems to be essentially vague. Whether or not a work is credible does not only depend on features that are intrinsic to the work, it also depends on conventions that hold in our social practices and on the reader, who participates in the game of make-believe – it is not, in other words, the world of the work, but the world of the game that needs to be credible.10 At times, a game world might appear implausible or unconvincing – and, thus, lose its attraction and alienate the reader – because she participates in it in a way that is not exactly authorized by the work. This might be the case of a reader who is not aware of genre-conventions. The description of a travel on a spaceship faster than light is acceptable for most readers, but might ruin the novel for a physicist who reads her

Reading (With) Others

119

first science-fiction novel and who is not aware that in this genre some laws of physics are occasionally suspended. A more interesting example is that of biased, tendentious, or superficial descriptions of a milieu or of the characters’ course of action or their inner life, which can create a serious obstacle for readers who have a genuine interest in the situations described and already bring a differentiated understanding to the text. These readers might share the author’s overall position or not; in the first case, they will likely find the work simplistic and overly didactic, in the latter case they might even feel offended or outraged. Both groups of readers are aware that they hold a work of fiction in their hands and that, in consequence, no part of the work stands in contradiction to their previously held views – after all, they are only invited to imagine the described scenarios, but not prescribed to believe that they have actually taken place. What they do struggle with is the way in which the scenarios or characters are described, which obstacles a differentiated reflection and a mature judgement of issues that are most relevant to them.11

V. Encounters With Concrete Persons This last consideration, I think, sheds an interesting light on the question of how, by imagining fictitious experiences, we can gain new insights and deepen our understanding of others and of ourselves. The hypothesis I want to submit is that works of fiction add to our understanding when they can be construed as genuine encounters with a recognizable perspective that can be attributed to a concrete person – where “concrete” must not be interpreted as “real” or “actually existent”.12 The term “perspective” is arguably a cluster concept that can be used in a wide variety of ways. I propose to understand it in a minimal sense that is tied to its most literal use as a “point of view” to which a specific way of seeing things can be attributed. A perspective, in this minimal understanding, is like an ideal mathematical point, a centre of gravity, around which propositions (beliefs, assumptions, etc.) and dispositions to act or react in given circumstances can be organized in a coherent and organic way. This happens in the form of third-person ascriptions that are subject to phenomena like an origin, orientation, occlusion, and distortion and have an indexical element (cf. Van Fraassen, 2008, p. 85). This understanding of perspective is more minimal than that of Elisabeth Camp, who defines it as “an open-ended disposition to notice, explain, and respond to situations in the world—an ability to ‘go on the same way’ in assimilating and responding to whatever information and experiences one encounters” (Camp, 2017, p. 78). Fictional characters can be credible or convincing for a reader even when they are not similar to her or do not share her general outlook on things. The way the characters act, feel, or react in a given (fictional) scenario might differ radically from the way the reader would act, feel,

120 Wolfgang Huemer or react in a comparable (real-life) situation. What is important for the characters to be credible is that the reader can “make sense” of them. Characters are, in other words, credible for a reader if their choices, their behaviour, and their inner lives are coherent with the goals, convictions, principles, background assumptions, and values that can be meaningfully ascribed to them in a third-person perspective. To do so, we need to imagine “what it is like” to be in a certain situation. Grasping a perspective in this minimal sense does not, however, require us to shift perspective (cf. Goldie, 2011, p. 305f), we do not need to imagine “from the inside” what it is like to be in this point or situation; nor does it require us to integrate the character’s disposition into our own – and thereby becoming, at least temporarily, another person, as Camp (2017, p.  94) suggests. I do not want to deny that these deeper engagements with perspectives presented in works of fiction are possible for readers who want to get involved and are willing to bring their empathic understanding to the text. What seems relevant to me, however, is that a mere grasping of another perspective is possible also in a more minimal sense. The specifics that characterize particular perspectives are dictated to a considerable extent by the work, but as fictional characters are notoriously incomplete and not determined in all their properties, there remain blanks – often substantial ones – that need to be “filled in” by the reader. Thus, the character’s credibility depends not only on the author’s craftsmanship but also on the way the work is received by the reader. The very fact that the credibility of a fictional character is an issue unveils our tendency to detect a concrete person in the characters, as well as our efforts to ascribe complete and self-contained perspectives to them. Not all works of fiction attribute a central role to the characters, though. Sometimes, a work describes familiar situations or phenomena in an unusual manner that sheds a new light on them and so allows the reader to see the world with “new eyes,” as it were. Thus, a concrete perspective can emerge not only from the characters but also from the narrator’s or the (implied) author’s voice or, more generally, from the way in which the fictional scenarios, characters, and events are presented.13 There are, thus, (at least)14 two ways in which concrete persons can emerge from a work of fiction. For this to happen, it is necessary that the characters’ actions and reactions – or the voice of the narrator or the implied author – point back to a focal point in which a concrete point of view, a coherent and self-contained perspective on the world, becomes manifest. The exact way in which this happens can vary widely from one work to another and depends also on conventions that hold for certain genres, periods, and traditions, etc. It seems important to point out, however, that these encounters with concrete (yet fictitious) persons are not substantially different from encounters with real persons who actually exist and with whom we have direct acquaintance. We never can have a full or comprehensive

