New Waves in Aesthetics

Edited by Kathleen Stock and Katherine Thomson-Jones New Waves in Aesthetics

June 21, 2008 10:20 MAC/NWA Page-i 9780230_220478_01_prexx New Waves in Philosophy

Series Editors: Vincent F. Hendricks and Duncan Pritchard

Titles include:

Jan Kyrre Berg Olsen, Evan Selinger and Søren Riis (editors) NEW WAVES IN PHILOSOPHY OF TECHNOLOGY Vincent F. Hendricks and Duncan Pritchard (editors) NEW WAVES IN EPISTEMOLOGY Thomas S. Petersen, Jesper Ryberg and Clark Wolf (editors) NEW WAVES IN APPLIED ETHICS Kathleen Stock and Katherine Thomson-Jones (editors) NEW WAVES IN AESTHETICS

Forthcoming:

Yujin Nagasawa and Erik Wielenberg (editors) NEW WAVES IN PHILOSOPHY OF RELIGION Boudewijn DeBruin and Christopher Zurn (editors) NEW WAVES IN POLITICAL PHILOSOPHY

Future Volumes

New Waves in Philosophy of Science New Waves in Philosophy of Language New Waves in Philosophy of Mathematics New Waves in Philosophy of Mind New Waves in Meta-Ethics New Waves in Ethics New Waves in Metaphysics New Waves in Formal Philosophy New Waves in Philosophy of Law

New Waves in Philosophy Series Standing Order ISBN 978–0–230–53797–2 (hardcover) Series Standing Order ISBN 978–0–230–53798–9 (paperback) (outside North America only) You can receive future titles in this series as they are published by placing a standing order. Please contact your bookseller or, in case of difficulty, write to us at the address below with your name and address, the title of the series and the ISBN quoted above. Customer Services Department, Macmillan Distribution Ltd, Houndmills, Basingstoke, Hampshire RG21 6XS, England

June 21, 2008 10:20 MAC/NWA Page-ii 9780230_220478_01_prexx New Waves in Aesthetics

Edited by

Kathleen Stock and Katherine Thomson-Jones

June 21, 2008 10:20 MAC/NWA Page-iii 9780230_220478_01_prexx Selection and editorial matter © Kathleen Stock and Katherine Thomson-Jones 2008 Chapters © the individual authors All rights reserved. No reproduction, copy or transmission of this publication may be made without written permission. No paragraph of this publication may be reproduced, copied or transmitted save with written permission or in accordance with the provisions of the Copyright, Designs and Patents Act 1988, or under the terms of any licence permitting limited copying issued by the Copyright Licensing Agency, 90 Tottenham Court Road, London W1T 4LP. Any person who does any unauthorized act in relation to this publication may be liable to criminal prosecution and civil claims for damages. The authors have asserted their rights to be identified as the authors of this work in accordance with the Copyright, Designs and Patents Act 1988. First published 2008 by PALGRAVE MACMILLAN Houndmills, Basingstoke, Hampshire RG21 6XS and 175 Fifth Avenue, New York, N.Y. 10010 Companies and representatives throughout the world PALGRAVE MACMILLAN is the global academic imprint of the Palgrave Macmillan division of St. Martin’s Press, LLC and of Palgrave Macmillan Ltd. Macmillan® is a registered trademark in the United States, United Kingdom and other countries. Palgrave is a registered trademark in the European Union and other countries. ISBN-13: 978–0–230–22046–1 hardback ISBN-10: 0–230–22046–0 hardback ISBN-13: 978–0–230–22047–8 paperback ISBN-10: 0–230–22047–9 paperback This book is printed on paper suitable for recycling and made from fully managed and sustained forest sources. Logging, pulping and manufacturing processes are expected to conform to the environmental regulations of the country of origin. A catalogue record for this book is available from the British Library. Library of Congress Cataloging-in-Publication Data New waves in aesthetics / [edited by] Kathleen Stock and Katherine Thomson-Jones. p. cm. — (New waves in philosophy) Includes bibliographical references and index. ISBN 0–230–22046–0 (alk. paper) — ISBN 0–230–22047–9 (alk. paper) 1. Aesthetics. 2. Arts—Philosophy. 3. Aesthetics, Modern I. Stock, Kathleen. II. Thomson-Jones, Katherine. BH39.N48 2008 111.85—dc22 2008016326 10987654321 17 16 15 14 13 12 11 10 09 08 Printed and bound in Great Britain by CPI Antony Rowe, Chippenham and Eastbourne

June 21, 2008 10:20 MAC/NWA Page-iv 9780230_220478_01_prexx Contents

Notes on the Editors vii Notes on the Contributors viii Introduction xi Katherine Thomson-Jones and Kathleen Stock

1 The Ontological Diversity of Visual Artworks 1 Sherri Irvin

2 New Waves in Musical Ontology 20 Andrew Kania

3 Metaphors and Musical Expressiveness 41 Saam Trivedi

4 Davidson, Metaphor and Error Theory 58 Andrew McGonigal

5 Artifact Expression 84 John Kulvicki

6 A Metaphysics of Creativity 105 Dustin Stokes

7 From Defining Art to Defining the Individual Arts: The Role of Theory in the Philosophies of Arts 125 Aaron Meskin

8 Imagining Fact and Fiction 150 Stacie Friend

9 Three Debates in Meta-Aesthetics 170 Elisabeth Schellekens

10 Aesthetic Ideals 188 Rafa¨el De Clercq

June 21, 2008 10:20 MAC/NWA Page-v 9780230_220478_01_prexx vi Contents

11 Configuring the Cognitive Imagination 203 Jonathan M. Weinberg

12 Personifying Art 224 Brian Soucek

13 Danto and Kant, Together at Last? 244 Diarmuid Costello

Index 267

June 21, 2008 10:20 MAC/NWA Page-vi 9780230_220478_01_prexx Notes on the Editors

Kathleen Stock is Senior Lecturer in Philosophy at the University of Sussex. Her research interests centre upon the nature of the imagination, and in particular problems raised by the relations between imagination and fiction. She has published articles on imaginative resistance, on the cognitive role of fiction, on mental images, and on definitions of art. Currently, she is writing a monograph on imagination and fiction. Her editing work includes Philoso- phers on Music: Experience, Meaning and Work (Oxford University Press, 2007), a collection of new writing on the philosophy of music. She is Secretary of the British Society of Aesthetics.

Katherine Thomson-Jones is an assistant professor at Oberlin College. Her research interests centre on various questions in the philosophy of film, particularly regarding our engagement with films, as well as aesthetic formalism and the value of art. She has published articles on ethical art criticism, formalism, narration, and empathy in film. She is also the author of Film and Aesthetics (Continuum Press, forthcoming).

vii

June 21, 2008 10:20 MAC/NWA Page-vii 9780230_220478_01_prexx Notes on the Contributors

Diarmuid Costello teaches Aesthetics at the University of Warwick. He has published a number of articles at the intersection of aesthetics and art theory, including papers on Kant, Benjamin, Heidegger, Wittgenstein, Lyotard, Danto, Greenberg, Fried, and Cavell. He is co-editor, with Jonathan Vickery, of Art: Key Contemporary Thinkers (Berg, 2007) and, with Dominic Willsdon, of The Life and Death of Images: Ethics and Aesthetics (Tate Publishing & Cornell University Press, 2008). He is completing a monograph, Aesthetics after Modernism.

Rafa¨el De Clercq holds a doctoral degree in Philosophy from the Katholieke Universiteit Leuven, Belgium, where he also spent six years working as a postdoctoral fellow of the Fund for Scientific Research—Flanders. He is now at Lingnan University, Hong Kong. His research interests are mainly in aesthetics and metaphysics. In aesthetics, he has published on aesthetic ineffability, aesthetic properties, and modern architecture.

Stacie Friend’s research has focused primarily on issues at the intersection of aesthetics, mind, and language, especially as these pertain to problems raised by our engagement with fictional narratives. She is a lecturer in Philosophy at Heythrop College, University of London. She received her PhD in Philosophy from Stanford University in 2002. Sherri Irvin received her MA and PhD in Philosophy from Princeton University, and an MS in Clinical Psychology from Rutgers University. She has taught at the University of Ottawa and Carleton University, and is now Assistant Professor of Philosophy at the University of Oklahoma. She is the 2005 winner of the John Fisher Memorial Prize awarded by the American Society for Aesthetics.

Andrew Kania received his MA from the University of Auckland, New Zealand, and his PhD from the University of Maryland, College Park. He is currently Assistant Professor of Philosophy at Trinity University in San Antonio, Texas. His main area of interest is philosophy of the arts, particularly music, film, and literature. He recently won the inaugural Essay Prize of the British Society for Aesthetics, for an essay on the methodology of musical ontology.

viii

June 21, 2008 10:20 MAC/NWA Page-viii 9780230_220478_01_prexx Notes on the Contributors ix

John Kulvicki got his PhD at the University of Chicago in 2001, was a postdoctoral fellow at Washington University in St Louis for two years and a postdoctoral fellow at Carleton University in Ottawa for a year before coming to Dartmouth in 2004, where he is Assistant Professor of Philosophy.

Andrew McGonigal completed his PhD at the University of Glasgow, and has held a lectureship at the University of Leeds since 2002. His current research interests include aesthetics, metaphysics, and metaethics, and he has published on metaphor, response-dependence, and epistemology. He is currently working on papers on the autonomy of aesthetic judgment, the role of metaphor in critical responses to art, and the nature of constitutive relations in metaphysics. He has presented papers at national and international conferences and colloquia, including the annual conferences of the British Society of Aesthetics, and the American Society of Aesthetics. He is Secretary of the British Society for Ethical Theory and Deputy Director of the Centre for Metaphysics and Mind.

Aaron Meskin is Senior Lecturer in Philosophy at the University of Leeds, UK. Before moving to Leeds, he taught at Texas Tech University in Lubbock, Texas. His research interests include the cognitive imagination, the epistemology of artistic value and beauty, and the art of comics and graphic novels. He is Aesthetics section editor for Philosophy Compass.

Elisabeth Schellekens completed her PhD at King’s College London in 2003. The main topic of her thesis was Aesthetic Objectivity. After working three years as a postdoctoral Research Fellow both at King’s College London and the University of Manchester, she is now Senior Lecturer in Philosophy at the University of Durham. Her main research interests include the epistemology of value judgments, Kant’s aesthetic theory, and aesthetic realism. She has published in all these areas, and is currently preparing a book entitled A Reasonable Objectivism for Aesthetic Judgements.

Brian Soucek has been a collegiate assistant professor and Harper Fellow in the University of Chicago’s Society of Fellows since receiving his PhD from Columbia University in 2005. In addition to various topics related to the personification of art—the subject of his dissertation—Soucek is also working on projects about Sartre, opera, and eighteenth-century aesthetics. An essay on Mozart and Strauss appears in The Don Giovanni Moment (Columbia University Press, 2006).

Dustin Stokes works on the philosophy of mind and aesthetics. His research interests include creativity, imagination, mental representation, and perception. He is a Mellon postdoctoral fellow in the Department of Philosophy, University of Toronto.

June 21, 2008 10:20 MAC/NWA Page-ix 9780230_220478_01_prexx x Notes on the Contributors

Saam Trivedi, was educated at universities in the US, Britain, and India, and is an Assistant Professor of Philosophy at Brooklyn College in the City University of New York. He has published articles on various topics in Aesthetics such as interpretation, musical expressiveness, ontology, and Tolstoy’s aesthetics, and is currently working on a project on music and the emotions.

Jonathan M. Weinberg received his PhD from Rutgers University in 2002, and is currently an assistant professor of Philosophy and a member of the Center for Cognitive Science at Indiana University in Bloomington. He works in the various intersections of the philosophy of mind, aesthetics, and epistemology, including work on the imagination, intuitions, and experimental philosophy.

June 21, 2008 10:20 MAC/NWA Page-x 9780230_220478_01_prexx Introduction Katherine Thomson-Jones and Kathleen Stock

In the 1989 anthology, Analytic Aesthetics, J.O. Urmson describes being an undergraduate at Oxford in the 1930s when ‘there was no provision whatsoever for instruction in or the study of philosophical aesthetics’. This situation did not seem to surprise anyone, since it was assumed, first, that aesthetics was concerned with only one question—the question of the objectivity of beauty—and, second, that the answer to this question would be ‘a mere consequence’ of the answer to questions about moral value. Fur- thermore, to British philosophers at the time, existing work in aesthetics seemed either ‘jejune’ or ‘unintelligible’.1 Over the next decade, however, the situation began to change. British philosophers came to realize that the traditional question of aesthetics, considered as centrally concerned with art, is poorly formulated as a question about beauty, since many great works of art are not beautiful. Even more importantly, they came to realize that this question, however it is formulated, is not the only one worth asking in aesthetics. There are also questions about, for example, ‘the differences between the arts, temporal and non- temporal, performing and non-performing, the nature of realism in art, the relation of score to musical performance and of script to dramatic performance’, none of which are mere analogs of questions in moral philosophy. By asking questions like these which focus on conceptual distinctions, British philosophers betrayed a new willingness to subject art and art criticism to rigorous analysis. In summing up this shift in attitude among himself and his colleagues, Urmson records how analytic aesthetics began:

Briefly we were attracted by a realization that there were questions in aesthetics beyond the rather moth-eaten topic of beauty and the eye of the beholder, and the recognition that these questions could be tackled by the same down-to-earth methods and in the same plain English which we had learned to use in other areas of philosophy.2

In the introduction to the same anthology that contains Urmson’s essay, Richard Shusterman lists what he takes to be ‘the most typical features and themes of analytic aesthetics’.3 According to Shusterman, analytic aesthetics conceived of itself as a second-order discipline charged with clarifying and refining key concepts in first-order art studies. Consequently, analytic

June 21, 2008 10:20 MAC/NWA Page-xi 9780230_220478_01_prexx xii Introduction aesthetics tended to focus on art rather than on the broader aesthetic domain which includes nature. Questions of evaluation also tended to be ignored, since such questions were considered to be part of a first-order inquiry. As a reaction against romantic aesthetics, analytic aesthetics rejected essentialism about art as well as a general tendency to segue from aesthetics into transcendental metaphysics. Focused precision, with minimal attention to social and historical context, also checked these tendencies toward grand systematizing. This characterization of analytic aesthetics tends to emphasize the approach’s limits and omissions. In fact, by contrasting analytic aesthetics with recent, more inclusive, approaches, Shusterman writes as though analytic aesthetics is all but over. This view appears to have been shared by Anita Silvers, who, around the same time, characterized analytic aesthetics simply as a particular program pursued between 1946, when Wimsatt and Beardsley published ‘The Intentional Fallacy’, and 1962, when Charles Stevenson published ‘On The Reasons That Can Be Given for the Interpretation of a Poem’.4 Both Shusterman and Silvers express relief at the demise of analytic aesthetics, which, as mere second-order analysis, failed both to engage directly with the arts and to actively shape art studies. In fact, both these criticisms of aesthetics as second-order analysis miss their mark. The first criticism fails because the clarification of key concepts such as ‘art’, ‘music’, ‘beauty’, ‘expression’, ‘imagination’, and so on, which is part of the business of aesthetics, is not somehow directly engaged with artistic concepts and only indirectly engaged with the arts themselves. Rather it is primarily concerned with what the objects which fall under such concepts are; conceptual clarification takes place via direct examination of the world as well as of our talk about it. The second criticism fails both because history shows the influence of such analytic aestheticians as Richard Wollheim, Arthur Danto, and Nelson Goodman on the practice of some contemporary art and art history, and because even if it had been true that analytic aesthetics was not a significant influence upon art practice or theory, it would not follow that the fault lay with analytic aesthetics rather than with those who ignored what it had to say. There is no doubt that contemporary aesthetics pursues a broader and more diverse program than the original analytic one. Very few aestheticians today would be satisfied with Arnold Isenberg’s conclusions in a 1950 report to the Rockefeller Foundation which defines aesthetics just as ‘an analysis of the concepts and principles of criticism and other aesthetic studies, such as the psychology of art’.5 As the chapters in this volume clearly demonstrate, aesthetics, and even aesthetics in the analytic tradition, is more than this. Despite the sense that we have abandoned the narrow analytic program, the questions that interested Urmson and his colleagues are still considered some of the most compelling in aesthetics. These are essentially meta-aesthetic questions concerning the individuation of art forms

June 21, 2008 10:20 MAC/NWA Page-xii 9780230_220478_01_prexx Introduction xiii and artworks, and the complex ontology of the performing arts. The continued interest in meta-aesthetics is seen in this volume with two chapters on ontology (Kania and Irvin), one on aesthetic judgments (Schellekens) and one on aesthetic properties (De Clercq). Given their complexity and degree of abstraction, it certainly seems sensible to tackle such questions, as our contributors do, with the ‘down to Earth methods’ and ‘plain English’ that Urmson recommends. Urmson’s questions also reflect an early recognition that different art forms raise different philosophical issues. This point was driven home in J.A. Passmore’s now-famous 1951 essay, ‘The Dreariness of Aesthetics’. According to Passmore, aesthetics becomes both ‘dreary’ and ‘wooly’ when an attempt is made ‘to impose a spurious unity on things, the spurious- ness being reflected in the emptiness of the formulae in which that unity is described’.6 Such dreary wooliness is best avoided by analyzing each art form separately, according to its intrinsic features. Although analysis serves as a common philosophical approach to all artworks, the result is not any kind of general theory of art but a system of distinctions based on the interesting properties shared by different kinds of art. The persistence of this particular- ist tendency is seen in several of the chapters in this volume—particularly in Friend’s chapter on literary fiction, and in Trivedi’s chapter on musical expressiveness. Some would argue that the post-war analytic program heralded by Passmore and Isenberg simply replaced one kind of dreariness in aesthetics with another. Today, however, it is clear that Anglo-American aesthetics, as it has evolved since the 1950s, is far from dreary. While retaining an emphasis on precision and clarity, analytic aesthetics has greatly expanded the scope and depth of its inquiry. Issues in metaphysics and the philosophy of mind that are raised uniquely or in distinctive ways by art now receive a great deal of attention. This is reflected in this volume with the aforementioned chapters on ontology, aesthetic judgments, and aesthetic properties, as well as with McGonigal’s chapter on metaphor, Weinberg’s chapter on imagination, and Friend’s chapter on fiction. Questions associated with romantic aesthetics that were rejected by the first generation of analytic aestheticians have been revisited, as indicated in this volume by Meskin’s chapter on definitions of art, Stokes’s chapter on creativity, and Kulvicki’s chapter on artifact expression. And new ground is always being broken, either through the use of hybrid approaches like the historicized analysis employed by Soucek in his chapter on personification in art, or through a reassessment of aspects of the history of aesthetics, such as that presented by Costello in his chapter on Danto and Kant. This volume contains chapters by some of the best up-and-coming aestheticians. Where there is overlap between some of the chapters, this reflects the degree of interest that particular questions in aesthetics have recently been receiving. Given that the volume still exhibits a great deal of diversity,

June 21, 2008 10:20 MAC/NWA Page-xiii 9780230_220478_01_prexx xiv Introduction however, the following is an attempt to give the reader an overall sense of its contents. In ‘The Ontological Diversity of Visual Artworks’, Sherri Irvin argues that contemporary visual art cannot be analyzed in terms of a single or just a few ontological categories, or at least not interestingly so; in fact, substantive ontological facts pertaining to contemporary visual artworks are likely to vary from case to case. In support, she cites the widely endorsed ‘critical practice constraint’, according to which ontological claims about artworks should fit with the conclusions drawn by mature reflection upon artistic practice. Mature reflection upon artistic practice, Irvin argues, reveals that there is ontological variation even between works within the category of contemporary art. This is because such reflection shows that properties of a given work of art relevant to its interpretation—including, for instance, meaning properties and presentational aspects, as well as ontological properties themselves—are determined by what Irvin calls ‘the artist’s sanction’: the artist’s intention, explicitly articulated or at least implicit in her art-making choices, that the artwork has those properties. Often, Irvin argues, sanctioned aspects of a work to do with meaning or presentation place constraints on an appropriate account of the work’s ontology. She illustrates this by detailed attention to Felix Gonzalez-Torres’s Untitled (Portrait of Ross in L.A.):apile of candies which spectators are invited to eat, and which is to be regularly replenished to meet a particular weight. Irvin argues that only if we consider this work to be non-identical to any particular candy, or even to a particular pile of candies, can we do justice to the sanctioned features of the work, as intimated in its title, aspects of its presentation, and its relation to the artist’s life history. Continuing with the focus on ontology, Andrew Kania’s wide-ranging chapter, ‘New Waves in Musical Ontology’, is a guide to the current state of one of the liveliest debates in aesthetics. Besides detailed critical analysis, Kania’s contribution to the so-called ‘fundamentalist debate’ is the view that classical musical works are non-spatial but temporal abstracta that exist as long as their instantiation is possible due to a prior act of composition. The main recommendation of this account is that it respects our intuition that musical works are created, which, Kania argues, continues to have force even after reflection. Having made this contribution, Kania then concedes that it is perhaps ‘higher-level,’ rather than fundamental, musical ontology that offers the richest ground for further research. This is because debates about the relations between works, performances and recordings, the comparative ontologies of various kinds of music, and the nature of such elements of musical works as melody, harmony, and rhythm, are independent of the fundamentalist debate and more likely to further our understanding of the value of musical works of art. Since, as Kania rightly points out, music is the art most discussed in analytic aesthetics, it should not surprise the reader that this volume

June 21, 2008 10:20 MAC/NWA Page-xiv 9780230_220478_01_prexx Introduction xv contains not only an essay on musical ontology, but also one on musical expressiveness. In ‘Metaphors and Musical Expressiveness’, Saam Trivedi responds to the common view that descriptions of musical expressiveness are metaphors. He provides reasons for thinking that metaphors are paraphraseable, thereby circumventing the claim that descriptions of musical expressiveness ineliminably involve metaphors. This allows him to argue that attempts to paraphrase such descriptions as metaphors involve a loss of meaning—specifically, emotional content, which in turn suggests that these descriptions are not metaphors in the first place. However, even if we grant that descriptions of our musical experience are metaphorical, on a proper understanding of the nature of metaphor, we discover that the ‘metaphorist’ view reduces to a non-metaphorist view. Trivedi claims that many metaphors point to resemblances and, on the basis of these resemblances, prescribe us to imagine one thing as, or in terms of, another. As a result, there is a deeper, non-metaphorical explanation to be given of musical expressiveness that is based on an explanation of metaphor itself. In ‘Davidson, Metaphor and Error Theory’, Andrew McGonigal also focuses on metaphor, this time in isolation from its invocation in other areas of aesthetics. In particular, he seeks to defend Davidson’s well-known theory of metaphor from a raft of published criticisms. Amongst other things, McGonigal denies that a metaphor on Davidson’s view must be something unjustifiable in terms of reasons, or something whose effects cannot be partly attributed to its syntax; argues that an account can be given of the behavior of metaphors embedded in non-assertoric contexts consistent with Davidson’s view; defends Davidson’s claims that metaphors are literally false; and endorses his rejection of simile theories of metaphor. Toward the end of this rich chapter, McGonigal considers, and rejects, a challenge to an error-theoretic approach to metaphor, extrapolated from the work of Crispin Wright: the objection that, since a minimalist conception of metaphorical truth is available, a view of metaphors as literally false is uncharitable and so unacceptable in comparison. In response, McGonigal argues that to retreat to a more minimal and so less demanding notion of metaphorical truth, as the minimalist would urge, undermines important commitments in our thinking about meaning in general. Whereas Trivedi’s analysis of expression is restricted to the case of music, John Kulvicki’s analysis of expression is restricted in a different way, not to another art form, but to artifacts in themselves as opposed to their contents. In ‘Artifact Expression’, Kulvicki proposes the following account of artifact expression: ‘An artifact expresses X if and only if (1) it seems to be X insofar as it seems to indicate that about itself and (2) for other reasons, it seems not to be X.’ A sad painting seems to be sad by seeming to indicate its own sadness, but it also seems not to be sad for a different reason, namely that a painting is not the kind of thing that can actually be sad. The account recommends itself for several reasons. It preserves indication, which is a characteristic

June 21, 2008 10:20 MAC/NWA Page-xv 9780230_220478_01_prexx xvi Introduction feature of human expression, but without ‘shackling’ the domain of artifact expression to the narrower domain of human expression. And it draws our attention to, even while explaining the elusiveness of, an important kind of artistic expression—the expressiveness of the work itself as opposed to the expressiveness of what it represents or its design. In ‘A Metaphysics of Creativity’ Dustin Stokes offers an analysis of another key aesthetic concept; namely, creativity, which, he argues, is primarily a process. It is not simply a feature of persons, since a creative person must be one who engages in a creative process. Nor is it simply a feature of products, as anti-intentionalists might have us believe, for, Stokes argues, creative products are necessarily those produced via creative processes. Understood as a process, creativity is not wholly located at any one temporal point, nor is it homogeneous (no description of the whole of a creative process applies to any part of it). More specifically, creativity is an accomplishment. An accomplishment, qua process, terminates in a product, the nature of which retrospectively inflects the nature of the process (that which has been produced affects what has been done in its production). Stokes goes on to offer three conditions as minimal constraints upon a creative process: it counterfactually depends on an agent’s causing it; on being novel, at least relative to the population to which the agent in question belongs; and on its being true that, relative to the ‘cognitive profile’ of the agent in question (his/her knowledge, skills, ability to have certain thoughts or imaginings, etc.), the process could not have been undertaken before the time it actually was. In his contribution to the volume, ‘From Defining Art to Defining the Individual Arts: The Role of Theory in the Philosophies of Arts’, Aaron Meskin considers both the project of defining art generally, and that of defining a given art form or genre. Taking arguments of Morris Weitz as a starting point, he expresses doubts, both about the possibility of success in either project and about the desirability of pursuing the latter. After eschewing one oft-cited grounds for anti-definitionalism—an inductive argument from the failure of attempts to define lexical concepts generally—he goes on to examine and defend from critics a second route to such a conclusion, in the form of a ‘psychological’ argument about the non-classical nature of concepts. Meskin then turns to the project of defining art forms or genres, suggesting that, with the exception of technical concepts, this more specific project is imperiled by precisely the sorts of problems which beset the more general definitional project. Such problems notwithstanding, he additionally argues that the goal of defining an art form, by identifying certain features as necessary and sufficient for it, is undesirable, since it tends to detrimentally remove from critical discussion consideration of the aesthetic impact of typical features of a form (which tend to be characterized as essential, and so aesthetically irrelevant) or atypical ones (which tend to be taken to show that the work does not belong to the form in question).

June 21, 2008 10:20 MAC/NWA Page-xvi 9780230_220478_01_prexx Introduction xvii

The issue of defining art is also at the centre of Stacie Friend’s chapter, ‘Imagining Fact and Fiction’. Specifically, Friend is concerned with the distinction between fiction and non-fiction texts and whether this distinction can be justified in terms of a definition of fiction. The kind of definition that Friend considers emphasizes the function of fiction in prompting make- believe or imagining on the part of the audience. Friend shows that, on any number of interpretations, make-believe is not uniquely prompted by fiction. Moreover, theorists are not justified in limiting their definitions to fictional statements as opposed to whole works. This is because the explanatory direction of the relation between statements and works is at least not obvious, and philosophers should be able to reveal what is at stake in actual debates about the classification of works. Friend does not wish to deny a special connection between fiction and imagination, however. Rather, she wishes to suggest that this connection results from our practices of reading, writing, and evaluating fictions, since these practices can be explained in part by their having the purpose of sustaining creative acts and imaginative responses. While most of the chapter is a detailed critique of make-believe theories of fiction, Friend concludes with a positive suggestion: in order to understand the nature of fiction, and appreciate why the classification of certain works as fiction or non-fiction is controversial, we must shift our focus from the work alone to the complex and mutable practices that surround the work. Returning to themes in meta-aesthetics, Elisabeth Schellekens takes a broad critical approach similar to Kania’s in assessing progress in the realism/anti- realism debate. In ‘Three Debates in Meta-Aesthetics’, Schellekens argues that the focus in meta-aesthetics has shifted from the traditional metaphysical question of whether aesthetic properties exist to questions about aesthetic judgments—specifically, whether such judgments attribute objective or subjective aesthetic properties, and consequently whether such judgments can be correct. Convergence in the debate is the result of a broad acceptance of the objectivity of aesthetic properties as response-dependent properties and the availability of a normative standard of correctness deriving from the fact that aesthetic experiences can be shared and revised. This convergence appears even more marked in relation to our understanding of the role of emotion in aesthetic judgment. For example, given the widespread acceptance of cognitive theories of emotion, the anti-realist has yet another reason to accept a standard of appropriateness for aesthetic judgments, and thus move closer to the realist. Staying in meta-aesthetical territory, in ‘Aesthetic Ideals’, Rafael¨ De Clercq examines four possible interpretations of a suggestive claim by Frank Sibley that the sortal concepts under which objects fall provide aesthetic ideals by which to judge such objects. He finds each of the interpretations wanting in some way: either uninteresting, or false, or unclear. He goes on to suggest that the special relevance of sortals to aesthetic judgments is not that they provide aesthetic ideals by which to judge objects which fall under them,

June 21, 2008 10:20 MAC/NWA Page-xvii 9780230_220478_01_prexx xviii Introduction but rather that they are associated with certain non-aesthetic ideals, to do with what is normal for their members, or what their members’ function is, knowledge of which is at least sometimes necessary for making appropriate aesthetic judgments about those members. He then considers the relevance of his findings to the question of whether beauty, in particular, is a sortal- dependent property, concluding that it is, at least in an epistemological sense, insofar as an object O counts as beautiful only if there is a sortal concept S such that knowing O is beautiful requires knowing that O falls under S. Since many of the contributors—for example, Friend and Trivedi—refer to imagination, it is fitting that the volume should contain a chapter dedicated to an analysis of this capacity. In ‘Configuring the Cognitive Imagination’, Jonathan Weinberg outlines a naturalistic theory of the imagination, proposing it as a means of illuminating, among other things, certain philosophical problems about our engagement with fiction. On the theory he endorses, imagining is a belief-like computational process: belief-like insofar as it shares the structure and format of belief, is processed by the same inferential and affect-producing mechanisms, and interacts with many of the same domain- specific systems. On the other hand, imagining is unlike belief in its relation to perception, action, and decision-making, and in the relation between its content and the will. This theory, with particular focus on its configurational claims about how the imagination and other parts of the mind can be brought to interact, is then used, among other things, to illuminate the question of what it is to master a literary genre; and of how best to characterize the source of the desire-like attitudes we take toward fictional scenarios. Turning to the history of aesthetics, in ‘Personifying Art’, Brian Soucek draws attention to the tendency of aestheticians to personify art: that is, to talk of artworks in terms usually employed to talk about persons. Soucek argues that whether personifying art is judged appropriate or not usually depends on what are thought to be the ontological facts about artworks in comparison to persons: whether or not they are sufficiently like one another. Yet, he argues, this gets things the wrong way round. Endorsing a suggestion from Kant, pursued and extended by Schiller, Soucek claims that personifying talk is due to a similarity between our ways of (morally) judging persons and our ways of judging artworks, rather than a similarity between persons and artworks themselves. Consciousness of this similarity between moral and aesthetic judgments gives rise to our viewing artworks in terms usually applied to persons, as a ‘by-product’. Not only, Soucek suggests, has this conception of the relation between moral and aesthetic judgment been historically influential in the rise of personifying talk about art, it offers a way in which we can understand such talk as justified. Another reassessment of important aspects of the history of aesthetics which also connects to the discussion of metaphor in earlier chapters is offered by Diarmuid Costello. Arthur Danto and Kant are usually thought of as occupying polar opposite positions on philosophical issues concerning

June 21, 2008 10:20 MAC/NWA Page-xviii 9780230_220478_01_prexx Introduction xix art. However, in ‘Danto and Kant, Together at Last?’, Costello seeks to show that Danto’s philosophy of art has certain important aspects in common with Kant’s, in particular the view that the meaning of a work of art is metaphorical. Though Danto has acknowledged this point only recently, Costello argues that, in fact, it was implicit in Danto’s writing as early as The Transfiguration of the Commonplace (1981). Acknowledgment of something like it is required in order to validate the distinction, central to Danto’s theory, between artworks and ‘mere representations’ such as maps or diagrams. Costello goes on to analyze the theories of metaphor endorsed by Danto and Kant respectively, and concludes that, in its details, Kant’s view, unlike Danto’s, has the resources to explain the inexhaustibility of artistic meaning. Perhaps surprisingly, this is so, Costello argues, even where the art concerned is conceptual art from the twentieth Century.

Notes

1. Urmson, ‘The Methods of Aesthetics’, 20. 2. Ibid., 22. 3. Shusterman, Analytic Aesthetics,3–4. 4. Silvers, ‘Letting the Sunshine In’, 137–138. 5. Isenberg, ‘Analytic Philosophy’, 128. The whole report has never been published. Excerpts can also be found in Callaghan (ed.), Aesthetics and the Theory of Criticism. 6. Passmore, ‘The Dreariness of Aesthetics’, 325.

Bibliography

Callaghan, W. (ed.) 1973. Aesthetics and the Theory of Criticism: Selected Essays of Arnold Isenberg, Chicago: University of Chicago Press. Isenberg, A. 1988. ‘Analytic Philosophy and the Study of Art’, Journal of Aesthetics and Art Criticism, 46.3: 125–136. Passmore, J.A. 1951. ‘The Dreariness of Aesthetics’, Mind, 60.239: 318–335. Reprinted in Elton, W. (ed.) 1954. Aesthetics and Language, Oxford: Blackwell. Shusterman, R. 1989. ‘Analysing Analytic Aesthetics’, Shusterman, R. (ed.), 1989. Analytic Aesthetics, Oxford: Blackwell. Silvers, A. 1987. ‘Letting the Sunshine In: Has Analytic Aesthetics Made Aesthetics Clear?’ Journal of Aesthetics and Art Criticism, 46.3: 137–149. Urmson, J.O. 1989. ‘The Methods of Aesthetics’, Shusterman, R. (ed.), 1989. Analytic Aesthetics, Oxford, Blackwell.

June 21, 2008 10:20 MAC/NWA Page-xix 9780230_220478_01_prexx This page intentionally left blank 1 The Ontological Diversity of Visual Artworks Sherri Irvin

Virtually everyone who has advanced an ontology of art has accepted a constraint to the effect that claims about ontology should cohere with the sort of appreciative claims made about artworks within a mature and reflective version of critical practice. In this chapter, I argue that such a constraint, which I agree is appropriate, rules out a one-size-fits-all ontology of contemporary visual art (and thus of visual art in general). Mature critical practice with respect to contemporary art accords artists a significant degree of stipulative authority regarding the features and boundaries of their works. As I will show, this results in ontological variation among visual artworks. Any claim to the effect that all works belong to the same ontological category will thus come out false or uninformative. Interesting, substantive claims about ontological status can be made only in relation to specific works; that is, we must consider the ontological status of each contemporary artwork individually. The only general ontological claim that can be made about visual artworks (and also about artworks in other forms) is that they belong to the sort of thing that artists create. But this is not a substantive ontological claim.

I. The critical practice constraint

There was a time, a hundred years ago or so, when visual artworks seemed to be pretty straightforward things, and the ontology of art might have seemed a correspondingly simple problem, especially to those preoccupied with painting. Artists made paintings by putting paint on a support, often a canvas. Like other products, paintings could be bought and sold: the seller would deliver the painted canvas into the hands of the buyer, who now had a property right in it. Paintings could also be destroyed, and the destruction of a painting was something like the destruction of your favourite party dress: the physical object did not necessarily cease to exist, but it lost at least some of the features that made it worth looking at. Similar stories can be told about drawings, carved sculptures, and so forth.1 Moreover, artists spend a great deal of time crafting objects, enthusiasts spend a great deal of

June 21, 2008 10:27 MAC/NWA Page-1 9780230_220478_02_cha01 2 The Ontological Diversity of Visual Artworks time viewing them, and museums spend vast resources exhibiting, storing, and conserving them. We refer to the works in question as visual artworks, suggesting that the primary appreciative act is one of looking; and what, if not a physical entity, can be looked at? For these reasons, the temptation has been strong to think that visual artworks are simply physical objects. What could give us grounds for rejecting this straightforward idea? The resistance to the identification of artworks and physical objects has typically been grounded, implicitly or explicitly, in some sort of critical practice constraint. A critical practice constraint is a methodological principle to the effect that

[a]rtworks must be entities that can bear the sorts of properties rightly ascribed to what are termed ‘works’ in our reflective critical and appreciative practice; that are individuated in the way such ‘works’ are or would be individuated[;] and that have the modal properties that are reasonably ascribed to ‘works’, in that practice.2

Whatever we take the artwork to be, then, it must be the sort of entity which possesses the properties we attribute to artworks, or otherwise makes sense of the claims we make, in appropriate practices of interpretation and appreciation. The critical practice constraint is an acknowledgement of the fact that artworks, unlike such things as stars and water molecules, have no existence independent of human interests and practices. As David Davies puts it, ‘the very notions of ‘‘art’’ and ‘‘artwork’’ are parasitic upon [critical] practice—artworks just are the things that play a particular kind of role in a particular kind of practice’.3 Davies and Gregory Currie argue at length that a central tenet of critical practice is that the artwork is something achieved by the artist. Judgements about the quality of an artwork, Currie argues, rely centrally on knowledge about what the artist achieved in creating it.4 Davies develops at length the idea that grasping the artwork involves recognizing an artistic statement that has been finely articulated in a medium.5 Davies and Currie agree that these insights about appreciation of artworks require us to attend carefully to the artist’s creative activity. Indeed, they hold that the artwork should be identified with that creative activity rather than with any object or structure the activity gives rise to. In Currie’s view, the true artwork is the action-type of discovering a certain structure through a certain process (which he refers to as a ‘heuristic path’).6 Davies, too, holds that the artwork is to be identified not with a structure produced or presented by the artist, but with the artist’s generative act. Davies argues that our intuitions and statements about an artwork cannot coherently be made true of one and the same entity. Some statements, particularly about what is achieved through the work, refer to the artist’s creative activity, while other statements, such as those about formal features, refer

June 21, 2008 10:27 MAC/NWA Page-2 9780230_220478_02_cha01 Sherri Irvin 3 to the ‘focus of appreciation’ (i.e., the physical object, performance, conceptual structure, etc., presented by the artist for our attention).7 Davies holds that it is only by attending carefully to the artist’s generative activity that we can appreciate the statement the artist has articulated. For this reason, in appreciating the artwork we are, in effect, appreciating that activity; and this shows that the artwork should be seen as equivalent to the activity, not to the focus of appreciation.8 It is quickly obvious that Currie’s and Davies’s views fail to preserve the truth of many intuitive statements about artworks. For instance, it will be false to say of Picasso’s artwork Les Demoiselles d’Avignon that it hangs in the Museum of Modern Art, that it has a patch of brown in the upper left corner, or that it depicts women in a brothel, since these are not properties of the action Picasso undertook in creating the painted canvas. However, Currie and Davies hold that their views provide the best way of making another important set of critical and appreciative statements come out true, namely such statements as ‘Les Demoiselles d’Avignon was revolutionary.’ Picasso’s action, which was committed in a certain context with certain considerations in mind and in response to particular art-historical developments, clearly has the characteristic of being revolutionary. But the characteristic of being revolutionary cannot be appropriately attributed to a piece of canvas with pigment on it, since an object with exactly the same physical properties could have been produced through a completely different generative act, perhaps with nothing revolutionary about it.9 An advantage of Currie’s and Davies’s views is ontological unity: in Currie’s view, every artwork belonging to whatever form or genre is an action-type, while in Davies’s view every artwork is an action-token. Thus even the distinction between singular arts (like painting and carved sculpture) and multiple arts (like most cast sculpture and music for performance) turns out to be insignificant as it concerns the ontology of artworks, for in every case the work is simply the artist’s (type or token) generative act, and a generative act belongs to the same ontological category regardless of what is thereby created or discovered. Moreover, as Currie points out, his view is ontologically conservative, assigning all artworks to a familiar category: actions belong to the category of events, a respectable category that seems necessary to any viable account of the world. I agree with Currie’s and Davies’s claim that critical practice makes evident a very intimate connection between the artwork and the activity of the artist. This is true also of standard institutional practice: upon acquisition of contemporary artworks, curators and conservators collect extensive information from artists to guide their practices of storage, conservation, and display. Even after a work has been in a collection for many years, the artist, if still living, may be consulted about restoration or invited to the institution to participate in installation.10 These practices reveal that the artist is taken to have special authority on the nature and disposition of the work. In what

June 21, 2008 10:27 MAC/NWA Page-3 9780230_220478_02_cha01 4 The Ontological Diversity of Visual Artworks follows, I will appeal to widespread institutional practices of individuation, treatment and display of artworks as well as to practices of appreciation and interpretation, which tend to depend upon and be responsive to the former. While agreeing that the artist’s activity is centrally important to the artwork, I disagree with the conclusion that we must therefore see the work as identical to that activity. It is in fact possible to account for all of the properties we commonly and appropriately attribute to artworks without referring them to two different entities. To account for achievement-related aspects of the work, even while identifying the work with the focus of appreciation, we can appeal to relational properties: every focus of appreciation has the property of having been created through a certain process. Thus everything required to account for achievement-related properties does belong to the focus of appreciation proper.11 Of course, if we identify the focus of appreciation with a particular physical object (in cases where such an object is available), then the problem of modal properties will arise: if, as seems plausible, the focus of appreciation has its process-related properties essentially, but the physical object does not, the identity and persistence conditions of the two kinds of entity will differ.12 But rather than driving us to the view that the ‘true’ artwork is an action, this observation should lead us to develop an account of the focus of appreciation that does not identify it with a physical object construed independently of its relational properties. It is to this task that I now turn.

II. Artworks as things the artist creates

An appropriate understanding of the critical practice constraint should lead us, then, to turn our attention to the focus of appreciation. I will present an account of the artwork, construed as a focus of appreciation, that incor- porates all the properties that are appropriately attributed to artworks in practices of interpretation and appreciation. As this account has been argued for in detail elsewhere,13 I will focus here on showing how the account applies to a particular contemporary artwork and some hypothetical variations on it and how it allows us to account for the idea that the work is something achieved by the artist while satisfying the widespread intuition that an artwork is an entity created and presented by the artist for our perusal, not the artist’s activity of creation and presentation. In my view, artworks acquire their features through a diverse array of art- making activities. In addition to making or selecting some physical object, structure, event, process, or state of affairs, the artist often specifies details of presentation, which may include acceptable venues and physical configurations. Even works in the traditional genre of painting are partly constituted by the specification of an acceptable configuration: every traditional painting has an orientation that counts as right side up (as was highlighted by Georg Baselitz when he stipulated that his paintings were to be hung such

June 21, 2008 10:27 MAC/NWA Page-4 9780230_220478_02_cha01 Sherri Irvin 5 that their representational content was upside down). The artist may also specify certain properties related to the meaning of the work, such as the title; and this will constrain how it is appropriate to interpret the work.14 I describe this determination of the work’s properties through the artist’s acts of art-making as the artist’s sanctioning of certain features of the work. Not all sanctioning of artwork features is done explicitly. Established artistic genres involve conventions that determine some of the work’s features. For example, a traditional painter did not have to declare that all future paint flaking is irrelevant to the nature of the artwork. Instead, traditional paintings fall under a strong convention to the effect that when arms break off sculptures and paint flakes off paintings, we do not come to see them as armless-artworks and flaking-artworks, or interpret their subjects as armless humans or landscapes punctuated by strange, irregularly shaped patches of emptiness. Ordinarily, unless the artist explicitly stipulates otherwise, damage to the objects the artist created is not interpretatively relevant. It is not, then, necessary for the artist to explicitly sanction every one of a work’s features either by giving the object certain characteristics or by saying something about whether the work has this or that feature; if it were, then no work would ever be completed, because an infinite number of possible features would have to be ruled out. The artist’s sanction determines the artwork’s features in the context of conventions that specify default artwork boundaries. Many such conventions are defeasible. It is perfectly open to a contemporary artist to stipulate that the flaking of a painting is a bona fide feature of the work. The Japanese Gutai artist Saburo Murakami did just that. In 1957, he created a series of Peeling Off paintings which are such that the flaking paint is interpretatively and evaluatively relevant:15 when we consider what the work accomplishes, we should consider such things as the statement that is made by embracing the flaking of the paint, and the beauty or ugliness of the surface that emerges as the paint flakes. It should be mentioned that not every statement made by the artist serves to sanction features of the artwork. I distinguish between the features of the work and its interpretation, where the latter is constrained but not wholly determined by the former. The artist has the authority to sanction the work’s features, but not (directly) to fix the appropriate interpretation of the work. When artists make statements about messages purportedly conveyed by their works, for instance, these statements typically have no more force than those of ordinary interpreters.16 It is also possible for an artist to try but fail to establish a sanction about something that does fall within the domain where sanctioning is possible. If the artist’s preference regarding configuration of the work is not expressed clearly, or is mistakenly expressed to a museum visitor who happens to resemble the curator, then no sanction has been established. And if the artist makes contradictory statements, or expresses preferences that would be dangerous

June 21, 2008 10:27 MAC/NWA Page-5 9780230_220478_02_cha01 6 The Ontological Diversity of Visual Artworks or impossible to carry out within the framework of the institution, it may be necessary to rely on art world or institutional conventions to resolve the issueofwhatweshouldtaketohavebeensanctioned.17 In typical cases, many of the work’s features are sanctioned simply through the artist’s act of making a particular physical object. Kandinsky’s application of paint to canvas was the act through which he sanctioned most of the features of his works. He also sanctioned a configuration of the artwork, namely that a particular orientation counts as right side up. But, as we will see, in some instances the creation of the work does not involve the artist’s making or even selecting a particular object. In such instances there is no plausible identity or constitution relation that could hold between the artwork and a particular physical object. The artwork, then, is the entity that realizes the parameters sanctioned by the artist. The artist’s acts of creation and presentation are acts of articulating this entity, and they shape the properties of the entity in ways that are relevant to appreciation and interpretation. The ontological status of the entity will differ from case to case: in some instances it may be a particular physical object, or a particular token event, while in others it may be any assemblage of objects that satisfies the conditions specified by the artist. In some instances, indeed, the artwork may be a generative activity that eventuates in some physical object or structure; some works of performance art, such as those of Yves Klein’s happenings during which paintings were created, are best described in this way. It might be thought a cost of my account that visual artworks do not end up belonging to a single ontological category. However, insofar as focuses of appreciation appear to belong to diverse ontological categories even in the views of Currie and Davies,18 it is not clear that my account ultimately requires more categories than theirs. Moreover, it is a virtue of my view that it allows us to distinguish between certain works of performance art, where what is presented for our attention is really the artist’s creative activity, and traditional visual artworks, where the entity presented for our attention is a physical object appropriately displayed.