Reading (With) Others

121

knowledge of others. Even if we know another person very well, there are aspects of her life of which we have to remain ignorant. In every encounter with the other we gain singular impressions which we combine to an overall picture by generously filling in the blanks. We interpret the (verbal and non-verbal) behaviour of the other as expressions of a concrete point of view, a coherent and self-contained perspective on the world. In short, we apply more or less the same strategy both in our encounters with real persons and in moments in which we try to individuate a concrete person in a fictional character or in a narrator’s voice. According to the line of reasoning I have sketched, an immersion in fictional worlds can impart a kind of knowledge to the attentive reader that she could also gain from actual encounters with real persons. Works of fiction, thus, are not (primarily) Freudian tools or revealing mirrors that unveil hidden traits of the reader’s personality. Rather, they allow the reader to get to know in detail the perspective and personality traits of another person – who does not have to be real, but needs to be credible. This raises the question, however, of why we should care to meet fictional characters, when there are so many real persons out there whose perspectives and personality traits we could explore? What is the specific value of encountering concrete perspectives in works of fiction? Several points come to mind. First and foremost, encounters with fictional characters are less demanding and require less attention and continuous efforts over time than encounters with real persons – and we do not risk to pass as indiscrete when exploring the darkest corners of the character’s personality. Moreover, we are less likely to run risks – not because we do not have to worry that the bad guys might win, as Walton (1990, p.  68) suggests, but because a work of fiction can provide us a more profound understanding of the bad guys and of how they see the world, but spares us a potentially unpleasant encounter in person. Finally, a continuous interest in works of fiction presents us a plenty of perspectives and diverse points of view that goes far beyond the ones we would likely meet in everyday life.

VI. Reading With Others The perspective on fiction that I have sketched in the earlier sections, according to which works of fiction allow for encounters with concrete (yet fictitious) persons, who view the world from a recognizable point of view, enables us to discern a function of fiction that is often overlooked in the debate: engaging in a work of fiction invites us to confront our own perspective with that of others. This is not only central for deepening our understanding of others and of ourselves, it has also an important impact on the social community of which we are part, in at least two ways: first, Walton’s emphasis on the social dimension of imagination sheds light on how it is possible that in nearly all cultures we find a body

122 Wolfgang Huemer of works that is familiar to (nearly) all members of the community and so, in turn, determines the moves they can make within the practice. This shapes the identity both of the group and of its members and contributes to its persistence over time. Second, encounters with others are of essential importance also on a more basic level. The direct contrast with the (concrete) perspective of others is a central element in a dynamic process of fine-tuning or calibration that allows us to conform one’s own forms of behaviour to that of the other members of our community. I will shortly address both points in turn. The first point is inspired by an appreciation of Walton’s suggestion that imagination is an activity in which we often engage together with others, a game we play together with other members of our community. We have noticed earlier that engagement in make-believe tends to be infectious (Walton, 2015b, p. 100). I take this to mean that people are naturally attracted to works they think other people, or better: to other members of their social community, have appreciated. It is an effect of this process that in most cultural communities a set of works has emerged with which most members are (more or less) familiar. These works – we might think of a canon in a weak normative sense15 – function as cornerstones or points of reference that circumscribe what we might call, in analogy to Sellars’s notion of a “logical space,” a “cultural space” within which members of a community can make their moves and perceive the moves of others. There is, thus, a mutual dependence between the works that constitute the cultural space and the social practice of imagination: the set of “canonical” works emerges from the practice, as it contains the works that are regarded as particularly valuable, as “good practice” by members of the community – and at the same time provides a shared frame of reference that shapes and enforces the rules that constitute the practice. In short, individual members, who participate in the practice, shape with their contributions the “cultural space” within which the practice is performed, and the cultural space determines, at the same time, the rules that constitute the practice. Moreover, the rules that determine the practice are also the rules that allow individual members to reflect about themselves and to express who they are and who they want to be.16 In consequence, in virtue of its social dimension, imagination has a considerable impact on shaping the identity of the persons who partake in it. Second, the relevance of engaging in encounters with other persons, be they fictional or real, becomes most obvious when we take the social nature of human beings into consideration. Social communities are held together by a number of interrelated rule-guided practices that Wittgenstein has called our form of life (cf. Wittgenstein, 2009). The rules that guide and constitute these practices, however, are never explicitly spelt out. In a way, we follow them blindly. In consequence, these systems of

Reading (With) Others

123

rules are highly dynamic and change considerably over time – a point that is often overlooked in the philosophical debate, which often tends to focus on codified systems of rules which, by their very nature, serve to provide stability and to perpetuate established systems. The fact that most of the rules that constitute our practice are not stated explicitly but hold implicitly explains their malleable nature. Moreover, they need to be adapted over time to a great range of new and unforeseen situations, which accounts for minor changes and adjustments in the system of rules and the ways in which they are applied. The slow and continuous changes in our rule-following behaviour is not the result of a process of deliberation. They occur organically as the rules are applied in ever new circumstances and by different members of the community. In light of this dynamism, it is essential for the integrity of the community over time that all of its members adapt equally to the changes, which are arbitrary and unpredictable. This is possible only if individual members of the community continuously fine-tune their own rule-following behaviour to that of others, which allows them to stay “in line” with the others and remain members of the community. Should they fail to do so, the community would eventually fall apart. Given the dynamic nature of rule-following, the mechanisms of fine-tuning and calibration, which are deeply rooted in our biological constitution, are conditions of possibility of a social community.17 Confronting one’s own ways of doing things, one’s convictions, principles, background assumptions, goals, and values with that of others is an essential aspect of this mechanism of fine-tuning – and engaging in works of fiction is, as we have seen, a most efficient way to carry it out.

VII. Conclusion These last two points can give us a more profound understanding of the value of literary works of fiction. Much of the debate in philosophy of fiction over the last three decades has focused on the question of whether works of fiction can add to our knowledge or advance our moral understanding. Kendall Walton’s theory of make-believe, which emphasizes the social dimension of imagination, allows us to appreciate the relevance of fiction within a broader perspective. If the line of reasoning I have sketched in this chapter is plausible, it shows that next to its cognitive and moral values, literary works of fiction also have an important role in the social community in which they are embedded: by allowing for processes of calibration and fine-tuning among members of the community, they contribute to its inner cohesion; by providing cornerstones that circumscribe our “culture space,” it provides us with the expressive means that determine who we are and allow us to reflect on who we want to be.18

124 Wolfgang Huemer

Notes 1. This is, of course, not a sociological claim, but only a personal observation of the cultures I am most familiar with. The point likely generalizes, so we might want to imagine alien anthropologists who observe us from some other planet – just to produce the intended sense of unfamiliarity. 2. I am alluding here to the effect Russian and Czech structuralists have called the poetic function of language (cf. Jakobson, 1960). As a linguist, Jakobson described a phenomenon that can be individuated texts. The same effect can arguably be found also in other forms of representational works (in paintings, movies, cartoons, etc., be they artworks or not) that foreground the techniques of representation. 3. While literary criticism and aesthetics often focus on aspects that are related to the stylistic dimension of the works, many philosophers of literature are concerned with linguistic form and the workings of language, especially with aspects like reference and truth. Incidentally, I am not suggesting that we can draw a clear-cut distinction between form and content of a work of art. 4. I am aware that this affirmation might be empirically wrong as there might be works, now or in the near future, created by intelligent machines that have not been programmed specifically to produce a work in exactly that way. It seems to me, however, that the assumption that works are created by real persons seems to be deeply anchored in our understanding of literature and art. If art is a practice or a series of intertwined social practices (Sedivy, 2014), it seems obvious that moves can be made only by members of the community. Should computer-generated art become more present, the community would drastically change, which would bring about a substantial transformation of the practice. 5. The analogy has been explored, in the context of the discussion concerning authorial intent, by Noël Carroll (1992). See also Greve (2012). 6. It can be doubted that works of fiction and, more specifically, literary works of art do transmit information in the first place; but even where they would do so, they notoriously would not “guarantee” for the veridicality of that information. 7. Critics often do, of course, appreciate works of art. When they perform their critical analyses, however, they need to keep distance to the work. They avoid getting “caught up” in the story or immersing into the fictional world and focus their attention on the props themselves and not on the fictional truths they generate (cf. Walton, 1990, p. 53). A similar point, incidentally, can be made concerning empathy: critics typically avoid taking an empathic stance when analyzing a work, as John Gibson has pointed out. The notion of empathy is “not helpful for explaining criticism and interpretation” (Gibson, 2015, p. 238). 8. The relevance of Walton’s theory of make-believe for the cognitive advancement in the sciences, especially for thought experiments and modelling, is discussed in Salis and Frigg (2020). 9. Minor contradictions my go unnoticed or can be interpreted away, but obvious contradictions, once noticed, can completely ruin the game for the reader. 10. For the distinction between the world of the work and the world of the game, refer to Walton (1990, p. 58). 11. We need to distinguish this kind of example from cases in which the “message” of the work stands at odds with the reader’s take on the relevant issues. These scenarios, though quite frequent, are less interesting for my overall argument.