III. The ontological diversity of artworks

If we take artworks to acquire their features through a process of sanctioning by the artist, as I suggest, we must conclude that artworks are ontologically diverse. This is because artists, implicitly or explicitly, sanction different kinds of entity in different cases. To illustrate, I will consider a particular contemporary artwork and a number of hypothetical variations on it. Felix Gonzalez-Torres’s Untitled (Portrait of Ross in L.A.) (1991) involves a pile of hard candies in colourful wrappers. The particular hard candies displayed on a given occasion are not essential, however: viewers are invited to consume them as desired, and gallery staff periodically replenish them, in accordance

June 21, 2008 10:27 MAC/NWA Page-6 9780230_220478_02_cha01 Sherri Irvin 7 with Gonzalez-Torres’s instructions. The size of the pile is ideally determined by weight: it should add up to 175 pounds of candies. But even this is not definitive. Whenever someone takes candy out of the pile, the weight will be less than ideal. And, presumably, gallery staff do not weigh all the candies when they add to the pile; they just attempt a visual approximation of the initial installation. Untitled relates to the death from AIDS of Gonzalez-Torres’s lover, Ross Laycock, and the ideal weight of the pile is related to Ross’s ideal body weight.19 So the work, while appearing fun-loving and light-hearted, in fact explores issues of mortality and of the squeamishness and fear surrounding AIDS: since this is a portrait of Ross, we are symbolically consuming him, although many would have been reluctant even to touch him during his illness. Untitled and other consumable candy works by Gonzalez-Torres are also compelling in an art context, even for those who know nothing about Ross and his death from AIDS: it can be wonderful, after spending a day in a museum looking hard at beautiful and perplexing and remote things, to be able to pick up a piece of an artwork and lay claim to it or, better yet, to be able to eat it and appreciate its sweetness. There is something generous about these works (and something interesting about the way they enlist the museum in this generosity). The candy works also raise questions about the nature of the experience we expect and receive from art. We tend to treat artworks as great cultural treasures, but is the enrichment we receive from them, especially after museum fatigue has set in, really greater than the fleet- ing but real and immediate enjoyment of a piece of candy? To what extent should artists play to the viewer’s desire for a reward? It should be immediately clear that this work is not like a traditional sculpture.20 Every particular physical component is replaceable; indeed, the work can survive 100% replacement of the candies. For traditional sculptures, on the other hand, every physical component is essential, and if a piece is lost it cannot simply be replaced by a look-alike.21 Gonzalez-Torres’s work, then, is not identical to the particular candies that are dumped on the floor the first day of the exhibition. One might think, however, that the work is identical to a particular pile of candies, where a pile of candies is a physical entity that can survive the gradual replacement of all the particular candies it contained when it was first constituted. This possibility is interpretatively attractive in that a pile of candies is similar to a human body in this respect: human bodies are physical objects that survive the gradual replacement of their physical components. In this way, then, the construal of the work as a pile of candies would connect it to the body of Ross.22 If the work were a pile of candies, however, the work would cease to exist whenever the pile ceases to exist. Were the curators negligent in restocking the candies on a busy weekend, the work might accidentally be destroyed. And were the museum to discard the remaining candies between exhibitions,

June 21, 2008 10:27 MAC/NWA Page-7 9780230_220478_02_cha01 8 The Ontological Diversity of Visual Artworks they would have destroyed the work. But in fact, there is nothing in critical or institutional practice to support the idea that the work can be destroyed in this way or that it is an entity with discontinuous existence. Works of this type, which involve the assembly of new materials for each display, continue to belong to museum collections, and are spoken of by critics in the present tense, even when years elapse between exhibitions.23 Moreover, regarding the work as existing discontinuously would make it out to be a very odd sort of thing indeed: a non-contiguous entity in four-dimensional spacetime made up of a series of piles. Gonzalez-Torres’s work thus has a complex relationship to any particular pile of candies through which it is presented: it is clearly not identical to the pile, and it is constituted by some pile or other at most partially and intermittently, like a soul moving from body to body. Before considering the actual work’s ontological status further, let us consider some alternatives, superficially similar to Untitled.

1) In selling the work to the museum, the artist delivers a particular 175-pound batch of candies, with instructions that, for display, the candies are to be placed in a pile in the corner of the room. No mention is made that the candies may be eaten. 2) The work resembles the actual case in all respects, except that the artist’s instructions specify that when all the candies have been eaten, they are not to be replenished. 3) The instructions are the same as in the actual case (the candies may be eaten and are to be periodically replenished), but the museum curator decides that visitors should not be allowed to eat the candies. Thus the presentation ends up being the same as that specified in 1. 4) The work is like that in case 2, except that the artist stipulates that the work may legitimately be presented anywhere, at any time, if his instructions are followed.24 When all the candies have been eaten, then, a new instance of the work may be constructed, just as a musical work may be performed again after a particular performance has ended. 25

How should we understand the work in cases 1–4? It seems clear that in case 1 we should see the work as a case of sculpture. The artist is presenting the pile of candies as an artefact to be put on display, stored between exhibitions, and so forth. In the absence of any indication to the contrary, the candies will be treated by the institution as essential to the work, and will thus be subjected to careful storage and conservation procedures. It is clear that this work functions, interpretatively, quite differently from the actual work: it does not offer the same sense of generosity; it does not challenge the taboos associated with AIDS in the same way; and, rather than poking fun at the distance that institutions often impose between artworks and viewers by subverting it and allowing the candies to be consumed, it

June 21, 2008 10:27 MAC/NWA Page-8 9780230_220478_02_cha01 Sherri Irvin 9 pokes fun at this distance by preserving it and directing it at an object with respect to which such distance seems completely ridiculous. The work still comments on many of the same issues, but coldly and sarcastically rather than in a playful and inviting way: since Ross has AIDS, we cannot physically engage even with his symbolic stand-in; since artworks are precious objects with which we are not permitted to engage, we cannot eat even a simple, easily replaceable piece of candy on the floor of a gallery as long as it is part of an artwork. This version of the work takes the institutional imposi- tion of distance as an unfortunate given, a lamentable convention that we are stuck with, whereas Gonzalez-Torres’s actual work, rather than adopting such a defeatist mode, proposes that this distance is something art can light-heartedly shed. The nature of the commentary that this work makes on matters of institutional distance depends upon its essential constitution relation to a particular physical object that will be protected and conserved by the museum. The work described in case 2 is different, in important ways, both from the actual work and from the work described in case 1. Because eating of the candies is permitted in case 2, this work does not maintain the sense of institutional distance present in case 1. But, whereas the actual work is characterized by inexhaustibility and generosity, and gives us a tribute in which the artist’s love for Ross is immortalized, the work in case 2 confronts us very firmly with the finiteness of things. Each time we take a piece of candy, we must recognize that we are hastening the work’s demise, just as Ross’s demise was hastened by his disease. Should we eat a piece and thus enjoy the full experience of the work, or should we be frugal, simply imagining what it would be like to eat one of the candies, so that the work can last longer? Who will eat the last candy, and thus consign the work forever to oblivion? Seeing the work as essentially constituted by a particular physical object allows us to understand it as making a particularly poignant commentary on mortality. However, as we see by comparing case 2 with case 1, ontological status is only one factor among others; works belonging to the same or similar ontological categories can have quite different interpretative contents. Now consider case 3, in which the instructions are exactly as in the actual case: the pile is ideally 175 pounds, but it is permissible for audience members to eat the candies; the pile is to be replenished indefinitely. However, a curator decides, for some reason, that audience members are not to be allowed to eat the candies; or perhaps museum guards are ignorant of the instruction that the candies may be eaten, and thus they prevent audience members from approaching them. Notice that what the audience will experience, in case 3, is just what they would have experienced in case 1. And the evolution (or lack thereof) of the pile of candies will also be the same as in case 1. But should we therefore interpret the work as being different than it is in the actual case? No: the fact

June 21, 2008 10:27 MAC/NWA Page-9 9780230_220478_02_cha01 10 The Ontological Diversity of Visual Artworks that a curator makes a decision to present things incorrectly does not change what the work itself is or how it should be understood.26 It might make it harder for audience members to grasp the work, just as hanging a painting upside down might make it more difficult for the audience to grasp the work. But it does not change what it is appropriate to say in the interpretation or evaluation of the work. Only the artist can make decisions that will change the work’s properties (unless the artist explicitly yields that power to others). For this reason, the work in case 3 is the same as in the actual case, despite the fact that the presentation differs. And the ontological status of the work in case 3 differs from that of the work in case 1, though the presentation is equivalent. Let us now turn to case 4. In this case, the artist’s sanction determines that the work is an abstract entity, since it may have more than one instance. The work is to this extent similar to works for performance, such as musical and theatrical pieces. However, since the work is in the genre of contemporary visual art, where most works are singular rather than multiple, the permis- sibility of multiple instances becomes interpretatively relevant. Because the work can always be newly instanced after a particular instance goes out of existence, the eating of the last piece of candy loses much of the poignancy it would have had in case 2 (since, in case 4, it is not really the last piece). The work becomes a meditation less on mortality and loss than, perhaps, on the way in which a person’s spirit can live on and be reconstituted in different forms through the survival of those who knew and loved him. The work in case 4 seems fundamentally more optimistic than that in case 2, and this interpretative fact is closely related to its ontological status as an abstract entity permitting multiple instances. These examples illustrate two important points. First, as I described in the previous section, the artist’s art-making role is not exhausted by the creation of a physical object. Gonzalez-Torres’s art-making activity did not involve making an object, or even specifying a particular object. Instead, he sanctioned the features of his work through acts related to presentation: by communicating instructions to the museum, giving the work a title, and so forth. Through these art-making activities, he made it the case that Untitled involves the presentation and periodic replenishment of a (roughly) 175- pound pile of candies that can be eaten by viewers. And by these same activities, he made it the case that the artwork is not identical to or essentially constituted by any presented pile. Through a different set of decisions, however, he could have made a work that was essentially constituted by a particular pile of candies, or a work that was an abstract entity, susceptible of multiple tokenings. Second, there is a fit between the ontological status of the work and the interpretative properties that are appropriately attributed to it. When the artist chooses, whether implicitly or explicitly, to make a work that is essentially constituted by a particular physical object, this makes available certain

June 21, 2008 10:27 MAC/NWA Page-10 9780230_220478_02_cha01 Sherri Irvin 11 kinds of interpretative content that could not have belonged to the work had the artist’s sanction determined that it was, instead, a multiply instantiable type. Ontological status, like other elements of a work’s form, is a resource artists can use to imbue their works with meaning. Acknowledging differences in the works’ ontological status allows us to acknowledge important differences in their interpretative contents as well.

IV. Artworks, physical objects, and parameters

I have suggested that because of the nature of the art-making process, the artwork can come unmoored from any particular physical object. Gonzalez- Torres’s actual work cannot plausibly be thought to be identical to, to be essentially constituted by, or to have as a component any particular physical object, even one construed loosely, such as a pile. Moreover, the nature of the process by which artworks are created makes it the case that artists have it in their power to create different kinds of entity. Gonzalez-Torres could have made an artwork on the model of a sculpture, as in case 1, or an artwork on the model of a musical work, as in case 4. But he in fact made neither of these, opting for an artwork that, while not a multiple, can survive the destruction of any particular physical object or assemblage through which it is presented. But can’t we, nonetheless, find some interesting and informative things to say about the ontology of such works as a class? Isn’t it still the case that the works are essentially tied to or constituted by or composed of physical objects in a way that would allow us to say something about what they all have in common ontologically? It seems to me that going down the path of trying to find unifying claims to make about the ontology of such works is likely to mislead us at best. We might be tempted to say, for example, that Kandinsky’s Improvisation No. 7 is a painted canvas, whereas Felix Gonzalez- Torres’s work is a pile of candies that is allowed to be eaten. But surely Kandinsky’s work is an object in a quite different way than Gonzalez-Torres’s work is. Gonzalez-Torres’s work, if it is a pile of candies, is not any particular pile of candies (even if we admit, as I think we should, that a pile of candies can survive a 100% replacement of its parts). Each time the work is put on display, a new pile of candies will be made. Moreover, as I mentioned earlier, the work continues to exist even when there is no pile of candies: even if the museum throws away all the old candies in between exhibitions, and has no objects in storage at all in relation to this work, the work itself has not been destroyed. The situation is quite different for Kandinsky’s work: this work centrally involves a particular object and cannot go on existing if that object is destroyed. Given widespread pre-theoretical intuitions, it seems attractive to maintain, whenever we can, that an artwork is a physical object. But even in the easiest sort of case, it will be at best misleading to say that the artwork is

June 21, 2008 10:27 MAC/NWA Page-11 9780230_220478_02_cha01 12 The Ontological Diversity of Visual Artworks simply a physical object. Kandinsky’s Improvisation No. 7, perhaps the most straightforward sort of case, is not simply a physical object; it is, at best, a physical object considered in a certain way, where the artwork cannot be grasped in the absence of such consideration. The object must be considered with the correct orientation, right side up rather than upside down or sideways. And, should some dirt begin to accumulate on its surface, or should a piece of paint flake off, it will not thereby become a dirty-work or a flaked-paint-work. Thus, if we wish to maintain that the work is a physical object, we must say that it is a physical object considered as though it were appropriately oriented and in pristine condition. The actual painted canvas that we see in the museum might not have all of these features; so gaining access to the artwork must involve going beyond what we see. Thus even a traditional painting is, at best, a physical-object-considered-in-a-certain- way; and the way in which it must be considered may well diverge from the way it actually is. A characterization offered by Amie Thomasson seems to fit such works: they are ‘individual concreta (bearing physical properties perhaps among others) constituted by physical objects, but not identifiable with their constituting matter’.27 However, I have suggested that Gonzalez-Torres’s work Untitled (Portrait of Ross in L.A.) is not a physical-object-considered-in-a-certain-way, since the work has a continuous existence, whereas no physical object that would be a candidate for an identity or constitution relation to the work exists continuously. To what ontological category, then, does the work belong? And if this ontological category is indeed viable, why can’t ontological unity be preserved by assigning all artworks to this category? The details of the case require that Gonzalez-Torres’s work be construed as an individual concretum not essentially constituted by a physical object.28 The work can continue to exist in the absence of any particular physical object or assemblage; however, it is not an abstract entity on the model of a type or kind, since it is not susceptible of having multiple tokens or instances. The nature of the concretum is determined by a set of parameters for presentation expressed by the artist: the work is such that presentation of it involves the construction of a 175-pound pile of candies that audience members are allowed to consume, and that the museum is to replenish periodically. It should be noted that even if it is possible to unite Untitled (Portrait of Ross in L.A.) with Kandinsky’s Improvisation No. 7 under the label ‘individual concreta’, this does not eliminate the ontological diversity of artworks. First, it is clear that ontologically diverse entities can be described as individual concreta: there is no reason to expect a common ontological account of the pebble caught in the tread of my boot, my present, occurrent thought about that pebble, and the sentence-token about the pebble that I am presently typing on my laptop. Second, as we saw earlier, it is open to a visual artist to create a work that is an abstract rather than concrete entity: Gonzalez-Torres

June 21, 2008 10:27 MAC/NWA Page-12 9780230_220478_02_cha01 Sherri Irvin 13 could have created a work that was susceptible of multiple instances, as in case 4, and the work in question would have been abstract. The recognition that some artworks are concreta partly constituted by physical objects whereas others are non-physical concreta or abstract entities is not aesthetically ineffectual. Creating different kinds of entities allows artists to imbue their works with different kinds of content. The ability to create a work that is a concrete non-physical entity allows Gonzalez-Torres to make a particular commentary on the nature of persons. Moreover, by creating a non-physical concretum that is presented through an assemblage of physical objects, Gonzalez-Torres bridges the domains of conceptual and traditional visual art, and this is an additional source of potential critical interest. To force works of this sort into the category of individual concreta identified with or essentially constituted by physical objects would rule out their possessing certain kinds of interpretative content that we appropriately attribute to them, and thereby violate the critical practice constraint. Above, I stated that Gonzalez-Torres’s creation of Untitled (Portrait of Ross in L.A.) involved the expression of a set of parameters for presentation. And, in general, my view about the artist’s sanction suggests that such specification of parameters, whether explicit or implicit, is a crucial part of the process of creating an artwork. Even Kandinsky’s creation of Improvisation No. 7 involved the expression of such parameters: correct presentation of his work requires that a particular painted canvas be hung flat on a wall with the painted side visible, and in a particular orientation. Why, then, can’t we identify the work with the set of parameters that specify acceptable presentations, such that all artworks are construed as abstract entities and are susceptible of being brought together within a common ontological category? On such an approach, visual artworks could be treated jointly with musical works, which are often construed as abstract entities involving parameters for acceptable performances.29 The critical practice constraint mandates rejection of the view that visual artworks in general are to be identified with sets of parameters expressed by the artist, despite the attractiveness of such a view when applied to musical works for performance. It seems that we admire particular musical performances as aesthetic objects in their own right, and that a substantial portion of our admiration for the musical work itself is by virtue of the possibilities it creates for interesting performances. Our appreciative practices, then, provide some warrant for the idea that what is appreciated, when we appreciate a musical work, is a set of parameters: the parameters are what create the possibilities for aesthetically appealing performances. Moreover, a work for performance is less aesthetically replete than any performance of it,30 just as any set of parameters not stipulating a maximally determinate set of values will be less fully specified than an object satisfying those parameters. Our practices of appreciating musical works, then, make sense of the idea that

June 21, 2008 10:27 MAC/NWA Page-13 9780230_220478_02_cha01 14 The Ontological Diversity of Visual Artworks a musical work is an abstract entity, perhaps aptly characterized as a set of parameters or norms for performance. Critical practice does not, however, treat most visual artworks as sets of norms or parameters. A traditional painting is not of interest for the range of possible presentations it creates. The parameters for its presentation are such as to make all permissible presentations fall within a very narrow range: the parameters will specify that a particular physical object be displayed in a particular orientation, so that viewers have a similar visual experience each time they encounter the object. Moreover, unlike many of the aesthetic properties of a performance of a musical work, the properties that are of aesthetic interest in a painting are not derived from a set of parameters specified by the artist; instead, most of the parameters (e.g., acceptable presentations of this work require the display of a particular painted canvas with such-and-such features) are generated implicitly through the application of paint to canvas. We do not look through the presentation of a painted canvas to infer the set of parameters the artist expressed, treating the parameters themselves as of substantial aesthetic interest; instead, we appeal to the parameters (such as those specifying correct orientation) so as to ensure that we are grasping the aesthetic features of the painted canvas appropriately. Moreover, it is not clear that a set of parameters, if it has any representational content at all, will have the same representational content as the object presented. The parameters expressed by the artist, then, are not of primary aesthetic interest in such traditional cases. To say that the artwork is simply the set of expressed parameters would be unjustifiably revisionist from the perspective of critical practice. The fact that traditional works of visual art are not sets of parameters, however, does nothing to show that no contemporary artwork is a set of parameters. Similarly, my objections above to the view that artworks are in general to be assimilated to action-types or action-tokens do not rule out the possibility that some artworks are action-types or action-tokens; I take it that there are works of contemporary performance art falling into each of these categories. The problem with views that attempt to shoehorn all visual artworks into a single category is that they deny us the resources to acknowledge the very real and aesthetically relevant distinctions among works that function differently and bear quite different relations to their art historical predecessors. By creating works belonging to different ontological categories, artists may express distinct artistic content, and only a view that acknowledges the ontological diversity of artworks can capture this fact.31

V. Grasping the artwork

I have suggested that some artworks are abstract entities while others are individual concreta, some of which are partly constituted of particular physical objects while others, though not essentially constituted by physical

June 21, 2008 10:27 MAC/NWA Page-14 9780230_220478_02_cha01 Sherri Irvin 15 objects, are presented by way of them. This raises the question, What is the relationship between what we see in a gallery on a given occasion and the artwork? The answer is that the artwork is never grasped exclusively through seeing: we never simply apprehend that a certain set of properties is interpretatively and evaluatively relevant by looking at some object, independently of rather detailed and specific background knowledge. For most artworks, many of their properties are grasped through vision. This is especially true for traditional paintings and sculptures. But invariably, even for traditional works, some properties must be grasped through cognition that is independent of vision. The correctness of a particular orientation of a painting cannot simply be grasped through vision, for it may be that the painting has been hung upside down, as with a large abstract painting I once saw in a Beijing museum. The relevance or irrelevance of the fact that paint is flaking off of the surface of a painting cannot simply be grasped through vision; one must know, independently of vision, either that the work falls under a convention making this fact irrelevant or that the artist has suspended the convention and made the flaking paint part of the work.32 Because the background knowledge that establishes which features of the object are relevant varies from case to case, there is no straightforward algorithm for inferring the artwork based on the physical object. One must attend directly to the details of what the artist has sanctioned in the particular case. It might seem unfortunate that we are forced to adopt a view that says artworks are not simply seen. But this fact may, in a sense, be a blessing, for it allows that even if the display is faulty, the audience’s ability to grasp the work may be preserved. As any museum conservator knows, parts of contemporary artworks disappear on a regular basis. I know of one case in which over 20 per cent of the pieces of an important contemporary artwork were stolen while the work was on loan from the museum that owns it. Indeed, I once saw a contemporary artwork that included many small plastic zip-top bags, each of which contained a small amount of material that had been stolen from someone else’s work. It is clear, then, that many art objects are missing some of their physical components. And then there are all the ways in which curators can get instructions slightly wrong, or be forced to install a work in a space for which it was not initially envisioned.33 There is the fact that a painting may not be hung perfectly horizontally, so the correct orientation for display is not quite realized. Because grasping an artwork always involves grasping some properties that are not accessible through vision, and because it often involves ignoring outright some of the features of the object on display, these cases do not present insurmountable difficulties. We can consider the painting as if it were properly hung, and the contemporary installation artwork as if all its components were intact. If we do this, then we will have grasped the work the artist created—and this, I think, is the fundamental activity of the art audience: to grasp the work through both careful looking

June 21, 2008 10:27 MAC/NWA Page-15 9780230_220478_02_cha01 16 The Ontological Diversity of Visual Artworks at and careful thinking about what the artist has made. If we succeed in this, then we will be in a position to interpret and evaluate artworks much more successfully.

Acknowledgements

I am grateful to Stephen Davies, Martin Montminy, and, especially,Kathleen Stock for comments on earlier versions of the manuscript.

Notes

1. Artworks with multiple instances, like prints and cast sculptures, have always complicated this picture. In such cases, the artwork cannot be identical to any particular physical object since it can survive the destruction of any particular instance (except, in some cases, the last instance). 2. Davies, Art as Performance, 18. In Art as Performance, Davies refers to this constraint as the ‘pragmatic constraint’. He employs the term ‘critical practice constraint’, which I favour, in ‘Artwork, Action, and Process’. As Davies notes, Levinson explicitly employs a similar constraint in ‘What a Musical Work Is’. Wollheim’s Art and Its Objects and Danto’s Transfiguration of the Commonplace are also deeply informed by an implicit critical practice constraint. 3. Ibid., 21. 4. Currie, An Ontology of Art, esp. 36–40. 5. Davies, Art as Performance, 52ff. 6. Currie, An Ontology of Art, esp. 66–71. By ‘structure’, Currie refers to such things as ‘a certain pattern of lines and colours, structure of sounds, or sequence of words’ (47). 7. Davies describes a focus of appreciation as the ‘outcome or product of a generative performance on the part of one or more individuals’ (Art as Performance, 26, emphasis in original). Though there are differences between Currie’s notion of a structure and Davies’s notion of a focus of appreciation, nothing hangs on this point for the sake of the present argument. I will use the term ‘focus of appreciation’ to describe the product of the artist’s activity in relation to both Currie’s and Davies’s views. 8. Davies’s view diverges from Currie’s in that Davies holds, on the basis of reasons I will not rehearse here, that it is preferable to see the artwork as a particular token of creative activity rather than an action-type. 9. For such an argument, see especially Currie, An Ontology of Art, 50–53. 10. For description of actual cases, see Irvin, ‘The Artist’s Sanction’ and ‘Museums’. 11. Kathleen Stock makes a similar point in her review of Art as Performance. 12. Robert Stecker argues that this concern can be dealt with by allowing that two distinct physical objects of different kinds can occupy the same spatiotemporal position. As we will see, though, some artworks cannot plausibly be construed as identical to or constituted by physical objects even on this more liberal notion. See Stecker, ‘The Ontology of Art Interpretation’, 177–191, section IV. 13. Irvin, ‘The Artist’s Sanction’. 14. However, simply giving the work a title is not sufficient to imbue the work with particular representational content. Whether or not the artist has made a picture of, say, a banyan tree will depend partly on our conventions and practices for

June 21, 2008 10:27 MAC/NWA Page-16 9780230_220478_02_cha01 Sherri Irvin 17

decoding pictorial representations, partly on what banyan trees actually look like, and so forth. 15. Schimmel, ‘Intentionality and Performance-Based Art’, 137. 16. For further discussion, see Irvin, ‘The Artist’s Sanction’. 17. Space constraints prevent me from discussing this important issue in detail. Some such cases are treated in Irvin, ‘Museums’. 18. I see no denial of this claim in Davies’s or Currie’s works. Moreover, insofar as ontological unity is to count as an advantage of their views, it must be the case that focuses of appreciation are ontologically diverse. 19. Sautman, Self-Guide,8. 20. Neither, though, is it purely conceptual. Comparison of Untitled (Portrait of Ross in L.A.) with other Gonzalez-Torres works, such as the Untitled (Placebo) (1991), involving 1000–1200 pounds of candies wrapped in silver and displayed in a neat rectangular array on the floor, shows that Gonzalez-Torres gives sculptural attention to the visual details of his work, and visual appearance is part of what viewers should and do consider in appreciation and interpretation. 21. See Sagoff, ‘On Restoring and Reproducing Art’, for discussion of this point in relation to the Vatican restoration of Michelangelo’s Pieta`. Of course, replacement parts can supply considerable information about what the original was like; but to encounter even an excellent replica is not to encounter the original work. 22. I am grateful to Jason Southworth for this point. 23. Another prominent example is Jana Sterbak’s Vanitas: Flesh Dress for an Albino Anorectic, exhibition of which requires that a dress be sewn from large pieces of flank steak. 24. Sol LeWitt’s wall drawings function in this way: he allows that anyone who follows his instructions precisely will have constructed a genuine LeWitt, whether or not LeWitt is aware of or authorizes the activity. 25. One might wonder whether the actual work should be regarded as an abstract entity, such that each pile of candies is a token or instance of the work. While I think that criticism and institutional convention regarding this and similar works support the idea that they are singular entities rather than multiply instantiable abstractions, I will leave this matter aside for the present. The four cases are sufficient to establish the ontological diversity of artworks, regardless of the verdict on the actual work. 26. It is possible for a curator-cum-artist to constitute a new work through appro- priation and (perhaps non-standard) display of items created by others, as when Garry Neill Kennedy proposed that all landscape paintings in a particular wing of the National Gallery of Canada be re-hung at his eye level. However, such interventions do not change the original work; Kennedy’s intervention did not make it the case that the original works he appropriated are now such that they must be hung at his eye level. For discussion of cases of incorrect installation, or conflict between the artist and the museum about appropriate display, see Irvin, ‘Museums’. 27. While regarding this as the most plausible construal of the view that artworks are closely associated with physical objects, Thomasson does not explicitly endorse this characterization. Thomasson, ‘The Ontology of Art’, 82. 28. In drawing the distinction between abstract and concrete entities in a way that allows that an entity may be concrete without being partly constituted by a physical object, I am informed by Gideon Rosen’s discussion of the Way of Abstraction in ‘Abstract Objects’. A particular marriage or a particular bank account would

June 21, 2008 10:27 MAC/NWA Page-17 9780230_220478_02_cha01 18 The Ontological Diversity of Visual Artworks

also count as a non-physical concretum on this account (although, as in the case of an artwork that is a non-physical concretum, many physical events and states of affairs will be relevant to its having come into existence and acquired the properties it possesses). In some ways of drawing the abstract/concrete distinction, of course, such entities, though singular, will be termed ‘abstract’ simply because they are not physically constituted. 29. For influential statements of such views, see Wolterstorff, Works and Worlds of Art; and Levinson, ‘What a Musical Work Is.’ Stephen Davies notes, in Musical Works and Performances, that some musical works, such as recordings of purely electronic music, are not for performance; such works may not be amenable to a construal as abstract entities. 30. As Davies notes in Musical Works and Performances (Ch. 1), this relation may vary from work to work, with some works being much ‘thicker’ than others. 31. Dominic Lopes offers an argument for the ontological diversity of architecture, based on the fact that sound critical practice varies cross-culturally in such a way as to have ontological consequences. A similar argument might offer independent support for claims about the ontological diversity of visual artworks. See Lopes, ‘Shikinen Sengu’. 32. Of course, once one knows that the orientation of the painting is correct or that the flaking paint is an element of the work, many of the work’s properties, such as the colours and forms on the surface of the painting and the spatial relations among them, will be directly seen. 33. For discussion of such cases, see Irvin, ‘Museums’.

Bibliography

Currie, G. 1989. An Ontology of Art, New York: St. Martin’s Press. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Davies, D. 1988. ‘Artwork, Action, and Process’, Acta Analytica 20: 131–153. Davies, D. 2004. Art as Performance, Malden, MA: Blackwell. Davies, S. 2001. Musical Works and Performances: A Philosophical Exploration, Oxford: Oxford University Press. Irvin, S. 2005. ‘The Artist’s Sanction in Contemporary Art’, Journal of Aesthetics and Art Criticism 63: 315–326. Irvin, S. 2006. ‘Museums and the Shaping of Contemporary Artworks’, Museum Management and Curatorship 21: 143–156. Levinson, J. 1980. ‘What a Musical Work Is’, Journal of Philosophy 77: 5–28. Lopes, D. M. 2007. ‘Shikinen Sengu and the Ontology of Architecture in Japan’, Journal of Aesthetics and Art Criticism 65: 77–84. Rosen, G. Spring 2006. ‘Abstract Objects’ in E. N. Zalta (ed.), The Stanford Encyclopedia of Philosophy, URL = http://plato.stanford.edu/archives/spr2006/entries/abstract- objects/. Sagoff, M. 1978. ‘On Restoring and Reproducing Art’, Journal of Philosophy 75: 453–470. Sautman, A. 2003. Self-Guide: Modern and Contemporary Art at the Art Institute,Chicago: Art Institute of Chicago. Schimmel, P. 1999. ‘Intentionality and Performance-Based Art’ in M. A. Corzo (ed.), Mortality Immortality? The Legacy of 20th-Century Art, Los Angeles: Getty Conservation Institute, 135–140.

June 21, 2008 10:27 MAC/NWA Page-18 9780230_220478_02_cha01 Sherri Irvin 19

Stecker, R. 2003. ‘The Ontology of Art Interpretation’ in S. Davies and A. C. Sukla (eds), Art and Essence, Westport, CT: Praeger, 177–191. Stock, K. 2005. Review of David Davies’ Art as Performance, Philosophical Quarterly 55: 694–696. Thomasson, A. 2004. ‘The Ontology of Art’ in P. Kivy (ed.), The Blackwell Guide to Aesthetics, Malden, MA: Blackwell, 78–92. Wollheim, R. 1980. Art and Its Objects, Second Edition, Cambridge: Cambridge University Press. Wolterstorff, N. 1980. Works and Worlds of Art, Oxford: Clarendon Press.

June 21, 2008 10:27 MAC/NWA Page-19 9780230_220478_02_cha01 2 New Waves in Musical Ontology Andrew Kania

Since analytic aesthetics began, around 50 years ago, music has perhaps been the art most discussed by philosophers. The reasons for philosophers’ attraction to music as a subject are obscure, but one element is surely that music, as a non-linguistic, non-pictorial, multiple-instance, performance art, raises at least as many questions about expression, ontology, interpretation and value as any other art—questions that often seem more puzzling than those raised by other arts. Musical ontology—the study of the kinds of musical things there are, and the relations that hold between them—has been discussed for as long as any other topic in analytic philosophy of music, placed center-stage by Nelson Goodman’s discussion in Languages of Art. Amie Thomasson has recently pointed out that the number of proposals offered for the ontology of art in general seems rather an embarrassment of riches,1 and this observation certainly holds for musical ontology in particular. The kinds of theories defended about the nature of just Western classical musical works include (1) nominalism—a work is a set of scores and/or performances,2 (2) idealism—a work is a particular, or type of, mental entity,3 (3) eliminativism—there are no musical works,4 (4) action theory—a work is a particular, or type of, action performed by the artist,5 (5) platonism—a work is an eternal abstract object,6 and (6) creationism—a work is a creatable abstract object.7 But the riches are even more embarrassing than this. For in addition to these theories about the fundamental metaphysical nature of the classical musical work (what I call the ‘fundamentalist debate’ below), there are various competing theories about (i) the nature of the relations between works, performances, and recordings, and (ii) theories about the similarities and differences between the ontologies of different musical traditions, both (a) within the Western classical tradition—Can we generalize about the ontology of classical music?8 Did the concept of a musical work arise

June 21, 2008 10:30 MAC/NWA Page-20 9780230_220478_03_cha02 Andrew Kania 21 relatively recently?9—and (b) between the Western classical tradition and other Western traditions, such as rock and jazz,10 and non-Western traditions such as Balinese gamelan.11 Finally, there have been a couple of recent defenses of skepticism about musical ontology—the view that most of these investigations are fruitless because born of a mistaken view of the nature of musical ontology.12 In keeping with the aims of this volume, I will say something about each of the three strands in musical ontology delineated above. First, I will say something about recent work on the fundamental nature of classical works; then I will discuss what I call ‘higher-level’ musical ontology, including comparative musical ontology and issues like the relations between works, performances, and recordings. After each of these engagements with the literature, though, I will address some skeptical views of musical ontology as it has been practiced heretofore.

I. The fundamentalist debate

David Davies has recently presented a thorough defense of an action-theory ontology of art in general.13 Discussion of this theory can be found in Dodd’s ‘Critical Study: Artworks and Generative Performances’ and Kania’s and Stock’s reviews of D. Davies, Art as Performance. A recent anomalous but able defender of a Goodmanian nominalism is Stefano Predelli.14 Ben Caplan and Carl Matheson also seem to be moving in a nominalist direction.15 Nonetheless, realism—the view that musical works are abstract objects— is the most widely accepted view, and the debate between platonism—the view that musical works are uncreatable abstracta—and creationism—that they are creatable abstracta—is still central in the literature. It is thus this debate that I will engage with here. One issue that is sometimes addressed within this debate is what kind of abstractum a musical work is—type, kind, or universal. However, nothing of metaphysical or musical consequence seems to turn on this point.16 More important, if works are acknowledged to be types, say, is the question of what exactly they are types of. When philosophers originally turned their attention to musical ontology in the middle of the twentieth century, though they nominally acknowledged the performance aspect of the Western musical tradition, their attention was not focused on the nature of performance. As a result, and perhaps for understandable reasons of beginning with as simple a model as possible, they tried to explain the nature of works and the relation between a work and its performances on the model of a simple property and its instances.17 This quickly led to the problem that either performances must be note-perfect in order to be of the works they purport to be of, or all works must be identical, since there could be no principled distinction between a perfect performance of one work and a hopelessly misguided performance of another.18

June 21, 2008 10:30 MAC/NWA Page-21 9780230_220478_03_cha02 22 New Waves in Musical Ontology

In the meantime, however, some theorists turned to a closer examination of the nature of musical performance.19 One of the things that came to light during these discussions was that there is an intentional relation between a performance and the composer’s specification of the work the performance is of. A performance can only be of a particular work if the performers intend to follow the instructions that the composer set down as to be followed in performing that work. Spelling out this necessary condition on work performance precisely is not a simple task. Mere wishing cannot count as the right kind of intention, or else I could perform The Well-Tempered Clavier. Questions of intentional content also arise, since I am performing the relevant work when I think I am playing Purcell’s Trumpet Voluntary,butam in fact playing a piece by Jeremiah Clarke (because I mean to follow these instructions, which I mistakenly take to have originated in a compositional act of Purcell’s).20 But what is important for our purposes here is just that some such intention is necessary for a performance to be of a work, that an account of such intentions could be given, and that such an account would allow for the possibility of someone’s meeting the intentional requirement, yet playing some wrong notes.21 For what it means to say that there is such an intentional requirement is in part to say that the type or kind of performance aworkis, is one that is intentionally related to the composer’s compositional act. Such a type of performance is still tokened when the performers mean to follow all the instructions in the score, yet make a few mistakes. Thus, I think the problem of imperfect performances is relatively easily solved once we have an understanding of the relation between a performance and the work it is of.22 This understanding is also helpful in considering the debate between musical platonism and creationism. The heart of this debate is whether a respectable conception of musical works as creatable abstracta is defensible. This can usefully be viewed in terms of an inconsistent triad of propositions:

(A) Musical works are created. (B) Musical works are abstract objects. (C) Abstract objects cannot be created.

Platonists reject (A) on the basis of arguments for (C). Creationists, then, must refute any arguments platonists have for (C), and ideally provide an account of how abstract objects can be created.23 Part of the debate has been a classic war of intuitions, with platonists attempting to soften the appearance of their position by talking up the possibility and value of creativity without creation, and our conception of composers as discoverers, while creationists emphasize the strength and centrality of the creation intuition to our conception of musical works. In my judgment, the outcome of this part of the debate is a small victory for the creationists.24 The arguments in favor of platonism often exhibit ignoratio elenchi—a creationist can accept their conclusions without

June 21, 2008 10:30 MAC/NWA Page-22 9780230_220478_03_cha02 Andrew Kania 23 budging from the position that musical works are created. The arguments in favor of creationism point out intuitions we could preserve only if platonism were false, but those intuitions alone are not strong enough to merit rejection of a substantive argument that platonism is the only defensible metaphysical view of musical works as abstract objects. This brings us to the substantive metaphysical debate. Platonism has recently found a very able defender in Julian Dodd, and it is his arguments for platonism and against creationism that I will now consider.25 While he contributes to the debate over the damage done to our creation intuitions by platonism, his more important contribution is an extended argument in favor of the view that abstract objects cannot be created. He summarizes the argument as follows:

(5) The identity of any type K is determined by the condition a token meets, or would have to meet, in order to be a token of that type. (6) The condition a token meets, or would have to meet, in order to be a token of K is K’s property-associate: being a k So (7) The identity of K is determined by the identity of being a k. So (8) K exists if and only if being a k exists. (9) Being a k is an eternal existent. So (10) K is an eternal existent too.26

A creationist can in principle attack any one of these premises in order to defuse the platonist attack, but the weakest point of the argument is premise (9), and it is on this premise and the reasons Dodd gives in support of it that I will focus. The main support Dodd gives for (9) is what he calls ‘an intuitive theory concerning the existence of properties. The theory in question, simply stated, is that the property being a k exists if and only if it is instantiated now, was instantiated in the past, or will be instantiated in the future.’27 As for its intuitiveness, when it comes to the metaphysics of properties I doubt we have any pre-theoretical intuitions beyond the existence of concrete particulars and their having properties.28 But Dodd also has some arguments in favor of the view. The first is that it steers us between a Scylla and Charybdis of alternative views. The Scylla is

the doctrine of transcendent properties: the view that the question of whether a property exists is utterly independent of the question of whether it is instantiated ...Charybdis, on the other hand, is the idea that properties exist only when instantiated: a view which has properties switching in and out of existence as they come to be, and then cease to be, instantiated.29

June 21, 2008 10:30 MAC/NWA Page-23 9780230_220478_03_cha02 24 New Waves in Musical Ontology

But without further argument, Dodd is open to the charge that these theories of the existence of properties are not a Scylla and Charybdis, but rather a false dichotomy. He, perhaps rightly, sees his theory of property existence as preferable to the two alternatives he considers, but he offers no argument for these three theories’ being the only games in town. A creationist may, for instance, argue in favor of a view that sees certain properties—for instance, those essentially involving contingent beings—as coming into being only when the contingent beings they involve come into being.30 This view would not entail that the existence of such a property is ‘utterly independent of the question of whether it is instantiated’, for it is linked to this question by way of questions about whether it could possibly be instantiated. Nor would it entail that ‘properties exist only when instantiated’. It suggests there may be different criteria for property existence than simply whether a property is, has been, or will be instantiated.31 This brief look at the kind of theory the creationist will need to provide in order to retain works’ creatability helpfully brings out two aspects of Dodd’s theory of properties. One is the principle of instantiation. This is the principle that there are no uninstantiated properties, that is, properties that are not instantiated at any point in time. The other is the principle of eternality.This is the principle that what properties there are exist eternally; properties do not come into or go out of existence. Note that in order to refute Dodd’s argument, it would be enough for the creationist to refute the principle of eternality. Let us look, then, at Dodd’s arguments for this principle. One argument that often seems implicit runs as follows:32

(1) Properties are abstract objects. (2) Abstract objects do not exist in space or time. (3) Causation is spatio-temporal—that is, the relata of the causal relation must exist in space and time. So (4) No properties are caused to exist.

This argument engages with some heavyweight issues in metaphysics, particularly the nature of abstract objects and the nature of causation. In a recent paper, Ben Caplan and Carl Matheson grapple with some of these issues in relation to Dodd’s views on musical works.33 On the nature of abstract objects, they argue that Dodd’s claims that abstract objects are non-spatio-temporal are more or less question-begging, since a creationist will not accept such a view of abstract objects. They claim that the platonist needs ‘another way of cashing out the distinction between abstract and concrete objects, one that is acceptable to [creationists and platonists] alike.’34 But this seems too much to ask of a philosophical opponent. On any charitable interpretation, Dodd does not simply pluck this characterization of abstract objects out of the air to suit his current purposes. It is a widely held metaphysical view that, many

June 21, 2008 10:30 MAC/NWA Page-24 9780230_220478_03_cha02 Andrew Kania 25 argue, is well motivated.35 The burden is on the creationist to provide an alternative conception of abstract objects if that is what he requires. On the issue of causation, Caplan and Matheson point out that one serious contender for a theory of causation posits events as the relata of causal relations, and holds events to be sets of a certain kind. But then abstract objects are the relata of the causal relation, rendering (3) false. This seems a weak argument. There are other theories of causation and events on offer, and Caplan and Matheson do not provide any arguments in favor of the views they describe. Now they seem to be in danger themselves of being accused of picking their metaphysics to suit their conclusion. As Caplan and Matheson are fond of saying, the settling of this issue requires ‘some serious metaphysical work’.36 Without that, it is open to a platonist simply to subscribe to another respectable theory of causation or events that does not result in abstract objects’ being the relata of the causal relation. On the other hand, as I noted above, Dodd’s argument for the principle of eternality is implicit, so it is not clear that the burden of proof lies at the creationist’s doorstep. I will have something to say about the status of these general metaphysical commitments in art-ontological debates near the end of this section. A better criticism Caplan and Matheson offer is that Dodd is inconsistent on whether he accepts premise (2).37 For although he insists that all properties are eternal, he grants that some abstract objects, namely some sets, come into and go out of existence, thus existing temporally. For example,

once the Eiffel Tower was built, the singleton containing the Eiffel Tower thereby came into existence, but the fact that such sets can come in and out of existence does not violate the principle of the causal inertness of abstracta: the causal process in this case involved people and bits of metal, the coming to being of the set being an ontological free lunch.38

Here, Dodd seems to reject (2) that abstract objects do not exist in space or time. Sets with temporally initiated concrete objects as members come into and go out of existence, and thus exist in time. To hold on to his conclusion (4), then, he must replace the general claims in premises (1) and (2) with a more specific claim:

(11) Properties do not exist in space or time. (3) Causation is spatio-temporal—that is, the relata of the causal relation must exist in space and time. So (4) No properties are caused to exist.

This replaces the issue of the nature of abstract objects with the issue of the nature of properties in particular. Dodd subscribes to David Armstrong’s theory of properties.39 This is an immanent, or ‘Aristotelian’, view of universals. It combines the principle

June 21, 2008 10:30 MAC/NWA Page-25 9780230_220478_03_cha02 26 New Waves in Musical Ontology of instantiation, mentioned above, with naturalism, a view that entails that universals exist only in their instances, that is within the spatio-temporal realm.40 This is opposed to a transcendent, or ‘platonic’ view of universals, whereby they exist outside of space and time, and not in their spatio- temporal instances.41 One consequence of such a view, which Dodd does not acknowledge, is that, in the absence of further argument, immanent universals would seem to exist very much in space and time. It seems quite natural, if universals exist only in their instances, to say that they begin to exist when they begin to be instantiated, and cease to exist when they cease to be instantiated. If Dodd wants to resist this conclusion (which appears to commit him to the Charybdis of intermittent properties discussed above), he must explain how universals’ existence when they are not instantiated in space and time is consistent with his naturalism. He may claim that the eternal existence of properties, like the existence of sets, is an ‘ontological free lunch’. But this is not a very satisfying response, since it is of a kind usually available to one’s opponents in some form in any metaphysical debate where its use is tempting.42 If properties can have temporal beginnings, as suggested above, then the creationist has made some headway. If there is a type for every property, and vice versa, and properties have temporal beginnings, then we are part of the way to an explanation of how types can be created. But Dodd might grant the temporal initiation of properties, and hence types, and yet resist their creatability, again on the basis of the nature of causation and abstract objects. For even if some abstract objects are brought into being by spatio-temporal events, as he admits the singleton containing the Eiffel Tower is, and as the above considerations suggest some properties are, Dodd might argue that this ‘being brought into being’ is not creation. That is, Dodd could retreat to the position that musical works are temporally initiated, but that it does not follow that they are caused or, a fortiori,created. Caplan and Matheson offer three ways to respond to this suggestion. The first is that the suggestion is inconsistent with the principle that things that come into existence must be caused to exist. Such a response seems to be an appeal to a version of the principle of sufficient reason. It is thus a weak response, since there are no compelling reasons either apriorior a posteriori to subscribe to this principle.43 The second response is that, on a counterfactual analysis of causation, the coming into being of an abstract object in the way described just is an instance of causation. Had people not acted in such a way as to construct the Eiffel Tower, the singleton containing it would not have come into existence. Therefore, the people who caused the Eiffel Tower to come into existence also caused the singleton containing it to come into existence. This is a weak response for the same reason that Caplan and Matheson’s appeal to the event-based theory of causation is weak. It is open to Dodd simply to appeal to some other theory of causation.

June 21, 2008 10:30 MAC/NWA Page-26 9780230_220478_03_cha02 Andrew Kania 27

But Caplan and Matheson’s third response is stronger, for it allows granting the retreating Dodd his claim that temporal initiation does not imply causation. The response is that even if we grant that the people who built the Eiffel Tower did not strictly cause its singleton to exist, the singleton still came into existence as a result of their actions; even if we grant that Beethoven did not strictly cause his Fifth Symphony to exist, it still would not have existed had he not engaged in his compositional activity. Thus this third response is a version of the second, without the counterfactual theory of causation.

[R]ecall that the creatability requirement is supposed to be motivated by untutored intuitions. Insofar as there is a distinction to be made between causing something to come into existence (in the strict and philosophical sense) and bringing it into existence (in the loose and popular sense), people do not have intuitions about what can, or cannot, be caused to come into existence; rather, they have intuitions about what can, or cannot, be brought into existence.44

If the argument thus far has been sound, then we have seen that properties, and hence types, are creatable in a sense that respects the creation intuition. But that is not enough to show that musical works are created by their composers during the act of composition. For a musical work, if a type, is a type of performance. The complex property that is the property-associate of the type is a property that performances instantiate, or possess. But if, as Dodd’s principle of instantiation implies, properties come into existence when they are first instantiated, then a musical work only comes into existence when it is first performed. And this would make the first performers of a work its creators, rather than the composer. Moreover, my remarks above suggest that properties come into and go out of existence as they are instantiated. This would have the further odd consequences (i) that a musical work does not persist through time, but pops into and out of existence, according to whether and when it is being performed, and (ii) that it is brought back into existence at each performance by whoever is performing it. All of these consequences seem to violate corollaries of the creation intuition. There is a further odd consequence of Dodd’s adherence to the principle of instantiation, one which can be turned into a reductio of his position. By subscribing to the principle of instantiation, he denies the existence of uninstantiated properties.45 But if musical works are types of performance, then works that go unperformed do not have their property-associate instantiated. It follows that neither the property nor its type-associate, the work, exist. But that means that the composer of the work could not have discovered it, composition being discovery for a platonist. Thus, on Dodd’s view, any work that has not, is not, and will not receive a performance has not in fact been composed.

June 21, 2008 10:30 MAC/NWA Page-27 9780230_220478_03_cha02 28 New Waves in Musical Ontology

I, however, am not committed to the principle of instantiation. I will now sketch an alternative metaphysical picture to the one Dodd provides. Its first element is the neglected third option for the mode of existence of abstracta, noted above. We are not compelled to choose between abstracta that exist outside of space and time and those that exist in their instances—firmly within space and time. Another option is that (some) abstracta exist in time, but not in space. This option is often ignored in surveys of the possible meaning of ‘abstract’,46 but not always.47 It is not obvious why temporal, but non-spatial, existence should be considered more problematic than non- spatio-temporal existence.48 Investigating this option might be seen as part of a program Amie Thomasson suggests of broadening the class of entities metaphysicians take interest in:

In short, if, rather than trying to make works of art fit into the off-the-rack categories of familiar metaphysical systems, one attempts to determine the categories that would really be suitable for works of art as we know them through our ordinary beliefs and practices, the payoff may lie not just in a better ontology of art, but in a better metaphysics.49

Subscribing to this view of the mode of existence of musical works (or other abstracta) immediately puts paid to the problems of intermittent existence. For, on this view, the work does not exist in its instances, thus there is no need for it to go out of existence when there are no instances of it around. Its existence in time, on the other hand, allows it to begin to be at a certain point—a key requirement for creatability. The second element of the view is a principle of possible instantiation, intended as an alternative to Dodd’s principle of instantiation. According to this new principle, a property exists at any time that it could be instantiated.50 If, as I suggested earlier, the type of performance a work is includes an intentional link to the compositional act of its composer, then such a performance becomes possible with that act of composition. So the other aforementioned problems with Dodd’s account do not apply to my suggestion—works come into existence with their composition, not performance, and unperformed works are easily individuated. I should say a little about the kind of possibility my principle invokes. There is a sense in which it is possible for there to be a performance of Beethoven’s Ninth Symphony right now. Perhaps no one has in fact done the requisite organizing, rehearsing, and so on, but someone could have. In this same sense of ‘possible’, it is not possible for there to be a performance of Beethoven’s Tenth Symphony right now. No amount of organization or rehearsal will be sufficient, because there is nothing to rehearse, nothing to organize a performance of. Why? Because, of course, Beethoven did not compose a Tenth Symphony. The reason this sense of possibility needs to be distinguished from others is that Beethoven might have composed a Tenth

June 21, 2008 10:30 MAC/NWA Page-28 9780230_220478_03_cha02 Andrew Kania 29

Symphony, had he lived longer, or had different priorities. And since he might have composed such a work, it might have been performed. (There are possible worlds where Beethoven composed a Tenth Symphony and then attended its premiere.)` Thus, it is the sense of ‘possible’ in which it is possible to perform Beethoven’s Ninth Symphony, but is not possible to perform Beethoven’s Tenth Symphony, which is appealed to in my principle of possible instantiation.51 I have articulated an alternative picture to Dodd’s, but I have not argued for it. At best I have shown that it is a coherent alternative. Why, then, should anyone subscribe to my view rather than Dodd’s? Precisely because my view respects the creation intuition. Dodd might respond that an intuition only counts as a reason to subscribe to a theory prima facie,orceteris paribus.He would claim that neither of these conditions is met by the creation intuition. Nobody aware of the debate can claim that we are still evaluating the creatability of musical works at first glance, and Dodd would say that, given his arguments, things other than the creation intuition are no longer equal. We cannot hold on to our creation intuition in the face of his arguments about the uncreatability of abstracta. My response to this begins with recalling my earlier arguments that Dodd’s case depends on controversial general, or fundamental, metaphysical theses—what might be called ‘technical’ points. There are two ways to avoid relying in a question-begging way on such theses. One is to wait for the more fundamental metaphysical disputes to be solved. Another is to solve them yourself. Taking the former path involves giving up musical ontology for the time being (and, if the history of metaphysics is anything to go by, for the foreseeable future). The latter path gives you something to do with the time you used to devote to musical ontology, but does not seem likely to bring the dispute to a close any earlier. In light of the fact that disputes such as those between Dodd and his critics52 are unlikely to be resolved anytime soon, given that their resolution depends on the resolution of more fundamental metaphysical disputes, we are back to the point where things other than the creation intuition are equal. Thus the intuition has force once more—not prima facie force but, perhaps even better, force after reflection.

II. Some skepticism about musical ontology

You might wonder whether the kind of debate engaged in above ought really to be classified as musical ontology—whether its proper place is really in a volume of aesthetics rather than a volume of metaphysics.53 For the issues being discussed seem to be those that are discussed in general metaphysics, which can give the impression that this is no more musical ontology than a metaphysical consideration of artifacts that takes musical instruments as its

June 21, 2008 10:30 MAC/NWA Page-29 9780230_220478_03_cha02 30 New Waves in Musical Ontology central example. One defense that might be given is that artworks, particularly musical works, are among the most metaphysically puzzling things, and thus that they are not mere examples, but centrally important ones for any general metaphysician.54 Anyway, even if this criticism is a good one, it does not object to this kind of metaphysics’ being done, only to its classification as part of the philosophy of music proper. Amie Thomasson has recently defended a limited skepticism about the ontology of art in general, which might be applied to the fundamentalist debate. She argues that in order to ground and re-ground the reference of an art-kind term, such as ‘string quartet’, users of the term must have a background ontological sortal in mind, about which they cannot be mistaken.55 For instance, when Shmarb proclaims ‘This is my sixth string quartet’, pointing to the pile of inked-up manuscript on his desk, if his baptism is to succeed, he must be thinking of a whole, multiply instantiable, sonic item— as opposed to a concrete particular (to which he is also ostensibly pointing), a spatial or temporal part of that concrete particular, the countersubject of the fugue in the fourth movement, and so on. Of course, he does this disambiguation in part by using the term ‘string quartet’, thus appealing to an established art-kind term. The grounders and re-grounders of this kind-term, like the (re-)grounders of any kind-term, face the same problem of disambiguation as Shmarb—the ‘qua problem’. In order for their (re-)grounding to be successful, they must have a more fundamental ontological sortal in mind which determines the identity and persistence conditions of members of the kind. As a result, and this is where Thomasson’s skepticism arises, there are only as many facts about the identity and persistence conditions of string quartets as can be derived from the sortal the (re-)grounders of the term have in mind. In other words, there may be questions about musical works which simply have no answer, since our conception of them is vague and incomplete.56 The creatability of multiple artworks is not one of the questions to which there is no determinate answer, according to Thomasson.57 It is one of the basic ontological components dictated by the sortal, about which we cannot be wrong. Yet, at the same time, she recognizes the importance of what David Davies calls ‘the pragmatic constraint’ on any ontology of art: ‘Artworks must be conceived ontologically in such a way as to accord with those features of our critical and appreciative practice upheld on rational reflection.’58 That is, the only way to discover our shared conception of an art-kind is by extracting its features from our shared artistic practice. The problem, though, is that our artistic practices often seem to point not just to vague or incomplete conceptions of artworks, but to downright contradictory ones, and the revisionary ontological views that Thomasson is at pains to rule out—such as Davies’s, that works are in fact the creative performances of their composers—are attempts to make the best sense of these contradictory practices. Dodd sees his view that musical works cannot be created as motivated in the same way: ‘I agree with Davies that there is no

June 21, 2008 10:30 MAC/NWA Page-30 9780230_220478_03_cha02 Andrew Kania 31 conception of artworks that manages to pass muster metaphysically without compromising some of our pre-theoretical intuitions; our disagreement concerns whether the performance theory does this better than the structuralism he rejects ...’.59 Thomasson might respond by pointing out that it is basic ontological conceptions that take priority here, since it is those that disambiguate and ground the reference of the term being used. Thus, other conceptions, such as the kind of property that might be attributed to a string quartet, should not be taken into account at this stage, as they are by Davies.60 Dodd’s arguments, though, are not of this kind. His strategy, rather, is to argue that there cannot be any such things as are described by our ontological conception of musical works. If he is right about this, it seems that Thomasson would have to conclude that ‘string quartet’ does not refer—that is, there are no quartets, or any musical works for that matter. For she is careful to note throughout her argument that though one cannot be wrong about the basic ontological category of the referent of the term one is grounding, one might be wrong about there being any such referent (as when one points to a feather duster and says ‘that kind of animal’): ‘All grounders are assured of ...is that, if there is any art-kind referred to by the terms they attempt to ground the reference of,ithas the ontological standing they commonly (if tacitly) understand and treat it as having.’61 Nonetheless, given her distaste for radically revisionary ontologies of art, I presume that Thomasson would resist the conclusion that there are no musical works. If it comes to this, Thomasson can, at the very least, put the burden back on the revisionist’s shoulders. For he is beholden to show that his proposed revision is as little revisionary as possible.62 In Dodd’s case this would mean showing that his theory of what a musical work is is the closest coherent competitor to our pre-theoretical conception of a musical work. Another option would be to reject Dodd’s metaphysically fundamentalist style of argument as inappropriate to cultural entities such as musical works. This would require a theory of the nature of cultural entities, and seems to be the direction in which Thomasson is headed. A further alternative would be to revive idealism, which has found few adherents in the analytic tradition. Georges Rey has recently defended a view of linguistic entities, such as sentences, words, morphemes, and phonemes, as ‘intentional inexistents’: ‘There’s no reason to believe in them over and above the reasons one has for believing that there are simply representations with the corresponding content.’63 The details of these views, and the extent of their similarity, I leave as avenues for future research.64 Where do all these leave my contribution to the fundamentalist debate— the suggestion that works exist (i) temporally but not spatially and (ii) when it is possible to instance them? I suspect Dodd and others involved in the debate will be unhappy with the lack of metaphysical rigor about my suggestion. Thomasson might like the basic thesis—that it is not possible to

June 21, 2008 10:30 MAC/NWA Page-31 9780230_220478_03_cha02 32 New Waves in Musical Ontology perform a piece before it has been composed—but is likely to think that the justification for such a claim is our artistic practice, rather than any high- falutin metaphysical argument. I leave it to the reader to decide in exactly which ways the contribution is inadequate.

III. Higher-level musical ontology

There has recently been growing interest among analytic philosophers of music in traditions other than Western classical music. The issues discussed in the literature on the ontology of other traditions, particularly rock and jazz, are quite different from those central to the fundamentalist debate. The fact that there are creatable, multiply instantiable pieces and recordings, and particular performance and playback events that instantiate them is taken more or less for granted. The focus of these debates is rather on the relationships between these things, and the roles they play in musical practices. For instance, Theodore Gracyk65 has argued that recordings are at the center of rock as an art form, and thus that they deserve the honorific ‘work of art’, while the songs they manifest, and live performances of those songs, are secondary. Stephen Davies disagrees, arguing that rock songs, like classical pieces, are works of art, and that they are merely created for a different kind of performance—studio performance.66 There is, rightly, no reference to the fundamentalist debate in this literature. That is because these issues lie at a higher (i.e., less fundamental) metaphysical level than the debate over the creatability of musical works. For instance, the thesis that rock works are recordings for playback, rather than works for performance, is neutral with respect to more fundamental metaphysical theories about the nature of the type–token relation. If the nominalists turn out to be right, then talk of types is simply a convenient way of talking about tokens. If the realists turn out to be right, type-talk is about quite different things from tokens. But which of them is right will not bear on the question of whether rock works are a different kind of type from classical works—for playback rather than performance. The independence of ontological theories at these different levels should come as no more of a surprise than the independence of scientific theories from fundamental ontological theories, or for that matter the independence of music analysis from musical ontology. The true fundamental ontological theory had better make room somewhere for our talk about planets, and indeed the most fundamental ontological theories are neutral with respect to the existence of planets.67 Theories of musical ontology had all better make room for melodies, and the most fundamental of them had better make room for works for performance. As with the first skeptical argument considered above, this is not an argument against engaging in the more fundamental debate, but it does seem clear that these higher-level ontological

June 21, 2008 10:30 MAC/NWA Page-32 9780230_220478_03_cha02 Andrew Kania 33 issues are more closely tied to other issues in musical aesthetics, such as those of interpretation and value, than the fundamental metaphysical issues. Higher-level ontology need not be comparative, however. The debate over the necessary and sufficient conditions on a performance’s being of a particular (classical) work is similarly neutral with respect to the fundamental metaphysical nature of performances and works. Another issue in this same boat is that of the nature of elements of musical works, such as melodies, harmony, rhythm, and so on, so far addressed in detail only by Roger Scruton68 and Stephen Davies.69 The beginnings of a different kind of musical analysis for playback works, such as electronic classical music and rock, raise questions about the nature of additional musical elements, such as the aural space of a recording and timbre, and how they can contribute to a musical whole.70 One meta-ontological conclusion that I draw from this range of topics in musical ontology is that there is no sharp line between philosophical ontology and musicology (broadly construed). Just as there is little sense in distinguishing the more abstract scientific writing about Quantum Theory from the applied philosophy of science on that topic, musical ontology at higher levels shades into musicology.71 On the other hand, though we are unlikely to confuse musicology and particle physics, as we descend to the fundamental ontological levels, despite our talking about quark flavors in one case and sound structures in the other, the issues can be the very same.

IV. Some more skepticism about musical ontology

Aaron Ridley has recently argued that musical ontology is a waste of time.72 The debate he discusses as a representative example of what he is arguing against is that over the necessary and sufficient conditions on a performance’s being of a particular work, a debate I classify as higher-level. Ridley’s attack has three parts. He argues that (i) there are no puzzling questions in musical ontology and, even if there were, nothing would be gained by considering them, since (ii) musical ontology has no consequences for musical practice or value. Despite what musical ontologists claim (iii) the ontological facts about music depend on facts about its value. There is no space to go into the details of Ridley’s arguments here.73 Instead, I will see if anything might be salvaged from Ridley’s conclusions. That is, I will discuss some more plausible theses that might be considered in the same ballpark as Ridley’s hyperbolic claims. His first claim is beyond redemption. Perhaps all those who have engaged in musical ontology have been logic-chopping simpletons, but it seems unlikely. Furthermore, Ridley strangely avoids simply stating the unpuzzling facts of musical ontology, and, despite his claims to the contrary, commits himself implicitly to a range of questionable ontological assumptions. Ridley’s second claim is also unsupportable, at least as a general claim. One cannot evaluate something without evaluating it as a particular kind of thing,

June 21, 2008 10:30 MAC/NWA Page-33 9780230_220478_03_cha02 34 New Waves in Musical Ontology and thus to evaluate it correctly, one must evaluate it as the kind of thing it actually is.74 Musical ontology is the study of the kinds of musical things there are, and thus it is an essential part of a complete theory of musical value.75 As we have seen, however, there are a number of different issues under the umbrella of musical ontology, and it may be that some of these are more relevant to questions of musical value than others. To give a non-musical example: to evaluate my new piece of designer furniture correctly, you must realize it is a coffee table rather than a chaise-longue. But I do not think you need to trouble yourself about whether it is a substance instantiating various universals or a bundle of tropes. Either way, it looks just as good in the living room. Similarly, Ridley’s conclusion may apply to questions about the fundamental nature of musical works (which he barely notes), though it does not apply to the question he focuses on—the nature of the work–performance relation—or other higher-level ontological questions.76 Could Ridley be wrong about the independence of value facts from ontological facts, but be right about the dependence of ontological facts on value facts? That is, could value facts and ontological facts be inter-dependent? In a sense, I think, the answer is yes. Perhaps the most widespread consensus amongst philosophers of art is that we are in the business of making sense of a very complicated, even messy, human cultural practice. Now, if there is some objective universality to this practice, as some argue on evolutionary- psychological grounds,77 we might be able to make some general claims about artistic practices that do not even exist yet, just as linguists can describe some features of any natural human language that has not yet been encountered.78 But there are other questions that cannot be answered without investigating the practice in question, and the higher-level questions of musical ontology are of this kind. Higher-level musical ontology, then, describes certain features of a pre-existing practice, one which has developed as a whole, with ontological, evaluative, interpretive, institutional, and all sorts of other features interacting with each other. Thus, in some sense, the ontology of a specific tradition might be the result to some extent of the evaluative practice in that tradition (and vice versa). But this is a long way from Ridley’s implausible idea that, confronted with a musical performance, first we settle the evaluative questions, and then we go on to answer the ontological questions on that basis.

V. Conclusions

None of the skeptical arguments we have looked at has provided a compelling reason to abandon musical ontology. However, the most promising avenues of research seem to lie outside the fundamentalist debate. First, there is the question of how cultural entities, such as musical works, are different from other kinds of objects, such as so-called natural kinds. Second, there are the higher-order musical ontological issues, such as the different ontologies of

June 21, 2008 10:30 MAC/NWA Page-34 9780230_220478_03_cha02 Andrew Kania 35 various kinds of music, the nature of the elements of various kinds of musical works, and the nature of the relations between different kinds of entities within these various traditions. If anything, these ontological questions seem to imply that there is even more to be discussed in the philosophy of music than has been considered so far.

Acknowledgments

I am grateful to Amie Thomasson and the editors for helpful comments on a draft of this chapter, and to Trinity University for financial support during its writing.

Notes

1. Thomasson, ‘The Ontology of Art and Knowledge’, 221. 2. See Goodman, Languages of Art; Predelli, ‘Against Musical Platonism’ and ‘Goodman and the Wrong Note Paradox’. 3. See Collingwood, Principles of Art;Sartre,The Psychology of Imagination. 4. See Rudner, ‘The Ontological Status of the Esthetic Object’. 5. See Currie, An Ontology of Art;D.Davies,Art as Performance. 6. See Kivy, ‘Platonism in Music’; Dodd, ‘Defending Musical Platonism’. 7. See Levinson, ‘What a Musical Work Is’; Howell, ‘Types, Indicated and Initiated’. The references given here are just illustrative samples. 8. See S. Davies, ‘Ontologies of Musical Works’. 9. See Goehr, The Imaginary Museum. 10. See Gracyk, Rhythm and Noise; Brown, ‘Musical Works’. 11. See S. Davies, Musical Works. 12. See Ridley, ‘Against Musical Ontology’; Thomasson, ‘The Ontology of Art and Knowledge’. 13. See his Art as Performance. 14. See his ‘Against Musical Platonism’, ‘Goodman and the Score’, ‘Goodman and the Wrong Note Paradox’, and ‘Musical Ontology and the Argument from Creation’. 15. See their ‘Can a Musical Work be Created?’ and ‘Defending Musical Perdurantism’. 16. S. Davies, Musical Works, 42. 17. See Goodman, Languages of Art; Wollheim, Art and Its Objects; Wolterstorff, Works and Worlds. 18. See Goodman, Languages of Art, 179–187. 19. For example, Levinson, ‘Evaluating Musical Performance’; S. Davies, ‘Authentic- ity’; Thom, For an Audience; Godlovitch, Musical Performance. 20. For a detailed discussion of these issues, see S. Davies, Musical Works, 163–166. 21. Predelli, ‘Against Musical Platonism’. 22. There is, of course, a lot more to say about what exactly that relation is. See, for example, S. Davies, Musical Works, 151–197. 23. In fact, this triad can be used to frame a debate between platonists, creationists, and nominalists. Nominalists, of course, deny (B) and try to explain away our apparent references to abstract works as references to concrete objects and events. 24. Kania, ‘Pieces of Music’, 61–74. 25. As this chapter goes to press, Dodd’s new book on musical ontology, Works of Music, is being published. He defends there the same view by means of largely the

June 21, 2008 10:30 MAC/NWA Page-35 9780230_220478_03_cha02 36 New Waves in Musical Ontology

same arguments as those I discuss here. The main difference is in his theory of property existence. This affects the argument of a few of the following paragraphs, as I note below. For some consideration of Dodd’s current view, see Kania, ‘Review of J. Dodd’ and ‘Methodology of Musical Ontology’. 26. Dodd, ‘Defending Musical Platonism’, 381–382. 27. Dodd, ‘Musical Works as Eternal Types’, 436. 28. Disturbingly, in Dodd’s new book this theory of property existence has been demoted from ‘intuitive’ to ‘[a] view that I once seconded’, ‘an uncomfortable result’, and ‘hard to motivate’ (Musical Works, 62), while an alternative theory is now touted as ‘highly intuitive’ in Dodd’s view (Musical Works, 60). The new view is that ‘a property F exists at t if and only if there is some time t* such that t* is either before, after, or identical with t, and at which it is (metaphysically) possible for F to be instantiated’ (Musical Works, 61). Though the view he now endorses is closer to that for which I argue below, the shifting sands of Dodd’s intuitions suggest that we should take their rational force with more than a grain of salt. 29. Dodd, ‘Musical Works as Eternal Types, 436. 30. Dodd briefly considers this view in his Musical Works, 63–65. 31. A version of this view—that property existence is tied to the possibility of something’s being the correlative way—is defended in Levinson, ‘Properties and Related Entities’ and ‘Review of D. M. Armstrong’, and appealed to by Dodd in Musical Works, 63. 32. The following formulation of the argument is mine, not Dodd’s. 33. Caplan and Matheson, ‘Can a Musical Work be Created?’ 34. Ibid., 118. 35. See, for example, Lowe, The Possibility of Metaphysics, 210–227, and the references given there. 36. Caplan and Matheson, ‘Can a Musical Work be Created?’, 119 and passim. 37. Ibid., 122–123. 38. Dodd, ‘Defending Musical Platonism’, 397. 39. Dodd, ‘Musical Works as Eternal Types’, 436, n. 18. 40. Thus, Dodd rejects this view in his Musical Works, along with the principle of instantiation, partly for the reasons I give here. His current view, however, while closer to that I defend below, still endorses the principle of eternality. 41. Armstrong, Universals. A neglected third option is that some universals may exist in time, but not in space (the question of their spatial location being a kind of category error). Such universals would not exist in their spatio-temporal instances. I discuss this option further below. 42. Thomasson, ‘Ontological Minimalism’. 43. Mackie, The Miracle of Theism, 82–87. 44. Caplan and Matheson, ‘Can a Musical Work be Created?’, 123. 45. Again, in Musical Works, Dodd rejects this principle, partly for the reasons I give here. 46. For example, Lowe, The Possibility of Metaphysics, 210–216. 47. For example, Rosen, ‘Abstract Objects’. 48. Compare spatial, but non-temporal, existence, which is hard to make sense of. I should also note here that my talk of ‘modes of existence’ is metaphorical. I take existence to be univocal. As Lowe puts it, ‘[t]o exist in space and time is not to have a special kind of existence—for the notion of existence, like that of identity, is univocal. Rather, it is just to have certain sorts of properties and relations— spatiotemporal ones’ (Lowe, The Possibility of Metaphysics, 212). The ‘mode of

June 21, 2008 10:30 MAC/NWA Page-36 9780230_220478_03_cha02 Andrew Kania 37

existence’ I am suggesting we consider is just something’s having temporal, but not spatial, properties. 49. Thomasson, ‘The Ontology of Art’, 90. 50. Dodd’s latest principle of instantiation, noted above, differs from mine in that he takes a property to exist at a given time if and only if there is some time at which it could be instantiated. This is what allows him to hold on to the principle of eternality. The implausibility of this view is suggested by the discussion which follows, but its full elaboration must come elsewhere. 51. This sense of possibility is closely related to the sense in which Saul Kripke suggests there could be no unicorns: ‘no counterfactual situation is properly describable as one in which there would have been unicorns’ (Kripke, Naming and Necessity, 156). No counterfactual situation is properly describable as one in which Beethoven’s Tenth Symphony is being performed. This sense of possibility might also be seen as related to Armstrong’s ‘inner sphere’ of possibility (A World of States of Affairs, 165–169), though that notion is much more technical, and thus less ready to hand for my purposes here. But see also David Lewis’s discussion of ‘alien possibilities’ (On the Plurality of Worlds, passim). 52. For example, Caplan and Matheson, as discussed above, Howell, ‘Types, Indicated and Initiated’, who defends creationism differently from Caplan and Matheson, but by similar reliance on controversial metaphysical theses, and Trivedi, ‘Against Musical Works as Eternal Types’. 53. Aaron Ridley wonders this, though for different reasons. I consider his concerns in Section IV. 54. This kind of defense of the ontology of art is made by Amie Thomasson, ‘The Ontology of Art’ and ‘The Ontology of Art and Knowledge’, though she might not endorse it in defense of the kind of debate engaged in Section I, as we shall soon see. 55. Thomasson, ‘The Ontology of Art and Knowledge’, 222–223. 56. Ibid., 227–228. 57. Ibid., 226–227. 58. D. Davies, Art as Performance, 23. 59. Dodd, ‘Defending Musical Platonism’, 86–87. 60. Of course, the question of what kind of properties a given kind of thing might possess is an ontological one. But perhaps an argument can be made that it is a less basic ontological part of our conception of a work, and thus trumped by more basic parts of that conception, such as creatability. 61. Thomasson, ‘The Ontology of Art and Knowledge’, 227, first emphasis added. 62. Thomasson, ‘Debates about the Ontology of Art’, 251–253. 63. Rey, ‘The International Inexistence of Language’, 250. 64. The similarity between Thomasson’s Rey’s views is intriguing, since Thomas- son has developed a substantial theory of fictional entities that harks back to Roman Ingarden’s idealism (Thomasson, Fiction and Metaphysics; Ingarden, The Literary Work of Art). For some recent thoughts on these issues, see Kania, ‘The Methodology of Musical Ontology’. 65. Gracyk, Rhythm and Noise, 1–36. 66. S. Davies, Musical Works, 30–36. 67. Despite what some of their adherents say. See Silberstein, ‘Reduction, Emergence, and Explanation’; Thomasson, ‘Ontological Minimalism’. 68. Scruton, The Aesthetics of Music, 19–79. 69. S. Davies, Musical Works, 47–71.

June 21, 2008 10:30 MAC/NWA Page-37 9780230_220478_03_cha02 38 New Waves in Musical Ontology

70. For examples of criticism and theory of this sort about rock music, see Daley, ‘Patti Smith’s ‘‘Gloria’’ ’; Zak, The Poetics of Rock, 48–96. 71. Note that this is consistent with my view stated earlier that conclusions at one level are independent of those at another. 72. Ridley, ‘Against Musical Ontology’. 73. For that, see Kania, ‘Piece for the End of Time’. 74. S. Davies, Musical Works, 203–205. 75. There is, of course, a huge literature on this topic, notwithstanding Ridley’s puz- zlement: ‘ ‘‘How, exactly, is a convincing ontological backdrop supposed to lend perspicuity to evaluative questions?’’ No one, so far as I am aware, has actually asked this: certainly no one has given any sort of explicit answer’ (Ridley, ‘Against Musical Ontology’, 210). For a couple of canonical examples, see Walton, ‘Cate- gories of Art’; Danto, The Transfiguration. Walton discusses a specifically musical example in his ‘The Presentation and Portrayal of Sound Patterns’. 76. Jerrold Levinson has argued that if musical works are not the creations of their composers, they are less valuable than we had thought (‘What a Musical Work Is, Again’, 218). If this is correct, even fundamental ontological issues have far- reaching consequences for musical value. 77. See, for example, de Sousa, ‘Is Art an Adaptation?’; Currie, ‘The Representational Revolution’. 78. Of course, they could not have the theory that enables these descriptions without having first investigated a range of natural human languages.

Bibliography

Armstrong, D. M. 1989. Universals: An Opinionated Introduction,Boulder,CO:Westview Press. Armstrong, D. M. 1997. A World of States of Affairs, Cambridge, MA: Cambridge University Press. Brown, L. B. 1996. ‘Musical Works, Improvisation, and the Principle of Continuity’, Journal of Aesthetics and Art Criticism 54: 353–369. Caplan, B. and Matheson, C. 2004. ‘Can a Musical Work be Created?’ British Journal of Aesthetics 44: 113–134. Caplan, B. and Matheson, C. 2006. ‘Defending Musical Perdurantism’, British Journal of Aesthetics 46: 59–69. Collingwood, R. G. 1938. Principles of Art, New York: Oxford University Press. Currie, G. 1989. An Ontology of Art, New York: St. Martin’s Press. Currie, G. 2004. ‘The Representational Revolution’, Journal of Aesthetics and Art Criticism 62: 119–128. Daley, M. 1998. ‘Patti Smith’s ‘‘Gloria’’: Intertextual Play in a Rock Vocal Performance’, Popular Music 16: 235–253. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Davies, D. 2004. Art as Performance, Malden, MA: Blackwell. Davies, S. 1987. ‘Authenticity in Musical Performance’ reprinted in Davies, S. 2003. Themes in the Philosophy of Music, Oxford: Oxford University Press, 81–93. Davies, S. 2001. Musical Works and Performances: A Philosophical Exploration, Oxford: Oxford University Press. Davies, S. 2003. ‘Ontologies of Musical Works’ in Themes in the Philosophy of Music, Oxford: Oxford University Press, 30–46.

June 21, 2008 10:30 MAC/NWA Page-38 9780230_220478_03_cha02 Andrew Kania 39 de Sousa, R. 2004. ‘Is Art an Adaptation? Prospects for an Evolutionary Perspective on Beauty’, Journal of Aesthetics and Art Criticism 62: 109–18. Dodd, J. 2000. ‘Musical Works as Eternal Types’, British Journal of Aesthetics 40: 424–440. Dodd, J. 2002. ‘Defending Musical Platonism’, British Journal of Aesthetics 42: 380–402. Dodd, J. 2005. ‘Critical Study: Artworks and Generative Performances’, British Journal of Aesthetics 45: 69–87. Dodd, J. 2007. Works of Music: An Essay in Ontology. Oxford: Oxford University Press. Godlovitch, S. 1998. Musical Performance: A Philosophical Study, London: Routledge. Goehr, L. 1992. The Imaginary Museum of Musical Works: An Essay in the Philosophy of Music, Oxford: Oxford University Press. Goodman, N. 1968. Languages of Art: An Approach to a Theory of Symbols, Indianapolis: Bobbs-Merrill. Gracyk, T. 1996. Rhythm and Noise: An Aesthetics of Rock, Durham, NC: Duke University Press. Howell, R. 2002. ‘Types, Indicated and Initiated’, British Journal of Aesthetics 42: 105–127. Ingarden, R. 1973. The Literary Work of Art: An Investigation on the Borderlines of Ontology, Logic, and Theory of Literature, with an Appendix on the Functions of Language in the Theater, G. G. Grabowicz (trans.), Evanston, IL: Northwestern University Press. (First published in German in 1931.) Kania, A. 2005. ‘Review of D. Davies, Art as Performance’, Mind 114: 137–141. Kania, A. 2005. ‘Pieces of Music: The Ontology of Classical, Rock, and Jazz Music’, Ph.D. dissertation: University of Maryland, College Park. Kania, A. 2008. ‘Piece for the End of Time: In Defense of Musical Ontology’, British Journal of Aesthetics 48: 65–79. Kania, A. 2008. ‘Review of J. Dodd, Works of Music’, Journal of Aesthetics and Art Criticism 66: 201–203. Kania, A. Forthcoming. ‘The Methodology of Musical Ontology:Descriptivism and its Implications’, British Journal of Aesthetics. Kivy, P. 1983. ‘Platonism in Music: A Kind of Defense’ reprinted in Kivy, P. 1993. The Fine Art of Repetition: Essays in the Philosophy of Music, Cambridge: Cambridge University Press, 35–58. Kripke, S. A. 1980. Naming and Necessity, Cambridge, MA: Harvard University Press. Levinson, J. 1978. ‘Properties and Related Entities’, Philosophy and Phenomenological Research 39: 1–22. Levinson, J. 1980. ‘What a Musical Work Is’, Journal of Philosophy 77: 5–28. Levinson, J. 1987. ‘Evaluating Musical Performance’ reprinted in Levinson, J. 1990. Music, Art, & Metaphysics, Ithaca, NY: Cornell University Press, 376–392. Levinson, J. 1990. ‘What a Musical Work Is, Again’, Music, Art, & Metaphysics, Ithaca, NY: Cornell University Press, 215–263. Levinson, J. 1992. ‘Review of D. M. Armstrong, Universals: An Opinionated Introduction’, Philosophical Review 101: 654–660. Lewis, D. 1986. On the Plurality of Worlds, Oxford: Blackwell. Lowe, E. J. 1999. The Possibility of Metaphysics: Substance, Identity, Time, Oxford: Oxford University Press. Mackie, J. L. 1982. The Miracle of Theism: Arguments for and against the Existence of God, Oxford: Clarendon Press. Predelli, S. 1995. ‘Against Musical Platonism’, British Journal of Aesthetics 35: 338–350. Predelli, S. 1999. ‘Goodman and the Score’, British Journal of Aesthetics 39: 138–147.

June 21, 2008 10:30 MAC/NWA Page-39 9780230_220478_03_cha02 40 New Waves in Musical Ontology

Predelli, S. 1999. ‘Goodman and the Wrong Note Paradox’, British Journal of Aesthetics 39: 364–375. Predelli, S. 2001. ‘Musical Ontology and the Argument from Creation’, British Journal of Aesthetics 41: 279–292. Rey, G. 2006. ‘The Intentional Inexistence of Language – But Not Cars’ in R. Stainton (ed.), Contemporary Debates in Cognitive Science, Oxford: Blackwell, 237–255. Ridley, A. 2003. ‘Against Musical Ontology’, Journal of Philosophy 101: 203–220. Rosen, G. 2001. ‘Abstract Objects’ in E. N. Zalta (ed.), The Stanford Encyclopedia of Phi- losophy, URL = http://plato.stanford.edu/archives/spr2006/entries/abstract-objects/. Rudner, R. 1950. ‘The Ontological Status of the Esthetic Object’, Philosophy and Phenomenological Research 10: 380–388. Sartre, J.-P. 1940. The Psychology of Imagination. Trans. unrecorded, New York: Philosophical Library. Scruton, R. 1999. The Aesthetics of Music, Oxford: Oxford University Press. Silberstein, M. 2002. ‘Reduction, Emergence, and Explanation’ in P. Machamer and M. Silberstein (eds), The Blackwell Guide to the Philosophy of Science, Malden, MA: Blackwell, 80–107. Stock, K. 2005. ‘Review of D. Davies, Art as Performance’, Philosophical Quarterly 55: 694–696. Thom, P. 1993. For an Audience: A Philosophy of the Performing Arts, Philadelphia, PA: Temple University Press. Thomasson, A. 1999. Fiction and Metaphysics, Cambridge: Cambridge University Press. Thomasson, A. 2001. ‘Ontological Minimalism’, American Philosophical Quarterly 38: 319–331. Thomasson, A. 2004. ‘The Ontology of Art’ in P. Kivy (ed.), The Blackwell Guide to Aesthetics, Malden, MA: Blackwell, 78–92. Thomasson, A. 2005. ‘The Ontology of Art and Knowledge in Aesthetics’, Journal of Aesthetics and Art Criticism 62: 221–229. Thomasson, A. 2006. ‘Debates about the Ontology of Art: What Are We Doing Here?’ Philosophy Compass 1: 245–255. Trivedi, S. 2002. ‘Against Musical Works as Eternal Types’, British Journal of Aesthetics 42: 73–82. Walton, K. 1970. ‘Categories of Art’, The Philosophical Review 79: 334–367. Walton, K. 1988. ‘The Presentation and Portrayal of Sound Patterns’ in J. Dancy, J. M. E. Moravcsik, and C. C. W. Taylor (eds), Human Agency: Language, Duty and Value, Stanford, CA: Stanford University Press, 237–257. Wollheim, R. 1980. Art and Its Objects, Second Edition, Cambridge: Cambridge University Press. Wolterstorff, N. 1980. Works and Worlds of Art, Oxford: Clarendon Press. Zak, A. J. III. 2001. The Poetics of Rock: Cutting Tracks, Making Records, Berkeley, CA: California University Press.

June 21, 2008 10:30 MAC/NWA Page-40 9780230_220478_03_cha02 3 Metaphors and Musical Expressiveness Saam Trivedi

I. Introduction

Many passages and indeed entire works of instrumental music without words or an associated story or program are often heard and described, by many musicians as well as laypersons, as expressive of mental states such as emotions, moods, and feelings.1 Thus it is not uncommon to come across musical passages or works being heard and described as sad, happy, anguished, serene, andsoon. However, as something that itself neither possesses mental states nor indeed is alive, music cannot literally be sad or happy. And this has led some to suggest that music is instead only metaphorically sad or happy,2 the alleged figurativeness of musical expressiveness being further explained in terms of notions such as metaphorical exemplification, metaphorical transference, and metaphorical hearing-as. In turn, many writers have exten- sively criticized various metaphor-based or metaphorist accounts of musical expressiveness.3 In this chapter, my main aim is to present several considerations against seeing musical expressiveness as metaphorical. Thus I will not say too much about my own positive resemblance-plus-imagination or imaginationist view of musical expressiveness that I have sketched and defended in three recent articles.4 I also leave aside formalist, arousalist, and resemblance-based approaches to musical expressiveness, all of which I have briefly discussed in the essays just mentioned, indicating therein why I am not persuaded by these positions. Here is a brief outline of this chapter. I begin in Section II by claiming that it is unclear what it means to say that music is metaphorically sad. And I reject the metaphorist theory of Nelson Goodman, whose view is one of the two most prominent accounts of metaphorism, the other being that of Roger Scruton. I then argue in Section III that metaphors have both a literal and a metaphorical meaning, rejecting Donald Davidson’s view that metaphors only have a literal meaning that is usually false,5 and I argue that metaphors

June 21, 2008 10:34 MAC/NWA Page-41 9780230_220478_04_cha03 42 Metaphors and Musical Expressiveness must be paraphraseable. Drawing on these conclusions in Section IV, I offer several considerations against metaphorism. Amongst other things, I argue that it is not clear that ‘The music is sad’ is metaphorical to begin with, for attempted paraphrases of this alleged metaphor fail to capture and explain its emotional content adequately without loss of meaning. In Section V, I switch gears in my argument against metaphorism. Assuming only for argument’s sake that metaphorists keep insisting, despite the arguments presented here, that ‘The music is sad’ is figurative, how does one proceed further? Rather than taking the notion of metaphor for granted as metaphorists do, I suggest one must dig deeper into the very nature of metaphors. Briefly exploring the nature of metaphors, then, my hunch is that metaphors, as used in ordinary language, involve resemblance plus imagination at the very least, and they may also involve a lot more, which I leave to be explored on other occasions, as I do not offer anything like a complete theory of metaphors. I also briefly and critically survey some central theories about metaphors, some of which also appeal to resemblance or imagination or both. In Section VI, I then use the conclusions of Section III (that metaphors have paraphraseable metaphorical meanings) and Section V (that metaphors involve resemblance plus imagination at least) to present more arguments against metaphorism, especially the metaphorism of Scruton. I argue that even if one grants only for argument’s sake that ‘The music is sad’ is a metaphor, a further analysis and explanation of this alleged metaphor via paraphrase shows that metaphorist views of musical expressiveness lead on to resemblance- plus-imagination views of musical expressiveness such as mine, which may thus underlie metaphorist views and be more basic. For if metaphors themselves involve resemblance plus imagination, as suggested in Section V, and if they have a metaphorical meaning that is paraphraseable, as suggested in Section III, then even if one grants only for argument’s sake that ‘The music is sad’ is a metaphor, when you paraphrase the metaphorical meaning of that alleged metaphor, that alleged metaphor itself involves resemblance and imagination and leads onto a resemblance-plus-imagination view of musical expressiveness, and in a way that eliminates the alleged metaphor, contra Scruton’s claim that metaphors are indispensable here.6 I also present other grounds for disagreeing with Scruton, and conclude by explaining why, appearances notwithstanding, I am not a metaphorist.

II. Against metaphorism (Part 1)

Here is a worry for metaphorist views in general. These views all seem to replace the mystery of musical expressiveness with another mystery, that of how music is or can be said to be metaphorically sad or happy or expressive of other mental states. They all need to explain, perhaps via an adequate paraphrase, what exactly it means to say that talk of musical expressiveness is only metaphorical, no matter whether one appeals further to metaphorical

June 21, 2008 10:34 MAC/NWA Page-42 9780230_220478_04_cha03 Saam Trivedi 43 exemplification, metaphorical transference, metaphorical hearing-as, or any other notion that essentially involves metaphors. How does one cash out or explain such allegedly metaphorical descriptions as ‘The music is sad’, given that metaphorists generally take the notion of metaphor for granted and do not offer theories of metaphors? What does it mean to use the word ‘sad’ with its ordinary meaning (rather than a special, metaphorical meaning which would need to be clarified anyway) to apply metaphorically to music, as Roger Scruton, for example, suggests, where it does not literally apply? The allegedly metaphorical use of the word ‘sad’ here must connect somehow to the ordinary use of ‘sad’, and while metaphorist views in general fail to make this connection clear, I will argue later that the connection involves resemblance plus imagination, and leads to an underlying and more basic resemblance-plus-imagination view of musical expressiveness. Perhaps metaphorists, such as Goodman and Scruton, generally think that talk of musical expressiveness is metaphorical because the huge influence of the linguistic turn in twentieth-century analytic philosophy makes them focus on linguistic entities. Relatedly and more importantly, perhaps they think music can only possess sadness metaphorically, because as something without life or mental states, it cannot possess sadness literally, and there can be no third option here. However, I submit that there is a third option: music can be and is imagined in not always highly foregrounded ways to possess sadness, as I have argued at length elsewhere. Pace metaphorists, it is experiences of musical expressiveness rather than linguistic descriptions of them that are primary, and experiences of musical expressiveness involve imagination. Put differently, if you do not see the problem of musical expressiveness in terms of linguistic descriptions, as metaphorists heavily influenced by the linguistic turn do, but instead in terms of experiences of musical expressiveness, then listeners imaginatively hearing music as sad makes perfect sense. Thus one can reject the false claim that music possesses sadness literally, and one can also reject the metaphorist’s unexplicated and mysterious claim that music metaphorically possesses sadness. A brief word now as to why I am not persuaded by Nelson Goodman’s view of expression or expressiveness as metaphorical exemplification.7 Applying this idea to locutions such as ‘The music is sad’ or ‘The music is expressive of sadness’, we get from Goodman that these amount to saying ‘The music metaphorically exemplifies sadness.’ But now the very notion of metaphorical exemplification, as used in this context, cries out for greater explication. Goodman tells us that exemplification is possession plus reference. It seems to follow that ‘The music metaphorically exemplifies sadness’ amounts to (i) ‘The music metaphorically possesses sadness’, plus (ii) ‘The music metaphorically refers to sadness’. Neither (i) nor (ii) gets Goodman very far. As regards (i), while it is clear that music cannot literally possess sadness, given that it is insentient and inanimate, what does it mean to say that music metaphorically possesses

June 21, 2008 10:34 MAC/NWA Page-43 9780230_220478_04_cha03 44 Metaphors and Musical Expressiveness sadness? Absent a clear answer to this question, it seems Goodman has replaced the mystery of musical expressiveness with another mystery, that of the metaphorical possession of sadness, which itself seems to be another metaphor that cries out to be cashed out, perhaps through a paraphrase. As for (ii), one must ask what metaphorical reference amounts to, and whether Goodman has not here given us another mysterious metaphor that needs further elucidation. Moreover, music does not, for the most part, literally refer to things outside itself the way words or gestures refer to things. Thus it seems usually false to say that music literally refers to expressiveness or to emotions, moods, and feelings such as sadness, and it seems mysterious to say music metaphorically refers to these things. Contra Goodman, instead of referring to these things, music is better and indeed usually described as being expressive of sadness and other such mental states. Music is imagined to be expressive, as I have argued elsewhere, and it seems at worst false and at best mysterious for Goodman to say that music metaphorically refers to sadness and other such mental states.

III. Metaphorical meaning and paraphraseability

In order to determine whether metaphors can be paraphrased, we must first determine whether metaphors only have a literal meaning (one that is usually false, except for some metaphors such as ‘No man is an island’, where the literal meaning is trivially true), as claimed by Davidson, or whether they also have a metaphorical meaning. But first, a word about the relevance of metaphorical meaning and paraphraseability to what follows. Amongst other things, I argue that if metaphors have a metaphorical meaning that is paraphraseable, when we try to paraphrase the metaphorical meaning of alleged metaphors such as ‘The music is sad’ we either end up in doubt about whether such locutions are metaphorical in the first place, as I argue in Section IV, or else we end up with resemblance-plus-imagination views of musical expressiveness such as mine which may underlie metaphorist views and thus be more basic, as I argue in Section VI. Davidson is right, I think, if he means that the individual words that constitute a metaphor do not have a special, metaphorical meaning but rather are used in their ordinary sense and in an imaginative way that points to similarities. Perhaps Davidson does mean this because he opposes Beardsley8 who thinks that in the metaphor ‘the spiteful sun’, for instance, there is a semantic opposition or tension, which forces a twist of meaning so that the word ‘spiteful’ gets a new special, metaphorical meaning which it may not have in other contexts. One must wonder against Beardsley here if instead the word ‘spiteful’ is not used in its ordinary sense, though in an imaginative and imagination-involving way, that points to some resemblance between the sun and spiteful people. Thus the metaphor has a metaphorical meaning

June 21, 2008 10:34 MAC/NWA Page-44 9780230_220478_04_cha03 Saam Trivedi 45 that can be paraphrased as saying roughly that the extremely hot sun is discomforting and harmful just as spiteful people are, and in virtue of this resemblance we are asked to and can imagine the sun as spiteful. Besides, Beardsley needs to explain what new meaning the word ‘spiteful’ has in this context, as he claims. However, I disagree with Davidson if he means the metaphorical utterance as a whole—as opposed to its individual words—does not also have a metaphorical meaning in addition to its literal meaning. The problem for Davidson is that he cannot allow that someone might grasp the literal meaning of a metaphor and yet not get the point of the metaphor, that is its metaphorical meaning.9 For in Davidson’s view, there is no such thing as metaphorical meaning that one can fail to grasp until it is revealed to one by correct paraphrase. Consider as an example Schelling’s famous metaphor, ‘Architecture is frozen music’. Suppose a non-native speaker or a Venusian takes it to mean that when you take the scores (whether manuscripts or printed, sheet music) of musical works and put them in deep freeze, the subsequent ice-patterns that form on paper are called ‘architecture’. What is it that our non-native speaker and Venusian have failed to grasp? The natural and perhaps only answer is metaphorical meaning, since they seem to have grasped the literal meaning. In contrast, someone who understands the metaphor correctly grasps its metaphorical meaning, which is the object of her understanding. It may be hard but it is not impossible to capture the metaphorical meaning of Schelling’s metaphor through paraphrase. Here is my best attempt: architecture is like music in that both art forms are architectonic where form or structure matters essentially, but whereas musical structure unfolds over time when a musical work is experienced, music being a temporal art, the structure of architectural works themselves is static, not dynamic, and does not itself unfold over time in our experience of them (even if we experience architectural works bit by bit over time and do not usually take them in all at once), and due to these similarities and differences we can imagine architecture to be music that is ‘frozen’ in time, as it were. There may be more to paraphrase here, since metaphors may be open-ended, but something like this paraphrased metaphorical meaning is, I suggest, the object of understanding for those who grasp the point of Schelling’s metaphor. So far, I have raised one worry for Davidson’s view that metaphors only have a literal meaning not a metaphorical meaning, a concern about meaning that the view that metaphors only have a literal meaning cannot explain adequately what it is that is understood or not understood when a metaphor is understood or otherwise. The point or metaphorical meaning of a metaphor can be grasped as expressed, without loss of meaning, by an adequate paraphrase. In particular, it is through an adequate paraphrase that conveys metaphorical meaning that one can explain the point of a metaphor to someone who has not grasped it.