Reading (With) Others

125

12. It is worth mentioning that this is one among many ways in which works of fiction can widen our cognitive horizons. The cognitive value of fiction and literature has been extensively discussed in the last decades (cf. Gibson, 2008) and several interesting proposals have been presented. The reflections I offer in the current chapter are not intended as an alternative to the others, but will at best complement them. 13. We could say that the work “speaks to us,” as we “have a direct sense of distinctively human utterance, of ‘man speaking’ and speaking in some characteristic tone of voice” (Walsh, 1969, p. 117). John Gibson makes a similar point when he observes that sometimes we seem to identify with the work or the work’s perspective (cf. Gibson, 2015, p. 241f). 14. I am, of course, oversimplifying here. Literary works of fiction and, more generally, literary works of art come in many different forms and adopt very different strategies; in consequence, the encounters with concrete persons in works of fiction can take place in many different ways. For an illuminating discussion of how lyric poetry invites for a personal engagement and, thus, invites for encounters of this kind, refer to Ribeiro (2009). 15. When I speak of a “weak normative” sense, I intend the body of works with which members of a cultural community are typically expected to have some basic acquaintance. This expectance puts a weak normative demand on them. Moreover, I take it as an empirical claim that most cultures, but even many more circumscribed social groups, do in fact share acquaintance with such a body of works. I explicitly do not want to subscribe to the elitist understanding of a canon as the body of works with which one has to be well acquainted in order to count as educated or refined. 16. It is, of course, part of our literary and aesthetic traditions that authors and artists – in an act of creativity – violate or overcome these rules. They can do so only locally, though, by violating isolated rules while respecting (by and large) all the other rules that are constitutive for the social practice. 17. For a more detailed discussion of this mechanism of calibration or finetuning, refer to Huemer, (2020) and (2021). 18. I thank John Gibson, Sonia Sedivy, Christopher Williams, Daniele Molinari, and Irene Binini for helpful comments on an earlier version of this chapter.

References Camp, E. (2017). Perspectives in Imaginative Engagement with Fiction. Philosophical Perspectives, 31(1), 73–102. https://doi.org/10.1111/phpe.12102 Carroll, N. (1992). Art, Intention, and Conversation. In G. Iseminger (Ed.), Intention and Interpretation (pp. 97–131). Temple University Press. Gibson, J. (2008). Cognitivism and the Arts. Philosophy Compass, 3(4), 573–589. https://doi.org/10.1111/j.1747-9991.2008.00144.x Gibson, J. (2015). Empathy. In N. Carroll & J. Gibson (Eds.), The Routledge Companion to Philosophy of Literature (pp. 234–246). Routledge. https://doi. org/10.4324/9781315708935-20 Goldie, P. (2011). Anti-Empathy. In Empathy (pp. 302–317). Oxford University Press. https://doi.org/10.1093/acprof:oso/9780199539956.003.0018 Greve, A. (2012). Fiction and Conversation. Philosophical Investigations, 35(3– 4), 238–259. https://doi.org/10.1111/j.1467-9205.2012.01477.x Huemer, W. (2020). Rule-governed Practices in the Natural World. Journal of Transcendental Philosophy, 1(1), 161–181. https://doi.org/10.1515/jtph-2019-0016