June 21, 2008 10:34 MAC/NWA Page-45 9780230_220478_04_cha03 46 Metaphors and Musical Expressiveness

Here is a second concern for Davidson’s view, this time about truth. Con- sider as an example the metaphorical expression ‘He is drowning in sorrow’, said of someone who is extremely sad. One possible literal meaning of this, though an obviously false one, is that the subject is sinking in a large body of liquid called ‘sorrow’. Someone who does not know English too well might grasp this literal meaning of the metaphor, and yet not have grasped the point of the metaphor. The point of this metaphor is its metaphorical meaning given by the paraphrase that the subject is extremely sad, so very sad that his sadness can be imagined to be a deep ocean that he is immersed in and unable to come out of. While the literal meaning of this metaphor is obviously false, one must wonder what it is about the metaphor that is true, what the vehicle or bearer of its truth-value is, given that it is said truly of someone who is extremely sad. The natural and perhaps only answer seems to be that it is the proposition that expresses the metaphorical meaning or content that is true. This answer is not available to Davidson, given his denial of metaphorical meaning. The problem seems especially acute because Davidson does allow for metaphorical truth, even though it is not clear what metaphorical truth would attach to on his view. Of course, Davidson might reply that the metaphor as a whole is true without allowing for such a thing as metaphorical meaning. But then the problem for Davidson is that it is not clear what makes the metaphor as a whole true, if not the metaphorical meaning or content of the metaphor and its fit or correspondence with the world. So much for Davidson. In passing, it should be noted here that metaphorical meaning and interpretation are dependent on literal meaning. It may be impossible to grasp metaphorical meaning unless one first grasps the literal meaning of a metaphor. Let us turn now to the issue of metaphors and paraphraseability, where note that by ‘paraphrase’ I only mean clarifying statements. That some metaphors may be hard to paraphrase fully in a way that commands universal assent does not show that it is in principle impossible to paraphrase or otherwise clarify them. The burden of proof is on those such as Max Black10 who claim that metaphors need not be paraphraseable to produce an example of a metaphor that allegedly cannot be paraphrased, which those in my camp would then try their best to paraphrase. As far as I know, Black’s camp has provided no such example, at least not so far. Absent such an example, I urge that metaphors must be paraphraseable if we are to be able to explain their cognitive content to those unfamiliar with them or unable to grasp their cognitive content. Otherwise, it is not clear how we would explain metaphors to those who do not understand them, and linguistic communication would break down at least when it comes to some metaphors. I grant, however, that there is a difference in effect and structure— especially communicative and rhetorical effect, and grammatical structure— between a metaphor such as ‘Architecture is frozen music’ and its paraphrase. Such differences may account for why it is felt by some that something is

June 21, 2008 10:34 MAC/NWA Page-46 9780230_220478_04_cha03 Saam Trivedi 47 lost in paraphrase, namely the ‘metaphorical mode of expression’, as Josef Stern puts it.11 But it is not clear that these differences in effect, structure, or mode of presentation must amount to a difference in cognitive content or informativeness between metaphors and their paraphrases, which is what needs to be shown by those who think metaphors can be unparaphraseable. It is also granted that it may often be hard to say exactly via paraphrase what metaphors in literature and poetry amount to. But the case of metaphors in ordinary language (such as ‘The music is sad’, allegedly) need not be the same, for the two kinds of metaphor need not be sufficiently similar.

IV. Against metaphorism (Part 2)

I now present another set of considerations against metaphorism. Let us assume, only for argument’s sake, that ‘The music is sad’ is indeed a metaphor. How might one explain this metaphor, especially its emotional content or the emotion-terms involved in it? How might one explain what it means to say that music is only metaphorically sad in a way that connects the ordinary use and meaning of ‘sad’ with its allegedly metaphorical application (or transference) to music, given that ‘sad’ cannot literally apply to music, which is inanimate and insentient? It seems the only way to explain metaphorical meaning here would be via paraphrase, assuming metaphors have paraphraseable metaphorical meanings, as argued in Section III. And there would seem to be only three candidates for paraphrase here: (i) in the layperson’s non-technical terms so that to say ‘The music is sad’ is to say ‘The music is slow in tempo, soft in volume, low in pitch, and so on’; (ii) in the technical terms of musical analysis so that to say ‘The music is sad’ is to say that ‘The music is in a minor key, uses diminished chords, and so on’; and (iii) in terms of a combination of (i) and (ii) so that to say ‘The music is sad’ is to say ‘The music is slow in tempo, in a minor key, soft in volume, low in pitch, uses diminished chords, and so on.’ None of these three possible paraphrases adequately captures the emotional content of the alleged metaphor ‘The music is sad’ without loss of meaning. In fact, instead of explaining musical sadness in emotion-terms, all the three paraphrases use non-emotional terms, whether technical or non-technical or both. The failure of these three paraphrases to explain how ‘sad’ applies only metaphorically to music in the sense in which we use that word ordinarily leads to doubts, contra metaphorism, about whether descriptions such as ‘The music is sad’ are metaphorical in the first place. So far at least, no other kind of paraphrase has been offered by metaphorists, and thus one must doubt whether locutions such as ‘The music is sad’ and ‘The music is happy’ are figurative, irrespective of whether one appeals further to metaphorical exemplification, metaphorical transference, metaphorical hearing-as, or some other notion that essentially involves metaphors. Second, contra the three paraphrases suggested above, it is conceivable that a musical passage or piece might be

June 21, 2008 10:34 MAC/NWA Page-47 9780230_220478_04_cha03 48 Metaphors and Musical Expressiveness sad without being slow, low, in a minor key, and so on. And, it is conceivable that a musical passage or piece might be slow, low, in a minor key, and so on, without being sad. For example, the theme in the rondo of Beethoven’s ‘Pathetique’ Sonata, Op. 13, is in a minor key and yet is expressive of a child- like playfulness.12 Third and relatedly, all the three candidates above for the appropriate paraphrase might be thought as giving us only a causal story at best of what, intentionally or otherwise, makes or causes some music to be, and be heard as, sad. They do not, however, tell us what it means, in an allegedly metaphorical sense, for music to be sad, which is what we want to know from an adequate paraphrase of this alleged metaphor. For it is possible that someone (a super-intelligent robot or an extra-terrestrial, say) might grasp that a musical passage or piece is slow, low, in a minor key, and so on, and yet not grasp what it is for the music to be metaphorically sad (or even what it is for it to be sad). Someone in this situation would thus grasp what makes the music sad, without knowing what it is for it to be metaphorically sad, for it remains to be clarified in what sense musical expressiveness and talk of it is metaphorical.13 Fourth, while we generally use metaphors deliberately, knowing fully and consciously that they are usually false when taken literally, in sharp contrast it is not clear that in hearing music and describing it in expressive terms we do so deliberately, knowing consciously that the emotion terms we are applying to music are literally false.14 Rather, I suggest that hearing music as expressive is often almost automatic, as we often animate music, amongst other things, and imagine willy-nilly, readily, and immediately that it is sad or happy, say, without necessarily realizing at the time that this is what we are doing. Fifth and finally, it seems that experiences of musical sadness and those of musical expressiveness in general are much stronger, phenomenologically speaking, than metaphorism suggests. If I am right, then in hearing music as sad we imagine in various not always highly conscious ways that the music is sad so that it is often felt as if the music itself is sad, as if the sadness is in the music, as if the sadness is an imagined property of the music itself, rather than being something that the music only has in a thin, metaphorical sense, whatever that obscure notion ultimately amounts to.

V. Metaphors, resemblance, and imagination

I believe the considerations of Sections II and IV above are decisive against metaphorism. However, metaphorists may not buy these considerations, for whatever reason, and may insist despite these considerations that ‘The music is sad’ is a metaphor. Does that lead us to an impasse? Not necessarily, for I believe one can open up a new front against metaphorism. To proceed further, in this section, let us first dig further into the very nature of metaphors rather than take them for granted as metaphorists usually do. And then let us see what follows if one grants only for argument’s sake that ‘The music

June 21, 2008 10:34 MAC/NWA Page-48 9780230_220478_04_cha03 Saam Trivedi 49 is sad’ is a metaphor. Looking ahead, in Section VI, I will draw on the conclusions of Section III—that metaphors have paraphraseable metaphorical meanings—and this section, and assuming only for the sake of argument that ‘The music is sad’ is a metaphor, I will inquire further into what this alleged metaphor involves, arguing ultimately that it reveals an underlying view of musical expressiveness that is more basic than metaphorism. So, what are metaphors? It is doubtful that this question can be answered in terms of a single, unified theory that applies to poetic as well as to more pro- saic metaphors. For it is not clear that metaphors in their everyday, ordinary language use are sufficiently similar to metaphors in poetry and literature to allow for a single description of all kinds of metaphors.15 Accordingly, I leave literary and poetic metaphors aside in what follows, and focus on the everyday use of metaphors in ordinary language. This narrower focus suffices for my main purpose here, which is to reject views that claim that everyday, ordinary language ascriptions of musical expressiveness such as ‘The music is sad’ and ‘The music is happy’ are only figurative. Though I do not have anything even remotely like a complete theory of metaphors, my hunch is that metaphors are used to point to or suggest (without necessarily asserting, as similes do) previously unnoticed and unexpected resemblances, or indeed lack thereof (often by denying class membership or identity, as in the case of metaphors such as ‘Life is not a bed of roses’, ‘No man is an island’, and ‘You, Dan Quayle, are no Jack Kennedy’), between two (or more) things in an imaginative way that involves the imagination of both those who (first) use a given metaphor as well as the reader or hearer. Our faculty of reason is asked and prescribed through the use of metaphors to imaginatively compare, see, and understand one thing at least partially as or in terms of another, or in some cases, not as or in terms of another. This imagining need not be conscious or even actual, and it is sufficient to be aware that we are being asked to imagine something by the use of a metaphor.16 I disagree with the ‘emotivist’ position that metaphors are merely expressions of emotion without cognitive content, and I also disagree with the view of the ‘decoration theorist’ that metaphors are merely pretty ornamental prose for what can be stated plainly. Instead, I claim that metaphors may serve unique communicative and rhetorical purposes. Because metaphors usually have a different communicative, rhetorical, and grammatical effect from their literal paraphrases or equivalents, they can be and often are used to point to things not yet seen or to help us see things in colourful, attractive, striking, and memorable ways. As a result, I am inclined to think that metaphors have both a cognitive content and an emotive significance, which often manifests itself in the surprise many hearers and readers feel on first encountering some metaphors. I fully grant there may be more to metaphors than resemblance and imagination, and leave it to other occasions to explore these additional elements. For the purpose of arguing against metaphorist views of musical expressiveness, I do not need to provide a complete theory of metaphors.

June 21, 2008 10:34 MAC/NWA Page-49 9780230_220478_04_cha03 50 Metaphors and Musical Expressiveness

I now briefly survey some of the central theories about metaphors, some of which also appeal to resemblance or imagination or both.17 On the analogy theory often said to derive from Aristotle, metaphors involve an analogy and a transfer of name, whereby we give one thing a name that belongs to another thing. I sympathize with the appeal to analogy or resemblance, though I would suggest that, at the very least, one must add imagination to this view due to the way that metaphors point to resemblances through an imaginative use of words. Amongst other things, what makes me wary of construing metaphors only in terms of resemblance is the fact that while resemblance is usually thought to be symmetrical, many metaphors are asymmetrical and irreversible. For example, ‘Man is a wolf’ makes a different point than its ordinary language reversal ‘A wolf is man’, if the latter has a point at all. A second position, the elliptical simile theory, sees metaphors as elliptical or condensed similes, whereby a comparison, which would be made explicit in a simile, is elliptically implied by a metaphor such that the metaphor and the simile assert the same thing. One problem with this view, however, is that it is not clear how a metaphor can make the very same assertion as its corresponding simile. For to say A is B seems to be different, in general, from saying A is like B. A second problem for this view is that metaphors seem to involve a greater appeal to imagination than do similes, as we are asked and prescribed to imagine not just that A is like B but rather imagine something more, to wit, that A is B. For example, the metaphor ‘Man is a wolf’ differs from its corresponding simile ‘Man is like a wolf’. While the simile explicitly points to similarities between men and wolves and asks our imaginations to figure out what the similarities are, the metaphor implicitly does the same but, in addition, I suggest, it asks and prescribes us to imagine, in virtue of the resemblances between men and wolves, that man is a wolf and is not merely like a wolf. In the case of ‘Man is a wolf’, in virtue of various similarities (of which there may be many), we are asked to imagine identity or class membership at least, while in the case of ‘Man is like a wolf’, the imagination is only asked to figure out the relevant similarities (of which there may not be very many), but we are not asked to imagine class membership or identity. A third view, the comparison theory, sees metaphors as conceptually comparing a primary subject to a secondary subject, so that we are asked to imaginatively compare various similarities between two (or more) things as suggested by the words that constitute the metaphor, and the context in which they are spoken or written. According to Garry Hagberg, the metaphorical expression ‘Juliet is the sun’ is fixed by the words and the context to mean something closer to ‘Juliet is necessary for life itself for Romeo’ than to ‘Juliet is a fiery ball of burning gas.’ As may be obvious from what I have briefly indicated above, I have some sympathies for this position too, and think it is not too far from the resemblance-plus-imagination view I favour. Indeed, it has been suggested that this comparison theory is one species of

June 21, 2008 10:34 MAC/NWA Page-50 9780230_220478_04_cha03 Saam Trivedi 51 the analogy or resemblance view discussed above that is said to derive from Aristotle.18 It is sometimes objected against this comparison theory that it assumes that the objects being compared exist, which need not be the case, as with the metaphor, ‘Sally is a dragon’.19 In response, we can say that the existence of the objects being compared need not be assumed but could only be imagined. So, in this example, dragons can be imagined to exist, Sally can be imagined to be like them in being ferocious, and indeed can thus be imagined as a dragon. Alternatively, one can just imagine Sally as a dragon without necessarily imagining first that dragons exist. And imagining existence can take various forms, including illusions such as mirages where we imagine water exists, and imagining Hamlet exists in the sense of entertaining this possibility without actually believing it.

VI. Against metaphorism (Part 3)

Drawing on the claims of Sections III and V, I now want to consider Roger Scruton’s metaphorism. In what follows, I argue that even if one grants only for argument’s sake that ‘The music is sad’ is a metaphor, the alleged metaphor can be eliminated, contra Scruton, via paraphrase in a way that shows that there is a view of musical expressiveness that underlies metaphorism and is more basic. There is a fair bit in Scruton’s theory that is in sync with my own resemblance-plus-imagination or imaginationist position, not on metaphor, but on musical expressiveness, which can be summed up as follows: musical passages and works are only imagined in various not always highly foregrounded or conscious ways to be sad or happy or expressive of other mental states, because they are perceived to resemble various things to do with emotions, moods, feelings, and other mental states such as their vocal and behavioural expression as well as their affective feel.20 I agree with Scruton that in hearing sounds, including musical sounds, we attend to them in a way that involves imaginative perceptions. I am also sympathetic to Scruton’s claim that we hear life and movement in (musical) sounds and situate them in an imagined space, though I would put the point in terms of our animating musical sounds, imaginatively endowing them with and projecting onto them life and life-like properties in not always highly conscious ways, and also imagining life and movement in music in other ways, such as imagining that the music is expressive of our own mental states, or imaginatively identifying with the music, or imagining an indeterminate musical persona in the music, and so on. Finally, I agree with Scruton that we irresistibly see expression where there is none, extending this habit to fish, plants, clouds, trees, landscapes, and the rest of nature, as well as to musical expression, though I would put the point in terms of our animating nature and music. However, I disagree with Scruton when he claims that metaphors are indispensable in some contexts because we are trying to describe how the world

June 21, 2008 10:34 MAC/NWA Page-51 9780230_220478_04_cha03 52 Metaphors and Musical Expressiveness seems from the active imagination’s point of view. By contrast, I claim that we can paraphrase metaphors in a way that involves the imagination, thus showing that metaphors can be dispensable. Scruton also claims that a complex system of metaphor lies in our most basic apprehension of music so that metaphor is ineliminable from descriptions of music; if you take the metaphor away, you cease to describe the experience of music. Through the use of an example below, I suggest that metaphors can be eliminated via paraphrase in a way that shows that there may be a view of musical expressiveness that underlies metaphorism and is more basic, and also captures how listening to music and hearing it as expressive essentially involves the imagination.21 For argument’s sake and only for argument’s sake, let us grant Scruton and other metaphorists that ‘The music is sad’ is a metaphor. Now the question is what this alleged metaphor means, and how the use of ‘sad’ in it connects to our ordinary use of ‘sad’. In other words, we want to know what it means to use the word ‘sad’ with its ordinary meaning to apply metaphorically to music, and how one would explain this alleged metaphor to someone who does not know what it means to say that music is metaphorically sad or expressive of sadness. If metaphors have paraphraseable metaphorical meanings, as suggested in Section III, and if they involve resemblance plus imagination at least, as suggested in Section V, then there is a natural candidate for a paraphrase of this alleged metaphor, one that explains the metaphor and how it connects to our ordinary use of ‘sad’: ‘The music is sad’ means that the musical passage or work in question resembles in its timbre, slow tempo, soft volume, low pitches, and so on, how sad people (and other animals) typically express their sadness vocally and through gestures, bodily movement, and behaviour (e.g., when they cry, walk slowly, talk softly in a low-pitched voice, droop physically, or hang their heads low), or else the music resembles in its sounds (e.g., in how it attempts to resolve its tensions, or attain intermediate and final goals) the affective component or feel of sadness and our mental life in general. In perceiving these resemblances we are asked and prescribed to imagine (in various not always highly conscious ways) that the music is sad, either by animating the music itself, or imagining a musical persona in it, or in some other way. In brief, very roughly, the music resembles something to do with sadness and our ordinary use of ‘sad’, in virtue of which we imagine that the music is sad. The above paraphrase thus gives us something like a resemblance-plus-imagination view of musical expressiveness, such as mine, which I suspect may thus underlie metaphorism and be more basic. Finally, I would add that we may imagine that the music is sad in the sense of fancying or supposing that it is sad, or in the sense of entertaining the possibility that it is sad without actually believing this, or in some other sense. It would seem, then, that metaphorist views lead on to a more basic and underlying resemblance-plus-imagination view of musical expressiveness, such as mine,

June 21, 2008 10:34 MAC/NWA Page-52 9780230_220478_04_cha03 Saam Trivedi 53 and in a way that eliminates alleged metaphors, contra Scruton. For even if one grants only for argument’s sake that ‘The music is sad’ is a metaphor, once we dig deeper into metaphors, we find there is perhaps no way to analyze and explain the metaphor in question and how its use of ‘sad’ connects to its ordinary use other than to paraphrase this metaphor as above, which reveals a resemblance-plus-imagination view of musical expressiveness as underlying metaphorism. Alternatively, a metaphorist might talk, as Scruton sometimes does, of musical expressiveness involving a metaphorical transference of concepts. To this, one must ask what metaphorical transference means, and what allows and underlies the transfer of terms like ‘sad’ from ordinary contexts of use to apply metaphorically (as alleged) to music if not resemblance and imagination in the manner just suggested, in which case we seem to be back to something like an underlying resemblance-plus-imagination view of musical expressiveness which is more basic. I turn now to a crucial passage from Scruton, which he uses partly to motivate his view of musical expressiveness:22

If we say the [musical work] actually possesses the sadness we hear in it, we face the question whether this sadness is the same property as that possessed by a person or another property. It surely cannot be the same property: the sadness of persons is a property that only conscious organisms can possess. But it cannot be another property either, since it is precisely this word—‘sad’—with its normal meaning that we apply to the music ...To say that the word ascribes, in this use, another property is to say ...that it is not used metaphorically but ambiguously ...It follows that the word ‘sad’ attributes to the music ...no property at all.

This passage nicely contrasts my view with Scruton’s. I agree with Scruton that musical sadness is not the same property as that possessed by persons, for music is without life, consciousness, or mental states. But I submit against Scruton that the sadness is an imagined, non-metaphorical (and unambigu- ous) property of the music, one it is only imagined to possess as listeners imaginatively, willy-nilly, readily, and immediately project it onto the music in various not always highly conscious ways due to the various perceived resemblances, as spelt out above, between music and mental states such as sadness and happiness. This imagined musical sadness connects via various perceived resemblances, as spelt out above, to sadness and the normal meaning of ‘sad’, and it is dependent on the music’s structural features, gestures, and the like, either in terms of supervenience, or emergence, or some other relation. If we imagine music as sad based on resemblances and musical sadness is not metaphorical, then what exactly is the property sadness that we imagine the music to have? In brief response, the imagined sadness of music is not a real property or feature of it, given that music is inanimate and insentient, but a feature or characteristic of it imagined variously by us (in terms of

June 21, 2008 10:34 MAC/NWA Page-53 9780230_220478_04_cha03 54 Metaphors and Musical Expressiveness animation, or a musical persona, or imaginative identification, and so on)23 somewhat in the way that we imagine wateriness (or being watery) to be a feature of the desert when we experience a mirage, even though this is not a real property of the desert. Thus I suggest it would not be far from the truth to say we are only ‘hearing things’ or ‘imagining things’ when we hear music as sad. It must be doubted then whether judgments of expressiveness such as ‘The music is sad’ are metaphorical. Thus, we must set metaphorism aside and look at other theories of musical expressiveness. However, though the considerations, especially of Sections II and IV, should make it clear that I am not a metaphorist, Section VI may give the appearance that I am a metaphorist. So I conclude by clarifying briefly why I am not a metaphorist. First, I deny ‘The music is sad’ is a metaphor to begin with. Second, while metaphorists appeal essentially to metaphors and to further metaphorist notions such as metaphorical exemplification and metaphorical transference when talking about musical expressiveness, I do not do so. Instead, my view of musical expressiveness appeals to resemblance and imagination as being basic and essential. Third, unlike metaphorists such as Scruton, I do not believe metaphors are ineliminable when it comes to musical expressiveness, and instead suggest they can be eliminated via adequate paraphrase. Fourth, metaphorists do not dig further into the very nature of metaphors and usually take them for granted, whereas I briefly explore the nature of metaphors in Section V. One final clarification: although it may seem that the considerations of Sections IV and VI are inconsistent, in fact they have an either/or character, and are meant to complement each other; if one does not work, the other should, and vice-versa. Either way, you have a problem for metaphorists: either ‘The music is sad’ is not a metaphor, or eliminating the alleged metaphor via adequate paraphrase gives you an underlying view of musical expressiveness that is more fundamental than metaphorism. A careful look at Section I and especially the start of Section V should reveal in fact that I switch gears in Section V, only supposing that metaphorists do not buy the considerations of Sections II and IV in order to open an alternate front against metaphorism in the remainder of the chapter.

Notes

1. Work on this project has been supported by a grant from the PSC-CUNY Research Foundation, who I duly thank. Thanks also to both editors, especially Katherine Thomson-Jones, for extensive feedback on earlier versions. 2. See, for example, Goodman, Languages of Art;Scruton,The Aesthetics of Music. 3. See, for example, Budd, Music and the Emotions, Chapter 2, and ‘Music and the Communication of Emotion’; S. Davies, Musical Meaning, Chapter 3; Matravers, Art and Emotion,Chapter6;Carroll,Philosophy of Art, Chapter 2.

June 21, 2008 10:34 MAC/NWA Page-54 9780230_220478_04_cha03 Saam Trivedi 55

4. See my ‘Expressiveness as a Property’, ‘The Funerary Sadness’, and ‘Imagination, Music, and the Emotions’. 5. Davidson, ‘What Metaphors Mean’. 6. What is the relation between my resemblance-plus-imagination view of musical expressiveness and my similar hunch about metaphors? First, the two views concern different domains (one deals with the realm of sounds and mental states heard in it, while the other deals with words and language) and are thus independent of each other. Second, while I admit that metaphors may involve more than resemblance and imagination, which I leave to be explored elsewhere, my view of musical expressiveness involves nothing beyond resemblance and imagination as basic and essential. Third, while I admittedly only have a hunch and not a complete theory about metaphors, my view of musical expressiveness is more developed and closer to completion even if not yet complete. 7. Compare Chapter 5 in this volume. 8. Beardsley, ‘The Metaphorical Twist’. 9. Compare Moran, ‘Seeing and Believing’; Levinson, ‘Who’s Afraid of a Paraphrase?’ 10. See Black, ‘Metaphor’ and ‘More about Metaphor’. 11. Stern, Metaphor in Context. 12. Compare Budd, ‘Music and the Communication of Emotion’; Davies, Musical Meaning. 13. Compare Levinson, ‘Musical Expressiveness’. 14. Compare Boghossian, ‘On Hearing the Music in the Sound’. 15. Compare Cohen, ‘Metaphor’; and Levin, ‘Language, Concepts, and Worlds’. 16. As Lakoff and Johnson (Lakoff and Johnson, Metaphors We Live By) memorably put it, metaphors unite reason and imagination and are thus imaginative rationality; though I would urge against them that metaphors are primarily a matter of words, even if they are right that concepts too can be metaphorical and we often think metaphorically. I also sympathize with Kendall Walton’s view (Walton, ‘Metaphor and Prop Oriented Make-Believe’) that metaphors in context imply or suggest or introduce or call to mind possible games of make-believe, though I would broaden talk of make-believe to talk instead of imagination, for make-believing or pretending to oneself is only one kind of imagining and metaphors may also involve other kinds of imaginings such as fancying, or supposing, or entertaining a possibility or a proposition without actually believing or affirming it, and so on. I briefly explore different kinds of imaginings in my ‘The Funerary Sadness’; see also Walton, Mimesis, Chapter 1. 17. See Hagberg, ‘Metaphor’. 18. Andrew McGonigal made this suggestion to me in personal conversation. 19. Searle, ‘Metaphor’. 20. My view in part synthesizes various resemblance-based views of musical expressiveness (e.g., Kivy, Sound Sentiment;S.Davies,Musical Meaning; Budd, Values of Art) with various imagination-based views of musical expressiveness (e.g., Walton, ‘What is Abstract?’ and ‘Listening with Imagination’; Levinson, ‘Musical Expressiveness’; Robinson, Deeper than Reason). 21. I agree with Scruton that a musical work’s expressive qualities are an important part of its content, but disagree that they are always the most important part; and I am puzzled by Scruton’s doubts (The Aesthetics of Music, 344–346) whether current theories of musical expressiveness could pass his four tests (the ‘semaphore test’, the understanding test, the value test, and the structure test), and submit that many of them (e.g., Peter Kivy’s resemblance-based theory, Kendall Walton’s

June 21, 2008 10:34 MAC/NWA Page-55 9780230_220478_04_cha03 56 Metaphors and Musical Expressiveness

make-believe-based theory, Jerrold Levinson’s persona-based theory), including mine, could easily pass these tests. 22. Scruton, The Aesthetics of Music, 154. See also Iseminger’s review; Levinson’s review; and Budd, ‘Aesthetic Realism’. 23. For more on this, see my ‘Expressiveness as a Property’, ‘The Funerary Sadness’, and ‘Imagination, Music, and the Emotions’.

Bibliography

Beardsley, M. 1962. ‘The Metaphorical Twist’, Philosophy and Phenomenological Research, 22: 293–307. Black, M. 1962. ‘Metaphor’, Models and Metaphors, Ithaca: Cornell University Press. Black, M. 1981. ‘More about Metaphor’ in A. Ortony (ed.), Metaphor and Thought, Cambridge: Cambridge University Press. Blackburn, S. 1993. Essays in Quasi-Realism, Oxford: Oxford University Press. Boghossian, P. 2002. ‘On Hearing the Music in the Sound: Scruton on Musical Expression’, Journal of Aesthetics and Art Criticism, 60: 49–55. Budd, M. 1985. Music and the Emotions, London: Routledge. Budd, M. 1989. ‘Music and the Communication of Emotion’, Journal of Aesthetics and Art Criticism, 47: 129–138. Budd, M. 1995. Values of Art: Pictures, Poetry, and Music, Hammondsworth: Penguin. Budd, M. 2005. ‘Aesthetic Realism and Emotional Qualities of Music’, British Journal of Aesthetics, 45: 111–122. Carroll, N. 1999. Philosophy of Art: A Contemporary Introduction, London and New York: Routledge. Cohen, T. 2003. ‘Metaphor’ in J. Levinson (ed.), The Oxford Handbook of Aesthetics, Oxford: Oxford University Press. Davidson, D. 1984. ‘What Metaphors Mean’, Inquiries into Truth and Interpretation, Oxford: Oxford University Press, 245–264. Davies, S. 1994. Musical Meaning and Expression, Ithaca: Cornell University Press. Goodman, N. 1976. Languages of Art, Second Edition, Indianapolis: Hackett. Hagberg, G. 2001. ‘Metaphor’ in B. Gaut and D. McIver Lopes (eds), The Routledge Companion to Aesthetics, New York: Routledge. Iseminger, G. 1999. ‘Review of Scruton’, Journal of Aesthetics and Art Criticism, 57: 375. John, E. and McIver Lopes, D. (eds) 2004. Philosophy of Literature, Oxford: Blackwell. Kivy, P. 1989. Sound Sentiment, Philadelphia: Temple University Press. Lakoff, G. and Johnson, M. 2003. Metaphors We Live By, Chicago: University of Chicago Press. Levin, S. 1993. ‘Language, Concepts, and Worlds’ in A. Ortony (ed.), Metaphor and Thought, Second Edition, Cambridge: Cambridge University Press. Levinson, J. 1996. ‘Musical Expressiveness’, The Pleasures of Aesthetics, Ithaca, NY: Cor- nell University Press. Levinson, J. 2000. ‘Review of Scruton’, Philosophical Review, 109: 605–606. Levinson, J. 2001. ‘Who’s Afraid of a Paraphrase?’ Theoria, 67: 7–23. Matravers, D. 1998. Art and Emotion, Oxford: Clarendon Press. Moran, R. 1989. ‘Seeing and Believing: Metaphor, Image, and Force’, Critical Inquiry, 16: 87–112. Ortony, A. (ed.) 1993. Metaphor and Thought, Second Edition, Cambridge: Cambridge University Press. Robinson, J. 2005. Deeper than Reason, Oxford: Oxford University Press.

June 21, 2008 10:34 MAC/NWA Page-56 9780230_220478_04_cha03 Saam Trivedi 57

Scruton, R. 1999. The Aesthetics of Music, Oxford: Oxford University Press. Searle, J. 1993. ‘Metaphor’ in A. Ortony (ed.), Metaphor and Thought, Second Edition, Cambridge: Cambridge University Press. Stern, J. 2000. Metaphor in Context, Cambridge, MA: MIT Press. Trivedi, S. 2001. ‘Expressiveness as a Property of the Music Itself’, Journal of Aesthetics and Art Criticism, 59: 411–420. Trivedi, S. 2001. ‘An Epistemic Dilemma for Actual Intentionalism’, British Journal of Aesthetics, 41: 192–206. Trivedi, S. 2003. ‘The Funerary Sadness of Mahler’s Music’ in M. Kieran and D. McIver Lopes (eds), Imagination, Philosophy, and the Arts, New York: Routledge, 259–271. Trivedi, S. 2006. ‘Imagination, Music, and the Emotions’, Revue Internationale de Philosophie 60: 415–435. Walton, K. 1988. ‘What is Abstract about the Art of Music?’ Journal of Aesthetics and Art Criticism, 46: 351–364. Walton, K. 1990. Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Walton, K. 1994. ‘Listening with Imagination: Is Music Representational?’ Journal of Aesthetics and Art Criticism, 52: 47–61. Walton, K. 2004. ‘Metaphor, Prop Oriented Make-Believe’, reprinted in E. John and D. McIver Lopes (eds), Philosophy of Literature, Oxford: Blackwell.

June 21, 2008 10:34 MAC/NWA Page-57 9780230_220478_04_cha03 4 Davidson, Metaphor and Error Theory Andrew McGonigal

Davidson’s error theory about metaphorical meaning has rightly commanded a lot of critical attention over the last 25 years or so. Each component of that theory—the case for anti-realism about metaphorical meanings, the diagnosis of the mistakes that led theorists to falsely ascribe such semantic properties to words and sentences, the suggested functional replacement of such talk in terms of the effects that metaphorical utterances bring about—has been examined, reformulated and criticized. The evaluation of the theory has been far from uniformly negative. It is widely recognized, even by realists about metaphorical meaning, that the ‘conventional wis- dom’ about ‘discerning two senses of the predicate term’ that Beardsley had adverted to three years earlier was shown to be misguided by the considerations that Davidson’s paper brought to bear.1 Contemporary recognition of the importance of elucidating the dependence of metaphorical language upon its literal base, and upon its context of utterance, can also be seen to have resulted from sustained critical engagement with Davidson’s article. Nevertheless, Davidson’s position as a whole has not commanded widespread acceptance. An error theory of whatever stripe is typically a revisionary enterprise, so it is to be expected that philosophers have been interested in investigating whether a more conservative position, consistent with the genuine insights of Davidson’s argument, might enable us to preserve our pre-theoretic practice of ascribing truth and meaning to metaphors, qua metaphors. In this chapter, I will outline and evaluate some recent, influential objections to Davidson’s theses. I do not pretend to completeness. My purpose is to present and address a selection of the concerns that the Davidsonian approach regularly provokes, many of which seem to me to admit of halfway decent answers. However, I do think that there is a serious worry about the Davidsonian position that has not been properly formulated or explored to date. I outline this objection in the final section of the chapter, and attempt to answer it there.

June 21, 2008 10:40 MAC/NWA Page-58 9780230_220478_05_cha04 Andrew McGonigal 59

I. Complaints about the ‘Causal Theory’

Davidson’s error theoretic approach fails to provide an acceptable functional replacement for our pre-theoretic talk about the workings of metaphor. His ‘Causal Theory’ suggests that metaphors work by brute causation, ‘nudging us’ into noticing similarities in a way that might just as well be brought about by a pill, or a bump on the head. But that is hopeless. In particular,

1. Such an account ignores the fact that metaphors are produced and interpreted within ‘the space of reasons’. Metaphors can be misinterpreted, supported by evidence, produced for reasons, rejected as unwarranted or false and so on.2 2. Such an account wantonly ignores the fact that we know very little about each other’s cognitive architecture and subjective associations, and thus can hardly be expected to predict the causal effects of producing a given metaphor. 3. Such an account cannot explain why the syntax of metaphors might be important, and thus cannot account for the clear metaphorical difference between, for example, ‘Surgeons are butchers’ and ‘Butchers are surgeons’.3 Such an account also blurs the difference between metaphors and strings of nonsense syllables, or word-salads, which might equally well causally bring about some recognition of similarities and so on.4

Davidson is often presented as endorsing a ‘Causal Theory’ of the way in which metaphor works its wonders, albeit one that is only sketchily developed. This is slightly misleading. In many ways, it would be more accurate to say that Davidson’s scattered remarks on the subject are an attempt to indicate why we should not expect to have any interesting explanatory theory of the workings of metaphor. It is true, however, that, like many error theorists, he does see the need to explain how what is genuinely valuable about our practices can outlast the discovery of the falsity of much of our theorizing about them. It seems to me, however, that the nature of this account, and in particular its relation to normativity, has often been misunderstood. Let us remind ourselves of the error that Davidson takes us to have fallen into in our pre-theoretic thought about metaphor. A misunderstanding of the nature of meaning in general—in particular, a failure to give proper weight to the conceptual connections between word-meaning on the one hand and the compositional, systematic and public nature of language on the other— has led us to make a useless theoretical posit, a cog that turns out to be driven by what it was intended to drive. Getting clearer on meaning and metaphor is supposed to let us see that to the extent that systematic enquiry into the latter is possible, its place is external to linguistic theory properly so-called, being concerned instead with the realm of extra-linguistic goals and perlocutionary effects. To help us get clearer on Davidson’s line of thought here, let us consider two analogous cases. First, take the case of warnings. A warning is something

June 21, 2008 10:40 MAC/NWA Page-59 9780230_220478_05_cha04 60 Davidson, Metaphor and Error Theory that can be achieved without recourse to natural language. For example, one might draw a picture of a bull, and leave it attached to a suitable fence. In that sense, warnings are not a linguistic phenomenon. Nevertheless, it is clear that warnings often are expressed linguistically, and that when they are, they make use of, for example, the syntactic and semantic properties of the language. Consider the case of warning somebody that a bull is in the field, by uttering the sentence ‘there’s a bull in that field’. Such a speech act can clearly be done for reasons, often very good ones in fact. It can be morally and practically evaluated, and one can be held to account for performing it, for example, on the basis of limited or irrelevant evidence. It can be misinterpreted by the unwary; as a dare, for example. One’s purposes in so acting can go unrecognized, remain unfulfilled or be challenged. Nobody thinks that the words ‘bull’ or ‘field’ change their meaning in the utterance, and the implications made by the utterance—that the bull is dangerous, for example—do not take on the status of a special ‘warning meaning’ of the utterance. Consider also the case of speaking pedantically. When one speaks pedantically—‘actually, you haven’t shown that asserting the consequent is invalid, just invalid in first order logic with identity’—one may aim to bring about quite definite effects. For example, one may wish to prompt one’s student into being careful not to confuse logic as a whole with the type of logic we teach at an elementary level. Those aims, again, may be well- or ill-advised, pursued with wilful disregard for pedagogical evidence and so on. Nobody thinks that the words ‘logic’ or ‘identity’ change their meaning in such utterances, and nobody confuses the aim of the utterance with what is conversationally implicated by one who makes it. Nor does anybody think that it is a particularly sensible project to seek a particular linguistic mechanism that would predict and explain how warnings and pedantic reminders about logic work their wonders. (Apart from anything else, the holism of evidence seems to imply that there will simply be too many ways in which we can come to believe that, for example, somebody is warning us of something.) A Davidsonian thinks of metaphor in a related way. In making a metaphor, one typically has some goal or purpose in mind. Often, the point of a metaphorical utterance is to get somebody to see similarities and analogies between two things, or two situations. Often, it is to get somebody to see or think of one thing as another; or to put it less convention- ally, if just as obliquely, to think of one thing through another. This goal is typically extra-linguistic in the sense that it is often something that, in principle, one could achieve without using the natural language. The purpose with which a metaphor is made can be evaluated, rejected or endorsed, and the means chosen to achieve that purpose can be criticized or applauded. People may fail to take up the metaphor in the

June 21, 2008 10:40 MAC/NWA Page-60 9780230_220478_05_cha04 Andrew McGonigal 61 intended sense, or may respond with ridiculously far-fetched comments and elucidations. This does not imply, however, that we should assimilate speaking metaphorically to a distinctive kind of speech act. Speaking metaphorically is most naturally contrasted with speaking literally, and speaking literally is not a distinctive speech act. To say something literally is a way of saying something, not an alternative to saying something. The same is true of making metaphors, although if Davidson is correct, our purpose in making metaphors only involves communicating a particular proposition in relatively atypical cases. Inviting someone to see philosophers as flies trapped in fly-bottles is something that could be done in a number of ways, only some of them linguistic; speaking metaphorically is one effective way to do it. In this respect, speaking metaphorically is more like the practice of speaking pedantically than that of giving warnings. Nevertheless, it ought to be clear that such acts typically take place within the ‘space of reasons’, on any reasonable understanding of that enigmatic phrase. An invitation is something that can be offered for good or bad reasons; its nature and purpose can be mistaken and misconstrued in a variety of familiar ways; it can be based on mistaken beliefs, or comprise an ineffective route to its goal. A Davidsonian simply need not think of metaphors as brute causal prompts, not amenable to intentional explanation, and to think that she must is to misread Davidson’s remarks on the topic. A worry might remain about the case of falsehood. Surely, we can often reject metaphors, deny them and so on, in ways that go beyond the clash of purposes here expressed. I will deal with this in Section II. For the moment, my concern has been to defuse a more general worry; that a causal account of metaphor assimilates it to brute psychological promptings, which do not admit of explication in terms of reasons, and whose effects it is unreasonable to expect a normal producer of metaphor to predict. I take this charge to be misguided.5 It seems to be this misunderstanding of the Davidsonian account that underlies the final objection outlined above.6 Stern claims,

in point of fact, Davidson’s own explanation of how metaphor works does not appeal to more than the separate literal meanings of the individual words in the sentence, ignoring any contribution made by the string syntactically or semantically structured as a sentence. For Davidson, there is no difference between a metaphor and a poem like T. S. Eliot’s ‘The Hippopotamus’: both are ‘devices that alert us to aspects of the world by inviting us to make comparisons.’ But Eliot’s poem works simply by the alternating presentation or display – the brute juxtaposition, as it were – of stanzas or clauses referring to hippopotami and the Church. Likewise,

June 21, 2008 10:40 MAC/NWA Page-61 9780230_220478_05_cha04 62 Davidson, Metaphor and Error Theory

Davidson would have us believe that metaphor works simply by way of the linear sequence of literal meanings of the individual words of the utterance, regardless of its sentential syntax.7

Like many commentaries on Davidson’s position, this is uncharitable and implausible in the extreme. Even if Davidson had said there is ‘no difference’ between the way metaphors work and the way that poems like Eliot’s work—which he did not—he might reasonably be taken to mean, given the immediate context, that they are both things that can alert us to similarities in the world without requiring special meaning-shifts.8 He would not have thereby committed himself to regarding them as equivalent in every respect. In particular, he need not hold that it is simply the ‘brute juxtaposition’ of the terms in the metaphorical sentence that ‘brutely causes’ the perception of certain similarities to spring to our minds. When dealing with the complexity and efficacy of our norm-governed practices, Davidson can easily afford to be far less brutish than that, without giving up on his basic claim: that we have mistakenly assigned to metaphorical meaning what belongs to the realm of perlocutionary effect. If Davidson’s account relied on treating the sentences used in metaphor as unstructured strings of words or sounds, which then served to prompt comparisons between two subjects, then he would be vulnerable to the surgeons/butchers objection, and to the worry that he cannot distinguish the case of metaphor from that of nonsensical strings. But it does not. A David- sonian can perfectly well hold that quite complex syntactic, semantic and pragmatic properties of sentences play a practically essential role in bringing about the intended perlocutionary effects, just as they do in giving warnings or being pedantic.9

II. Metaphor and other linguistic phenomena

Davidson’s error theoretical approach makes a mystery of some incontrovertible linguistic data related to metaphor. This undermines the claim to have provided a suitable functional replacement for ascriptions of metaphorical meaning. In particular,

(1) Error theory about metaphorical meaning cannot account for the behaviour of metaphors which are embedded in non-assertoric contexts.10 (2) Syntactic facts about metaphor demand the positing of metaphorical meanings which are independent of speaker intentions.11 (3) Such an account is committed to treating metaphorical sentences as false when they are taken literally, when in fact many are neither true nor false, but rather semantically anomalous.12

The Frege-Geach problem—the demand for an explanation of how utterances that lack cognitive content can embed in non-assertoric linguistic contexts, such as negated or conditional sentences—has been a deep and

June 21, 2008 10:40 MAC/NWA Page-62 9780230_220478_05_cha04 Andrew McGonigal 63

painful thorn in the side of expressivist accounts of, for example, morality, modality and aesthetics. Stipulate that metaphors do not express distinctively metaphorical propositional contents, whether semantically or pragmatically construed. The question then becomes how we should understand linguistic contexts like

(M1) It’s just not the case that, as the Wittgensteinian quietists seem to think, philosophy is just getting flies out of fly bottles. (M2) If the garden was a slum of bloom, then it could hardly have been winter that Stevens was writing about in ‘Banal Sojourn’.

It is clear, first, that Davidson can avoid the immediate Frege-Geach problem faced by expressivists. For Davidsonian error theorists metaphorical utterances are typically straightforwardly false, and thus truth-valued, when taken literally. They thus possess the resources to give a straightforwardly truth- conditional account of non-assertoric contexts. Unlike, for example, the (crude) expressivist about moral sentences, the Davidsonian can appeal to the literal content of metaphors to explain why they can embed in more complex sentences at all. The Davidsonian is not quite off the hook, however. The deeper challenge is to give an account of what is going on when we react to metaphors by apparently negating them, or making conditional inferences from them or reporting beliefs concerning them. The realist about metaphorical meaning has a relatively straightforward account. For the ‘semantic’ realist, the metaphorical sentence comes to express a new proposition, distinct from the literal, which can then be negated, conditionalized upon, featured in belief reports and so on. For the ‘pragmatic’ realist, the implicated proposition is available to play a similar role. Assume that someone asserts

(a) Susan has two children thereby implicating (b) Susan has exactly two children The following seem perfectly normal responses to (a) (c) No, she has three children (d) Well if she has two children, we’ll only need two presents. (e) You may believe that she has two children, but I think you’re wrong—I’m sure she has three.

In each case, however, it is the implicated proposition, and not the proposition that is literally expressed, negated, conditionalized and so on. (Or at least, if there is a more general problem here about how merely implicated propositions can apparently be negated and so on, the speaker-meaning

June 21, 2008 10:40 MAC/NWA Page-63 9780230_220478_05_cha04 64 Davidson, Metaphor and Error Theory theory of metaphor has only one problem to solve. The Davidsonian apparently has two: how negation can govern implied rather than expressed propositions, and how negation can govern metaphors when no proposition is implicated.) The problem is not one that the Davidsonian can duck. Responses like (M1) and (M2) are completely central and everyday aspects of our dealings with metaphor. Any attempt to write them off as the result of an unfortunate mythology of meaning would entail that the error theoretic position was much more radically revisionary of our practice than Davidson suggests or makes the case for. Such a view would require significantly more in the way of detailed argument than has been offered to date. Let us return to the case of warnings. If, in response to your warning about the bull in the field, I reply with any of the following—

(i) If the bull’s in the field, we had better walk around the long way. (ii) It might be in the field, or it might be penned in there at the back. (iii) Yes, I had thought that it would be.

—the force of your utterance is not preserved in the new contexts. Similarly, although, on the Davidsonian account, Wittgenstein and Stevens may have put forward their metaphors as a means of encouraging us to see the end of philosophy as escape, or flowers as slum-dwellers, the illocutionary force of such speech acts is presumably not preserved in (M1) and (M2).Sowhatis going on in those cases? The problem is a difficult one, and it is not an issue that I have resolved to my own satisfaction. But there is perhaps room for the Davidsonian to take something like the following line. Although the illocutionary force of a warning, or an invitation, might not survive into non-asserted contexts, pedantry and literality certainly can.

(iv) John hasn’t submitted a list of his research publications, but rather a list of his publications along with those of his articles that have been accepted for publication but have not yet appeared. (v) If John hasn’t submitted a list of his research publications, but rather a list of his publications along with those of his articles that have been accepted for publication but have not yet appeared, then we must point it out to him. (vi) John—and I’m speaking literally here—researches the sex life of pot plants. (vii) She had a sudden insight into university life when she realized that John—and I’m speaking literally here—researches the sex life of pot plants.

June 21, 2008 10:40 MAC/NWA Page-64 9780230_220478_05_cha04 Andrew McGonigal 65

Illocutionary force might not hold up well under embedding, but ways of speaking certainly do. In a similar vein, I do not think that it is unreasonable to construe, for example:

(M1) It’s just not the case that philosophy just aims at getting flies out of fly bottles. (M2) If the garden was a slum of bloom, then it could hardly have been winter that Stevens was writing about in ‘Banal Sojourn’. as

(M3) It’s just not the case that philosophy just aims at—to put it metaphorically—getting flies out of fly bottles. (M4) If the garden was indeed—as in Stevens’ metaphor—a slum of bloom, then it could hardly have been winter that Stevens was writing about in ‘Banal Sojourn’. How does this help?