126 Wolfgang Huemer Huemer, W. (2021). Two Pillars of Institutions: Constitutive Rules and Participation. in: L. Townsend, P. Stovall & H.B. Schmid, _The Social Institution of Discoursive Norms_, (pp. 177–193). Routledge. Jackson, F. (1982). Epiphenomenal Qualia. The Philosophical Quarterly, 32(127), 127–136. https://doi.org/10.2307/2960077 Jakobson, R. (1960). Linguistics and Poetics. In T. A. Sebeok (Ed.), Style in Language (pp. 350–377). MIT Press. Kind, A. (2013). The Heterogeneity of the Imagination. Erkenntnis, 78(1), 141– 159. https://doi.org/10.1007/s10670-011-9313-z Ribeiro, A. C. (2009). Toward a Philosophy of Poetry. Midwest Studies in Philosophy, 33(1), 61–77. https://doi.org/10.1111/j.1475-4975.2009.00185.x Salis, F., & Frigg, R. (2020). Capturing the Scientific Imagination. In A. Levy & P. Godfrey-Smith (Eds.), The Scientific Imagination (pp. 17–50). Oxford University Press. https://doi.org/10.1093/oso/9780190212308.003.0002 Sedivy, S. (2014). Art From a Wittgensteinian Perspective: Constitutive Norms in Context. The Journal of Aesthetics and Art Criticism, 72(1), 67–82. https://doi. org/10.1111/jaac.12043 Van Fraassen, B. C. (2008). Scientific Representation: Paradoxes of Perspective. Oxford University Press. Walsh, D. (1969). Literature and Knowledge. Wesleyan University Press. Walton, K. L. (1990). Mimesis as Make-Believe: On the Foundations of the Representational Arts. Harvard University Press. Walton, K. L. (2015a). Empathy, Imagination and Phenomenal Concepts. In In Other Shoes: Music, Metaphor, Empathy, Existence (pp. 1–16). Oxford University Press. Walton, K. L. (2015b). Existence as Metaphor? In In Other Shoes: Music, Metaphor, Empathy, Existence (pp. 89–117). Oxford University Press. Walton, K. L. (2015c). Fictionality and Imagination: Mind the Gap. In In Other Shoes: Music, Metaphor, Empathy, Existence (pp. 17–35). Oxford University Press. Walton, K. L. (2015d). Spelunking, Simulation and Slime: On Being Moved by Fiction. In In Other Shoes: Music, Metaphor, Empathy, Existence (pp.  273– 287). Oxford University Press. Wittgenstein, L. (2009). Philosophical Investigations (P. M. S. Hacker & J. Schulte, Eds.; G. E. M. Anscombe, P. M. S. Hacker, & J. Schulte, Trans.; Rev. 4th ed.). Blackwell.

8

The Puzzle of Fictional Morality Stuart Brock

I. Hume’s Puzzles In 1757, David Hume made public the following observation about our responses to narratives: Whatever speculative errors may be found in the polite writings of any age or country, they detract but little from the value of those compositions. There needs but a certain turn of thought or imagination to make us enter into all the opinions, which then prevailed, and relish the sentiments or conclusions derived from them. But a very violent effort is requisite to change our judgment of manners, and excite sentiments of approbation or blame, love or hatred, different from those to which the mind from long custom has been familiarized. And where a man is confident of the rectitude of that moral standard, by which he judges, he is justly jealous of it, and will not pervert the sentiments of his heart for a moment, in complaisance to any writer whatsoever. (Hume, 1757/1985, Paragraph 33) For most of the intervening 250 years, this passage has languished, receiving little critical attention from philosophers or art critics. Part of the reason for this neglect has to do with the paragraph’s lack of any natural ft with the rest of Hume’s now famous work Of the Standard of Taste, within which this passage is contained. Comparatively recently, though, philosophers have begun for the frst time to focus their attention on Hume’s aside, to appreciate its depth, and to understand the various problems it raises. The result has been a watershed. Since 1990, when Kendall Walton refected on the diffculty Hume’s observations pose for standard accounts of truth-in-fction (cf. Walton, 1990, pp. 154–156), there have been scores of articles published on Hume’s elusive comments, and many more are likely to come out.1 As a consequence, our understanding of literature, and the role played by the reader and author in our interpretation of it, has been greatly enhanced. DOI: 10.4324/9780367808662-9