The basic problem, remember, was not to account for how metaphorical utterances can embed in logical and propositional attitude contexts, but to give an account of what is going on when we do so. The first step that might be made here is to hold that such a treatment of (M1)and (M2), taking them to be saying implicitly what their counterpart utterances say explicitly, makes space for thinking of their component sentences as governed by some kind of modifying operator, qualifier or marker.13 Our mutual recognition that such an element was, in some sense, modifying the component sentences might help to explain why we react to such utterances in a fairly structured and predictable way. The problem now is to see how appeal to such a modifier can avoid surrendering the error theoretic character of Davidson’s account. To help identify one potential such course, I suggest we return to David- son’s positive characterization of metaphorical practice, slender though that is,

if I show you Wittgenstein’s duck-rabbit and I say, ‘It’s a duck’, then with luck you see it as a duck; if I say, ‘It’s a rabbit’, you see it as a rabbit. But no proposition expresses what I have led you to see ...Seeing as is not seeing that. Metaphor makes us see one thing as another by making some literal statement that inspires or prompts the insight.14

Let us set aside bizarrely literalistic interpretations of this passage that take Davidson to be proposing an account of metaphor which relies wholly on the phenomenon of aspect perception.15 Davidson’s thought is clearly that, via the making of some wholly literal statement, we are brought to some sort of imaginative insight into a situation. The metaphor presents us with a kind

June 21, 2008 10:40 MAC/NWA Page-65 9780230_220478_05_cha04 66 Davidson, Metaphor and Error Theory of lens or prism through which we think of one object, event or situation in the light of another. It is difficult to spell out what such a thought amounts to, although it seems fairly clear that such a process (i) is not intrinsically linguistic, (ii) is often passive, in the sense that a given metaphoric point of view can force itself upon us, as when I am suddenly struck by the sheer childishness of a colleague, and (iii) often draws on the rich resources of the imagination, in both propositional and experiential ways. Perhaps despite its vagueness, we might appeal to such a conception of the effects metaphor can have upon us, opening up the possibility that we might think of embedded sentential contexts as somehow relating not to the content of the metaphor, but to the content of such an imaginative engagement. We might take the qualifying marker —‘metaphorically speaking’—as a pragmatic cue that such a shift in interpretive approach is required.16 A preliminary and crude attempt at this might be to take

(F) Metaphorical utterance SM (in part) metaphorically-means* that P iff according to an apt act of imaginative insight I that SM , in context, itself prompts or suggests, P.

Consider an example. Wittgenstein’s metaphor encourages me to imaginatively construe philosophers as flies trapped in fly-bottles. Such an activity might, inter alia, lead me to think of philosophical problems as essentially dissoluble puzzles that prevent us living in a fulfilling manner. Such a course of thought is, arguably, mandated by the metaphor itself in context, given Wittgenstein’s view of philosophical endeavour. So (F) warrants my claim that the famous metaphor metaphorically-means* that philosophical problems as essentially dissoluble puzzles that prevent us living in a fulfilling manner. But that claim is consistent with my holding that, strictly speaking, there are no metaphorical meanings. It is true that expressions will become associated with metaphorical-meanings*, but that is simply a useful fiction; a convenient way of relating metaphors to the content of the imaginative acts they prompt. Nothing deserves to be called a metaphorical meaning, since nothing is suitably meaning-like, and suitably related to the metaphorical character of the relevant utterance. But we might mock up a simulacrum of our talk about metaphorical meaning, projecting a certain kind of perlocutionary effect of the metaphorical utterance back onto the metaphor itself. In this way, we can hope to avoid the charge that we confuse what the metaphor makes us notice with its meaning, while hoping to explain the inferential linkages that embedded contexts reflect.17 What is it for a proposition to hold according to an apt act of imaginative insight? This is hard to spell out, although I hope the intuitive idea is clear enough. Basically, it is for that proposition to comprise part of the content of the imagining; the way the imagining presents things in general as being. This intuitive characterization allows for representations drawing on

June 21, 2008 10:40 MAC/NWA Page-66 9780230_220478_05_cha04 Andrew McGonigal 67 expressive resources from both the imagined world and the real world to enter into the propositional content of the imagining. Thus, we can say both that Wittgenstein characterizes philosophical talk as being a useless, frustrated, buzzing, and that he thinks that it is not a worthwhile intellectual pursuit. The former type of propositions will often be presented in sentences that are themselves seemingly metaphorical in character. Those who believe in the eventual dispensability of metaphorical modes of speech will be able to hold that the content of the imagining will eventually be able to be fleshed out in wholly literal terms. Those who do not may be reassured that the ‘nonreductive’ element of Davidson’s account is preserved by such a fictionalist addendum.18 Given such a translation from the imagined world, the Davidsonian might try to develop a suitable semantics. I shall not try to outline such a semantics in detail here, but rather simply continue to sketch how one might begin on such a task. Imagine that we divide the token sentences of our language up into literal sentences like

(1) Juliet is the sun and metaphorical sentences like (2) Juliet—metaphorically speaking—is the sun.

Metaphorical sentences are not a syntactic kind, obviously. Rather, they contain expressions that are correctly recognizable as having been uttered metaphorically. Sentences like (1) are typically straightforwardly false, whereas metaphorical sentences can be said (in an extended, fictional sense) to be true, when things are as a related apt act of imaginative insight I suggests they are. In more detail,

(F) Metaphorical sentence SM (in part) metaphorically-means* that P iff according to an apt act of imaginative insight I that SM , in context, itself prompts or suggests, P. Now let

*[SM , P1-Pn] abbreviate

S: SM metaphorically-means* that P1,&SM metaphorically-means* that

P2,&...SM metaphorically-means* that Pn Such a notation simply lets us conjoin the various particular propositions that (F) allows us to ascribe to the metaphor. Then we have

(1) SM is true with respect to Pi iff*[SM , Pi]&Pi

(2) SM is true simpliciter iff SM is true with respect to each member Pj of some

suitably weighted subset of {P1,P2, ...,Pn} and then the standard semantics for non-atomic sentences

June 21, 2008 10:40 MAC/NWA Page-67 9780230_220478_05_cha04 68 Davidson, Metaphor and Error Theory

(3) ‘A & B’ is true iff A is true and B is true (4) ‘not-A’ is true iff A is not true. (5) ‘If A then B’ is true iff not-A is true or B is true. and so on. Such a semantics clearly does not commit the Davidsonian to realism about metaphorical meanings, since we are distinguishing them from the merely fictional metaphorical-meanings*. However, it seems that such a semantics can help the error theorist set out a realist-style semantics for embedded contexts in a straightforward way. Take, for example, (M1) It’s just not the case that philosophy just aims at getting flies out of fly bottles. Contextual clues help us identify this as equivalent to (M3) It’s just not the case that philosophy just aims at—to put it metaphorically—getting flies out of fly bottles.

This has the form not-SN , where

SN = ‘Philosophy just aims at—to put it metaphorically—getting flies out of fly bottles’. Now we know that

Metaphorical sentence SM (in part) metaphorically-means* that P iff according to an apt act of imaginative insight I (that SM , in context, itself suggests), P.

Imagine that the relevant act of imaginative insight IN has as its content: philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner. Then by semantic axiom (1) we have

SN is true iff (according to IN, philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner), & (philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner)

So by employing semantic axiom (4),wecanderive not-SN is true iff it’s not the case that [(according to IN, philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner), & (philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner)]

June 21, 2008 10:40 MAC/NWA Page-68 9780230_220478_05_cha04 Andrew McGonigal 69

Since, by hypothesis, the first conjunct of the right-hand-side is true, the above biconditional is equivalent to not-SN is true iff It’s not the case that (philosophy’s proper aim is to dissolve persistent and frustratingly complex conceptual problems that impede our progress through life, and prevent our living in a fulfilling manner) and since

SN =‘Philosophy just aims at—to put it metaphorically—getting flies out of fly bottles’ we have more or less the desired result. The metaphorical sentence is correctly rejected just in case things are not the way the imaginative viewpoint they prompt presents them as being. Similarly, from

(M2) If the garden was a slum of bloom, then it could hardly have been winter that Stevens was writing about in ‘Banal Sojourn’ we move to

(M4) If the garden was indeed—as in Stevens’ metaphor—a slum of bloom, then it could hardly have been winter that Stevens was writing about in ‘Banal Sojourn’ which has the form If SP then P, where the obvious abbreviations are made.

Axiom (5) tells us that this is true iff not-SP is true or P is true. And this, modulo the standard worries about the paradoxes of the material conditional, is exactly what we would expect. The conditional comes out as false when things are as the metaphor prompts us to see them, and yet the interpretation presented in the consequent does not fit with them. So we have some explanation of the use of the conditional, and its place in our talk and thought about metaphor. In rejecting metaphors, we reject the imagined view of the world that they invite. In extending them via conditional thought, we implicitly move within such a worldview, and when we interpret them, we examine the actual world through the intellectual and experiential lens that they provide. It seems then, that in tackling metaphors which occur in, for example, negated contexts, mixed conditionals, belief-attributions and so on, the Davidsonian has some room to manoeuvre.19 Perhaps the above account, sketchy and provisional as it is, will ultimately prove to be unsatisfactory. Nevertheless, the general approach should surely seem attractive. The great appeal of such a fictionalist strategy is that it seems to provide the error theorist with a means of appropriating realist-style semantics, while rejecting realist ontology.20 Such an approach makes it far easier to provide functionally equivalent replacements for rejected discourses than would

June 21, 2008 10:40 MAC/NWA Page-69 9780230_220478_05_cha04 70 Davidson, Metaphor and Error Theory otherwise be the case. Of course, success is far from guaranteed, and must be earned, via the provision of a plausible translation scheme in and out of the fiction. A fully defensible version of fictionalism about metaphorical meaning can scarcely be attempted here. Nevertheless, I hope to have indicated one promising strategy for the Davidsonian, and given the beginnings of an outline of how such an account might go.21 What then, about the charge that Davidson cannot give a plausible account of clearly anomalous sentences? Philosophers like White and Stern, who are in general sympathetic to Davidson’s position, or at least willing to accept that he is right about words retaining their usual literal meaning in metaphor, resist the idea that the sentences composed out of those words need have any literal meaning at all. Such philosophers are willing to give up the idea that a syntactically well-formed sentence composed of significant parts must also be significant, insisting instead that metaphors like ‘they donated his face to the wildlife fund’ are semantically anomalous, or nonsensical. Let W be a thing that it is supposedly nonsensical to imagine donated to a wildlife fund—my face, say—and let sentence S be an instance of the form ‘W is donated to the wildlife fund’. Then it seems that Davidson can run through instances of the following schematic argument:

(1) It is apriorithat, for all x, if x is not the kind of thing that can be donated to wildlife funds, then it is not the case that x is donated to the wildlife fund. (Assumption) (2) It is apriorithat W is not the kind of thing that can be donated to wildlife funds. (Assumption) (3) So, it is apriorithat it is not the case that W is donated to the wildlife fund. (From 1 and 2, Universal Quantifier rules and Modus ponens) (4) It is not the case that W is donated to the wildlife fund. (From 3, the factivity of the a priori) (5) Line 4 follows from aprioritrue statements by truth-preserving steps, and is thus true. (A prioricity of 1, 2) (6) Line 5 is the sentential negation of S (Definition of S) (7) For all sentences X,ifX is the sentential negation of Y,andY is true, then X is false. (Definition of classical negation) (8) S is false (from 5, 6, 7, &-I, Universal Quantifier rules and Modus ponens) (9) If S is false, then it is not nonsensical. (Assumption) (10) S is not nonsensical. (From 8, 9, Modus ponens)

Thus it seems that Davidson has a halfway plausible argument that proceeds from relatively uncontroversial steps, from plausible or shared assumptions, to the conclusion that metaphorical sentences are false when taken literally.22

June 21, 2008 10:40 MAC/NWA Page-70 9780230_220478_05_cha04 Andrew McGonigal 71

III. Metaphor and reduction

(1) Davidson’s objection to simile theories of metaphor illegitimately characterizes such theories as committed to the idea that metaphors be reduced to explicit literal similes.23 (2) Davidson’s account illegitimately demands that a linguistic theory ought to issue specifications of content in only literal terms. 24

Davidson makes a series of objections to simile theories of metaphor. Famously, he argues that everything is like everything else, so if metaphors were literal comparisons, they would be trivially true, instead of interestingly significant, as the supporters of metaphorical meanings hold. Recently, Fogelin has argued that this reply overlooks the obvious; the simile theorist should identify the meaning of the metaphor with the figurative meaning of the corresponding simile. Fogelin sketches an account of the difference between literal similitude and its figurative counterpart in terms of varying standards of salience, ‘modes of relevance and evaluation governing the likeness claim’.25 The reply masks a deeper worry, which is the sheer underdevelopment of the simile theory, whether literal or figurative. The theory does not demonstrate how each metaphor can be parsed into a corresponding simile. For example, we have not been shown how to turn:

(a) The garden was a slum of bloom

into simile form. Perhaps, ‘The garden was like a slum of bloom’? But what’s a slum of bloom? It seems that we have to understand the metaphor in order to make the comparison, so rephrasing as a comparison scarcely casts light upon the metaphor. Similarly with, for example,

(b) The Dean gloated into the office. (c) The Dean did SOMETHING into the office that was like gloating.

Grammar and insightful analysis tend to come under strain, if we take Fogelin’s approach seriously. A more interesting question is whether Davidson implicitly appeals to reductivist presuppositions that can be challenged even by non-simile theorists. Perhaps the idea that a systematic linguistic pragmatics requires the possibility of literal paraphrase can be challenged. If I understand her correctly, that is what Alison Denham aims to do in her Metaphor and Moral Experience and elsewhere.26 There, Denham is interested in what she terms the conceptual autonomy of the metaphorical:

(DEN) Metaphorical utterances are conceptually autonomous if and only if they are not reducible to literal paraphrase, and do not admit of non- circular analysis in non-figurative terms.

June 21, 2008 10:40 MAC/NWA Page-71 9780230_220478_05_cha04 72 Davidson, Metaphor and Error Theory

The category of metaphors Denham is concerned with are those she terms phenomenological metaphors. These, she holds, have the following properties:

(a) They function catachrestically, in that they introduce a new use for an old concept, to fill a lexical gap. (b) They ‘take on a representational role which would otherwise be fulfilled by a response-dependent concept, i.e., a role which would be filled by a response-dependent concept were one available in the speaker’s lexicon’.27 (c) They are minimally truth-apt. (d) They are conceptually autonomous, in the sense expressed by (DEN).

Moreover, given that she holds that a metaphor implicitly proposes some relation of similarity to obtain between its topic and vehicle, she suggests

(e) If any metaphors are of this kind, then their truth or assertibility conditions will consist in the obtaining of some relation of similarity that (i) is capable of yielding determinate truth assignments, (ii) cannot be expressed in non-figurative terms and (iii) is a property that would otherwise be identified by a response-dependent concept. Denham introduces the notion of austere similarity, to describe such a relation, and argues that it is best thought of as being constituted by the responses of suitably situated subjects. She draws an analogy with the perceptual case. Defenders of response-dependent accounts of colour hold that the similarity between two shades of, for example, pink is not reducible to any ‘deeper’ metaphysical feature that the two shades have in common, but consists merely in the fact that epistemically well-placed subjects judge that they look the same. In the same way, argues Denham, the relations of austere similarity expressed by phenomenological metaphors do not pick out a set of common features that we might in principle enumerate. Rather, such relations obtain in virtue of the fact that certain external states of affairs, picked out by the vehicle of the metaphor, strike us as brutely similar to certain of our experiential states. We may not be capable, even in principle, of setting out any exhaustive list of common features ‘underwriting’ such judgements, but only of appealing to a limited and provisional set of seemingly relevant features. In extreme cases, we may even be debarred from doing much more than recording the proposed similarity in the very same terms as the metaphor, merely making explicit that the relation is one of similarity by inserting the requisite ‘like’ or ‘as’. Even if the reader is able to proffer some non-circular elucidation, that elucidation may fail to be given in wholly non-figurative terms—it may be that his best prospects lie in offering further metaphors and further figurative similarity claims as elaborations of the target one.28

June 21, 2008 10:40 MAC/NWA Page-72 9780230_220478_05_cha04 Andrew McGonigal 73

Denham’s theory thus bears some resemblance to Fogelin’s theory as outlined above.29 Fogelin’s is a non-reductive simile theory, since it does not attempt to assimilate metaphors and similes to cases of literal comparison.30 Denham’s theory of phenomenological metaphors is doubly non-reductive, however. Not only does she claim that the simile obtained from the metaphorical utterance is an (implicit) non-literal comparison, but she casts doubt on the idea that either explicit or implicit figurative comparisons can be fully spelled out in literal terms. Fogelin takes similarity to consist in a shared set of salient features, and differentiates between literal and non-literal comparisons by postulating a contextually variable parameter for salience. Denham, in contrast, holds that the facts about similarity that constitute the metaphor’s truth condition need not hold in virtue of any shared set of features, except, trivially, their striking us as similar. To summarize, Denham seems committed to the truth of the following schema: (P1) Phenomenological metaphor q means that phenomenological state X is like Z where Z is a placeholder for some actual or possible type of state of affairs picked out by the vehicle of the metaphor q, and where the relevant concept of likeness is one that can be fully captured via the following (aprioriand necessary) provisional equation: (P2) For all subjects S in ideal epistemic circumstances, (Phenomenological state X is like state of affairs Z iff S judges that X is like Z) The truth conditions for the metaphor q can thus be given as follows (P3) q is true iff for some X and Z, q means that phenomenological state X is like Z, and all subjects S in ideal epistemic circumstances would judge that X is like Z. Since the truth conditions essentially include a type of state of affairs picked out via a figurative comparison (‘X is like Z’) phenomenological metaphors are conceptually autonomous in the Denham sense. Is an analysis of this kind adequate? As we can see, Denham’s pre- ferred account of the truth conditions for phenomenological metaphors involves appeal to ideal subjects. The judgements of such subjects constitute the facts about whether a given experiential state is primitively similar to the state of affairs picked out by the vehicle of the metaphor. Now, there are well-known problems with explicating metaphorical meaning in terms of likeness, some deriving from the fact that αs can be like βs in irrelevant respects, and some from the fact that αs can fail to be like βs and yet be falsely believed to be so. For example, Wittgenstein is like Iggy Pop in virtue, inter alia, of being human, but the metaphor ‘Wittgenstein was a real Iggy Pop’ is not thereby appropriate. Similarly, gorillas are shy,

June 21, 2008 10:40 MAC/NWA Page-73 9780230_220478_05_cha04 74 Davidson, Metaphor and Error Theory retiring creatures, but ‘The bouncer was a gorilla’ does not convey that he was shy and retiring. If Denham is to avoid these problems carrying over to her account, she needs some principled way of excluding the possibility that her supposedly ideal subjects might make their judgements as a result of false or irrelevant beliefs.31 It might seem that Denham’s appeal to austere similarities would circum- vent this problem. After all, that appeal was supposed to allow Denham to reject the idea that whether a given metaphor is true ultimately depends on the obtaining of some set of similarities which might, in principle, be stated in literal terms. However, it is important to be clear that this move does not remove the need for an account of the ideal subject that explains why such subjects would not have false or irrelevant beliefs. Denham agrees with Wright that any judgement-dependent account of some putative realm of facts must give a substantive account of the nature of the subjects and conditions that allow constitutive judgements to be made.32 Denham is, of course, aware of the old worries about similitude. She writes,

As Goodman and Davidson are fond of pointing out, all similarity statements are trivially true: they cannot fail to be true, since everything is in some respects similar to everything else ...If I were concerned with absolute truth conditions attaching to sentences in null contexts this would be problematic, for it would suggest that the truth conditions of metaphors (viewed as implicit similes) were hopelessly underdetermined. However, the level of meaning at which I have located the determination of truth conditions is context relative (Grice’s utterance type occasion meaning), and so that worry need not detain us: we can rely on our knowledge of the background discourse and situational context in which the metaphor occurs to constrain the range of possible similarities to a class of relevant ones.33

The idea seems to be, then, that the ideal subjects whose judgements constitute the facts about whether X austerely resembles Z can be apriori guaranteed not to do so on irrelevant grounds, since their background linguistic knowledge will always constrain the range of similarities to relevant ones. Let us represent such putative background knowledge as I, and the linguistic knowledge that entitles us to constrain the ‘range of possible similarities’ as a conditional of the form I → F, where F stands for the class of similarities relevant to a given utterance of, say q. It seems that now an incoherence emerges in Denham’s theory. For if a given phenomenological metaphor q is to be conceptually autonomous in the sense of being irreducible to literal paraphrase, then F must include either the corresponding simile s or some other irreducible figurative expressions f1-fn. Otherwise, we would have identified some set of similarities which could provide a literal paraphrase of q,

June 21, 2008 10:40 MAC/NWA Page-74 9780230_220478_05_cha04 Andrew McGonigal 75 namely the very set specified in the consequent of the relevant conditional I → F. The class of similarities F, if it is not to provide exactly the type of literal paraphrase that Denham aims to avoid, must therefore itself include figurative elements. Now, however, we are back where we started. For the problem of how we narrowed down the set of possible similarities, excluding irrelevant ones in a principled manner, was supposed to have been solved by citing our mastery of such conditionals. Since it seems that, in order to master such conditionals, we must already have identified the relevant set—given that the conditionals involve either the very figurative expression in question or some other for which the original problem arises—it seems that we have merely deferred the hard question: why those similarities? Moreover, this requirement seems inconsistent with one of the defining characteristics of phenomenological metaphors, namely the fact they represent an instance of catachresis, plugging a lexical gap by introducing a new use for an old word or phrase. Denham writes,

Phenomenological metaphors are used to represent phenomenally characterized states of affairs and so used because (to the knowledge of informed and competent speakers) no equally suitable word or expression occurs in the language ...they are one way of attempting to conceptualize phenomenology in contexts in which the character of an experience or its objects is too fine-grained or too idiosyncratic to be represented by our standard repertoire of linguistic concepts—our standard lexicon.34

But if, ex hypothesi, a phenomenological metaphor is being introduced to make up for the lack of a linguistic concept in our standard repertoire, how could we have already mastered conditionals like I → F,whichuse either the very concept that was supposed to be being introduced or some other, for which analogous problems arise? To summarize, Denham’s theory seems committed to both

(a) construing speaker competence with a given language L as involving grasp of a conditional which includes some irreducibly metaphorical element q and (b) holding that the metaphor q can be introduced into a language for the first time, ‘to fill a lexical gap’, by means of some linguistic rule that only draws on the existing conceptual resources related to mastery of the language L.

But the two theses seem inconsistent. Since it seems that q (or some other figure for which a parallel problem arises) must be a part of the pre-existing

June 21, 2008 10:40 MAC/NWA Page-75 9780230_220478_05_cha04 76 Davidson, Metaphor and Error Theory linguistic rule that supposedly enables us to master its truth-conditions, (a) and (b) are mutually exclusive. I conclude that we have, so far, no reason to believe that phenomenological metaphors are both cognitive and conceptually autonomous, on either of the resolutions of that notion. So Denham’s attempt to elaborate a theory of metaphor that allows metaphors both to express distinctive cognitive contents and to be resistant to literal paraphrase is, I believe, unsuccessful. Of course, rejecting an instance of a theory-type is clearly not to undermine the approach in general. But the Davidsonian can only address rival theories as they are developed. My purpose here has simply been to show that the idea that a satisfactory linguistic pragmatics can be given in non-literal terms is not straightforward, and has yet to be set out in the kind of coherent detail that might pose a serious threat to the error theoretic approach.

IV. Wright’s argument against error theories

The theoretical parsimony and prima facie defensibility of an error theoretic account places a significant burden of proof on Davidson’s realist rivals. However, such a position still faces an important challenge. Crispin Wright has presented a dilemma for error theoretic positions in general, an argument that, if good, would successfully serve to undermine the motivation for Davidson’s position. I conclude by outlining Wright’s challenge, and arguing that distinctive features of the debate about metaphorical meaning allow the Davidsonian to answer it in a satisfactory way. Wright is a minimalist and a pluralist about truth. His truth minimalism consists of two claims, relating to (1) truth and (2) truth aptitude. The former amounts to the claim that it is sufficient for a given predicate T to meet the conceptual requirements on truth that it accords with a series of platitudes. A richly metaphysical or epistemic account of truth is not secured merely by reflective appeal to such uncontroversial principles. Minimalism about truth aptitude comprises two sub-claims: (a) that possession of truth aptitude is co- extensional with possession of assertoric content and (b) that it suffices for an utterance or discourse to be assertoric that it meets constraints regarding (i) syntax and (ii) discipline. Roughly, if a sentence embeds appropriately under negation and so on, and there are publicly shared standards which determine when it is correct and incorrect to employ it, then the sentence possesses assertoric content. Finally, Wright’s pluralism results from his view that although truth is conceptually minimal, it may be metaphysically robust, in a variety of kinds and degrees, across different subject areas. Truth can be defined in minimal terms; but what falls under or realizes that definition in particular cases may be metaphysically substantial.35 Wright argues that such a conception of truth makes life difficult, if not impossible, for error theorists. Discussing the question of whether scope for error theory is simply closed off by a minimal theory of truth, he writes,

June 21, 2008 10:40 MAC/NWA Page-76 9780230_220478_05_cha04 Andrew McGonigal 77

The position, I think, is that error-theoretic proposals remain theoretically feasible but that their development is interestingly constrained. Suppose a philosopher denies that anything or much of what we say within a given discourse is true but grants that its truth predicate admits interpretation as superassertibility by its standards of assertoric warrant.36 Then he has to produce reason to deny that anything, or very much, of what we say is superassertible by the lights of those standards. And that will be a commitment to denying that any, or very many, of the statements of the discourse are even assertible by those standards.

Wright goes on to identify two possible strategies for the error theorist who aims to show precisely this:

One is to contend that while the standards in question are perfectly coherent, little of what we accept really complies with them; our ‘error’, on this view, will consist in a propensity to apply those standards erroneously. The other is to argue that the discourse is not governed by coherent standards of assertoric warrant, and consequently that nothing is genuinely warranted by those standards ...The great question, it seems to me, is not the coherence of error theory but its motivation; why insist on construing the discourse in terms of a notion of truth which has us in massive error when the alternative of superassertibility is prima facie available and would avoid the charge?37

Wright’s question is an important one, and pressing in the present instance. Davidson’s theory of metaphor apparently represents everyday talk and thought as being in massive error. When we call metaphors true or false, or talk about what they say, or signify or mean, we are completely mistaken. How can this be the most charitable interpretation of our practice? Isn’t the minimalist account—that represents us as being answerable to standards of warrant which are internal to the practice of making/interpreting metaphors, and which can, therefore, secure the defensibility of our standard ways of thinking about metaphorical meaning—by far the more plausible position? This seems especially worrying in light of the ‘mocked-up’ account of truth and falsity that I sketched on behalf of the Davidsonian in Section II. If some such account is available, then why not simply identify the norms governing our attributions of metaphorical meaning with the norms of the alleged fiction? The error theorist seems compelled to explain why he takes the standards of correct use that govern our way with metaphors as having been set so high that we almost always fail to accord with them, when he himself is committed to the existence of a much lower standard, whose demands we manage regularly to respect. This challenge to the Davidsonian about metaphor seems to me not to have been properly formulated or pressed

June 21, 2008 10:40 MAC/NWA Page-77 9780230_220478_05_cha04 78 Davidson, Metaphor and Error Theory to date, and seemingly provides a serious challenge to such an error theoretic approach. The idea that minimalist theories of metaphorical meaning have a certain default status has been defended in the literature, notably by Denham.38 She writes,

How does my account of phenomenological metaphor answer the objection that irreducible metaphors do not express truth-apt judgements? Can it show that phenomenological metaphors are held accountable to sufficiently stable standards of correctness in the absence of any non-circular analysis in literal terms? I aim to do that somewhat indirectly, by showing how we might assess whether or not a subject understands a given metaphor; I take it that if we have to hand standards or norms which are sufficiently disciplined to determine attributions of understanding, we have also standards or norms which are sufficiently disciplined to confer the status ‘truth-apt’.39

I actually agree with many of the suggestions Denham makes regarding the criteria for ‘manifesting understanding’ of a metaphor: the capacity to elaborate a metaphor using a mixture of literal commentary and further figuration; the fact that one is able to pick out paradigm instances of objects, events and so on that would fall under particular metaphorical characterizations; the ability to extend the metaphor in the natural way—all these suggestions seem eminently sensible. So it may seem that Wright’s minimalist challenge to the error theorist can be reinforced by the development of a set of plausible proposals describing the nature of the more modest standards the minimalist appeals to. However, such an argument is not, I think, nearly as attractive in the case at hand as it is in, for example, that of moral, aesthetic or mathematical talk. Consider an analogy. It ought to be uncontroversial that error theory about phlogiston, or the humours, is the appropriate position to take. When scientists posited the existence of such substances, it was hoped that they would thereby play a genuinely explanatory role in accounting for certain natural phenomena. When it turned out that such explanatory work could not be delivered in a way that was consistent with the other commitments of natural science, the original motivation for believing in such entities simply lapsed. The good standing of assertions about phlogiston depended upon their being sanctioned by the existing, reflectively endorsed norms governing scientific practice and discussion. It would have been wholly inappropriate to think of retreating to a less demanding norm of correctness for, for example, phlogiston assertions. The point of the standards that govern scientific discourse as a whole would simply have been undermined by such a move. The same, the Davidsonian should maintain, is true of the case of metaphorical meaning. Better-founded, more fundamental commitments of

June 21, 2008 10:40 MAC/NWA Page-78 9780230_220478_05_cha04 Andrew McGonigal 79 our thought about meaning count against our taking talk of metaphorical meaning seriously. It is no doubt arguable that talk about meaning is less robustly objective than theoretical science, although this is scarcely uncontroversial. To assign dialectical significance to this would be to mistake the point of the analogy with outdated scientific theory, however. The point is rather that whatever the realist standing of the constitutive norms governing a given discourse,itmust always be a live possibility that assertions which appeared to be warranted by such norms might turn out not to be. So much is written into the notion of a genuine norm or standard. A standard that could not, even in principle, fail to be lived up to could provide no conceivable influence on action or belief, and thus would hardly be playing a normative role at all. One cannot sincerely intend to bring one’s actions into accord with something that will sanction them no matter what. Susceptibility to error is built into the notion of respecting a norm. So even if semantic discourse is accountable to merely minimal standards of correctness—which is itself, of course, hotly debatable—there ought still to be an open possibility that we have mistakenly taken certain of our judgements to be warranted by those standards, when in fact they ought to have been rejected. That, in effect, is the contention of the position I have been recommend- ing on behalf of the Davidsonian. Such a position does not accept that discourse about the meaning of metaphors is independent of semantic discourse more widely construed. Rather, it holds that constitutive elements of that wider discourse about meaning—requirements of compositionality, explanatory efficacy, accord with the goals of a pragmatic theory and so on—provide the resources for a demonstration that no suitable account of metaphorical meaning is genuinely forthcoming even by the potentially minimal lights of discourse about meaning and communication. Just as in the scientific cases, principles and data that we count as better grounded, or at least more fundamental, are taken to overthrow the apparent warrant for committing to a given theoretical entity. In a sense, then, the Davidsonian error theorist takes the first of the two dialectical strategies that Wright offers:

to contend that while the standards in question are perfectly coherent, little of what we accept really complies with them; our ‘error’, on this view, will consist in a propensity to apply those standards erroneously.

However, this has to be understood in a qualified manner. The error theorist will typically concede that much of semantic discourse is well founded in the light of its proper standards, rejecting only those parts of it that mistakenly extended to the story of ‘how metaphor works its wonders’. The intended lesson of the phlogiston example was that the debate regarding the existence of metaphorical meaning is a debate which is internal to semantics.

June 21, 2008 10:40 MAC/NWA Page-79 9780230_220478_05_cha04 80 Davidson, Metaphor and Error Theory

Theorists like Denham, who aim to give a ‘self-standing’ minimalist account of the metaphor, simply fail to give sufficient regard to the internal connections between superficially ‘different’ discourses. Metaphorical meaning, if it lives at all, lives in the shadow of meaning proper, and inconsistency with key commitments of that grounding discourse spells doom for realism about metaphorical content. Mere surface discipline is insufficient in this kind of case.40

V. Conclusion

I have attempted to set out and evaluate several important objections to a broadly Davidsonian error theory about metaphorical meaning. My discussion here has been, of necessity, incomplete and provisional. I hope, however, to have offered some reason to think that an error theoretic account might deserve to be taken more seriously than it currently is. Whether such a position can ultimately be made defensible is something that I take no stand on here.

Notes

1. See Beardsley, ‘Metaphor and Falsity’. 2. See Lycan, Philosophy of Language, 212; Nogales, Metaphorically Speaking, 121. 3. Stern, Metaphor in Context, 47–48; Nogales, Metaphorically Speaking, 118. 4. Lycan, Philosophy of Language, 211. 5. This misreading of Davidson seems to have been fostered by the construal of his account offered by Richard Rorty. See in particular Rorty, ‘Unfamiliar Noises’, 167, where he compares metaphors with ‘anomalous non-linguistic phenomena like platypuses and pulsars’ that can prompt new insights in a wholly unpredictable manner. Much of what Rorty says in that article is strictly correct, if expressed in a way which is likely to mislead, but some of it seems to me to be straightforwardly incorrect. For example, ‘live metaphors can justify belief only in the same metaphorical sense in which one may ‘‘justify’’ a belief not by citing another belief, but by using a non-sentence to stimulate one’s interlocutor’s sense organs’ (169). This is about as plausible as holding that my recognition of the fact that somebody is warning me that a bull is in the field cannot help justify my belief that a dangerous bull is in the field. 6. Or at least, if it is not this, the objection seems to me to be wholly unmotivated. 7. Stern, Metaphor in Context, 47–48. 8. Davidson, ‘What Metaphors Mean’ (256) comments on the poem as follows. ‘Here we are neither told that the Church resembles a hippopotamus (as in simile) or bullied into making this comparison (as in metaphor), but there can be no doubt the words are being used to direct our attention to similarities between the two. Nor should there be much inclination, in this case, to posit figurative meanings, for in what words or sentences would we lodge them?’ 9. The above quotation from Stern immediately follows discussion of an objection of White’s and Margalit’s to the effect that metaphors taken literally are often nonsensical. Stern may think that the suggested feature of Davidson’s account

June 21, 2008 10:40 MAC/NWA Page-80 9780230_220478_05_cha04 Andrew McGonigal 81

enables him to sidestep this latter objection, but if so, it is clearly a poisoned chalice. I discuss the White/Margalit objection below. 10. Cohen, ‘The Semantics of Metaphor’, passim; Stern, Metaphor in Context, 69–71; Moran, ‘Metaphor’, 260. 11. Stern, Metaphor in Context, 69. 12. White, The Structure of Metaphor, 204–226; Stern, Metaphor in Context, 47; Margalit and Goldblum, ‘Metaphors in an Open-Class Test’, 234–237. 13. The Mthat operator of Stern, Metaphor in Context, would provide one such marker. However, my own inclination is to resist Stern’s claim that such an operator must be psychologically real, or syntactically represented at some deep sentential level. It seems possible that some theory of ‘asides’ will allow such modifiers to guide interpretation, without thereby becoming genuine parts of the string. They could rather be thought of as metalinguistic self-interruptions, intended to ease interpretation. 14. Davidson, ‘What Metaphors Mean’, 263. 15. Kemp, ‘Metaphor and Aspect Perception’, rightfully takes such theories to task. 16. Similarly, we might take ‘I’m guessing here’ to defeat the normal presupposition of informed inference in ‘If John comes then—I’m guessing here—Jane will leave’. 17. This move may seem very non-Davidsonian in spirit. After all, doesn’t David- son explicitly distance himself from such a suggestion? He says, remember, ‘If what the metaphor makes us notice were finite in scope and propositional in nature ...we would simply project the content the metaphor brought to mind onto the metaphor’ (Davidson, ‘What Metaphors Mean’, 263). So it might well seem that he takes the kind of fictionalist projection suggested above to be ruled out by the limitless, non-propositional character of metaphor. My concern here is less Davidson exegesis than an examination of whether an error theorist who rejects metaphorical meaning when speaking strictu sensu can nevertheless give some structured account of embedding. But it is worth noting that the following remark can be taken in a way that accords fairly well with the suggested strategy: ‘Once we understand a metaphor we can call what we grasp the ‘‘metaphorical truth’’ and (up to a point) say what the ‘‘metaphorical meaning’’ is. But simply to lodge this meaning in metaphor is like explaining why a pill puts you to sleep by saying it has a dormative power.’ (Davidson, ‘What Metaphors Mean’, 247). This suggests that we do grasp something when we understand a metaphor, but that this is the effect, rather than the explanation, of the metaphor’s success. 18. Why not take the metaphors to be direct expressions of such imaginative states, rather than the more roundabout route canvassed above? While there is no doubt room for an expressivist or quasi-realist development of such a theory, I will not explore it here. It is worth noting that one prima facie advantage of the current proposal over such an account is that the above account does not require that anybody had such thoughts prior to reflection on the metaphor itself, whereas the expressivist account would tend to see the metaphor as in some sense the result of some prior imaginative state. But there is no doubt room for skirmishing, and for compromise solutions. My primary concern here is with the standing of an error theoretic account in the light of worries about embedding. 19. Such an approach might also be employed to deal with an interesting problem recently raised by Stern regarding the resistance of metaphors to certain kinds of natural language conjunctions. I do not have space to discuss this here. See Stern, Metaphor in Context, 70–71, for the problem.

June 21, 2008 10:40 MAC/NWA Page-81 9780230_220478_05_cha04 82 Davidson, Metaphor and Error Theory

20. I do not mean to suggest that the semantics sketched above works exactly like a standard fictionalism, say in modality or morality. The point is merely that the introduction of pseudo-meaning promises the best strategy for a Davidsonian who wishes to give some account of embedded contexts. 21. I do not pretend originality in linking metaphorical content and imaginative construal. The above account has obvious affinities to ideas set out in, for example, Walton, Mimesis as Make-Believe, ‘Metaphor and Prop-Oriented Make-Believe’; and Yablo ‘Does Ontology Rest on a Mistake?’. However, Yablo takes his account to underwrite a realism about metaphorical content, rather than the type of pseudo- fictionalism mooted here. 22. Thanks are due to Gary Kemp for the argument presented here (inter alia!). 23. Fogelin, Figuratively Speaking, Chapter 2. 24. Denham, Metaphor and Moral Experience, 259. See 318–328 for an extended discussion. 25. Fogelin, Figuratively Speaking, 75–76. 26. See Denham, ‘Metaphor and Judgements of Experience’ and Metaphor and Moral Experience, passim. 27. Ibid., 236. 28. Ibid., 249–250. 29. Fogelin, Figuratively Speaking. 30. See Tirrell, ‘Reductive and Nonreductive Theories of Metaphor’. 31. It is not enough to simply rule this out by fiat, since that would render (P2) trivial. 32. See Wright, ‘Reply to Jackson’, 112, Denham, ‘Metaphor and Judgements of Experience’, 45–49. 33. Denham, Metaphor and Moral Experience, 292–293. 34. Ibid., 283. 35. See Wright, Truth and Objectivity; Wright, ‘Reply to Jackson’, passim. 36. A sentence is superassertible, if we are epistemically justified in asserting it, given our current state of information, and we would remain so justified given any way that that state of information might be ‘enlarged or improved’. 37. Wright, Truth and Objectivity, 86–87. 38. Denham, Metaphor and Moral Experience, Chapters 7–9. 39. Ibid., 318–319. 40. Similar arguments could be adduced against, for example, minimalism about the analytic/synthetic distinction. If it turns out that the notion of analyticity is thrown into confusion by central facts about meaning in general, then adverting to the surface discipline of ascriptions of analyticity could not—pace minimalist descendants of Grice and Strawson—suffice to rehabilitate it.

Bibliography

Beardsley, M. 1958. Aesthetics: Problems in the Philosophy of Criticism, New York: Harcourt, Brace and Co. Beardsley, M. 1976. ‘Metaphor and Falsity’, Journal of Aesthetics and Art Criticism 35: 218–222. Davidson, D. 1984. ‘What Metaphors Mean’ Inquiries into Truth and Interpretation, Oxford: Oxford University Press, 245–264. Davidson, D. 1986. ‘A Nice Derangement of Epitaphs’ in E. Lepore (ed.), Truth and Interpretation, Oxford: Oxford University Press, 433–446.

June 21, 2008 10:40 MAC/NWA Page-82 9780230_220478_05_cha04 Andrew McGonigal 83

Denham, A. 1998. ‘Metaphor and Judgements of Experience’ in Casati and Tappolet (ed), European Review of Philosophy Volume 3: Response Dependence, Stanford, CA: CSLI, 225–253. Denham, A. 2000. Metaphor and Moral Experience, Oxford: Oxford University Press. Fogelin, R. J. 1988. Figuratively Speaking, New Haven, CT: Yale University Press. Kemp, G. 1991. ‘Metaphor and Aspect Perception’, Analysis 51(2). Lycan, W. 2000. Philosophy of Language, London: Routledge. Margalit, A. and Goldblum, N. 1994. ‘Metaphors in an Open-Class Test’ in J. Hintikka (ed.), Aspects of Metaphor, Dordrecht and Boston: Kluwer, 219–241. Moran, R. 1997. ‘Metaphor’ in Wright and Hale (eds) A Companion to the Philosophy of Language, Oxford: Blackwell. Moran, R. 1989. ‘Seeing and Believing: Metaphor, Image, and Force’, Critical Inquiry 16: 87–112. Nogales, P. D. 1999. Metaphorically Speaking, Stanford, CA:CSLI. Rorty, R. 1991. ‘Unfamiliar Noises: Hesse and Davidson on Metaphor’, Objectivity, Rel- ativism and Truth: Philosophical Papers, Volume 1, Cambridge: Cambridge University Press, 162–172. Stern, J. 1991. ‘What Metaphors Do Not Mean’, Midwest Studies in Philosophy XVI: 13–52. Stern, J. 2000. Metaphor in Context, Cambridge, MA: MIT Press. Tirrell, L. 1991. ‘Reductive and Nonreductive Theories of Metaphor’, Journal of Philosophy 88(7): 337–358. Walton, K. 1990. Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Walton, K. 1993. ‘Metaphor and Prop Oriented Make-Believe’, European Journal of Philosophy 1: 39–56. White, R.M. 1996. The Structure of Metaphor, Oxford: Blackwell. Wittgenstein, L. 1998. Culture and Value, Revised Second Edition with English Translation, Oxford: Blackwell. Wright, C. 1994. ‘Reply to Jackson’, Philosophical Books 35: 169–175. Yablo, S. 1998. ‘Does Ontology Rest on a Mistake?’ Proceedings of the Aristotelian Society Supplementary Volume LXXII: 229–261.

June 21, 2008 10:40 MAC/NWA Page-83 9780230_220478_05_cha04 5 Artifact Expression John Kulvicki

Expression is a notion that finds its home with people and their mental states. We express, paradigmatically, emotions and moods. In some related sense, we express beliefs, desires, and other states of mind. Artifacts seem to express things too. Paintings, for example, express sadness, anxiety, joy, and much else. It is natural to wonder how artifact expression relates to expression in humans. It may be that the term ‘expression’ applies in some primary sense to people, and only derivatively to artifacts. But it is an open question in what respects artifact expression resembles its human counterpart. This chapter suggests that artifact expression and human expression share a symbolic structure. Nelson Goodman identified artifact expression with a symbolic structure, though it was one that human expression does not share.1 Understanding expression in these terms has been out of favor since Goodman.2 It is a shame because some of the ties that bind aesthetics to other areas of philosophy are found precisely here, and the ground seems quite fertile. Earlier work of mine reasserts the significance of symbolic structure for understanding representational kinds—pictures, images, diagrams, and so on—and the present work makes a similar plea for expression.3 There is a sense in which any account of expression in artifacts is an account of symbolic activity. So it is not that competing accounts ignore symbolic structure altogether. The worry is that most approaches de-emphasize it in favor of focusing on the following domain desideratum: artifacts can express only what humans can express. It is unclear why this is a desideratum, but one cannot respect it by focusing on whether artifact expression shares a symbolic structure with human expression. The current proposal throws its hat in with those who have taken the domain of artifact expression to be open-ended, in the hopes of shedding new light on what expression can be. Typically, artifacts express emotions and moods, but that is not essentially so: ‘a painting may express heat, a musical composition may express color or fragility.’4

June 21, 2008 10:45 MAC/NWA Page-84 9780230_220478_06_cha05 John Kulvicki 85

Motivation for the current view comes not just from intuitions about artifact expression’s domain. If recent work on the topic is any guide, few philosophers share these intuitions. The present account captures a phenomenon overlooked by those committed to the domain desideratum and it is a more plausible account of many instances of artifact expression than other accounts are. That is not to say it is the account of artifact expression, because it seems like expression is more a family or genus of kinds of aboutness than a specific species thereof.5 Most competing accounts of expression, as we will see, delineate phenomena that are interesting aesthetically and otherwise. The question is whether and to what extent each phenomenon is suitable as a candidate for artifact expression, and in this regard the present account seems superior to others. The first section discusses human expression with an eye toward its symbolic structure. The second section provides an account of artifact expression that mirrors the symbolic structure of human expression without being shackled to human expression’s domain. The third section distinguishes artifact expression from other phenomena with which it might be confused and considers some close competitors.

I. Human expression

One way to think of human expressions is that they are Gricean natural signs of people’s mental states.6 A bona fide expression of sadness is some feature of a person that indicates sadness, in the way that smoke indicates fire or thunder indicates lightning. A rather restrictive reading of indication suggests that under the right circumstances, necessarily, if there is thunder then there is lightning.7 Less restrictively, we can say that under the right circumstances, if there is thunder then it is (sort of, moderately, very) likely that there is lightning. Expressions are (decent, good, excellent) evidence for what they express.8 I do not propose to discuss which approach to indication is best since, for present purposes, nothing of note depends on it. This vague proposal seems to draw the boundaries of expression too broadly in some respects and too narrowly in others. On the one hand, we tend to identify expressions as a much more limited class of people’s features than those that are natural signs of their mental states. Facial configurations and hand gestures are fine but it is a stretch to call galvanic skin response an expression, as far as ordinary use of the term goes. On the other hand, facial configurations identified as expressions often do not naturally mean what they are taken to be expressions of. We smile to be polite and act dissatis- fied to get a better deal. This latter point was Goodman’s stated reason for abandoning the suggestion that human expression is a kind of indication.9 It is certainly true that we tend to call something an expression only if it is the kind of thing others can notice and use as a sign of some inner

June 21, 2008 10:45 MAC/NWA Page-85 9780230_220478_06_cha05 86 Artifact Expression machinations. But it is difficult to see why being an expression requires being usable in such a straightforward manner.10 Paul Ekman’s ‘micro-expressions’ are things that most people tend not to notice, and it takes some training to see them at all.11 With our ordinary perceptual equipment we are also unable to notice what a spectral analysis of someone’s voice reveals, but the latter features can indicate and in that sense be expressive of dishonesty, forthrightness, and so on. There are doubtless interesting facts about which expressions we can and cannot notice and thus use as signs of our own or others’ inner lives, but they are not so much central to an account of expression as they are central to an account of how we use expression to understand ourselves and others. Just as we should not let facts about which expressions we can use or consume govern an account of what expression is, we should not let the fact that we can intentionally produce states superficially like genuine expressions distract us from their true nature as natural signs. The number of rings in a tree might fail to indicate the tree’s age, and smoke does not always mean fire. We are able to identify natural signs on their own terms, independently of what they indicate. Dominic Lopes helpfully calls these ‘expression-looks.’12 It makes perfect sense that we can on occasion take something as a natural sign of age, fire, or even sadness when in fact it is not.13 Expression-looks can fail to be expressions, and because we are in decent control of our bodily comportment we can even deploy some expression-looks duplicitously. We express a lot about our mental lives. In addition to emotions, which are the popular example in aesthetics, we express moods, states of mind like confusion, diffidence, and confidence, as well as dispositions such as cautiousness, recklessness, and patience. ‘It’s raining’ expresses the belief that it is raining and, depending on the context and the way it is said, ‘I believe it’s raining’ expresses cautious acceptance of the claim that it is raining, the belief that it is raining, and/or the belief that one believes it is raining.14 Other acts, such as opening an umbrella before venturing outdoors, can similarly express our beliefs, though we rarely use such acts as expressions. Expressions are indexical in two interesting and significant ways. In Ruth Millikan’s terms, they have two representationally significant self-referring variables.15 As indicators of inner states, they are specific to the person having the expression-looks. Your expression-looks express your mental states, not anyone else’s. This link might seem too obvious to mention, but it also might seem to break down under limited circumstances. For example, some expression-looks are contagious: we smile when we see that others are smil- ing, and assume a dourer expression-look when others do the same. When this happens, there is a sense in which my expression-look is a better indicator of someone else’s expression-look, and perhaps thereby their mental state, than it is of my mental state. In these cases, expression-looks can be deceiv- ing because we take them to express states of the individual having them. Interestingly enough, my expression-look is one among many causes of the

June 21, 2008 10:45 MAC/NWA Page-86 9780230_220478_06_cha05 John Kulvicki 87 emotional state it normally expresses. This is one mechanism by which emotions, and not just expression-looks, can be contagious, which means that one’s expression-look will tend to become an expression. Smile if you want to be happy.16 Expressions also have representationally significant self-referring temporal parts. An expression of sadness now indicates sadness now,justas‘now’ indicates the time at which it is uttered. That look on your face does not express yesterday’s confusion. It might not express anything: you might just be practicing your expression-looks. You might be vividly remembering yesterday’s confusion too, and this could result in such a look. In the latter case, you might be expressing a confusion-like state that you undergo during such a vivid recall of the day before, but you are not expressing yesterday’s confusion in any but some derivative sense. The temporal indexicality of expression might seem to break down when the full representational power of language is recruited in the service of expression. ‘I was so scared yesterday’ is a good indicator of one’s earlier mental states. The claim is also a good indicator of one’s current mental state, however: one believes and is remembering that one was scared. We generally think of the latter as being expressed while the former is not. Expression is not merely indication in general, but indication of the current mental state of the person doing the expressing. An expression consists of an expression-look at a time in a person that indicates some mental state of that person at that time. So far, not much has been said about what properties of people are responsible for indicating facts about their mental lives. They certainly vary, and there are doubtless many stories to be told about just why any given facial configuration, galvanic skin response, change in blood pressure, or spectral pattern in one’s voice indicates what it does. Some are under our control, some not. Some can be deployed variously across cultures and times, while others are more universal.17 In all cases of expression, some feature of a person at a time indicates one of their current mental states. The mechanisms that support the expression vary. In addition, not much has been said about how we know that certain features express others. In certain cases, the correlation seems hard-wired into the way we perceptually engage with the world, while ritualized gestures suggest that others are simply learned, though we may be wired so to learn. It might be that expression plays a rich role in our lives because some expressions have the natural function of indicating what they do. That is, there mightbeacoreofnatural expressions—indicators whose biological purpose is to be and be used as indicators of internal states—from which the more general notion adumbrated here derives. Galvanic skin response is unlikely to fit the bill as such a core case of expression, but because it has the indicative and indexical features characteristic of natural expressions, it counts by extension. Lopes limits expression to natural expression, requiring that human expressions have the natural function of indicating what they do.18

June 21, 2008 10:45 MAC/NWA Page-87 9780230_220478_06_cha05 88 Artifact Expression

For reasons that will become clear, I think the more expansive notion is better suited to helping us understand artifacts.