128

Stuart Brock

To illustrate Hume’s problem, Brian Weatherson asks us to entertain the following story: Death on a Freeway. Jack and Jill were arguing again. This was not in itself unusual, but this time they were standing in the fast lane of I-95 having their argument. This was causing traffic to bank up a bit. It wasn’t significantly worse than normally happened around Providence, not that you could have told from the reactions of passing motorists. They were convinced that Jack and Jill, and not the volume of traffic, were the primary causes of the slowdown. They all forgot how bad traffic normally is along there. When Craig saw that the cause of the bankup had been Jack and Jill, he took his gun out of the glovebox and shot them. People then started driving over their bodies, and while the new speed hump caused some people to slow down a bit, mostly traffic returned to its normal speed. So Craig did the right thing, because Jack and Jill should have taken their argument somewhere else, where they wouldn’t get in anyone’s way. (Weatherson, 2004, p. 1) Hume thought that in cases like these, something goes wrong. When the author’s explicit moral evaluations within a fction confict with the evaluations we would naturally make in such cases it grates with our emotions, we judge that the novel is aesthetically blemished, we fall short of imagining what we’re invited to imagine, and we attribute a mistake of sorts to the author. The puzzle as Hume conceived it, then, is to explain why we have such reactions when confronted with deviant moral evaluations within a text. Part of the explanation for the recent interest in Hume’s passage is the clarity gained from disentangling the myriad of problems it alludes to. Walton (1994) noticed that there is not just one puzzle here; there are at least three. He distinguished between the aesthetic puzzle – how is it that descriptively deviant claims aren’t considered aesthetic flaws per se, but deviant moral claims are by their very nature aesthetic flaws in a work?; the puzzle of imaginative resistance – why do we resist or find ourselves unable to imagine worlds where the moral facts are different given that we have no trouble imagining worlds where the descriptive facts are different?; and the puzzle of fictional morality – why do authors have such difficulty making deviant moral propositions true in their fiction, when they can make other propositions true fairly straightforwardly? Brian Weatherson gave voice to a further puzzle to be found in Hume’s passage, what he calls the phenomenological puzzle: why do we feel uncomfortable about claims that conflict with our moral beliefs when we have no such reaction in response to descriptive claims that conflict with our other beliefs? In each case, the puzzle involves explaining an asymmetry in our attitudes towards implausible evaluative claims made within

The Puzzle of Fictional Morality

129

a story, and our attitudes towards the implausible descriptive (or nonevaluative) claims made within that same work. Walton, in his various writings on the puzzles (1990, 1994, 2006), takes special care to avoid appealing to the notion of make-believe in his characterisation of them. This is because he thinks that such an appeal encourages us to conflate the fictionality puzzle, the puzzle of imaginative resistance, and potentially some of the other puzzles. Recognising this allows us to develop different solutions to the different puzzles, while also recognising that there is a connection between them (see Walton, 2006, n. 8). In this chapter I will be concerned to examine only the fictionality puzzle, but what I say here is likely to be relevant to the other puzzles. The fictionality puzzle has been called variously ‘the puzzle of fictional morality’, ‘the alethic puzzle’, and ‘the fictionality puzzle’. This puzzle might be stated more fully thus: how can we explain the asymmetry in our judgements about the descriptive propositions an author can make true in her fiction and the moral propositions an author can make true in her fiction? In the former case, an author has carte blanche. Matters are different, though, when it comes to ethical claims made within a story. An author may explicitly state and fully intend that some state of affairs is acceptable (impermissible, or obligatory), but when such claims deviate from the moral evaluations we would naturally make in such circumstances, the author loses control of the content of her fiction. In such cases, we simply deny that the author has accurately described the fictional world she has created. Like Walton, I believe that the fictionality puzzle ‘is the most perplexing of the bunch’, even though it has received less attention than many of the other puzzles. In fact, this chapter is an extended argument for the following claim: the puzzle of fictional morality is really puzzling. In Section II, I consider a challenge to the intuitions that give rise to the puzzle. In Section III, I outline some striking empirical evidence in support of the intuitions shared by Hume, Walton, Weatherson, and others. And in Section IV, I reflect on the implications of this evidence.

II. A Puzzle About the Puzzle The puzzle of fictional morality is the problem of accounting for the apparent tension between the following two propositions: 1. An author can easily make any descriptive (or non-evaluative) claim true in a fiction whether or not we believe the claim is actually false. 2. An author cannot easily make a moral claim true in a fiction if the moral claim is one that we believe is actually false. Many philosophers who write in this area take propositions (1) and (2) as basic data that need to be accounted for. That is, they assert (1) and