II. Artifact expression

Artifacts are neither humans nor persons. They have no mental states. Most of them do not change over time or interact significantly with their environ- ments. Portraits age and in that sense change over time, but not, for example, in the manner of Dorian Grey’s. Also, compared to people, artifacts possess a greater ratio of surface to depth. It is not just that artifacts lack mental states, but that they often lack internal states altogether, leaving little for their external features to indicate. What would count as a painting’s internal states? At best, features of the concealed canvas and the stretchers that give it form. But these features are bad candidates for what a painting expresses. These points of disanalogy undermine the most basic of proposals:

Artifacts express X iff they have features that indicate that they are X.

Without being X—and indeed, most artifacts that express X cannot be X— there is little hope of having features that are natural signs of being X. Artifact expression cannot be bona fide indication with the indexical structure characteristic of human expression. Where do we go from here? The road forks initially in three directions and it helps to sketch what each option looks like, though it is beyond the scope of this chapter to discuss these proposals in detail. First, one could claim that artifacts express X when they exhibit features that indicate that some person, actual or fictional, is X. For some, artifacts are expressions of artists’ emotions.19 For others, they express the emotions of fictional personae.20 The literature provides some reasons for concern about such proposals.21 My worry is that they save the wrong phenomena. Both versions of this approach have an easy time with the domain desideratum: artifacts can express only what people can express. The first, older version, according to which artifacts are expressions of artists’ emotions, also saves the indicative nature of expression. Artifact expressions indicate artists’ feelings. But it abandons the indexical structure of that indication: artifacts indicate that something else had an emotion at some other time. The persona view abandons both the indexical structure of human expression and its indicative nature. Since fictional personae do not exist, an artifact cannot really indicate anything about them. The indication is merely apparent. As we will see, the approach favored in this chapter also appeals to merely apparent indication, but it preserves the indexical structure of that (merely apparent) indication. The other road much traveled is to claim that artifacts are experienced as resembling either human emotional states or expressions of human emotional states.22 A piece of music is experienced as resembling a happy gait, a

June 21, 2008 10:45 MAC/NWA Page-88 9780230_220478_06_cha05 John Kulvicki 89 dejected shrug, or a pattern of feeling, for example. This proposal removes worries about fictional personae, but it also gives up on the indicative nature of artifact expression altogether. Artifacts neither indicate nor seem to indicate anything about themselves or others: they are merely experienced as resembling indicators or emotional states.23 This proposal limits the domain of expression to emotions or mental states more generally, but it does so in an ad hoc fashion. That is, the domain restriction can be abandoned by this theory and what is left might still look like an account of an interesting, perhaps expressive phenomenon. A proposal like this will be discussed below. The third approach focuses on the symbolic structure of artifact expression. Goodman explicated expression as a way of symbolizing properties that is quite different from its human counterpart. Artifacts express X when they metaphorically possess X and they thereby exemplify X. Exemplification is a way of symbolizing a property, as a tailor’s swatch picks out a kind of fabric.24 There will be more to say about Goodman in what follows. Lopes suggests that we can understand artifact expression in terms of indication, rather than exemplification, thus rendering the artifact case more like human expression than Goodman did.25 Lopes’s approach is closest to the present one, but it will become clear that there is distance between us. These three forks in the road lead away in different directions from the failed, basic proposal mentioned above. To wit:

Artifacts express X iff they have features that indicate they are X.

The goal is to see how this can be improved upon via the third fork, not the first two. The basic proposal fails in no small part because artifacts can express X even if they cannot be X. They express emotions, for example, even though they have no emotional lives. Possession is one important condition on successful expression in humans, especially if we think of expression as indication. We want an account of artifact expression that distinguishes successful from failed expression as well. Sometimes artifacts are recognized as failed attempts at expression, but success and failure in the artifact case cannot depend on whether the artifact expresses what it possesses. Keeping that in mind, consider a simple emendation of the basic proposal:

Artifacts express X iff they have features that seem like indicators of their being X.

This is related to the second approach discussed above, favored by Hospers, Kivy, and Davies, but it does not insist that artifacts exhibit features that are experienced as resembling human expressions. In effect, it jettisons the ad hoc restriction to mental states and human expression-looks. Perhaps on occasion the explanation for why certain features seem like indicators of X (be it

June 21, 2008 10:45 MAC/NWA Page-89 9780230_220478_06_cha05 90 Artifact Expression an emotional state or something else) is that they are experienced as resembling human expression-looks, perhaps sometimes it is because they in fact resemble human expression-looks, and perhaps there are other things at work altogether. This proposal thus has the advantage of broadening the potential range of artifact expression beyond that of human expression. It still faces the worry that it divorces expression from indication in an implausible manner, and it faces two further problems as well. First, we are often unable to isolate features of an object that are responsible for its expressing what it does, so features of an object that expresses X need not seem like indicators of X. Here the focus on perceptible properties of the artifacts seems misplaced since they are not readily taken as indicating anything. Second, it is not obvious how this approach can accommodate some kinds of expressive failure. In particular, it would seem that an artifact can have features that seem like indicators of X, even to the appropriate population, without actually expressing X, just as people can smile unconvincingly. That is to say, there are two kinds of failed expression. In one case, someone tries to make an artifact’s features seem like indicators of X and fails: they do not seem that way. The present proposal can deal well with this type of failure. In the other case, one makes an artifact’s features seem like indicators of X, but the artifact nevertheless fails to express X. This kind of failure seems common: ‘I recognize features of the song that seem like expressions of joy, but it just doesn’t work.’ It is difficult to make sense of this failure given the present proposal because resembling indicators of joy should be sufficient for expression. Dominic Lopes makes a proposal related to the previous one, which avoids some of its problems:

Artifacts express X iff they have features that function to indicate X.26

Certain features of a painting’s surface, at least in some contexts, are supposed to indicate X, even though they might not in fact do so. An artifact can express some emotion without genuinely indicating the emotion’s presence, in a fictional persona or otherwise, since some features can have the function of indicating emotions without actually doing so. It is unclear what confers functions upon artifacts’ features, however, and Lopes is reluctant to reveal much about this.27 If having the function of indicating X requires that an artifact’s features seem to do so, then Lopes is faced with the worries about expressive failures accompanying the previous proposal. His other options seem grim as well. It cannot be that artifacts with such features usually indicate X since artifacts that express X generally cannot be X. And it cannot be that by some process of natural selection artifacts come to have such a function; they certainly do not evolve in the manner of organisms. Conditions on successful artifact expression must depend in no small part on their human consumers—they are artifacts, after all—but it

June 21, 2008 10:45 MAC/NWA Page-90 9780230_220478_06_cha05 John Kulvicki 91 is far from clear how this is so. Lopes’s decision to avoid giving an account skips over one of the important puzzles here. In addition, Lopes mentions nothing about whether artifacts function to indicate features of themselves or other things, at the present time or at some other time. As we will see, these are the important aspects of artifact expression. What is lacking in the above proposals is an appropriate focus on what is expressed, as opposed to what does the expressing. So it is time to propose an admittedly controversial claim that will help clear the way to a new account of artifact expression:

Artifacts express X only if they seem to be X.

It may be obscure just why a given artifact seems to be X. One may be unable to specify the features that lead one to have that sense about it, but psychologically speaking, our sense of what an object expresses seems rather immediate. It just seems sad, for example.28 This seems to fit the phenomenology of noticing expressive properties and it fits well with the way we talk about at least many properties that are expressed. Things that express joy seem joyous, those that express resolve seem resolute. Notice that this makes it easy to explain the problematic expressive failures discussed above. An artifact can have features that some artifacts expressive of sadness also have—these features can seem like indicators of sadness, for example.— but fail to express sadness because such an artifact can fail, despite having those properties, to seem sad. Similarly, songs that are obvious attempts at expressions of joy can fail: they do not seem joyous. There are many cases, however, in which objects seem to be a certain way even though they do not express that property. A red painting seems red, and indeed it is red, but that is not to say it expresses redness. In addition, unlike humans, artifacts can express X if they are not X, and even if they cannot be X. Expression is not so much instantiation as indicated instantiation, and even then the full, success-laden notion of indication is misplaced. Here, then, is a first stab at an interesting, necessary condition on artifact expression:

Artifacts express X only if they seem to indicate that they are X.

It is important to understand seeming to indicate being X as implying seeming to be X. When artifacts are expressive of X there is a way in which they seem to be X, namely by seeming to indicate that they are X. None of this is meant to pick out a psychological order to states of seeming. It is not that first an object seems to indicate that it is X and it then comes to seem, via some inference, as though it is X. Psychologically speaking, objects just seem to be X. That may very well strike one first, even though none of the object’s features is readily identifiable as either X or an indicator of X. It might also be that a sense for what the artifact expresses comes only after careful scrutiny

June 21, 2008 10:45 MAC/NWA Page-91 9780230_220478_06_cha05 92 Artifact Expression of its features. Something similar seems true of human expression. People can seem to be happy even though it is not clear why. Something about their perceptible features seems to indicate this about them. This proposal does not suggest that artifacts really indicate what they express, since they typically do not possess what they express, but such a proposal is doomed anyway. This proposal does carry over the indexical structure of artifact expression from the human case, however: they seem to indicate that they are X, not that something else is X or that they were X. The kind of apparent indication mentioned above is not something we can simply identify with representation ordinarily conceived, though it is certainly a kind of aboutness. Consider:

This inscription is lonely.

The inscription represents itself as being lonely without being expressive of loneliness. It certainly does not seem lonely, at least. Despite representing itself as being so, it does not seem to indicate that it is lonely in the sense we have been considering. We come to know that the inscription represents itself as lonely by interpreting what the inscription says. This may prompt us to inspect it anew, trying to decide whether it does seem to be lonely, but there is no deep connection between interpreting the inscription and the inscription seeming lonely. If a representation expresses loneliness, inspection reveals that it seems to be lonely. Inscriptions and the like can certainly be expressive, but it is not solely in virtue of what they linguistically represent that they manage to do so. Similarly, one can depict a person as lonely without making a picture that expresses loneliness, at least in the sense being considered here. All of this suggests that the kind of expression articulated herein does not, in general, characterize written works of narrative fiction. So if such works are expressive—and it seems as though they can be—they aresoinsomeotherway. If we were to isolate the above inscription in the middle of an otherwise empty sheet of paper, it might, in that context, express loneliness. Writing the inscription in a different color of ink from a myriad of inscriptions surrounding it could accomplish a similar effect (though it would also be silly). In both cases, the representational content of the inscription can put our minds in the ballpark of what is expressed, and thus get us looking at the inscription in the right way, but the representational content does not fully explain what the artifact expresses. Related examples, like ‘this inscription is friendly,’ would also express loneliness in related contexts. Similarly, ‘fragile’ does not in and of itself express fragility. Inscribe ‘fragile’ on a box, however, particularly among other boxes not similarly marked, and it might do some expressive work. It is not merely that the term represents the property, but that placed on the side of the box it makes it seem as though the box and its contents are fragile. The box seems

June 21, 2008 10:45 MAC/NWA Page-92 9780230_220478_06_cha05 John Kulvicki 93 fragile, and you treat it as such, because it seems to indicate that about itself. Since having a representational content is one among many of an object’s features, it can help determine what gets expressed. Consider E.E. Cummings’s poem, l(a.29 The poem consists of the word ‘loneliness’ combined with the phrase, ‘a leaf falls.’ It is arranged vertically, with ‘loneliness’ being broken up into four parts: l-one-l-iness. The phrase appears in parentheses between the first two fragments of ‘loneliness,’ itself broken up into vertically arranged parts: (a-le-af-fa-ll-s). At first sight, the poem looks like a column of letters that splays slightly toward the bottom. There is nothing particularly expressive about this arrangement of letters. Also, the word ‘loneliness’ represents loneliness, but inscriptions of that word are rarely expressive of loneliness. Ditto for the phrase ‘a leaf falls.’ The poem as a whole, nevertheless, expresses loneliness in the sense that on reflection it can seem to be lonely. The representational content of the poem and the arrangement of its letters are co-conspirators in bringing about the poem’s expressiveness. A random selection of letters, similarly arranged, would not do the work of this poem. But the arrangement plays no small part in the poem’s expressing what it does. It matters that ‘l’s are at the top and bottom of the ‘leaf fall’ column, with alternating ‘af’ and ‘fa’ in between, suggesting a downward, solitary spin. Robert Stecker suggests that lyric poetry is an interesting example of expression because we can usually imagine such a poem’s words being read out by someone, and thus attribute its expressive content to the expressive content of someone’s utterance.30 We might do this from time to time with some poetry, music, and narrative fiction.31 But this activity does not capture artifact expression so much as a way in which artifacts can be records of human expression. The hope has been that we are closing in on a notion of artifact expression that is not a derivative product of human expression—as the first fork in the road mentioned above suggests—so much as something that has a symbolic structure like human expression. After all, the Cummings’s poem is expressive, but it cannot be so because we imagine someone reading it out. You simply cannot read that out, just as some ways in which things that can be said simply cannot be written out. l(a is not your standard poem, but the hope is that non-standard examples such as this one can shed light on a phenomenon present in more standard examples as well, even though it may be difficult to discern. It is easy enough to say that a poem expresses what we imagine someone would express were she to say it aloud with feeling, but that might not be all that there is to expression in artifacts. Similarly, Abstract Expressionist paintings might shed light on expression in representational paintings by revealing a phenomenon that, in the representational case, is too easily overlooked because of the focus on other phenomena, like what is expressed by what is represented.

June 21, 2008 10:45 MAC/NWA Page-93 9780230_220478_06_cha05 94 Artifact Expression

None of the above is inconsistent with artifacts that express X also seeming as though they are not X. In fact, any account of artifact expression must deal with the fact that

Generally, artifacts that express X are not X, and they often seem not to be X.

We fully realize that most paintings expressive of sadness are not sad. But we should not confuse the inconsistency of seeming to be X and not seeming to be X with the perfectly common conjunction of seeming to be X and seeming not to be X. Usually, artifacts seem not to be X in that they seem to have properties incompatible with being X. We know that the canvas cannot be sad because being a painted canvas is incompatible with having an emotional life. In that sense the canvas seems not to be sad. But the painting also seems to be sad in that it seems to indicate that it is sad. So far, we have a plausible but incomplete story about artifact expression. Artifacts express X only if they seem to indicate being X, and this is consistent with their seeming not to be X. Is there another necessary condition lurking around that will make this a full account of the phenomenon? In the service of finding one, it helps to notice why the necessary condition is not sufficient. Artifacts indicate many things about themselves, but they express at most some of them. In particular, given that objects expressive of X tend not to be X, many features of an artifact that expresses X may very well indicate that it is not X. For example, that something is a piece of canvas covered with oil paint suggests that it has no emotions. Does this mean that the painting expresses a lack of emotions? Paintings can express emotionlessness, but that is an achievement, not merely the result of being made of paint and canvas. Also, the artifact’s dimensions indicate its temperature without being thermally expressive. The cracks in the painting indicate that the painting is old, but it is not expressive of age. These latter examples may not conflict with what the object expresses—in the way that emotionlessness conflicts with being sad—but they are cases of indication without expression nonetheless. Here we find some motivation for limiting the scope of expression at the outset to mental states or even, more strictly, to emotions. Lopes, whose account is closest to the one offered here, limits his discussion to the expression of emotion. He does not suggest that nothing else can be expressed—he only offers an account of what it is to express emotions— but he does not suggest that expression is any broader than that, either. The hope here is to be more inclusive, taking to heart the suggestion of Arnheim, Beardsley, and Goodman that the domain of artifact expression outstrips that of human expression. This hope runs into a challenge similar to the one that Goodman faced, so it helps to consider briefly his own account.

June 21, 2008 10:45 MAC/NWA Page-94 9780230_220478_06_cha05 John Kulvicki 95

Goodman understood expression of X in terms of possession of X: artifacts express only those properties that they (at least metaphorically) possess. Since there are indefinitely many properties any given object possesses, Good- man needed to explain why only some of them are expressed. He did this by identifying the properties expressed with the properties that an artifact (metaphorically) exemplifies. Just what properties any given object exemplifies depends on our symbolic practices. The tailor’s swatch exemplifies a color, weight, pattern, and material of fabric, but not a size or shape of a piece of fabric, even though it possesses all of those properties.32 There are many reasons why any given object exemplifies some properties and not others, and these reasons may change over time and from place to place even if they also depend in part on our natural tendencies to take one thing as standing for something else. The first response to the challenge facing the present view mirrors Good- man’s move. Notice that an artifact can indicate many things, even about itself, without seeming to indicate those things and thus expressing them. So, an artifact’s dimensions might indicate its temperature, but they do not strike us as doing so. This response only gets us so far, because there is a sense in which paintings and the like really seem not to have emotions, and indeed they seem to indicate this about themselves. Similarly, the cracks in the painting seem to indicate the painting’s age. To complete the response, we make a move that mirrors metaphor’s role in Goodman’s account. Goodman did not think the tailor’s swatch expresses being houndstooth or woolen.33 It really possesses and exemplifies those properties. By contrast, expression’s domain includes only those properties metaphorically possessed by objects. Paintings that express sadness are not really sad; they are only metaphorically sad. When the painting is metaphorically sad and this fact stands out, thus referring us to sadness as exemplars do, we get the expression of sadness. The present proposal, in terms of seeming indication, does not require that paintings be metaphorically sad. It only requires that they seem to indicate that they are sad. Goodman’s appeal to metaphor might be problematic, so the fact that this account avoids metaphor is welcome.34 What plays the role of metaphor in the current account? That is, what helps to sort the things that seem to be indicated into those that are expressed and those that are not? Notice that among the things that an artifact seems to indicate about itself, some of them are surprising in a way that others are not. For example, we know and fully expect that paintings do not have mental states, so it is unsurprising that they seem not to have them. By contrast, as mentioned above, artifacts that express X tend not to be X and are often the kinds of things that cannot be X. That they seem to be X stands out, often against a contrary backdrop of what we know to be true of such objects. For Goodman, it is not just that metaphorical possession is distinguished from literal possession, but that metaphor is surprising: ‘an

June 21, 2008 10:45 MAC/NWA Page-95 9780230_220478_06_cha05 96 Artifact Expression attraction that overcomes resistance.’35 We need not appeal to metaphor, or exemplification, to accommodate the surprising nature of artifact expression, but Goodman’s moves point in the right direction. Overall, then, the proposal is that

An artifact expresses X if and only if (1) it seems to be X insofar as it seems to indicate that about itself and (2) for other reasons, it seems not to be X.

The painting can seem to be sad and seem not to be sad, but in the latter case it is not just because it seems to indicate that about itself: paintings are not the kinds of things that can be sad. In fact, paintings are the kinds of things that cannot have emotions at all. In that sense, all paintings will seem to be emotionless. A painting that expresses emotionlessness, however, indicates a lack of emotion on the part of something that should be capable of feelings. This distinguishes the achievement of expressing emotionlessness from the simple fact that all paintings seem to lack emotions. Nor need we limit artifacts’ expressive potential to mental states. They are prime candidates for expression because they are within the domain of human expression and the same mechanisms responsible for apprehending such expressions are likely at work when we engage with artifacts. But a painting or sculpture can express fragility or warmth, even if in some sense it seems neither fragile nor warm. Also, though not impossible, getting a red painting to express redness would be an impressive accomplishment. It is hard to get a red painting to seem not to be red, and it is hard to get it to seem red insofar as it indicates that about itself. If anything, it is easier to express colors with musical compositions, if they can be expressed at all. Artifacts that express X seem to be X, but to whom do they so seem? Cer- tainly, we should not expect artifacts that express X to seem X to anyone who happens to look or listen. Appropriately trained, experienced, attentive, and knowledgeable crowds are required, one assumes. Of course, the trick is in unpacking appropriateness for each of these conditions, but that task is beyond the scope of this chapter. One point in this regard is that our ability to notice what an artwork expresses sometimes relies on our being pointed in the right direction. Jeffrey Mumford’s composition The Focus of Blue Light depends in part on its title to get one in the right frame of mind to notice what it expresses. It also helps to know that many of his compositions have titles that refer to light and color. Ernst Gombrich thought that such ‘pointers’ were essential, and I agree with him.36 It is not that expression is guaranteed to succeed as long as one points. Indeed, many might find it implausible that Mumford succeeds. Without pointers, however, we could not count on people to put themselves in the right ballpark to assess whether the artifact’s maker has realized his/her ambitions.

June 21, 2008 10:45 MAC/NWA Page-96 9780230_220478_06_cha05 John Kulvicki 97

III. Isolating artifact expression

The kind of artifact expression explicated herein can be elusive, for two reasons. First, artifacts can represent people or other things as expressing something, and this can obscure whether and to what extent the artifact itself seems to be this way or that. Second, expressive attributions, which are fundamentally part of interpreting an artifact, overlap with and have the same form as some evaluative claims. First, many paintings, drawings, etchings, photographs, sculptures, and the like, represent people as being sad, for example, in that they are represented as expressing sadness. One might think that there is a sense in which such artifacts eo ipso express sadness, but that would not be in the sense explicated above. An artifact can represent someone as expressing sadness without itself seeming sad. In this respect, the current account agrees with Goodman’s claim that ‘To describe, as to depict a person as sad or as expressing sadness is not necessarily to express sadness.’37 Sometimes, indeed often, a painting that depicts an expression of sadness also expresses sadness, but this is by no means assured. Artifact expression is elusive because it is often present in representations of people expressing something, which tempts one to think that expression in such cases is wholly explicable as a feature of the artifact’s representational content. There are many interesting issues concerning the successful representation of human expressions in pictures, but they cannot be identified with the problems of successful expression by pictures. In fact, the discussion of the Cummings’s poem suggests that what an artifact represents can play a role in what that artifact seems to indicate about itself. Sometimes the fact that a picture represents someone as expressing sadness can contribute to the picture’s seeming sad. There is even reason to suspect a more intimate connection between what a picture represents and what it expresses than one finds, for example, between the Cummings’s poem and what it expresses. Elsewhere, I have argued that pictures are interesting in part because they manifest at least some properties that they represent their objects as having. To a limited extent, pictures are mimetic representations. For some properties, P, pictures represent their objects as being P if and only if they are P.38 A picture of a sad person clearly does not manifest the property of being sad, or of being a person, but it does manifest certain features of the spatial organization of a sad person’s face. Given that, it is plausible that there is substantial overlap between the picture’s features that are responsible for it representing a sad person and those responsible for it seeming to indicate of itself that it is sad. So, looking at a picture of a sad person can result in a kind of vacillation between the picture itself seeming sad and the person depicted seeming sad. There is a similar back-and-forth between seeing a picture as a pattern of pigment and seeing it as an instance of what it represents.39 Since other representations, like descriptions, tend not to manifest properties that figure in their contents,

June 21, 2008 10:45 MAC/NWA Page-97 9780230_220478_06_cha05 98 Artifact Expression it would be at best fortuitous that the features responsible for the description representing what it does are those responsible for it expressing what it does. Part of what makes the Cummings’s poem interesting is his exploitation of precisely those features toward an expressive end. These remarks are little more than a sketch, but hopefully they show that once the current account of artifact expression is on the table it opens up some interesting avenues for investigating ways in which representation and expression interact in different media. Lopes suggests that there are at least three kinds of artifact expression— figure expression, scene expression, and design expression—and one might think that this can explain the elusiveness of artifact expression. ‘In figure expression, a person is depicted as expressing an emotion.’40 More specifically, figure expression is ‘an expression that is wholly attributable to a depicted person or persons.’41 For Lopes, to the extent that a picture depicts someone as expressing an emotion, it figure expresses that emotion. Lopes would agree that figure expressing an emotion does not entail either of the other two kinds of expression. It is unclear whether figure expression and its close cousin, scene expression, are kinds of artifact expression. Figure expression, at least, seems to capture a fact about the representational content of some pictures. Lopes agrees with the latter claim, but thinks that it is nevertheless an important part of artifact expression. ‘[A] picture may express an emotion by depicting a figure or scene as expressing the emotion.’42 If this boils down to a fight over terminology, so be it. The substantive worry with Lopes’s proposal is that figure expression seems not to capture much about what the artifact expresses. This chapter has been at pains to unpack a notion of artifact expression that is true to expression’s symbolic structure, and thus one that captures the ways in which expressions indicate features of their bearers at the time in question. Calling figure expression bona fide artifact expression gives up on that goal, and is thus unappealing. Similar remarks apply to scene expression, which for Lopes is expression ‘attributable at least in part to a depicted scene and is not wholly attributable to a depicted person or persons.’43 None of this suggests that scene expression and figure expression are aesthetically uninteresting, or that there is no such thing as the phenomena he describes. It is just to suggest that they do not capture the core sense in which an artifact can be said to express something. It might seem that Lopes is closest to the present view concerning what he calls ‘design expression’: ‘an expression that is wholly attributable to a picture’s design or surface and not to any figure or scene it depicts.’44 Is artifact expression as explicated herein, modulo some differences mooted above, just a kind of design expression? Not really. For Lopes, design expressions are distinct from figure and scene expressions because they depend exclusively on features of the picture’s design. On my account, artifact expression is distinct because the artifact seems to indicate that it is a certain way now and

June 21, 2008 10:45 MAC/NWA Page-98 9780230_220478_06_cha05 John Kulvicki 99

Lopes says little about the indexical structure of design expressions. Also, it is plausible that an artifact can seem sad, fragile, worried, tense, or what have you, in part because the artifact represents certain things. As mentioned earlier, what an artifact represents is one of that artifact’s features, and it can play an important role in determining what the artifact expresses. Cer- tainly, an artifact’s design contributes both to what it expresses and to what it represents. But for something to qualify as design expression in Lopes’s sense, it must be expression that is not attributable to ‘any figure or scene it depicts.’ Sometimes artifact expression is attributable to the fact that an artifact depicts a certain figure or scene. If Lopes were to modify his notion of design expression to include features of the artifact, including properties like representing such and such, then this might bring Lopes closer to the current proposal but it would undermine the spirit of design expression. For Lopes, kinds of expression are sorted based on what properties are responsible for what gets expressed. It is unclear, however, why design expression as Lopes envisions it is the theoretically interesting category that we should be after.45 It seems more appropriate to sort out kinds of expression based on the objects that do the expressing rather than the features responsible for it. Artifact expression is expression by artifacts, which depends on the artifact’s myriad features. Figure expression is expression by depicted figures, but not expression by the artifacts that depict them in any but a derivative sense of the word, and similarly for scene expression. Keeping these kinds of expressions separate is difficult, and because it will often happen that an artifact that represents someone expressing an emotion also artifact expresses that emotion, the kind of expression explicated here can be elusive. The second reason artifact expression can be elusive is that it comes tantalizingly close to another topic in aesthetics, from which it must be distinguished. We are often willing to say of artifacts that they are sad, happy, joyful, fragile, pathetic, strong, powerful, sharp, muted, lithe, graceful, awkward, and so on. Goodman, who countenances the broadest range of expressive potential in artifacts, says: ‘For instead of saying the picture expresses sadness, I might have said that it is a sad picture.’46 George Dickie thinks that an expansive notion of expression, which goes beyond human mental states, makes at least partial sense out of the range of such attributions.47 Perhaps all of the foregoing attributions are expressive attributions. Even those who think that expression is limited to emotions and the like agree that we call an artwork sad when it expresses sadness.48 It is worth noting that the present account makes sense out of this phenomenon. If artifacts express X only if they seem to be X, it is unsurprising we say works that express X are X. Not all claims that paintings are sad, graceful, or pathetic are expressive attributions, however. When we call some painting sad, graceful, or pathetic, we might be evaluating it aesthetically. That is, we sometimes ascribe

June 21, 2008 10:45 MAC/NWA Page-99 9780230_220478_06_cha05 100 Artifact Expression aesthetic properties to works of art, and these ascriptions, among other things, are expressive of aesthetic approval or disapproval, but they need not be ascriptions of expressive properties.49 We tend to think that the graceful things are aesthetically good at least to the extent that they are graceful, while pathetic things are bad at least to the extent that they are pathetic. Awkward or garish things are aesthetically disfavored for being so. Garishness is not so often something expressed as something possessed, but this is true for all of the properties mentioned in the previous paragraph when they are used evaluatively. Notice, a painting that expresses patheticness, awkwardness, or even garishness can be quite good while few pathetic, awkward, or garish paintings are worth our time. It can be quite difficult to sort out when a term is being used to pick out an expressive quality of a work or an aesthetic quality. Sometimes the uses overlap or one can find both in a given statement about an artwork, and it might be that some expressive features are generally evaluated positively, or negatively.50 None of these is meant to suggest that a painting’s expressive features do not figure in evaluations of it. It is just to say that when we say a painting expresses sadness, awkwardness, power, and so on, we are not eo ipso evaluating it. The expressive features of artifacts are, in this sense, akin to their representational features. It is difficult to fault a painting purely on the grounds that it represents X, just as one cannot fault it just because it expresses Y.

IV. Conclusion

Without doubting that there are many faces to expression in artifacts, the hope has been to capture a fairly central instance of it, and indeed one that is easily overlooked. Artifacts express X when they seem to indicate that they are X even though, for other reasons, they seem not to be X. It is an interesting fact that artifacts like paintings can seem to be sad, and that pieces of music can seem to be blue, or fragile, or warm, or what have you. The exact domain of artifact expression is open, at least so far, and this helps to explain why trying to make expressive artifacts is such a compelling activity. Gombrich was fond of John Constable’s claim that paintings are experiments into how we can perceive the world. It pays to remember that Constable was thinking both of representation and expression.51 And though the account proposed here departs considerably from Gombrich’s discussion of expression, it nevertheless suggests many avenues of experimentation.

Acknowledgments

Dorit Bar-On and Mitch Green were kind enough to invite me to present a version of this chapter at their workshop on expression at UNC-Chapel Hill

June 21, 2008 10:45 MAC/NWA Page-100 9780230_220478_06_cha05 John Kulvicki 101 in November 2006. I thank the workshop participants—especially Maura Tumulty, who prepared quite useful written comments—for a very helpful discussion. I also thank Hong Yu Wong, Stacie Friend, and the other attendees of the London Aesthetics Forum for the invitation to present this work and for a very stimulating discussion. Finally, thanks to Kate Thomson-Jones for helpful comments and to both editors for inviting me to contribute to this volume.

Notes

1. Goodman, Languages of Art,Ch.2. 2. Lopes (Sight and Sensibility, Ch. 2) is an exception to this rule. 3. Kulvicki, On Images. 4. Goodman, Languages of Art, 46. Arnheim (Art and Visual Perception) and Beardsley (Aesthetics) agreed, but few have pursued this line since then. Vermazen (‘Expres- sion as Expression’, 214–215) gives some interesting reasons favoring the domain desideratum, and he might deserve the blame for its prominence. 5. I get the family-genus-species terminology for representational kinds from Hauge- land’s ‘Representational genera’. Budd (Values of Art) suggests that a narrower kind of pluralism is essential to understanding artifact expression and Trivedi (‘Expressiveness as a Property’, 412) follows him. 6. See Grice, ‘Meaning’. Those who suggest something along these lines include Hampshire (Freedom of Mind, 143), Wollheim (On Art and the Mind,Ch.4), Vermazen (‘Expression as Expression’), and Davis (Meaning, Expression, and Thought, Ch. 3), among others. 7. See, for example, Dretske, Knowledge and the Flow of Information, 1981. 8. Davis, Meaning, Expression, and Thought, 45. 9. Goodman, Languages of Art, 47. 10. Vermazen, ‘Expression as Expression’, 197. 11. Ekman, Telling Lies, 1985. 12. Lopes, Sight and Sensibility, 70; see Wollheim, On Art and the Mind,Ch.4. 13. See, for example, Dretske’s discussion of ‘channel conditions.’ (Knowledge and the Flow of Information,Ch.5) 14. See Grice, Logic and Conversation;Davis,Meaning, Expression, and Thought. 15. Millikan, ‘Biosemantics’. 16. See James, ‘What is an Emotion?’; Prinz, Gut Reactions, 55–60. 17. Ekman, ‘Universals and Cultural Differences’; Baxandall, Painting and Experience; Lopes, Sight and Sensibility. 18. Lopes, Sight and Sensibility,Ch.2. 19. Tolstoy, What is Art?; Collingwood, Principles of Art. 20. For example, Vermazen, ‘Expression as Expression’; Levinson, The Pleasures of Aesthetics. 21. For example, Davies, ‘Artistic Expression’; Stecker, ‘Expressiveness and Expression’. 22. Langer, Philosophy in a New Key; Hospers, ‘The Concept of Artistic Expression’; Kivy, The Corded Shell;Davies,Musical Meaning and Expression; Budd, Values of Art. 23. See also Trivedi’s ‘Expressiveness’, 412, which makes a similar charge against Budd in particular. 24. Goodman, Languages of Art, 57. 25. Lopes, Sight and Sensibility,Ch.2.

June 21, 2008 10:45 MAC/NWA Page-101 9780230_220478_06_cha05 102 Artifact Expression

26. Lopes, Sight and Sensibility, 78. This is not a quote from Lopes, but my own (incomplete) reconstruction of his proposal for present purposes. I mention more about Lopes’s view below and in a brief review of Sight and Sensibility (Kulvicki, ‘Review’). 27. Lopes, Sight and Sensibility, 82–86. 28. Cf. Trivedi, ‘Expressiveness’, 414–416. He claims that we imagine a piece of music to have mental states, among which is a certain emotion that the music itself is expressing. We might do this now and again. The current proposal does not require that we imagine anything in particular, unless whenever something seems to be X we imagine it to be X. Trivedi’s interesting view deserves more attention than space permits. 29. It is relatively easy to find a copy of this poem online or in Cummings’ complete poems (Firmage, ed.), and it is impressively difficult to get permission to reproduce it so the following description will have to suffice. 30. Stecker, ‘Expression’, Sec. 3. 31. On music, see Levinson, ‘Music and Negative Emotion’. 32. Goodman, Languages of Art, 57; see Trivedi’s treatment of Goodman’s view in ‘Metaphors and Musical Expressiveness’ (Chapter 3). 33. Goodman, Languages of Art, 86–87. 34. See Davies, ‘Artistic Expression’, 183–184 and The Philosophy of Art, 150; Trivedi, ‘Metaphors and Musical Expressiveness’, Sec. IV. 35. Goodman, Languages of Art, 69–70. 36. Gombrich, Art and Illusion, 368. 37. Goodman, Languages of Art, 92. 38. Kulvicki, On Images, Chs 3 and 4. 39. Gombrich, Art and Illusion; Wollheim, Art and Its Objects. 40. Lopes, Sight and Sensibility, 78. 41. Ibid., 51. 42. Lopes, Sight and Sensibility, 56, 86–87. John Bender, by the way, agrees: ‘Exempli- fication and Expression’, Sec. 1. 43. Lopes, Sight and Sensibility, 52. 44. Ibid., 57. 45. See also Kulvicki, ‘Review’. 46. Goodman, Languages of Art, 50. 47. Dickie, Introduction to Aesthetics, 120–124. 48. Hospers, ‘Art as Expression’, 174. 49. Sibley, ‘Aesthetic Concepts’. 50. Goldman, ‘Aesthetic Qualities’, 28–29. 51. Gombrich, Art and Illusion, Ch. 11.

Bibliography

Arnheim, R. 1954. Art and Visual Perception, Berkeley: University of California Press. Baxandall, M. 1988. Painting and Experience in Fifteenth-Century Italy, Second Edition. Oxford: Oxford University Press. Beardsley, M. 1958. Aesthetics: Problems in the Philosophy of Criticism, New York: Harcourt, Brace and Co. Bender, J. 2005. ‘Esemplificazione ed Espressione: Le Teorie di Goodman sono Sbagliate’, Discipline Filosofiche, XV. 2: 81–96. Budd, M. 1995. Values of Art: Pictures, Poetry, and Music, Hammondsworth: Penguin. Collingwood, R.G. 1938. Principles of Art, New York: Oxford University Press.

June 21, 2008 10:45 MAC/NWA Page-102 9780230_220478_06_cha05 John Kulvicki 103

Davies, S. 1994. Musical Meaning and Expression, Ithaca: Cornell University Press. Davies, S. 2006. ‘Artistic Expression and the Hard Case of Pure Music’ in M. Kieran (ed.), Contemporary Debates in Aesthetics and the Philosophy of Art, Malden, MA: Blackwell. Davies, S. 2006. ThePhilosophyofArt, Oxford: Blackwell. Davis, W. 2003. Meaning, Expression, and Thought, Cambridge: Cambridge University Press. Dickie, G. 1997. Introduction to Aesthetics, New York: Oxford University Press. Dretske, F. 1981. Knowledge and the Flow of Information, Cambridge, MA: MIT Press. Ekman, P. 1972. ‘Universals and Cultural Differences in Facial Expressions of Emotions’ in J. Cole (ed.), Nebraska Symposium on Motivation, 1971, Lincoln: University of Nebraska Press, 207–283. Ekman, P. 1985. Telling Lies. New York: W.W. Norton. Feagin, S. and Maynard, P. (eds) 1997. Aesthetics, New York: Oxford University Press. Firmage, G.J. (ed.) 1994. E.E. Cummings: Complete Poems, New York: Liveright Publishing. Goldman, A. 1990. ‘Aesthetic Qualities and Aesthetic Value’, Journal of Philosophy, 87.1: 23–37. Goldman, A. 2006. ‘There are no Aesthetic Principles’ in M. Kieran (ed.), Contemporary Debates in Aesthetics and the Philosophy of Art, Oxford: Blackwell. Gombrich, E. 1961. Art and Illusion, Second Edition, Princeton: Princeton University Press. Goodman, N. 1976. Languages of Art, Second Edition, Indianapolis: Hackett. Grice, H.P. 1957. ‘Meaning’, Philosophical Review, 66.3: 377–388. Reprinted in Grice, H.P. 1989. Studies in the Way of Words, Cambridge, MA: Harvard University Press, 213–223. Grice, H.P. 1967. Logic and Conversation. Reprinted in Grice, H.P. 1989. Studies in the WayofWords, Cambridge, MA: Harvard University Press, 1–144. Hampshire, S. 1971. Freedom of Mind, Princeton: Princeton University Press. Haugeland, J. 1991. ‘Representational Genera’ in W. Ramsey, S. Stich, and D. Rumelhart (eds), Philosophy and Connectionist Theory, Hillsdale, NJ: Lawrence Erlbaum. Hospers, J. 1954. ‘The Concept of Artistic Expression’, Proceedings of the Aristotelian Society, 55: 313–344. Hospers, J. 1967. ‘Art as Expression’ in The Encyclopedia of Philosophy, Macmillan Library Reference USA. Excerpt in S. Feagin and P. Maynard (eds), 1997. Aesthetics, New York: Oxford University Press, 172–174. James, W. 1884. ‘What is an Emotion?’ Mind, 9: 188–205. Kieran, M. (ed.) 2006. Contemporary Debates in Aesthetics and the Philosophy of Art, Oxford: Blackwell. Kivy, P. 1980. The Corded Shell: Essays on Musical Expressiveness, Princeton: Princeton University Press. Kulvicki, J. 2006. On Images: Their Structure and Content, Oxford: Clarendon Press. Kulvicki, J. 2007. ‘Review of Dominic McIver Lopes: Sight and Sensibility’, Dialogue, 46.2 412–414. Langer, S. 1942. Philosophy in a New Key, Cambridge, MA: Harvard University Press. Levinson, J. 1982. ‘Music and Negative Emotion’, Pacific Philosophical Quarterly, 63: 327–346. Levinson, J. 1996. The Pleasures of Aesthetics: Philosophical Essays, Ithaca, NY: Cornell University Press. Levinson, J. 2006. ‘Expressiveness as Hearability-as-Expression’ in M. Kieran (ed.), Contemporary Debates in Aesthetics and the Philosophy of Art, Oxford: Blackwell.

June 21, 2008 10:45 MAC/NWA Page-103 9780230_220478_06_cha05 104 Artifact Expression

Lopes, M. 2005. Sight and Sensibility: Evaluating Pictures, Oxford: Clarendon Press. Millikan, R. 1989. ‘Biosemantics’, Journal of Philosophy, 86: 281–297. Mumford, J. 1993. The Focus of Blue Light, Composer’s Recordings Incorporated. Prinz, J.J. 2004. Gut Reactions: A Perceptual Theory of Emotion, New York: Oxford University Press. Sibley, F. 1959. ‘Aesthetic Concepts’, Philosophical Review, 68: 421–450. Stecker, R. 2001. ‘Expressiveness and Expression in Music and Poetry’, Journal of Aesthetics and Art Criticism, 59.1: 85–96. Tolstoy, L. 1996. What is Art? A. Mundale (trans.), Indianapolis: Hackett. Trivedi, S. 2001. ‘Expressiveness as a Property of the Music Itself’, Journal of Aesthetics and Art Criticism, 59: 411–420. Vermazen, B. 1986. ‘Expression as Expression’, Pacific Philosophical Quarterly, 67: 196–234. Wollheim, R. 1974. On Art and the Mind, Cambridge, MA: Harvard University Press. Wollheim, R. 1980. Art and Its Objects, Second Edition Cambridge: Cambridge University Press.

June 21, 2008 10:45 MAC/NWA Page-104 9780230_220478_06_cha05 6 A Metaphysics of Creativity Dustin Stokes

I. History, mystery, and neglect

What is special about art? One of the first answers to this question will involve mention that artworks and artists are creative. Set to one side for the moment what this fact amounts to, just assume that there is some truth to it. This fact is at tension with another fact: aestheticians, at least of the analytic school, have said very little about creativity relative to other special features of art. Upon quick perusal of collections in aesthetics from my bookshelf, I find only six entries out of 258 focused centrally on creativity, and all of them except for one focused on genius rather than creativity generally.1 One will, however, find scores of entries on definitions of art, ontology, aesthetic value, and interpretation, among others. If in fact ‘creativity’ is one of the first things that rolls off the tongue in ordinary and critical conversations about art, why is it so grossly overshadowed by these and other topics?2 Part of the answer is found in the history of thought on creativity. Com- mon to both ancient and modern explanations of creativity is a central if not exclusive emphasis on genius. Famously, Plato took the master poets to be conduits for divine inspiration. Homer knew nothing of real charioteer- ing but rather reported whatever his muse inspired him to report. Works of genius derived not from the expertise or skill of the artist, but rather from the divine inspiration they were lucky to have. In the early eighteenth century, Joseph Addison, following the ideas of third-century AD critic Longinus, endorsed a notion of natural genius.3 The natural genius is unconstrained by artistic rules or conventions. In fact, as Peter Kivy recounts Addison’s notion, the natural genius is outside all conventional realms, creating art without any knowledge, a kind of creative primitive, if you will. Addison distinguishes this kind of genius, which echoes the Platonic version, from a learned genius who, lacking the innate capacities of the natural genius, must learn and master his art. Although Addison explicitly claims the contrary, he favours the natural genius as superior since,

105

June 21, 2008 10:51 MAC/NWA Page-105 9780230_220478_07_cha06 106 A Metaphysics of Creativity among other reasons, it is only the natural genius who may create something truly original.4 This marks the importance of novelty for those who follow Addison but at a cost, namely, requiring absolute novelty of creative genius. Kant’s model of artistic genius is developed in his Critique of Judgment.The definition he offers at the start of his discussion of genius is telling. ‘Genius is the innate mental predisposition (ingenium) through which nature gives the rule to art’.5 Genius is thus a natural ability to create artworks of the highest quality, namely, ones which give the ‘rule’ to art. Two points to note on rules for Kant. First, geniuses give the rule to art by creating works from which rules for the (imitative) making of later works may be extracted. And second, geniuses do not—and this is an analytic point for Kant—use rules to create such works; there are no rules for creating works of genius. Rather, such works must be original, giving the rules rather than following them. So Kant too, like Addison, endorsed absolute novelty as a condition on creative genius. Were we to continue through the eighteenth and into the nineteenth century, we would find much of the same. Most theories of creativity from this time are, like those of Plato, Addison, and Kant, theories of genius. From the ancient roots in Plato and Longinus, to the German idealists after Kant, to the romantics, there is an emphasis on radical originality, innate cognitive capacity, and irrationality. It is an understatement to say that these philosophers offered insights into the creation of art: much of their work is essential for the development of modern philosophical aesthetics. However, they tended, explicitly and otherwise, to mystify creativity in a way that thwarts further analysis. The Platonic view chalks creativity up to divine inspiration, stripping the responsibility from the creator and tagging creativity as no more explicable than divine intervention. On neo-Longinian views such as Addison’s, creativity results from a native disposition towards genius. Kant’s view rejects creative use of rules or constraints, requiring absolute novelty. If one were to take any of these views as a kind of explanatory metric for creativity, the prospects for explanation would look grim. They leave us with little illumination regarding what the phenomenon of creativity is, and which features of the phenomenon are the ones that underwrite its importance to art, science, and the lot of human life. And the features that do get the attention are treated in such that mystery is compounded rather than removed. Here are three common features. Creation ex nihilo: Creative ideas, tradition has often had it, come from nowhere. This derives, it seems, from the fact that creative Fs are novel Fs and the supposition that novelty, if it is genuine, is entirely new. It is, of course, another step or two to the inference that novel Fscomefrom nowhere. Suffice it to say that theorists of creativity—Addison and Kant are both examples—have in fact made such inferences, and studies of creativity have suffered (or simply not occurred) as a result.