130

Stuart Brock

(2) without citing any independent evidence in support of them, but nonetheless fnd them pretheoretically intuitive. After making these intuitions explicit, they take their task to be one of explaining this apparent asymmetry in an author’s control. To illustrate the role such intuitions are playing in this context, consider the following: We easily accept that princes become frogs or that people travel in time, in the world of a story, even, sometimes that blatant contradictions are fictional. But we balk—I do anyway, in some instances, and it is evident from the literature that I am not alone—at interpretations of stories or other fictions on which it is fictional that female infanticide is right and proper. (Walton, 2006, p. 140, my emphasis) I shall not be providing arguments that this asymmetry exists. There seems to be enough agreement that it does exist for it to be worth considering. If any readers do not recognise it at all, they should simply think themselves lucky and stop reading this chapter (Matravers, 2003, p. 92). The trouble is that others don’t share these intuitions. Although clearly not part of the philosophical mainstream, a vocal minority deny this supposed asymmetry. To appreciate the various ways one might deny this orthodox philosophical opinion, let me introduce the notion of a ‘Hume case’. A Hume case is any story containing an explicit statement such that the proposition it expresses does not form part of the content of the story. One way the accepted philosophical wisdom might turn out to be false is if stories that contained deviant moral claims – like Death on a Freeway – turned out not to be Hume cases at all. Such a view is held, for example, by Cain Todd (2009), who expresses his disagreement with the philosophical orthodoxy thus: It is not at all obvious that authorial authority breaks down in the supposedly puzzling cases. . . . I suggest that our initial intuitions here are not at all as clear as they are purported to be. Confronted with this story, my initial reaction, at least, is simply to be slightly baffled as to why the author has stated that; to wonder what the intention behind such an initially counterintuitive statement could possibly be. We should not (and generally would not) immediately strip the author of her authority in the establishment of fictional truths in the story. (p. 208) Todd is not the only philosopher to make such a claim. Michael Tanner (1994) and Mary Mothersill (2003) make similar claims. It is important to note, however, that there is another way to challenge the philosophical intuition. One can, pace Todd, Mothersill, and

The Puzzle of Fictional Morality

131

Tanner, accept that stories like Death on a Freeway are genuine Hume cases while at the same time denying that the evaluative content of the fiction is responsible for the author’s loss of control. That is, one can accept that the explicit moral claim made by the author isn’t true in the story while denying that it is the moral aspect of the statement that is giving rise to the trouble. It is this more subtle and concessive response to the philosophical mainstream that I want to investigate in this chapter. In particular, I will test the hypothesis that it is the ironic nature of such claims, rather than their evaluative character, that is responsible for the author’s loss of authority. It is interesting to note that in every published discussion of the problem of fictional morality, the examples used to illustrate the asymmetry can be interpreted as involving an element of verbal irony. Whether or not it is constitutive of the ironic, many cases of irony arise when statements are used to express something opposed to their literal meaning and something that is unexpected given a commonly accepted set of background facts.2 The claim that Craig did the right thing by shooting Jack and Jill because they should have taken their argument elsewhere in the context of the story in which it is embedded seems paradigmatically ironic in this sense. Yet, the cases used to illustrate the author’s complete control over the descriptive content of her fiction never seem ironic or open to an ironic interpretation. Matching the cases together in this way (the ironic/evaluative and the non-ironic/descriptive) thus leaves open the possibility that there is no asymmetry in authorial control over the evaluative and descriptive content of a fiction; instead, what we are observing is the uninteresting and understandable difference in the comparative difficulty a writer has in making a statement literally true in her fiction when the statement seems ironic. According to this alternative hypothesis, then, there is nothing special about the moral nature of the claims an author loses control over; instead, Hume cases arise only when a statement within the fiction can be interpreted ironically. The aim of this chapter is to test this hypothesis.

III. The Experiment A sample of 540 first-year students from Victoria University of Wellington in New Zealand were each asked to consider one of four suites of three short stories in a 2x2 factorial study design. (All of the surveys are collated in Appendix 2.) The participants were randomly allocated into four approximately equal groups based on the type of stories within the suite. The order of the stories within each suite was also varied appropriately to avoid any possibility of an order effect contaminating the results. The four groups were as follows: Descriptive/Ironic (DI); Descriptive/ Non-ironic (DNI); Evaluative/Ironic (EI); Evaluative, Non-ironic (ENI). For the most part, the EI and DNI stories were selected or constructed

132

Stuart Brock

using examples from Walton (1994) and Weatherson (2004). The ENI stories were constructed using examples that were expected to give rise to cross-cultural and intercultural moral disagreement. The explicit evaluations within the story were an articulation of one commonly expressed moral point of view. As such, readers would not be justified in inferring that either the author or the narrator did not believe the literal content of the evaluations and so could not naturally interpret them ironically. Story 1 from this suite made a moral claim about the moral permissibility of euthanasia, a claim that conservative students were expected to disagree with. Story 2 from this suite made a moral claim about the moral permissibility of seriously injuring someone in order to prevent an abortion taking place, a claim that liberal students were expected to disagree with. The excerpt from ‘Somali Story’ made a moral claim about the moral value of female circumcision, a claim that participants were predicted to disagree with and yet recognise that the moral opinions of people from Somalia might diverge from theirs. After presentation of each story, participants were presented with a question of roughly the following form, where q is something explicitly stated in the story: Each of the following statements (except the last) says (A) something about what’s true for us, and (B) something about what’s true in the said story. Tick the box next to the statement you agree with most. Tick only one box. (1) – (A) not: q, but (B) in the story q. (2) – (A) not: q, and (B) in the story not: q. (3) – (A) q, and (B) in the story q. (4) – (A) q, but (B) in the story not: q. (5) – None of the above. The question that interests us is: ‘What will participants say about the content of the story when they don’t believe what is explicitly stated by the author is actually true?’. The only responses that are revealing in this context are the frst two, and these will be the focus of the analysis. Participants who ticked any of the other boxes either indicated agreement with what was explicitly stated or declined to commit themselves either way, and so were not of interest. Recall that the only short story types considered in the literature are either of the DNI or EI kind. As a consequence, demonstrating a tendency on behalf of participants to tick the first response in DNI cases but to tick the second response in EI cases doesn’t help us determine which variable – irony or moral content – is responsible for the asymmetry. In order to determine which variable is responsible for the asymmetry, and to what degree, we need to consider the DI and ENI cases. If irony is primarily responsible

The Puzzle of Fictional Morality

133

for the asymmetry, we should expect a tendency of participants to tick the second response in DI cases and to tick the first response in ENI cases. If, on the other hand, the moral content is mainly responsible for the asymmetry, we should expect a tendency of participants to tick the first response in DI cases and to tick the second response in ENI cases. In order to evaluate the data in a meaningful way it was decided to determine the participants’ overall response across the three stories for each story type, rather than for the individual stories themselves. Because we are interested only in cases where the reader did not believe what was explicitly stated, we are interested only in comparing the frequency of Response 1 and Response 2 across the different story types. A participant’s overall response was determined to be 1 when 1 was a more common response than 2. A participant’s overall response was 2 when 2 was a more common response than 1, and a participant’s overall response was 0 otherwise. The results, summarised in the following pages and outlined in Tables 8.1–8.5 in Appendix 1, were interesting. Table 8.1 shows demographic data for the 532 participants, together with results of tests of association between the demographic variables and the story type. There was no evidence of a significant difference among story types with respect to the distribution of gender, ethnic group, and previous philosophy enrolment. Table 8.2 shows the number and percentage response for each story. All responses between 3 and 5 were recoded as a 0, since all these responses indicated no disagreement with the relevant explicit claim made within a story. In general, Response 1 (what I call the Uninteresting Case) was most common for both ironic and non-ironic descriptive stories. The exception was the first descriptive ironic story (38.1%), in which Response 0 was most common. Preference for Response 1 was more evident in the non-ironic stories (average 82.8% across three stories) than in the ironic stories (average 58.5%). For the evaluative stories, Response 2 (what I coined a Hume Case) was most common for both the ironic and non-ironic stories, with the exception of the first non-ironic story (18.2%). Preference for Response 2 was more evident in the ironic stories (65.9%) than in the non-ironic stories (47.0%). Overall Opinion Across Three Stories Table 8.3 shows the assumed overall view about the story types for the remaining 500 participants. The results in Table 8.3 show an overall preference for Response 1 for the descriptive stories (more evident for non-ironic stories) and an overall preference for Response 2 for the evaluative stories. The results are illustrated in Figure 8.1. The uninteresting cases are cases where a participant’s overall response was 1 (suggesting

134

Stuart Brock

Percentage of respondents’ opinions

100% No disagreement Uninteresting case Hume case

90% 80% 70% 60% 50% 40% 30% 20% 10% 0% Descriptive, Ironic

Descriptive, Non-ironic

Evaluative, Ironic

Evaluative, Non-ironic

Type of story

Figure 8.1 Percentage of the overall opinion across three stories of each story type

a tendency to allow the author’s stipulations to determine the content of the story even when the stipulation conflicts with our beliefs about reality). Hume cases are cases where a participant’s overall response was 2 (suggesting a limitation on the author’s control over the content of the story). No disagreement cases are cases where a participant’s overall response was 3, 4, or 5 (suggesting that the author’s stipulations are in no obvious conflict with the participant’s beliefs). A chi-square test was used to determine whether participants’ opinions were similar across the different story types as shown in Table 8.3. The results suggest that the differences in the overall opinion of the participants across story types were statistically significant (chi-square test, P < 0.0001), with participants mainly selecting Response 1 for descriptive stories and Response 2 for evaluative stories in both the ironic and nonironic case. Further tests were carried out to determine whether there were significant differences between participants’ opinions about ironic and non-ironic claims (Table 8.4) and between descriptive and evaluative claims (Table 8.5). The results suggested that participants’ opinions were similar for ironic and non-ironic stories (chi-square test, P = 0.29), but differed between descriptive and evaluative stories (chi-square test, P