June 21, 2008 10:51 MAC/NWA Page-106 9780230_220478_07_cha06 Dustin Stokes 107

Flash phenomenology: Creative ideas come to us, just like that,aswesay.This phenomenology is what motivates the Platonic view and its contemporary adaptations. Ideas that come to us in a flash are not ones that we deliberately form and thus are not ones for which we are responsible. Who or what is? Gods. Muses. A euphoric drug trip. And so on. This is discouraging if one wants an analytic or naturalistic explanation of creativity. Incubation: According to a four-stage model of creativity endorsed by a number of natural scientists and psychologists—including Hermann von Helmholtz, Henri Poincare,´ Arthur Koestler, Graham Wallas, and Jacques Hadamard—the initial conscious, preparatory stage of creativity is followed by an unconscious stage of cognitive processing.6 The phenomenon is a familiar one. One is consciously struggling with a problem and then leaves it aside for something else. Upon return to the problem ... Eureka! ...one has the insight needed for a solution. Something important, it seems, happens during this unconscious period of incubation. The creative insight from this stage is then subjected to evaluation and criticism in the final stage. Analyses of incubation appeal to Freudian egos or unconscious automata, typically focusing on the seemingly random and uncontrolled combination of ideas.7 There certainly is truth to the thought that incubation occurs and that it is important to creativity. Tagging it as the work of the unconscious self, however, blocks theoretical angles from analytic aesthetics, naturalistic philosophy of mind, and much of cognitive science.8 There are also some theoretical choices which exacerbate the task of analysing creativity. First, as already indicated, theorists often take as their examples, geniuses or masterminds. This is understandable: someone asked to explain creativity may just list figures like Bach, Beethoven, Picasso, Van Gogh, and so on. Creativity and genius are importantly related concepts. However, taking a genius like Bach or Picasso as the departure point for an analysis of creativity increases the complexity of the task right from the gates. Second, creativity, and this too is understandable, is typically explained from within the context of some particular artistic (or scientific) domain. So many a book has been written on musical genius or, even more specifically, the creativity of Bach or Mozart. This too may be problematic, since it is difficult to distinguish the properties that are specific to the creativity (whatever that should turn out to be) from those that are features of the artistic domain or genre. Creativity does not occur in a vacuum, so context- specificity is not unmotivated. But one does better to isolate the general phenomenon, if there is one. Finally, theorists often fail to make an important distinction. To what categories of things do we attribute ‘creativity’? We talk about creative artworks. We talk about creative artistic processes. And we talk about creative artists. Are creativity attributions the same, no matter the kind of attributee? And is there one kind of attribution that is fundamental?

June 21, 2008 10:51 MAC/NWA Page-107 9780230_220478_07_cha06 108 A Metaphysics of Creativity

The following account attempts to identify the fundamentals of creativity and its situation in a broader theory of aesthetics. The underlying spirit is minimal, beginning at the bottom rather than the top, and pragmatist, taking actual critical and appreciative practice as a theoretical constraint. A theory of art must take as its objects of enquiry and analysis whatever our best practices of criticism and appreciation take as their objects. ‘Ontology of art is in this way answerable to epistemology of art’, as David Davies puts the point.9 This is endorsed as a general methodological principle. Call it, following Davies, the pragmatic constraint.10 The account centres around answering the question of attribution. To what category or categories of things do we properly attribute creativity? The suggestion is that creativity is, most basically, attributed to a process which culminates in an accomplishment. The next step, then, is to say what typ- ifies the artistic creative process, and how this is informed by and informs philosophical aesthetics. Just these few steps take us far.

II. The question of attribution

Is there one category of thing most basic for predication of creativity? We have three candidates: persons, products, and processes.

Persons As suggested in the introduction, historically, the bulk of studies of creativity have been studies of radically creative persons, geniuses. This is perhaps reasonable given our ordinary usage of ‘creativity’ and its cognates. If I were to ask you what ‘creative’ means, you might provide an ostensive definition, ostending the masters before their masterworks.11 There is another intuitive reason theories focus on genius: geniuses, however we try to explain them, are intriguing and inspiring. There are few things more romantic and mysterious than the artistic genius at work. So why focus on the banal when we have Bach and Beethoven? However, there are problems with analyses that focus on creative persons as the central subject of creativity, some of them already alluded to. From Plato to at least the nineteenth century, with few exceptions, the creative genius is portrayed as a tortured, irrational, divinely inspired soul. The abilities to create masterful artworks are, on such accounts, often innate and out of the control of the genius, creative insights often coming in dreams or reveries. Finally, these are the artists that make the rule for art, in Kant’s terms, and thus the level of novelty required for a work of genius is maximal. All of this makes for challenging explanatory work, as it takes only a little magic to make a lot. It is thus that for most theories of genius, the explanation goes no further. The exceptional features of the genius are tagged, but we are offered no deeper insights into how a genius, even if divinely inspired, goes about creating. So one might acquiesce in an inspirationalist or nativist

June 21, 2008 10:51 MAC/NWA Page-108 9780230_220478_07_cha06 Dustin Stokes 109 model, but for many of us this does not suffice as an explanation; we are left wanting more. Perhaps geniuses are inexplicable. But this may just be all the worse not for theories of creativity in general, but for theories of creativity that focus on genius. Let us be more charitable. Imagine, for the moment, that we have a truly informative theory of radically creative persons. Why would we be interested, either from an aesthetic or philosophical standpoint, in such a theory? Put another way, why do we value geniuses? Whatever our answer is, it seems that we do not value geniuses, as it were, for genius’s sake. Geniuses (or creative persons) possess no intrinsic value qua geniuses. Rather, we value geniuses for one of two reasons. We may value a genius for the artworks she creates. Alternatively, but not exclusively, we may value a genius for the artistic processes used in creating art. And so, to answer the first question, we would expect from a theory of genius some illumination on the connection between geniuses and their products and/or between geniuses and their creative processes. An account of creative persons, geniuses or less, must therefore place some emphasis on creative products and creative processes. Perhaps then, one can give an informative theory of creativity that focuses on genius. But if that theory is to shed light on the things for which we value genius, a focus just on the person will not suffice; the theory will have to analyse either the creative products or the processes, if not both.

Products We, no doubt, talk about artworks as creative, and it seems appropriate to do so. Innovations like Monet’s impressionism, Picasso’s cubism, the poetry of E.E. Cummings or Emily Dickinson, and Gaud´ı’s arches are nothing if not creative. But what do we mean when we say this? According to a view in philosophical aesthetics from the middle of the last century, associated most closely with Monroe Beardsley, one thing we do not mean is that the works are aesthetically or artistically valuable because they resulted from some creative process. Beardsley argued just the opposite, namely that the value of a work of art consisted solely in the formal properties of the manifest work and the experience of those properties.12 The implication here is twofold. First, we enjoy and value artworks not for the modes of production that generate them. To think otherwise is to commit what Wimsatt and Beardsley famously call the Intentional Fallacy. In evaluating or interpreting an artwork, appeals to an artist’s intentions, designs, biography, or the context of presentation of the work in question are fallacious.13 Beardsley does not stop with aesthetic evaluation, but extends his anti-intentionalism to creativity. ‘The true locus of creativity is not the genetic process prior to the work but the work itself as it lives in the experience of the beholder.’14 This second claim carries normative import: we should attribute creativity to artworks as products and not to the processes that begot them. The strength of this claim is not to be understated. The value

June 21, 2008 10:51 MAC/NWA Page-109 9780230_220478_07_cha06 110 A Metaphysics of Creativity of an artwork, Beardsley writes, ‘is independent of the manner of production, even of whether the work was produced by an animal or by a computer or by a volcano or by a falling slop-bucket’.15 Gaut and Livingston take Beardsley’s strong anti-intentionalism to be one of the movements of twentieth-century aesthetics and criticism which con- tributed to a revitalization of formalism and, in turn, to the ‘occlusion of creativity’.

In the New Criticism’s break with both common-sense biographical criticism and those versions of biographical criticism based on existentialism, hermeneutics, and phenomenology, a leading idea was that an appropriate form of aesthetic appreciation requires the critic to focus entirely on the finished text’s or other artistic structure’s inherent, artistically relevant features ...Facts about the text’s provenance were to be set aside, especially whenever such facts were a matter of the ‘‘private’’ psychology of the creator, held to be unknowable or irrelevant.16

Structuralism and post-structuralism, although at odds with New Criticism in other ways, nonetheless shared this, one of its central tenets: the artist and the circumstances of artistic creation are not aesthetically or critically relevant. This broad anti-intentionalism long dissuaded the attention of aestheticians from the creation of art. We should distinguish product approaches to aesthetic value from product approaches to creativity.17 Anti-intentionalism is more typically presented as a theory of the former kind.18 And indeed it is as a theory of aesthetic value, sometimes termed ‘aesthetic empiricism’, that it is more plausible.19 There is an obvious sense in which the work may become overshadowed, in criticism or evaluation, by biographical information about the artist and her modes of production. So there is something to be said for isolating just the work and its perceptible properties, as both the anti-intentionalist and formalist suggest.20 The relevant question, of course, is whether this is the correct or best method of aesthetic evaluation and appreciation. This question, at the centre of a long-standing debate, will not be addressed here. There is, however, one type of view in opposition to a product approach to aesthetic value that provides some insight into how one might respond to product approaches to creativity. A number of ontologies of art take artworks to be, in some sense, events. According to Gregory Currie, for example, artworks are the discovery, by an artist, of a structure (of colours, sounds, words, and so on) by way of a certain heuristic path. As such, the mode of discovery, how the artist produces the work in question, is partly constitutive of the work; artworks are thus event-types.21 David Davies argues instead that artworks are event-tokens; they are performances which include the generative actions that culminate in the end product. For Davies, the latter provides the ‘focus of appreciation’, which embodies the creative achievement of the

June 21, 2008 10:51 MAC/NWA Page-110 9780230_220478_07_cha06 Dustin Stokes 111 artist.22 Both ontologies directly oppose any broad product approach, since the artwork is not exhausted by the formal properties of the finished product. How that product was created is constitutive of the artwork and thus is part of what we aesthetically value. This general lesson should be kept in mind in considering product approaches to creativity. How, then, does a product approach do as a theory of creativity? It is hard to see how it could do very well. Rather, it looks like a forced consequence of a view like anti-intentionalism. If one concludes with Beardsley that facts about the process of artistic production are not aesthetically relevant then one faces the following dilemma. Either (a) creativity includes features of artistic process and production and is therefore not aesthetically relevant, or (b) creativity does not include features of artistic process and production and therefore may be aesthetically relevant. Neither (a) nor (b) are attractive options. Horn (a) violates the pragmatic constraint: its abandonment of creativity as aesthetically irrelevant is inconsistent with critical practice and appreciation. Horn (b) denies basic conceptual intuitions about creativity. Beardsley opts for (b), arguing that creativity should instead be located in the finished work (and perhaps the experiences that audiences have with that artefact). Beardsley’s position seems to be a consequence of his theoretical commitments, as captured by the following line of reasoning.

(P1) If some F is a feature of or includes features of artistic process or intention then F is not aesthetically relevant. (P2) Creativity is relevant in aesthetic evaluation and critical discussion of art. (C) So, creativity is not a feature of and does not include features of artistic process or intention.

(P1) is just a rough articulation of Beardsley’s anti-intentionalism. (P2) is motivated by the fact—which Beardsley acknowledges—that art appreciators and critics alike talk about creativity in their evaluations of artworks. This fact also survives rational reflection, thus meeting the pragmatic constraint. The conclusion—which is the negative component of horn (b) of the dilemma— follows straightforwardly from (P1) and (P2). The argument is thus valid, but the conclusion is counterintuitive. Let us explore some cases that underwrite that counterintuition, and then reconsider the premises. We are standing before an early impressionist painting, say Monet’s Impres- sion, Sunrise. You say to me, among other things, that the work is genius, truly creative. I ask why, that is, what makes it creative? You might, in your early response manage to report features of the painting itself and how they are especially novel relative to the prior history of painting. So you might note the emphasis on light and shadow, the vivacity of the colours, the

June 21, 2008 10:51 MAC/NWA Page-111 9780230_220478_07_cha06 112 A Metaphysics of Creativity fact that the sun is of nearly the same luminance as the surrounding grey clouds. In justifying your attribution of creativity, however, it is likely that you would describe impressionist techniques. You are likely to mention the short, loose brushstrokes used; the use of pure (unmixed) paints side by side (so that the viewer does the mixing, as it were, to create the impressions of mixed colours); the placing of wet paint on wet paint. All of these features, among others, are typical of the process of impressionist painting; they are the innovations of the artistic movement. And that is just the point: in giving reasons for attributing creativity to the Monet painting you have, quite naturally, invoked features of Monet’s process of creation. And not only is this explanation natural, it or something like it is needed. Without mention of these features, your explanation would fall flat. But by invoking them, you have justified your attribution. This case is not particularly special. We might have a similar kind of discussion about a cubist or fauvist painting, or Bach’s Brandenburg Concertos or Well-Tempered Clavier, a Henry Moore sculpture, George Martin’s four-track recordings of the Beatles, Kurosawa’s Seven Samurai,Capote’sIn Cold Blood, and so on. To make sense of the creativity of these works, one must say something about the generative processes from which they resulted. To overlook details of the generative process is to overlook the creativity in the works. It is thus hard to swallow the conclusion that the artist’s creative process is not (at least partly) constitutive of the creativity of her artwork in the light of cases like these. (C) is thus, as initially expected, to be doubted. Observing our practices, creativity does include and indeed perhaps just is the features of artistic production and intention. The anti-intentionalist, however, reaches (C) as a natural extension of his view, as embodied in (P1). And here lies the dilemma. Coupling (P2)—the rationally considered intuition that creativity does matter to aesthetic evaluation—with (P1) the anti-intentionalist is stuck with (C). He thus must embrace horn (b): creativity is not located in artistic process but solely in the finished work and our experience with it. The product approach may instead deny (P2) and opt for horn (a). This is to admit that creativity involves features of artistic process and intention while denying the aesthetic relevance of those features, and thus, of creativity. This choice comes at the cost of inconsistency with appreciative and critical practice. The anti-intentionalist is stuck with one or the other option in virtue of his commitment to (P1). This looks problematic for a product approach to creativity and, more broadly, for a product approach to aesthetic (or artistic) value. Thus, what began as an analysis of anti-intentionalism qua theory of creativity results in a reductio of the broader position, since it only takes acknowledgement of the plausible (P2) coupled with the basic thesis of the view (P1) to support the unintuitive consequence (C). The only way out, it seems, is to deny the problematic (P1). And this is just to abandon anti- intentionalism.

June 21, 2008 10:51 MAC/NWA Page-112 9780230_220478_07_cha06 Dustin Stokes 113

Process Let us return, one last time, to the initial question: when we talk about creativity what features of an artwork and its generation are we talking about? Consider once more the Monet example discussed above. Reflection on this example revealed that in giving reasons for an attribution of creativity to the Monet painting, one would invoke features of Monet’s generative process: how he actually painted Impression, Sunrise. One might mention features of Monet’s brushstrokes, his use of the medium of paint, his use of light and colour. One might also invoke features of Monet’s thought process: perhaps he intended to capture an impression, an instant rather than just a place; he thus chose an instant before him and then attempted to depict it is as seen, in its natural light. One might also mention how Monet broke from his own tradition, both in intention and action: the goals of impressionists were consciously at odds with artistic tradition and their works manifest this fact. Albeit a rather hefty rational construction, this is the kind of story one should tell if one wants to give a thorough explanation for the creativity of Monet’s piece. Note the various kinds of events and properties that one has invoked in doing so: how does one choose among them a locus, in Beardsley’s terms, of creativity? One does not have to. Instead, this variety should be maintained: artistic creativity is a process, and a variegated one at that. This should come as no great surprise. The creativity that resulted in Impression, Sunrise was not one intention, imagining, brushstroke, study, choice of subject. It was a process which involved each event and more. We do, of course, sometimes speak of singular creative thoughts or creative actions, but our appreciation of artistic creativity takes wider scope. It would be unsatisfying to be told only that Impression, Sunrise is creative because, say, Monet made a decision to do no pre-mixing of paints or because he formed an intention to capture the luminance of the sun as he saw it at that instant. Each event may be essential to Monet’s achievement, but as much as each of them is needed, all of them are needed. Without that decision and the corresponding action, and without that intention, among several other thoughts and actions, Monet would not have made the work he did. We should think of creativity not in terms of single events or properties, but rather in terms of process. Processes are continuous: sequentially structured in stages or phases. They are thus particulars, but with both temporal and spatial parts. Processes are the kind of thing that perdure; they are not wholly located at any one time. Events, on most accounts, have temporal parts. Perhaps the ontology of events will illuminate an ontology of processes. Metaphysicians distinguish different kinds of events. Zeno Vendler distinguishes achievements from states.23 Achievements are instantaneous: one

June 21, 2008 10:51 MAC/NWA Page-113 9780230_220478_07_cha06 114 A Metaphysics of Creativity reaches the top of the mountain or finishes the 5 lb. hamburger. States— like loving, knowing, hoping—last for some period of time. States are homogeneous—any part of a state satisfies the same description as the whole. Neither achievement terms nor state terms denote a procession of stages, as in a process; they lack, as Vendler puts it, continuous tenses. They are thus not processes. Activity terms and accomplishment terms, however, do possess continuous tenses. These two types of events are distinguished along two dimensions: homogeneity and culmination. Activities are homogenous. And activities are non-culminating: there is no terminus upon which the truth of the predication of the activity term depends. Accomplishments differ in both respects. They are non-homogeneous: the description of the whole event will not appropriately apply to any sub-part of the whole. And accomplishments are culminating, proceeding towards a terminus the occurrence of which justifies predication of the accomplishment term. Vendler compares the activity of running with the accomplishment of running a mile.

If it is true that someone has been running for half an hour, then it must be true that he has been running for every period within that half-hour. But even if it is true that a runner has run a mile in four minutes, it cannot be true that he has run a mile in any period which is a real part of that time, although it remains true that he was running, or that he was engaged in running a mile during any substretch of those four minutes ...It appears, then, that running and its kind go on in time in a homogeneous way; any part of the process is of the same nature as the whole. Not so with running a mile or writing a letter; they also go on in time, but they proceed toward a terminus which is logically necessary to their being what they are. Somehow this climax casts its shadow backwards, giving a new color to all that went before.24

Both activities and accomplishments are processes in the sense that they involve successive stages. However, processes require, by definition, that something is processed, that is, that there is some output or terminus. Given this last desideratum on processes, accomplishment is the event category most adequate to characterize processes. And in fact, artistic creative processes best fit the category of accomplishments, albeit imperfectly. ‘Cre- ating an artwork’ denotes something non-homogeneous. By analogy with running, if it is true that an artist has created an artwork over a year’s time, it will not be true that she has created the work at any proper part of that time (although we can naturally say that she was engaged in creating during those times). And a creative process is culminating: only when an artist has succeeded in making a work, can we say that she has created art. That is, it is only then that the accomplishment ‘casts its shadow backwards’ upon the process that begot it, and then that we can attribute creativity to the artist’s activities.25

June 21, 2008 10:51 MAC/NWA Page-114 9780230_220478_07_cha06 Dustin Stokes 115

Understanding artistic creativity in this way has its flaws. Accomplish- ments, and more generally events, are ordinarily interpreted as independent or separate occurrences: she accomplished such-and-such, this (event) occurred. So our speech at least indicates that we take such events to be the results of processes, not the processes themselves. Moreover, describing an F as a process, as contrasted with describing it as an event, contextual- izes the parts of F; it is thus understood as having stages, each of which is contextually situated between prior stages and stages forthcoming. Second, accomplishments necessarily involve a terminus. And at least typically, this terminus is consciously targeted: writing a letter, running a mile, and bak- ing a cake all have a clear culmination occurrently tokened in the mind of the accomplisher. The latter is less true for artistic creative accomplishment, if accomplishment is the right category. An artist will rarely have in mind such a clearly defined terminus. She may have some emotion he wishes to express, or idea to represent, or medium to explore, but the finishing point or product, when it arrives, is largely unforeseen. It is an open question whether this makes creativity a special kind of accomplishment, or of a distinct category. The foregoing analysis clarifies some of the special, fundamental features of artistic processes and the terms used to describe them. The identified features imply some desiderata for an ontology of creativity. Creativity is not a homogeneous object, property, or event. Nor is it wholly located at one time or other. Finally, it involves both the culminating event and the stages that lead up to that event. ‘Process’ likely remains the best choice of term, but ‘accomplishment’ brings out the fact that creative processes are ones that process towards some end. Without the end, the process is not a creative one; and without the process, there is no end. Artistic creativity is thus a spatio- temporal package, perduring towards and until the culminating artwork is made. This approach comports well with our practices of criticism and appreciation, and for the same reasons that the product approach failed. We do not, when attributing creativity to some work of art, attribute it only to the product before us, that is, the manifest properties. Our appreciation is (partly) of the artist’s having undergone a certain process to make that product. And moreover, our practice acknowledges the fact that the product possesses interesting properties because there was a special process that generated it. Thus we value generative processes not merely as instrumental to some end. As Richard Wollheim has it, appreciating a work of art is to attempt to retrieve the process that generated it. When engaging with an artwork, we ask, suggests Denis Dutton, ‘what has been done here?’26 Wollheim and Dutton are suggesting a general mode of appreciation. It is debatable whether all appreciation of art is backward looking in this way. Whatever the truth of such suggestions, the only point made here is this: in appreciating creativity, we do ask Dutton’s question. And this is the question we answer in justifying an attribution of creativity.

June 21, 2008 10:51 MAC/NWA Page-115 9780230_220478_07_cha06 116 A Metaphysics of Creativity

III. Conditions on creative process

Creativity is a process, but what kind of process is it? One can read studies and theories of creative process written by artists, art theorists, critics, art teachers, philosophers, psychologists, business management firms, and self-help publications, among others. It might be hard to see, if one selects two books from two of these approaches, how they are even talking about the same phenomenon. How, then, can we say something general about the processes to which we rightly attribute creativity, given volumes of such theoretical variety? The answer: toss them out. Reinvention has its advantages. Rather than incorporate case studies of geniuses and analyses of creativity in narrow artistic and scientific domains, we do better to begin with minimal conceptualizations of creative process.

Agency What, at minimum, characterizes a creative process? First, a creative process needs a processor, and one with agency: we do not attribute creativity in the absence of a responsible agent. Consider our linguistic intuitions on the following. If you were to say to me, ‘That sunset is creative’, I might pour you another drink or nod out of charity, but I could rightly take issue with your use of the term ‘creative’. Sunsets may be beautiful or vibrant, or stunning, but not creative. Your utterance is thus conceptually problematic, and for the same reason that the following is problematic. ‘The Starry Night is creative but no one is responsible for it.’ After a moment’s reflection, the misuse of ‘creativity’ reveals itself. Creative artworks are things that are done and made, and for which we praise their makers. The processes that generate them involve intentional agency and it is this process, at least in part, for which we praise the agent. This implies that the process must depend in some non-trivial way upon that agency. We do not appropriately praise (any more than we blame) agents for processes out of their control. We capture this intuition with a simple condition on creativity, call it the agency condition. Some F is creative only if F counterfactually depends upon the agency of an agent A. The notion of agency at work and the relation between F and A will, for a complete analysis of creativity, need to be made precise. Philoso- phers typically require a cognitive and deliberative capacity for agency, while cognitive scientists require less, perhaps only self-governing or self-moving autonomous behaviour. The strength of agency at work in the agency condition depends upon what one thinks about creativity and responsibility, and how cognitive a phenomenon one takes it to be. Given an interest primarily in artistic creativity, how responsible are artists for their work? Do they fore- see the end results in detail? Do they have a clear problem in mind? A clear strategy or method? How many accidents do we allow before we strip the attribution of creativity and call it a mere happy accident instead? Answers

June 21, 2008 10:51 MAC/NWA Page-116 9780230_220478_07_cha06 Dustin Stokes 117 to some of these questions may be obvious; others are not. But they are the questions that a theory of creativity must answer to build a minimal model of creative process. Agency is fundamental to creative processes, artistic, or otherwise. An agency condition is thus specified as one necessary condition for the concept, without being sharpened further.27

Novelty Considered intuitions secure another feature of creative process: novelty.Cre- ativity implies novelty, that is, novel ways of making and doing. Artists succeed by working in new and original ways and thinking new thoughts. Thus, an F is creative only if F is novel. Identifying novelty as a condition on creativity, however, only introduces a new task of analysis: what is novelty? One might think novelty is just newness simpliciter.SosomeF is creative only if F has never occurred before. Perhaps, on first glance, this characterizes paradigmatic cases of radical creativity or genius. But specifying categorical novelty as a condition on general creativity has its problems. First, the suggestion is simply incomplete. One needs to specify the sense in which a creative F is novel simpliciter, answering at least the following questions. Is every event in or property of F novel? Or is it merely necessary that some number or percentage of properties be novel? Or just one? Just essential ones? Assume these questions can be addressed and set them to one side. Margaret Boden provides a relevant and useful distinction between historical novelty and psychological novelty.28 An F (Boden speaks of ideas) is historically novel if and only if F is novel with respect to the entire history of ideas. An F is psychologically novel for some agent if F is new relative to the mind of that agent (i.e., the agent tokening the relevant thought or performing the relevant action). The first is categorical novelty, the second, relative novelty. Does categorical novelty exhaust our interest in artistic creativity? Consider an example. Imagine a young artist in the mid-1960s, working on an island somewhere in the eastern part of the globe. The island is technologically modern but culturally remote, its population ignorant of the various artistic movements of the western world. Call the artist Q. Q has developed his own style, which can be characterized as follows: On very large canvasses, Q paints blurred rectangular blocks in two or three vibrant colours typically positioned on a monochrome background. The blurred blocks seem to float on the canvass, dividing it into sometimes opposing, sometimes complementary spaces. The viewer seems invited into the painting, almost hypnotized by its contrast of depth and simplicity. Now, if we were to come upon the work of Q, we would instantly think of Mark Rothko. Given that Rothko had already established the ‘multiform’ style as his signature in the 1950s, we would withhold an attribution of historical novelty to Q’s work. We would, however, find Q’s work novel, and presumably creative, in rich ways, and this in spite of the fact that Q’s style is not novel simpliciter.First,Q’s

June 21, 2008 10:51 MAC/NWA Page-117 9780230_220478_07_cha06 118 A Metaphysics of Creativity accomplishments are psychologically novel in Boden’s sense. But they are novel in a broader sense, namely, relative to the population of which Q is a part. Q’s artistic achievement is what we might call population-relative novel. We, even if members of the elite New York art world of Rothko’s time, would (given the appropriate information about Q’s circumstances) recognize this and would praise Q accordingly. Categorical novelty thus is not the only kind relevant in appreciation of art. This acknowledgement is consistent with our practice. When praising an artist’s creative achievement, we will say that so-and-so’s work was ‘so new for her time’ or it ‘was an innovation of the period’ .An attribution of creativity is often relativized to context, and that for the reason (at least in part) that the novelty is so relativized. In Kendall Walton’s well-known hypothetical society of Guernica makers, a two-dimensional Guernica (like Picasso’s) may be novel, but not for the reasons that it is novel in our art world. It would be novel not for its sharp, jagged, and violent figures (as it is in our society) which would be standard properties of such works in the Guernica making society, but for its flatness, a variable property for the Guernica category. Thus, as Walton suggests, Picasso’s Guernica would strike this hypothetical art world as ‘cold, stark, lifeless, or serene and restful, or perhaps bland, dull, boring—but in any case not violent, dynamic, and vital’.29 And so the same productive behaviour may be counted as novel relative to one population for one set of reasons, and novel (or not) relative to another population for another set of reasons. A creative process, then, requires novelty, where the degree or kind of creativity mirrors that degree of novelty. This is not yet enough. For example, I can step away from my computer and out into the department hallway, hop on one leg and tug on the opposite ear while repeatedly yelling ‘ ‘‘Nixon’’ IS a good name for a dog’. This behaviour results from my own agency and is novel relative to, well, just about anything. Goofy behaviour it is; creative it is not. So the agency and novelty conditions are insufficient to characterize a creative process.

Value? A number of theorists have considered creativity to be an obviously value- laden concept, and have thus made value a central or the central criterion for creativity.30 There are a number of ways that one might incorporate such a criterion. Creative processes may involve evaluation on the part of their agents. This, though likely, is hardly informative. Insofar as creative agents are thinking when creating, and evaluating a situation is a way of thinking about that situation, creative processes involve evaluation. A more promising inclusion of value in a theory of creativity is to make value a condition of the process and/or product, from the perspective of the audience. So an F is creative only if F is valued or to be valued.

June 21, 2008 10:51 MAC/NWA Page-118 9780230_220478_07_cha06 Dustin Stokes 119

This may be right, and indeed it would be consistent with the pragmatic analysis offered to this point. One worry is that this commitment invites a number of very challenging open questions: What kind of value are we talking about? Artistic? Aesthetic? Cognitive? Are creative Fs objectively valuable, or mind-dependent? Intrinsically or instrumentally valuable? Endorsement of a value criterion thus incurs a heavy analytical burden. This, of course, is no reason to reject the suggestion: one does not reasonably make theoretical decisions based on the ease of analysis. There is, though, a more fundamental reason to be wary of the theoretical purchase of a value criterion. Assume one did endorse something like the bald value condition suggested just above. What would one know about the nature of creative process that one did not already know with the agency and novelty condition in hand? Knowing that something is valuable or to be valued does not by itself reveal why or how that thing is valuable. By analogy, being told that a carburator is useful provides no explanatory insight into the nature of a carburator: how it works and what it does. And it is the latter kind of story that is needed, at least for a start, in theorizing creativity. If one finds the value criterion plausible, one may endorse it and with it the obligation to address the open questions mentioned above and to explain further why creative Fs are valuable. If one does not find the value criterion plausible, one may reject it and maintain a descriptive account of creativity. The present, modest analysis is non-committal to but consistent with either choice, closing with a final suggestion aimed at explaining why creative processes are what they are.31

Cognitive change and possibility Agency and novelty are not sufficient for creativity. An analysis of creative processes thus needs an additional condition. As just discussed, some have opted for a value condition. Set value to one side, however, and consider the following observations. Thought is generally systematic. What contents one has tokened and how one has tokened them depends upon one’s broader cognitive profile. Fur- ther, what one can think depends upon this profile. Finally, the actions one performs, and indeed the actions that one can perform, depend upon one’s cognitive profile and the skills that one possesses (among other situational and environmental circumstances). Now, creativity requires novelty. And novel thoughts and actions may require the acquisition of new knowledge or skills, the imagining of a hypothetical situation or consequence, the visu- alization of actions, and so on. When one makes the latter kinds of changes, one changes with it one’s modal profile. Thus what one can do and think changes with what one has done and thought. One might take this relationship between cognitive novelty and cognitive change to imply modal facts about creative processes. So, if creative processes involve thoughts and/or actions novel for some agent A—and, given

June 21, 2008 10:51 MAC/NWA Page-119 9780230_220478_07_cha06 120 A Metaphysics of Creativity the systematicity of thought and action, this novelty requires of A certain cognitive and behavioural changes—then the novel thoughts and actions could not have been tokened by A before the time they in fact were tokened. They were, for that agent, nomologically impossible (made possible only by the relevant changes). Put most strongly, one might set up a modal condition as follows. An F is creative only if F could not, relative to the cognitive profile of the agent in question, have been done or performed before the time it actually was. Such a modal condition accommodates the intuition that a creative process seems typified by changes in an actual cognitive profile which, in turn, enable changes to the corresponding modal cognitive profile. A cognitive profile can be individuated at lesser or greater fineness of grain. The level of mental tokens (and individual actions) is likely too fine, as some thoughts or actions will be relevant to the possibility of a creative advance, while others may be causally efficacious but clearly not necessary. More coarsely, cognitive profiles can be individuated at the level of overall organization, how a set of thoughts and actions relate with one another, and with certain circumstances. This would be in line with the general process conceptualization of creativity; for any creative process, there will be an organization of thoughts and actions essential to that process, a kind of heuristic path that leads to the culminating accomplishment. A modal condition could be finessed along a number of dimensions, but some condition like it, broadly construed, looks promising. Com- bined with agency and novelty, it provides a minimal characterization of a creative process, emphasizing that creative accomplishment requires cognitive change on the part of an agent. This implies—as the next step for an analysis of the creative process—the development of a cognitive architecture of creativity. Contrary to much of theoretical and popular tradition—which talks about the creative process and the stages of creativity— creative thinking does not consist in one cognitive capacity or one set of capacities. Instead, a creative process involves a complex of cognitive capacities and skills. An architecture of creative cognition would begin to identify the possible roles, structures, and relationships that compose that complexity.

Acknowledgements

Special thanks to Dom Lopes and Kathleen Stock for helpful comments. This paper was initially drafted while a member of the research project Creativity, Cognition, Computational Intelligence, and Aesthetics: A Multidisci- plinary Investigation, funded by the AHRC and centred at the University of Sussex.

June 21, 2008 10:51 MAC/NWA Page-120 9780230_220478_07_cha06 Dustin Stokes 121

Notes

1. Texts consulted: Alperson, The Philosophy of the Visual Arts; Feagin and Maynard, Aesthetics; Gaut and Lopes, The Routledge Companion to Aesthetics; Kivy, The Black- well Guide to Aesthetics; Korsmeyer, Aesthetics: The Big Questions; Lamarque and Olsen, Aesthetics and the Philosophy of Art. 2. See Gaut and Livingston on the ‘occlusion of creativity’, ‘The Creation of Art: Issues and Perspectives’, 1–6. 3. Addison and Steele, The Spectator. The relevant text from Longinus is On the Sub- lime. However, according to Peter Kivy, the notions of genius adopted by Addison are likely not those of Longinus but rather of an earlier critic who has become known as ‘pseudo-Longinus’. See Kivy, The Possessor and the Possessed, 13. Much of the above discussion is indebted to Kivy’s text, which provides a discussion and analysis of the ideas of Addison and Longinus. 4. See Kivy, 22–36. 5. Kant, Critique of Judgment, 174. 6. Hadamard, An Essay on the Psychology of Invention in the Mathematical Field; Helmholtz, Vortrage¨ und Reden;Koestler,The Act of Creation; Poincare,´ Science and Hypothesis; Wallas, The Art of Thought; Mendolsohn, ‘Associative and Attentional Processes in Creative Performance’, Martindale, Cognition and Consciousness and ‘Creativity and Connectionism’. 7. For example, Poincare,´ as cited by Koestler (The Art of Creation, 164). See also Poincare,´ Science and Hypothesis. 8. There has been recent naturalistic work on incubation. For experimental research, see Smith and Blankenship, ‘Incubation Effects’ and ‘Incubation and the Persis- tence of Fixation in Problem Solving’. From the angle of philosophical psychology, see Stokes ‘Incubated Cognition and Creativity’. 9. Davies, Art as Performance, 18. 10. Many, perhaps most, analytic aestheticians endorse such a principle. For example, Currie, An Ontology of Art; Danto, Transfiguration of the Commonplace;Davies,Defi- nitions of Art; Levinson, Music, Art, and Metaphysics;Stecker,Artworks; Thomasson, Fiction and Metaphysics, among others. 11. Note that ordinary usage is not limited to any of the three suggested candidates. In ordinary circumstances, we call artworks themselves creative. And we describe certain artistic processes as creative. So ordinary usage, insofar as it ever does so, is not going to adjudicate between the adequacy of person vs product vs process approaches. 12. Beardsley’s discussion shifts between talk of aesthetic experience to talk of the value of artworks and, in some places, even to talk of criticism of art. Given his aesthetic empiricism and general philosophy of art, ‘aesthetic value’ is the relevant term. If, however, one makes a distinction between aesthetic value and artistic value, one may safely substitute ‘artistic value’ or the disjunction ‘artistic or aesthetic value’ for ‘aesthetic value’ in the discussion and critique of the product view. 13. Wimsatt and Beardsley, ‘The Intentional Fallacy’. 14. Beardsley, ‘On the Creation of Art’, 302. 15. Ibid., 301. 16. Gaut and Livingston, ‘The Creation of Art: Issues and Perspectives’, 3. 17. Creativity may be an aesthetic value, but it is certainly not the only one and, moreover, may be characteristically different from other aesthetic values. Or, creativity

June 21, 2008 10:51 MAC/NWA Page-121 9780230_220478_07_cha06 122 A Metaphysics of Creativity

may not be an aesthetic value at all (this is, in some sense, how one might read Beardsley after all), in which case all the more reason to separate the treatments of general aesthetic value and creativity, vis-a-vis` the product approach. 18. In what follows, take Beardsley’s anti-intentionalism as the paradigm of a product approach, for both aesthetic value and creativity. 19. For a critique of ‘aesthetic empiricism’, see Currie, An Ontology of Art.Seealso Davies, Art as Performance, 25–49. 20. Note, however, that putting the point this way begs one of the relevant questions, namely just what a work of art is. On some accounts, as discussed below, the artwork is, in part, the genetic process which led to the manifest object or event. One must be careful, then, when talking about the artwork. When needed for clarity and/or theoretical neutrality, the terms ‘manifest work’ or ‘work’ will be used to denote the physical artefact. 21. Currie, An Ontology of Art. 22. Davies, Art as Performance. 23. Vendler, ‘Verbs and Tenses’. See also Ryle, Concept of Mind; Casati and Varzi, ‘Events’. 24. Vendler, ‘Verbs and Tenses’, 145–146. 25. This is a simple but elusive point. The running analogy may help once more. If S is engaged in the activity of running, S can stop at any time and the following proposition will be true (at that time): ‘S has run’. However, if when attempting to run a mile S stops before finishing, the following proposition is false: ‘S has run a mile’. The same goes for creating artworks. If the artist were attempting to create an artwork, but were to stop before the terminus, then we cannot say of her (for any of the relevant times) that ‘She created an artwork’. The truth of this proposition requires creative accomplishment, not merely artistic activity. 26. Wollheim, Art and its Objects; Dutton, ‘Artistic Crimes’. For further discussion of this feature of appreciation, see Davies, Art as Performance, 27–28. 27. Note that this is no trivial matter. Specification of an agency condition puts the present analysis at odds with much of the tradition, namely inspirationalist accounts like Plato’s, nativist accounts like Longinus’ and Addison’s, and irra- tionalist accounts like Schopenhaeur’s (see Schopenhaeur, The World as Will and Representation). 28. Boden, The Creative Mind. 29. Walton, ‘Categories of Art’, 347. 30. This is true of both philosophers and scientists. See, among others, Boden, The Creative Mind; Novitz, ‘Creativity and Constraint’ and ‘Explanations of Creativity’; Gaut, ‘Creativity and Imagination’; Gaut and Livingston, ‘The Creation of Art’, 10–11; Martindale, ‘Biological Bases of Creativity’; Poincare,´ ‘Science and Hypoth- esis’; Sternberg and Lubart, ‘An Investment Theory’, ‘Buy High and Sell Low’, and ‘The Concept of Creativity’. 31. One might worry that this value neutrality is inconsistent with the arguments offered in Section II above. Recall that the process approach was partly justified by appealing to the fact that in appreciating and criticizing art, we value processes intrinsically, not merely instrumentally (and moreover, the person approach was criticized oppositely). But one can maintain that creative processes are valued for their own sake, without committing to value being a necessary condition or a constitutive feature of creative processes. The process is just whatever it is (and a few plausible conditions are being suggested here), and may thus be valued because

June 21, 2008 10:51 MAC/NWA Page-122 9780230_220478_07_cha06 Dustin Stokes 123

it is a thing of that kind (i.e., meets the conditions specified). This is distinct from making value itself a constitutive condition for a concept of creativity.

Bibliography

Addison, J. and Steele, R. 1879. ‘On Genius’ in A. Chalmers (ed.), The Spectator,New York: D. Appleton. Alperson, P. (ed.) 1992. The Philosophy of the Visual Arts, Oxford: Oxford University Press, 129–130. Beardsley, M. 1965. ‘On the Creation of Art’, Journal of Aesthetics and Art Criticism, 293: 291–304. Boden, M. 2004. TheCreativeMind, 2nd edn, London: Routledge. Casati, R. and Varzi, A. 2006. ‘Events’ in E. Zalta (ed.), The Stanford Encyclopedia of Philosophy, URL = . Currie, G. 1989. An Ontology of Art, New York: St. Martin’s Press. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Davies, D. 2004. Art as Performance, Malden, MA: Blackwell. Davies, S. 1991. Definitions of Art, Ithaca: Cornell University Press. Dutton, D. 1979. ‘Artistic Crimes: The Problem of Forgery in the Arts’, Journal of Aesthetics and Art Criticism, 19: 304–314. Feagin, S. and Maynard, P. (ed.) 1997. Aesthetics, New York: Oxford University Press. Gaut, B. 2003. ‘Creativity and Imagination’ in B. Gaut and P. Livingston (eds), The Creation of Art, Cambridge, MA: Cambridge University Press. Gaut, B. and Livingston, P. (eds) 2003. ‘The Creation of Art: Issues and Perspectives’, The Creation of Art, Cambridge: Cambridge University Press. Gaut, B. and McIver Lopes, D. (ed.) 2005. The Routledge Companion to Aesthetics,Second Edition, New York: Routledge. Hadamard, J. 1954. An Essay on the Psychology of Invention in the Mathematical Field, New York: Dover. Helmholtz, H. von. 1896. Vortrage¨ und Reden, Brunswick: Friedrich Viewig und Sohn. Kant, I. 1987. Critique of Judgment, W.S. Pluhar (trans.), Indianapolis: Hackett. Kivy, P. 2001. The Possessor and the Possessed: Handel, Mozart, Beethoven, and the Idea of Musical Genius, New Haven, CT: Yale University Press. Kivy, P. (ed.) 2004. The Blackwell Guide to Aesthetics, Oxford: Blackwell. Koestler, A. 1964. The Act of Creation, New York: Macmillan. Korsmeyer, C. (ed.) 1998. Aesthetics: The Big Questions, Oxford: Blackwell. Lamarque, P. and Olsen, S. 1992. ‘Intention and Interpretation: A Last Look’ in G. Iseminger (ed.), Intention and Interpretation, Philadelphia: Temple University Press. Lamarque, P. and Olsen, S. (eds) 2004. Aesthetics and the Philosophy of Art: The Analytic Tradition, Oxford: Blackwell. Levinson, J. 1990. Music, Art, and Metaphysics: Essays in Philosophical Aesthetics, Ithaca, NY: Cornell University Press. Longinus 1995. On the Sublime, W. H. Fyfe (trans.), in S. Halliwell (ed.), Aristotle: The Poetics, Longinus: On the Sublime, Demetrius: On Style, Second Edition, Loeb Classical Library, Cambridge, MA: Harvard University Press. Martindale, C. 1981. Cognition and Consciousness, Homewood, IL: Dorsey Press. Martindale, C. 1995. ‘Creativity and Connectionism’ in S. Smith, T.B. Ward and R. Finke (eds), The Creative Cognition Approach, Cambridge: MIT Press.

June 21, 2008 10:51 MAC/NWA Page-123 9780230_220478_07_cha06 124 A Metaphysics of Creativity

Martindale, C. 1999. ‘Biological Bases of Creativity’, in R. Sternberg (ed.), Handbook of Creativity, Cambridge, MA: Cambridge University Press. Mendolsohn, G. 1976. ‘Associative and Attentional Processes in Creative Performance’, Journal of Personality, 44: 341–369. Novitz, D. 1999. ‘Creativity and Constraint’, Australasian Journal of Philosophy, 77: 67–82. Novitz, D. 2003. ‘Explanations of Creativity’, in B. Gaut and P. Livingston (eds), The Creation of Art, Cambridge: Cambridge University Press. Poincare,´ H. 1982. ‘Science and Hypothesis’ The Foundations of Science, Washington, D.C.: University Press of America. Poincare,´ H. 1982. ‘The Value of Science’ The Foundations of Science, Washington, D.C.: University Press of America. Poincare,´ H. 1982. ‘Science and Method’ The Foundations of Science, Washington, D.C.: University Press of America. Ryle, G. 1949. The Concept of Mind, London: Hutchinson. Schopenhaeur, A. 1958. The World as Will and Representation, 2 Volumes, E.F.J. Payne (trans.), Indian Hills, CO: Falcon Wing’s Press. Smith, S. and Blankenship, S. 1989. ‘Incubation Effects’, Bulletin of the Psychonomic Society, 27: 311–314. Smith, S. and Blankenship, S. 1991. ‘Incubation and the Persistence of Fixation in Problem Solving’, American Journal of Psychology, 104: 61–87. Stecker, R. 1997. Artworks: Definition, Meaning, Value, University Park: Pennsylvania State University Press. Sternberg, R. and Lubart, T. 1991. ‘An Investment Theory of Creativity and its Development’, Human Development, 34: 1–31. Sternberg, R. and Lubart, T. 1992. ‘Buy Low and Sell High: An Investment Approach to Creativity’, Current Directions in Psychological Science,1:1–5. Sternberg, R. and Lubart, T. 1999. ‘The Concept of Creativity: Prospects and Paradigms’ in R. Sternberg (ed.), Handbook of Creativity, Cambridge, MA: Cambridge University Press. Stokes, D. 2007. ‘Incubated Cognition and Creativity’, Journal of Consciousness Studies, 14.3: 83–100. Thomasson, A. 1999. Fiction and Metaphysics, Cambridge, MA: Cambridge University Press. Vendler, Z. 1957. ‘Verbs and Tenses’, Philosophical Review, 66: 143–160. Wallas, G. 1926. The Art of Thought, New York: Harcourt, Brace, and World. Walton, K. 1970. ‘Categories of Art’, The Philosophical Review, 79: 334–367. Wimsatt, W. and Beardsley, M. 1946. ‘The Intentional Fallacy’, Sewanee Review, 54: 468–488. Wollheim, R. 1980. Art and Its Objects, Second Edition, Cambridge: Cambridge University Press.

June 21, 2008 10:51 MAC/NWA Page-124 9780230_220478_07_cha06 7 From Defining Art to Defining the Individual Arts: The Role of Theory in the Philosophies of Arts Aaron Meskin

Introduction

What is the definition of art? Not too long ago, this might have been properly said to be the central issue in philosophical aesthetics. Not anymore.1 But philosophers of art are unlikely to give up on the pursuit of definition that easily. Most of us have invested far too much time and energy learning how to go about the game of definition and counterexample to simply let it go just like that. Might there not be some other questions of definition that could structure the field in this new century? What, after all, is a comic book? Or a film? Or a poem? Or a dance? Perhaps if we pursued definitions of the individual arts or art forms we might get somewhere. In this essay, I shall offer some sceptical thoughts about that pursuit—sceptical thoughts inspired by the paper that inadvertently triggered interest for more than five decades in the definition of art, namely, Morris Weitz’s ‘The Role of Theory in Aesthetics’. A little more than 50 years ago, Weitz claimed that ‘aesthetic theory’— understood primarily as a traditional definition of art (i.e., a true and informative statement of the necessary and sufficient conditions for being a work of art)—is ‘never forthcoming in aesthetics’.2 Needless to say, his Wittgenstein-inspired arguments did not dissuade many from pursuing the definitional project. In the succeeding years, we were offered a range of sophisticated proposals for defining art—institutional definitions,3 aesthetic definitions,4 Danto’s neo-representational definition,5 disjunctive definitions,6 and intentional-historical definitions,7 to name some of the most prominent. It can honestly be said that the publication of Weitz’s article was one of the key factors in generating the late twentieth-century interest in the definition of art. So one central role Weitz’s paper has played was to encourage aestheticians to focus on the project of defining art, to pursue ‘theorizing in the grand manner’ as Peter Kivy put it in his 1992 Presidential Address to the American Society for Aesthetics.8 In fact, even many of those who have argued against

125

June 21, 2008 10:57 MAC/NWA Page-125 9780230_220478_08_cha07 126 From Defining Art to Defining the Individual Arts the possibility of defining art, such as Berys Gaut, can be seen as engaged in the broad project of pursuing grand aesthetic theory (i.e., offering a characterization of the concept of art), albeit in non-definitional form.9 This was, I believe, unfortunate. While this will be tendentious, I think contemporary aesthetics has been most successful either when it has followed Kivy’s advice, and pursued the ‘careful and imaginative scrutiny of the individual arts and their individual problems’,10 or, on the other hand, when it has investigated distinctive phenomena (e.g., imagination, emotional arousal, expression, depiction) as they are manifested across a range of the arts. In fact, it has largely been the success of those sorts of aesthetic projects and the failure of most definitional projects to provide much insight into the arts—that is to say, not anti-definitional neo-Wittgensteinian arguments— that have moved the definitional question out of the centre of the field.11 At least that is how I read the recent history of philosophical aesthetics. But one way that ‘intensive and special study of the separate arts’12 might be pursued would be to attempt to define the individual arts or art forms. That is, one natural way of following Kivy’s suggestion would be to pursue definition on a small scale—to theorize, as it were, in the formal or medium or generic modes. While it cannot be said that this has been anywhere near the central project of those philosophers who have explored the individual arts, it is not an uncommon part of their larger projects. So, for example, Jerrold Levinson has proposed a definition of music; Peter Lamar- que and Stein Olsen, as well as Robert Stecker, have offered definitions of literature; Noel¨ Carroll has proposed a kind of definition of the moving image and Trevor Ponech has recently argued for an alternative account of the cinematic; Greg Hayman and Henry John Pratt have offered a definition of comics; Anna Ribeiro has recently proposed a definition of poetry; and Monroe Beardsley offered a definition of dance.13 For analytic aestheticians trained in the techniques of conceptual analysis but frustrated with its lack of success in the case of art, the possibility of definition at the level of the arts or art forms holds out some attraction. It might seem that pursuing such definitions is likely to be both more feasible and more valuable than the traditional project of defining art. The individual art forms are not as internally diverse as the category of art itself. And insofar as heterogeneity underwrites the intuitive appeal of the anti- definitional position (‘how could such a hodgepodge admit of definition?’), this suggests that anti-definitionalism will have less appeal at the level of the individual arts or art forms. To take a fairly obvious example, the ontological diversity that art in general seems to display does not seem to be exhibited by many of the individual arts.14 And while straightforward functional definitions of art are implausible since art may have so many different functions, it is tempting to think that the individual art forms might each be characterizable in terms of distinct functions or purposes. It might also be thought that the apparent fragmentation of the concept of art—the existence

June 21, 2008 10:57 MAC/NWA Page-126 9780230_220478_08_cha07 Aaron Meskin 127 of distinct concepts of art in different traditions, cultures, and subcultures— is not exhibited by the concepts associated with the individual arts.15 While there may be many concepts of art itself, there do not appear to be many concepts of poetry or theatre. There is not the same history of failure with respect to philosophical attempts at defining the individual art forms as there is with respect to the project of defining art, so local inductive arguments against the possibility of defining the various art forms look even less attractive than an inductive argument against defining art itself. It might appear that we are at least halfway there with respect to the definition of the various art forms, since it seems that we know that an object belongs to an art form only if it is an art.16,17 (But it is not the case that something is a work of art only if it belongs to an art form. Some experimental works of art belong to no art form. Other works of art that are now properly seen as initiating new art forms might not have belonged to any art forms at all had the course of art history gone differently—that is, if further works had not followed their lead. But they would have still been art.) And while a definition of art seems unlikely to provide normative, evaluative, interpretive, or appreciative guidance—as Malcolm Budd put it, ‘Given what has happened to the concept of art, especially in this century, it would be fruitless to proceed in this way: an account of artistic value cannot be extracted from the present concept of art’18—it is tempting to think that a definition of film or literature or the comic book could provide some such help.19 As I have already suggested, I think we should resist the temptation. The pursuit of art’s definition did not provide the payoff that was promised, and I doubt whether the pursuit of definition at the level of the arts or art forms will provide much either. Systematic theorizing about the individual arts and art forms may be crucial to making progress in philosophical aesthetics, but I suspect the product of this theorizing will not look much like traditional explicit definitions.20 As I shall suggest below, there may even be an aesthetic cost to the pursuit of such definitions. And I think Weitz—despite the weak- ness of his explicit arguments—was onto this. For Weitz did not just argue against the possibility of defining art in ‘The Role of Theory in Aesthetics’. He also suggested that many of art’s various sub-concepts (i.e., the art form and genre concepts) cannot or should not be defined, and he offered some reasons why this is the case. So I think we can still learn something important by thinking about Weitz’s arguments, and that is what I propose to do for most of the remainder of this chapter. At the end, I shall briefly reconsider the putative attractiveness of the definitional project writ small.

I. Weitz’s argument

If we focus on the aspects of Weitz’s original article that have been central to contemporary philosophical discourse, we will not get much insight into the questions that concern us about defining the arts or art forms. Most

June 21, 2008 10:57 MAC/NWA Page-127 9780230_220478_08_cha07 128 From Defining Art to Defining the Individual Arts discussions of Weitz’s article have focused on what I shall call his ‘creativity’ argument, which is captured in the following statement: ‘What I am arguing, then, is that the very expansive, adventurous character of art, its ever-present changes and novel creations, makes it logically impossible to ensure any set of defining properties’.21 This singularly unpersuasive argument focuses on the concept of art (or of art work) and suggests that there is a distinctive reason why definition of that concept is not forthcoming. The argument fails because there is no inconsistency between definition on the one hand and expansiveness or adventurousness on the other, and because it appears that Weitz exploits an ambiguity in shifting from the practice concept of art to the work concept.22 The creativity argument, then, does not establish the indefinability of the concept of art, and it is just as ill-suited to establish the indefinability of the concepts that pick out the various arts. Moreover, the question of whether art can or cannot be defined is logically independent of whether its various sub-concepts can or cannot be defined. Issues about the possibility and value of defining the various art forms are, therefore, not settled at all once we have discarded (or, for that matter, embraced) the creativity argument. And, as I read him, Weitz does not explicitly offer the creativity argument as a reason for thinking we cannot define the various sub-concepts of art. But the creativity argument is not all there is to Weitz’s paper. Right before his introduction of the creativity argument, Weitz addresses the question of the definition of the various ‘sub-concepts’ of art—not as a part of an argument for the indefinability of art but rather as part of an attempt to ‘illustrate’ his thesis. And after arguing that the question of whether something is or is not a novel cannot be answered descriptively by providing necessary and sufficient conditions for the concept of the novel, he states that ‘what is true of the novel is, I think, true of every sub-concept of art: ‘‘tragedy,’’ ‘‘comedy,’’ ‘‘painting,’’ ‘‘opera,’’ etc., of ‘‘art’’ itself’.23 So there is at least a statement of anti-definitionalism about the various art forms and genres. And this is a place to begin. Moreover, Weitz provides other anti-essentialist arguments. He offers a distinct Wittgensteinian-inspired ‘look and see’ argument against the definability of art, and he presents a family resemblance style account of what goes on in the case of the individual arts. He suggests that even artifactuality is inessential to art status, and he also presents a general argument for the indefinability of art and its sub-concepts based on the putative indefinability of all non-logical/non-mathematical concepts. (I shall refer to this as the general argument.) Weitz has, then, quite understandably been understood as arguing—in a variety of ways—that art and its various sub-concepts cannot be defined. And although there has not been uniformity on this matter, this has been another role of Weitz’s argument. It has focused the attention of philosophers on the question of whether the definition of art or its

June 21, 2008 10:57 MAC/NWA Page-128 9780230_220478_08_cha07 Aaron Meskin 129 sub-concepts is possible. What other questions about the definition of art could there be? One other question that has been asked is what reasons we might have to define art. What might we achieve by defining art?24 Another related question is whether we need (in some sense or other) to define art in order to achieve some important goal that we might have.25 My interest—although related to these two questions—is not quite the same. Well there is another intriguing aspect of Weitz’s essay. At a number of points, Weitz admits that various concepts, even various sub-concepts of art, can be defined.26 So, for example, Weitz writes that ‘there are legitimate and serviceable closed concepts in art’.27 He mentions ‘(extant) Greek tragedy’, but I think that more interesting cases of this are found when we consider the concepts associated with the closed poetic forms (e.g., VILLANELLE, SESTINA, and SONNET).28 And, in fact, Weitz suggests that those ‘empirically- descriptive and normative concepts’ that do not, in some important sense, admit of definition in terms of necessary and sufficient conditions can be defined if ‘we arbitrarily close them by stipulating the range of their uses’.29 Perhaps even more tantalizingly, Weitz suggests—right after he offers the creativity argument for anti-definitionalism—that ‘we can, of course, choose to close the concept. But to do this with ‘‘art’’ or ‘‘tragedy’’ or ‘‘portraiture,’’ etc., is ludicrous since it forecloses on the very conditions of creativity in the arts’.30 I am especially interested in these latter aspects of Weitz’s thought—(1) his allowance that there are ‘legitimate and serviceable’ classically structured concepts to be found among art’s sub-concepts, and (2) the normative tenor of some of the aforementioned comments. For he appears to be suggesting that though we could, in some sense, define the concepts in question, it would be better (at least in most cases) if we resisted; that closing those concepts would (to some extent at least) be a bad thing; hence, that we ought not do so. The suggestion that we can but should not define art and at least some of its sub-concepts is very different from the claim that we cannot. And it is worth considering what might be done with that suggestion. Before I move on, let me make one final point. Weitz’s arguments are often dismissed for failing to take into account the possibility of defining art by means of non-exhibited, relational properties (e.g., institutional or historical relations).31 This is a legitimate criticism. But it is worth noting that it would still be interesting if Weitz had provided devastating general arguments against the possibility (or advisability!) of non-institutional and non-historical accounts of art and the various art forms. After all, such theories are still with us. The aesthetic theory of art, for example, still has its defenders.32 For similar reasons, it would still be of interest if he had merely provided sound general arguments against the advisability of treating certain kinds of features as necessary conditions for membership in the category of art or some various art forms.

June 21, 2008 10:57 MAC/NWA Page-129 9780230_220478_08_cha07 130 From Defining Art to Defining the Individual Arts

In the next section I shall look at Weitz’s general argument which, if it were to succeed, would seem to scotch any attempt to define art or the art forms. I shall explore two issues: (1) what reason we might have for thinking that no (or very few) concepts can be defined; and (2) how a version of this argument might deal with the apparent counterexamples from the closed poetic forms. After that, I shall turn to the normative argument.

II. The general argument against definition

The general argument A page before the creativity argument, Weitz offers a distinct argument against the possibility of art theory.

If necessary and sufficient conditions for the application of a concept can be stated, the concept is a closed one. But this can happen only in logic or mathematics where concepts are constructed and completely defined. It cannot occur with empirically-descriptive and normative concepts unless we arbitrarily close them by stipulating the ranges of their uses.33

We have here, of course, the basis for a very simple two premise argument for the impossibility of non-arbitrarily defining the concept of art or any of its sub-concepts:

P1: No non-logical/non-mathematical concepts admit of non-arbitrary definition. P2: ART and its sub-concepts (PAINTING, SCULPTURE, MUSIC, etc.) are neither logical nor mathematical concepts. C: ART and its sub-concepts do not admit of non-arbitrary definition.

Since the argument is valid, and P2 seems non-tendentious, the success of the argument depends on the truth of P1. Now Weitz offers no argument for P1 in the paper in question, and this—along with the apparent definability of at least some empirical concepts (e.g., some non-lexical concepts like EXTANT GREEK TRAGEDY, the concepts associated with the closed poetic forms mentioned above, the bachelor concept)—is certainly one of the reasons that philosophers of art have tended to overlook it. But since there are things to be said in favour of P1 (or something like it), it is worth the consideration of philosophers of art. And, of course, they have paid some attention to it, but I will suggest philosophers of art need to think a bit more about the general argument. What basis could we have for thinking P1 (or some relevantly close suc- cessor to it) is true? I take it there are at least two general reasons one might be generally sceptical of definition: (1) induction on the history of attempts

June 21, 2008 10:57 MAC/NWA Page-130 9780230_220478_08_cha07 Aaron Meskin 131 to define various concepts, and (2) psychological theories about the nature of concepts. Let me address these concerns in turn.

The inductive argument Are there inductive reasons to think that no lexical concepts (or no lexical concepts of interest to philosophers) can be defined?34 I think it untenden- tious that the history of attempts to provide definitions for various concepts of philosophical interest has—to put it mildly—not been one of unrivalled success. And the inductive argument has had its fans.35 But I hesitate to appeal to the general inductive argument against the possibility of defining art. In the first place, it is plausible that some lexical concepts have been successfully defined. I have, for example, already mentioned the concepts associated with the closed poetic forms. Of course, this might just encourage the fan of the inductive argument to limit the scope of the induction. (Or, perhaps, to argue that despite appearances, such concepts do not admit of satisfactory definition. See below for relevant discussion). Perhaps there is a special category of concepts such that all observed attempts at defining them have failed: folk concepts, ‘commonsense’ concepts,36 ‘abstract’ concepts,37 those concepts of ‘most philosophical interest’,38 or concepts about which there is not widespread explicit knowledge of their definition. Still, in reply to any version of the inductive argument, the friend of definition may reply that there may have been successful but unrecognized definitions in the relevant category.39 Perhaps more importantly, they may point to the fact that the friend of the inductive argument needs to provide some reason for thinking that the category of concepts in question forms some sort of natural kind. For unless we have reason to think we are dealing with some sort of natural kind (in even a loose sense), we will have little reason to think that there is much to be said for the inductive argument. If we are dealing with a gerrymandered category (i.e., if the category of concepts picks out a motley), then we have reason to be sceptical of making inductive inferences from a selective sample. Since I think there are some reasons for being suspicious of the claim that concepts are a natural kind,40 I hesitate to appeal to the general inductive argument.41

The psychological argument The inductive version of the general argument looks problematic. It relies on two arguable assumptions: (1) that there have been no successful but unrecognized definitions in some class of concepts, and (2) that concepts (or some significant category of them) belong to a natural kind. Since both assumptions are questionable, I think we must put the general form of the inductive argument aside. But what about the challenge to the definitional project from contemporary psychology—a sort of empirically based reason for being sceptical of the definability of ordinary concepts? I take there to be a serious issue

June 21, 2008 10:57 MAC/NWA Page-131 9780230_220478_08_cha07 132 From Defining Art to Defining the Individual Arts here for anyone engaged in the definitional—or more specifically, the analytic—project. In a 2003 article in the British Journal of Aesthetics, Jeffrey Dean argued (along the lines of earlier discussions by William Ramsey and Stephen Stich): (1) that there are reasons to think that the classical theory of concepts (roughly, the view that most lexical concepts are composed of other concepts which encode necessary and sufficient conditions for the application of the former concept) is mistaken, and (2) that this calls the traditional project of defining art into question.42 The failure of the classical account to offer an explanation of typicality and atypicality judgements, the dearth of non-controversial analyses of any of the concepts philosophers have been interested in, the apparent commitment of the classical account to a questionable analytic-synthetic distinction, the trouble the account has for making sense of certain sorts of error and ignorance, and the failure to find evidence of the psychological reality of the conceptual complexity to which the classic account seems committed have pushed many psychologists and philosophers away from the classical account of concepts to embrace some sort of prototype account of conceptual structure. Such an account seems in tension with the definitional project. Why exactly? Prototype theories hold that concepts are structurally complex, but that they are not composed of sub-concepts which encode necessary and sufficient conditions; rather, concepts are composed of representations which are probabilistically related to the application of the concept. But the very practice of coming up with philosophical definitions seems to make most sense on the assumption that concepts have a sort of classical structure rather than a prototype structure. Insofar as the definitional project in its analytic mode depends on intuition-based category judgements, then the possibility of nice and neat counterexample-proof necessary and sufficient conditions for the application of the concept seems to depend on the truth of the classical account of conceptual structure.43 If concepts are prototypes, and if categorization is driven by similarity judgements rather than the application of necessary and sufficient conditions, then we should not expect the product of compiling category judgements about actual and counterfactual cases to result in neat traditional definitions. Aestheticians—and philosophers in general—have been underwhelmed by such arguments. While Noel¨ Carroll mentions the challenge from prototypes in his introductory book,44 it does not seem to have stopped philosophers of art from engaging in the definitional project. And, in fact, there are two recent responses to Dean’s arguments, from Thomas Adajian (2005) and Robert Stecker (2005).45 I believe that these responses miss the point of the challenge, and that there is a real issue here for philosophers of art interested in the definitional project. There seem to be two general responses to the argument from prototypes. The first is that the prototype theory itself is problematic.46 I think that is

June 21, 2008 10:57 MAC/NWA Page-132 9780230_220478_08_cha07 Aaron Meskin 133 right, and I will not get into the details here.47 But I think this misunderstands the real challenge to definition as it is normally pursued. That challenge is not rooted in the plausibility of the prototype account; it is rooted in the implausibility of the classical account. So, for example, if atomism is true about concepts (i.e., the view that they do not have internal structure) then conceptual analysis and the definitional project seem doomed.48 So too with other non-classical theories of concepts. As Adajian himself says, ‘other well known theories of concepts ...exist, and compete with PT [prototype theory]’.49 But it is the failure of the classical account of concepts that generated the problem for definition, not merely the truth of the prototype view. So appeal to problems with the prototype account (and the existence of competing non-classical accounts of concepts) does not seem to answer the challenge. Note also that the argument does not rely on a commitment to a single natural kind to which the concepts belong. The key to the argument is the claim that no (or very few—see below) concepts are structured classically. The second response (emphasized by Adajian who is heavily influenced by Rey) is to accuse the argument from psychology of confusing metaphysics and epistemology.50 The psychological results tell us about how we, as a matter of fact (rightly or wrongly), categorize. But they do not tell us about that nature of those categories. They do not tell us about what there is in the world. And the definitional project is, Adajian suggests, a metaphysical project—one concerned with determining the necessary and sufficient conditions for being a work of art—not merely the ways that we think about art or our methods of categorizing such works. Now I am not sure that Adajian is right in thinking that philosophers of art have been interested in a purely metaphysical project. Many philosophers of art seem interested in giving an account of our concept of art, where the target of their investigation seems to be a mental or linguistic entity rather than what Adajian calls ‘extra-psychological conditions’.51 To take one important example, Jerrold Levinson’s work on definition—work which has been central to the contemporary debate—is clearly targeted at providing an account of ‘our present concept of art’ and the concept of music ‘possessed by twentieth-century Westerners’.52 This does not sound like a search for extra-psychological conditions. But more importantly, the methodology that philosophers of art tend to use—the focus on getting the extension right and trying to accommodate intuitions about actual and counterfactual cases—seems ill-suited to investigating extra-linguistic reality. As Robert Stecker puts it, ‘the main thing we have to go on in defining art is our classificatory practice’.53 If this is right, and I believe it is a fair account of the resources that most philosophers appeal to when they pursue the definitional project, then the appeal to the epistemic/metaphysical distinction will not have much force against the psychological argument in the case of art. For if our classificatory practice is driven by prototypes or exemplars or theories rather than necessary

June 21, 2008 10:57 MAC/NWA Page-133 9780230_220478_08_cha07 134 From Defining Art to Defining the Individual Arts and sufficient conditions, and our classificatory practice is the main thing that underwrites our search for definition, then we have little reason to think we will find a satisfactory definition if we continue to pursue the project of defining art as it is typically pursued. As William Ramsey suggests, ‘any distinction between a prototype representation that facilitates ordinary categorization judgments and a more definitional representation comprised of features discovered and relayed by empirical investigators . . . is not going to carry much weight when we consider concepts defined by traditional conceptual analysis’.54 So, while Adajian is right to distinguish between the epistemological and metaphysical and to raise concerns about whether we can infer metaphysical conclusions from a theory about our epistemic practices, the implications for the traditional approach to defining art are less clear. We cannot, in general, infer metaphysical conclusions about category membership from psychological information about how we categorize. So the truth of the conclusion of the general argument—that the concept of art cannot be defined—would not settle the issue of whether a ‘real definition’ of art is forthcoming (i.e., a ‘definition’ of what it is to be art). It does not entail that there are no informative necessary and sufficient conditions for being art. But this conclusion should not be especially reassuring to most philosophers of art, for the methods they use seem ill-suited to getting at any such necessary and sufficient conditions. Consider also that in the case of the putatively natural kinds (e.g., the biological, chemical, and physical kinds), we typically think that there is a substantive gap between our quotidian methods of categorization and the real essence of the relevant categories. One standard and very plausible story about such cases involves appeal to presuppositions of some sort of hidden essence, a practice of scientific investigation into the nature of that essence, and a practice of deference to experts with respect to the extension of various terms.55 The problem for philosophers of art is that it is not at all obvious whether the various phenomena that allow for and explain the epistemic/metaphysical gap in the case of the natural kinds are in place when it comes to their primary objects of study.56 There certainly does not appear to be a practice of scientific investigation into the nature of the putative essence of art or its sub-concepts, and while it might be suggested that there is some practice of deference to experts in the case of art, our judgements about the extension of art’s sub-concepts seem to display significant autonomy. (Do we really defer to experts about whether some questionable work is music or sculpture or a comic book? I think this is far from obvious.)57 Moreover, even if we do defer to experts about the extension of ‘art’, it is not clear that the way this works in the case of art mirrors the standard natural kind cases. It might be that we defer, not because they know something about art that we do not know, but because their very claim that something is art is criterial for art status.58

June 21, 2008 10:57 MAC/NWA Page-134 9780230_220478_08_cha07 Aaron Meskin 135

So suppose the psychological argument does provide evidence against the view that the concept of art and its sub-concepts can be defined. This does not tell us whether there are or are not necessary and sufficient conditions for being art. But it does suggest that the methods ordinary philosophers of art use in pursuit of a definition are unlikely to be successful. The best explanation for the methods used by philosophers of art in their pursuit of definition is that there is a tacit assumption in place that there is not, in the case of art, a robust epistemic/metaphysical distinction. That is, a charitable reading of the work on defining art by Levinson, Stecker, Carroll, Gaut, and many others suggests that—at least in the case of art—the gap between the epistemic and the metaphysical has narrowed. As Stecker writes,

Artworks are the artefacts we classify using the label ‘art.’ Their having an essence depends on the possibility of giving a lexical definition of ‘art,’ that is, on there being a uniform set of properties – properties necessary and sufficient for being art – that invariably informs the classifications of those who understand the meaning of ‘art’. (italics mine)59

My suggestion, then, is that there is a widespread assumption in the philosophy of art that we can, in fact, ‘infer what it is to be a P from premises about the features by which people decide’ under certain sorts of conditions ‘whether something is a P’.60 While this assumption might be mistaken, it is not completely implausible, and positing it would explain much of what goes on in the debates about the definition of art. If so, and if the assumption can be bolstered, then there would be a reasonable inference from the psychological argument to metaphysical conclusions about the non-existence of real definitions of art and its sub-concepts.

The case of closed concepts What about those closed concepts that I have mentioned earlier—those concepts expressed by terms such as ‘extant Greek tragedy’ or ‘villanelle’ or ‘sonnet’? Does the existence of definitions of these concepts undercut the psychological version of the general argument? Perhaps not. EXTANT GREEK TRAGEDY is not a lexical concept (i.e., not a concept that corresponds to a lexical item in natural language), and it seems open to the prototype theorist to admit that there are some non-lexical concepts that did not exhibit a prototype structure. But what about the lexical concept VILLANELLE? This poetic form (most notably exemplified by Dylan Thomas’s ‘Do Not Go Gentle Into That Good Night’) seems straightforwardly definable as a 19-line poem consisting of 5 tercets and a concluding quatrain with a rhyme scheme of a1/b/a2 a/b/a1 a/b/a2 a/b/a1 a/b/a2 a/b/a1/a2. One possibility is to simply resist the claim that this concept and other similar concepts associated with closed poetic forms do admit of definition. Suppose the candidate poem

June 21, 2008 10:57 MAC/NWA Page-135 9780230_220478_08_cha07 136 From Defining Art to Defining the Individual Arts was created by someone unfamiliar with villanelle form. Would the formal features be enough to make it count as a villanelle? Perhaps not, but this might be handled by adding an intentional condition to the definition. What about a poem in which, for example, a few words in one instance of one of the refrains are changed? Would ‘Do Not Go Gentle Into That Good Night’ have failed to be a villanelle if one word had been replaced in the last line? What about verse in villanelle form written solely for a greeting card? While resisting the claim that VILLANELLE can be straightforwardly defined is a tempting route to pursue, an alternative response would be to argue that it is a technical concept (i.e., one originally constructed for a distinct although non-arbitrary purpose). If so, it is distinct from the folk concepts that have been of primary interest to philosophers and that appear to be least susceptible to definition. In fact, informal surveys of friends and colleagues suggest that the concept is not an especially widespread one. Most do not seem to possess the concept at all, and even fewer would recognize it if they were confronted with one. Moreover, those people who have heard of the form tend to have explicit knowledge of its definition or be willing to defer to experts about the concept’s extension. So VILLANELLE is, then, a very different sort of concept than ART, COMIC, or FILM. It is plausible that it exhibits many of the features that our natural kind concepts have (e.g., deference, expertise). But then it would not be strange to recast the general argument in the following way (it should be understood that we are quantifying only over lexical concepts):

P1*: No non-logical/non-mathematical/non-technical concepts admit of non-arbitrary definition. P2*: Neither ART nor many of its sub-concepts (e.g., PAINTING, SCULP- TURE) are logical, mathematical, or technical concepts. C: ART and many of its sub-concepts do not admit of non-arbitrary definition.

There is still the old chestnut of BACHELOR to worry about. It is neither a logical nor a mathematical nor a technical concept, but the standard story is that it can be non-arbitrarily defined. This can be denied.61 It might also be emphasized that the putative definition of BACHELOR is known explicitly by just about everyone who is familiar with the term. But this seems to suggest that it too is a very different sort of concept than that of ART or MUSIC or COMIC. Perhaps the right thing to say is that it is a ‘one-criterion concept’62 and that such concepts behave differently than those in which philosophers are interested. In any case, no fan of the psychological argument should be too troubled by BACHELOR. I conclude that that the general argument for the indefinability of art and its sub-concepts must still be taken seriously.

June 21, 2008 10:57 MAC/NWA Page-136 9780230_220478_08_cha07 Aaron Meskin 137

III. The normative argument against definition

Weitz, as I mentioned earlier, does not always seem committed to the view that ART and its various sub-concepts cannot be defined. He mentions the existence of some ‘legitimate and serviceable’ sub-concepts of art, and he suggests that we could define both ART and its sub-concepts, but that this would be ‘ludicrous’ to do. What should be made of this? It is natural to think that both the ‘closed’ sub-concepts and the possibility of closing ART itself depend on the aforementioned possibility of arbitrary stipulation. Perhaps. But, if so, such stipulation sometimes results in legitimate and serviceable concepts. And Weitz says that the boundaries of those closed concepts ‘have been drawn for special purposes’, which suggests that such stipulation is not always purely arbitrary. I am tempted to read these statements as suggesting that although we could—in some sense or other—define art and its various sub-concepts, there is a pro tanto reason not to do so. And what I would like to do for the remainder of the chapter is see if I can make any sense of that thought, for it has struck me more than once as an intriguing one. Here is one way of thinking about injunction to resist defining art. Weitz’s suggestion that we can but should not define art and its sub-concepts might be understood as directed at the meta-level project of examining and theorizing about ordinary concepts and their uses. Understood this way, the injunction is directed towards philosophers and not the ordinary folk (or, at least, only towards ordinary folk in their occasional role as philosophers of art). It is an injunction to resist engaging in the project of pursuing definition, even if it can be achieved. But what would be the basis for such an injunction? While it is true that not everything that can be done should be done, it is also unclear what reason there would be for resisting the pursuit of definition if it were possible. Dennis Dutton offers some suggestive remarks that might shed some light on the issue: ‘The obsession with accounting for art’s most problematic outliers, while both intellectually challenging and a good way for teachers of aesthetics to generate discussion, has left aesthetics ignoring the center of art and its values.’63 And while there might be more to be said about this line of thought, I want to pursue other issues. So, here is another way of thinking about the suggestion that we ought to resist defining art. There are, it seems to me, plausibly a range of concepts of art available to us. Perhaps we have available to us in some sense an institutional concept, a historical concept, an aesthetic concept, and a neo- representational concept, not to mention a pure craft/skill concept (i.e., the ‘art as techne’ concept), and a robustly evaluative concept too. What do I mean by calling them concepts of art? For one thing, each one of those concepts—at least the first four—picks out roughly the same extension. They each classify paradigm cases of both art and non-art in the same way. And

June 21, 2008 10:57 MAC/NWA Page-137 9780230_220478_08_cha07 138 From Defining Art to Defining the Individual Arts what do I mean by saying that they are available to us? One of the two things—either they are actual extant concepts that can be found in the wild as it were or, perhaps, just that we could use them. That is, we could cognize the world by means of those concepts. But then my suggestion is that Weitz might be interpreted as suggesting that while we could use any one of those concepts, we would be better off—in dealing with those objects that all of those concepts aim at picking out— using a concept that did not encode necessary and sufficient conditions. Are there a range of concepts of art? It seems to me plausible that there are. Compare Frank Jackson’s following suggestion about what we might have learned about the concept of knowledge:

There is a weak and a strong sense in which you might hold that knowledge is unanalysable. On the weak sense, the view is, as we put it above, that there is no unitary concept of knowledge but rather a series of interlocking and to some extent vague notions. Each individual notion has its analysis but there is no illuminating formula that covers the lot. The Gettier literature has, on this view, been an exercise in displaying the variety of notions and its failure to induce consensus gives us defeasible reason to hold that there is no unitary notion underlying our folk use of ‘knowledge’.64

What I am suggesting is that we might think of the recent history of philosophical aesthetics—especially the history of ‘grand theorizing’—as suggesting that there is no unitary folk concept of art but rather a series of interlocking and to some extent vague notions. (Similarly with our concepts of the various art forms. More on this below.) The failure to induce consensus gives us some defeasible reason to think there is no unitary notion underlying our folk use of ‘art’. This is no argument against defining art. If true, it means only that there are a range of distinct concepts of art to define. But we need not leave things there. For if there is this variety of notions, we might ask which among those notions is most useful or valuable. So Weitz might be understood as arguing that the various non-classically structured (i.e., family resemblance, prototype, ‘open’) notions are more useful than the classically structured ones. On this Weitz-inspired approach, then, the question becomes not whether we can define art, but, first of all, what sort of concept of art will be most useful or valuable and, second of all, whether a classically definable/defined concept will be more useful or valuable than an undefined one. The question of use or value here cannot be divorced from the purposes for which we are to use the concepts. So we might want to think of the various purposes and functions the concepts in question might serve. It strikes me that we might think broadly of creative functions (i.e., the function that the concept plays in the production of artistic works), appreciative functions

June 21, 2008 10:57 MAC/NWA Page-138 9780230_220478_08_cha07 Aaron Meskin 139

(i.e., the function that the concept plays in ordinary engagement with various works), and critical functions (i.e., the functions that the concept plays in critical practice). Weitz’s creativity argument might be best understood as focusing on the creative aspect of the concept. His thought is that a classically structured concept will be problematic if we want a concept that will allow for or encourage creativity and novelty. Since we want creativity and novelty from the art world, the use of such a restrictive concept would be unattractive. Still, the argument does not work. A perfectly closed concept of art will allow for—perhaps even encourage—creativity and adventurousness. (Perhaps this is the very purpose of the rather strict definition of VILLANELLE that I have discussed above. Necessary conditions provide a framework in which an artist can display creativity and skill in achieving some difficult task.) But you can see—I hope—how the story might go. My discussion has been, up until this point, rather abstract. I have suggested that Weitz may be read as suggesting that, although we can define art we ought not, and I have also suggested that we might make sense of this claim in one of two ways. Moreover, I have suggested that the more interesting way to approach the suggestion is to think of it in the light of the possibility (I think actuality) of there being many concepts of art available to us. If there are, surely some are definable; for example, OBJECT WITH AN AESTHETIC FUNCTION,65 or Dickie’s institutional concept.66 But it is plausible that some such concepts are not so structured. Weitz may be read as suggesting we ought to use the latter, not one of the former. And even if Weitz’s creativity argument does not work, there is still something to be said for thinking of the issue this way. Before attempting to apply the argument to the case of one of ART’s sub- concepts, I would like to address a concern. It is neatly expressed by Nick Zangwill’s pointed remarks on defining art:

What we are in search of is a good theory of the nature of art. This issue does not concern the word ‘art’ or the category or concept of art. We want to know about a range of objects and events, not about the words or concepts that we use to talk about those things. We are interested in objects, not concepts – the world not words.67

I have already suggested that this might not be an entirely fair characterization of what aestheticians have been interested in. The very fact that they frequently talk of definition and concepts suggests otherwise. But even if it is the case, I am not sure it undercuts my suggestion that we might shift from a purely descriptive task to some sort of normative task. For even at the level of objects and events, we can and must choose which things to investigate and theorize about. Zangwill says: ‘we must set up the target and then shoot at it’,68 but then we face the question of which target to set up and where. He

June 21, 2008 10:57 MAC/NWA Page-139 9780230_220478_08_cha07 140 From Defining Art to Defining the Individual Arts suggests we look at the set of things that have an aesthetic function and theorize about those. It is true that there might be some interesting things to say about that category of things. But not all categories are equally interesting (e.g., the category of things that are blue and have an aesthetic function is not likely to be an especially fruitful one to investigate). Understood metaphysically, then, Weitz can be interpreted as suggesting that we will be better off focusing on and theorizing about a class of things (i.e., a class that contains the paradigm cases of art and many or most of the things that are widely agreed to be art), in which membership is not governed by necessary and sufficient conditions. This will have greater payoff than attending to a class in which membership is governed by simple non-enumerative necessary and sufficient conditions. If we have to set up a target to shoot at, we should set up an open one. Well then, why might a non-definable concept of art (or its various sub- concepts) be more useful or valuable than a definable one? My proposal is to focus on the sub-concepts of art. It strikes me that we can say a bit more about what we want from these concepts than from the concept of art itself. I will focus on one art form in particular (the comic) and while I will not argue that it cannot or should not be defined, I will try to make a case that thinking normatively rather than descriptively about the issue of definition may be fruitful.

IV. Defining an art form: Are narrative, pictures, and speech balloons essential to comics?

While not many philosophers have paid attention to the art of the comic, there has been a small amount of discussion of the definition of the art form. In the only book-length study of the subject by a philosopher, David Carrier offers three putatively essential conditions for being a comic: ‘the speech balloon, the closely linked narrative, and the book-size scale’.69 And Greg Hayman and Henry John Pratt claim that something is a comic iff it is a ‘sequence of discrete, juxtaposed pictures that comprise a narrative, either in their own right or when combined with text’.70 So Carrier, quite tenden- tiously, thinks that speech balloons are a necessary feature of comics, and he agrees with Hayman and Pratt that narrative is another such necessary feature. Moreover, Hayman and Pratt are committed to the essentially pictorial nature of comics. I believe all these views are mistaken. More to the point, I believe that there is a good reason for not deploying a concept structured in such a way. But how to make the case? It is fairly easy to find examples of things that are commonly thought of as comics that lack speech balloons. In fact, Hayman and Pratt provide two of their own,71 and the comics literature is filled with other cases. The non-narrative comic is harder to find, although I have suggested elsewhere that it is easy to imagine and make sense of possible cases, and that Robert

June 21, 2008 10:57 MAC/NWA Page-140 9780230_220478_08_cha07 Aaron Meskin 141

Crumb’s oeuvre provides some examples.72 In addition, some instructional comics are plausibly non-narrative, and single panel cartoons which do not narrate are often characterized as comics. And many art forms that we think of as predominantly or typically narrative are not essentially narrative (e.g., literature, film, television). So we have reason to think neither speech balloons nor narratives are necessary features of comics. Non-pictorial comics are an even trickier case, but I have argued that the existence of non-pictorial instances of the typically pictorial arts (e.g., films that traffic in mere images rather than full-fledged pictures) suggests that they are possible.73 But, of course, the argument does not end there. It is not implausible that there are a range of concepts at play here. Hayman and Pratt make a good case for the importance of the concept of the sequential pictorial narrative, and Carrier has some things to say in favour of the view that the class of narratives with speech balloons is an interesting kind. Moreover, it would be open to either to claim that my putative counterexamples to their accounts are simply red herrings—non-comics masquerading as comics. How might we settle the issue? It is hard to see how appeal to ordinary intuitions could do so. Most ordinary comics readers have not thought about non-narrative cases let alone non-pictorial ones. As I mentioned above, a common move in the discussion of the definition of art is to suggest that more than one concept of art is at play in contemporary thought and discourse. If so, it is not too much of a leap to think that there are a range of concepts of the comic at play in contemporary discourse about that particular art form.74 So one thing that might happen with the debate about the definition of comics is an agreement that the concept is fragmented—we can all then be perfectly correct about the specific concepts we are interested in. But, following Weitz, we might also want to ask which concept of comics is better or more useful—that is, which one of the concepts we have most reason to use. My suggestion is that there are some good art-critical and art-appreciate reasons for not using the concept of the sequential pictorial narrative or the concept of the narrative with speech balloons. What are those reasons? In short, the central problem with those classically structured concepts is that they treat certain aesthetically relevant features of comics as necessary; that is, they treat features such as the possession of speech balloons or the exemplification of narrative as ‘too standard’ to play any important aesthetic role. The guiding idea here is that works of art are appreciated, evaluated, and interpreted with reference to the art categories in which they are judged to fall.75 And, in particular, judgements of categorial typicality and atypicality play a central role in our engagement with art. Seeing a work of art as atypical for a salient art category (or at least as having an atypical feature for that category) often leads us to judge it very differently than if we were to have judged it as typical for that category (or for another salient category).

June 21, 2008 10:57 MAC/NWA Page-141 9780230_220478_08_cha07 142 From Defining Art to Defining the Individual Arts

Moreover, this fact is central to a wide range of contemporary art practices. That is, the aesthetic use of atypicality is a central strategy in contemporary art. And the use of typicality (or ‘standardness’) is central to a wide range of artistic practices both contemporary and non-contemporary. But treating membership in those art categories as governed by strictly necessary conditions will work against our seeing certain works as atypical instances of salient categories and, instead, encourage us to see them as typical members of other salient categories. Similarly, a commitment to necessary conditions will lead us to see those features as something more than typical—that is, they will be, as Kendall Walton puts it, ‘too standard’ to play an important aesthetic role.76 If this is right, then we have a pro tanto reason to use the concepts of the art forms (and genres and styles) that allow for maximal typicality and atypicality effects. Classically structured concepts can allow for some such effects, but (by definition!) they will not allow for typicality and atypicality when it comes to putatively defining features. Squares are not typically four-sided; they are essentially four-sided. Divisibility by some number other than itself and one is not atypical of primes; it is sufficient for not being a prime. A concept of comics in which narrative is essential does not allow for robust judgements of typicality and atypicality with respect to the presence of narrative. Non-narrative items are, ipso facto, excluded from the category of comics—they are not atypical comics; they are non-comics. Ditto for the case of the non-pictorial. But this means that we will lose out on various aesthetic effects if we treat comics as essentially narrative or pictorial; similarly with treating comics as essentially including speech balloons. If we stay at the level of the purely descriptive, I do not see how the dispute about comics and their relation to narrative can be resolved. But if we shift to the normative question—and ask what concept of comics will pay off critically and appreciatively—I think we have reason to choose a concept that does not treat narrative (or the speech balloon) as an essential condition of the concept.77 Can an argument of this form be used to make a full-fledged normative case for anti-essentialism? I do not think so. It is hard to see how to apply it to institutional and historical theories. They have their own troubles. I do, however, think it might be marshalled against any putative definition of an art form that appealed to an aesthetically relevant feature of that art form. And even if it cannot, it looks to have some payoff in helping us settle some disputes about the nature of the various art forms.

Conclusion

The pursuit of definition at the level of the arts or art forms might seem promising, but I have suggested a few reasons to be sceptical of such a

June 21, 2008 10:57 MAC/NWA Page-142 9780230_220478_08_cha07 Aaron Meskin 143 project. On the one hand, there are broadly empirical reasons for thinking that most concepts will not admit of definition—at least if our methods of pursuing definition rely primarily on categorization judgements and intuitions. This leaves open the possibility of finding a real definition of art in some other way, but there is some reason to question the applicability of standard approaches to solving this sort of problem (i.e., scientific investigation), and it is not clear what the alternatives are. I have also suggested that we may have pro tanto reasons to resist defining certain art form concepts—or, more precisely, to resist using concepts of those art forms that are governed by certain kinds of necessary and sufficient conditions. But what about the attractive features of the definitional project writ small that I mentioned in the introduction? Do they not give us good reason to pursue that project? I think not. The fact that art form concepts encompass less diversity than the art concept itself is irrelevant to the question of their susceptibility to definition. Moreover, the concepts of painting, sculpture, and music turn out to contain more diversity than one might initially think. It is plausible that there are ontological distinctions within art form categories (e.g., between recorded, performed, and improvised music), and many of the individual arts display almost as wide a range of functions as art.78 AndIhave already mentioned the possibility that art form concepts exhibit a significant degree of fragmentation; that is, that there are many extant concepts of the various art forms. It is true that we do not have much of an inductive base for scepticism about the possibility of successfully defining many of the art forms, but this lack of evidence provides us with no positive evidence that they can be defined. It is not obvious that ‘being halfway there’ with respect to the definition of the various arts actually gives us much reason to think a full-fledged definition is possible. As Timothy Williamson has pointed out, the fact that we can establish one necessary condition X for being F does not, by itself, give us reason to think that we can fill out the definition non-circularly by conjoining some other condition to X.79 Perhaps the most significant attraction of the definitional project writ small is the promise it holds out for providing us with guidance regarding the interpretation, evaluation, and appreciation of works of art. The most plausible of the contemporary attempts to define art (i.e., institutional and historical theories) provide us with no guidance with respect to these crucial activities. So one cannot help but hope that definitions of theatre or film or the comic book will provide such help. Perhaps, although I doubt it. In any case, definition is unnecessary. We do not need definitions of art or the individual art forms in order to engage in legitimate interpretation, evaluation, and appreciation. We have been doing just fine without having those definitions—warranted evaluation, interpretation, and appreciation have never waited on philosophical definitions of the various arts. Just so long as we know enough about how the various art forms, genres, and styles work (e.g., what purposes they can and have been put to, what their standard,

June 21, 2008 10:57 MAC/NWA Page-143 9780230_220478_08_cha07 144 From Defining Art to Defining the Individual Arts variable, and contra-standard features are), we have all we need. We may need theories of the arts, but I do not see any reason to think that we need definitions.

Acknowledgements

Thomas Adajian commented on an earlier version of this chapter, and I discussed the ideas in it with Stacie Friend, Matthew Kieran, Andrew McGonigal, Kathleen Stock, and Jonathan Weinberg. I owe thanks to all of them.

Notes

1. For a nice discussion of why the ‘What is art?’ question cannot be considered the ‘grand basic question’ of aesthetics, see Walton, ‘Aesthetics’, especially pp. 147–150. 2. Weitz, ‘Role of Theory’, 27. 3. See Dickie, Art and the Aesthetic and Art Circle. 4. See Beardsley, ‘An Aesthetic Definition’; Zangwill, ‘Are there Counterexamples?’ 5. See Danto, Transfiguration of the Commonplace. 6. See Stecker, Artworks, ‘Is it Reasonable?’, and Aesthetics. 7. See Levinson, ‘Defining Art’, ‘Refining Art’, and ‘Irreducible Historicality’. 8. Kivy, ‘Differences’, 131. 9. See Gaut, ‘Cluster Concept’, and ‘Cluster Account Defended’. In Meskin, ‘Cluster Account Reconsidered’, I suggest that Gaut’s continued pursuit of ‘art theory’ in the broad sense is the source of some problems with his account. 10. Kivy, ‘Differences’, 131. 11. Arthur Danto’s work (most significantly his Transfiguration of the Commonplace)is the great exception here, and it would be interesting to consider why this is the case. 12. Passmore, ‘Dreariness’, 355. 13. Levinson, ‘Concept of Music’; Lamarque and Olsen, Truth, Fiction, and Litera- ture; Stecker, ‘What is Literature?’; Carroll, ‘Moving Image’; Ponech, ‘Substance of Cinema’; Hayman and Pratt, ‘What are Comics?’; Ribeiro, ‘Intending to Repeat’; Beardsley, ‘Dance?’. 14. But see Currie, Ontology of Art, and Davies, Art as Performance, for arguments in favour of a unified ontology of the arts. 15. For critical discussion of the fragmentation thesis as applied to the concept of art, see Stecker, ‘Is it Reasonable?’, 54–56. 16. Some of the terms we ordinarily use to pick out the art forms (e.g., ‘comic’, ‘film’, ‘dance’) are often used in ways that their proper application does not entail art status. In fact, it is plausible that there are a great many comics that are not art (e.g., instructional comics and comics used in advertising), and there are certainly many films that are not art (e.g., Department of Defense film productions). In fact, many such terms are ambiguous between picking out a medium (a complex communicative technology) and an art form (an institutionalized practice of artmaking within a medium or a set of media). Only on the latter use does the necessary condition seem to hold. 17. See below for further discussion. 18. Budd, Values of Art,3.

June 21, 2008 10:57 MAC/NWA Page-144 9780230_220478_08_cha07 Aaron Meskin 145

19. The medium-specificity thesis—the view that each art form is embodied in a distinctive medium, and that the nature of the medium provides norms about what should be attempted in that art form—holds an attraction for many theorists of the individual arts, and it seems to offer a way of connecting definition with a theory of value. See Carroll, ‘Specificity of Media’, for a discussion and criticism of the thesis. Consider also David Carrier’s claim that ‘to interpret an art, we need to know its essence’. Aesthetics of Comics,7. 20. Might it not be possible to construct implicit definitions out of such theories? Perhaps. It would be worthwhile to pursue the potential of treating ‘art’ and its sub-concepts as theoretical terms and attempting to analyse them according to the Ramsey–Carnap–Lewis approach. It must be said that this has not been the explicit approach in any of the important work on art’s definition, but for one of the very few discussions of how such a project might go, see Scholz, ‘Rescuing the Institutional Theory’. (Andrew McGonigal has also suggested such an approach to me in personal conversation.) While I am sceptical of this project (e.g., the problem of multiple realizers looks especially serious in the case of art and appeals to naturalness seem unlikely to be of much help), in this chapter I shall ignore it and, instead, focus on standard attempts to provide straightforward explicit definitions. 21. Weitz, ‘Role of Theory’, 32. 22. See Carroll, Philosophy of Art, 218–219, for a statement of this latter problem. 23. Weitz, ‘Role of Theory’, 32. 24. See Eaton, ‘A Sustainable Definition’, 141–142, for a brief discussion of some practical reasons why we might want to pursue art’s definition. Many other authors have touched on the issue in their discussions of the topic. 25. See Carroll, ‘Art, Practice, and Narrative’, ‘Historical Narratives’, ‘Identifying Art’, and PhilosophyofArt, for some expression of scepticism about whether we need to define art to achieve the central purpose for which definition has been pursued in the current era—the identification of controversial cases. 26. So is there a tension in Weitz’s arguments, that is a tension between a claim of indefinability on the one hand and a claim about the inadvisability of defining on the other? In short, yes. One might try to resolve the tension in various ways, but I am simply happy to explore both claims since both are interesting. It is no part of my brief to defend the coherence of Weitz’s paper or, in fact, any of his particular arguments. 27. Weitz, ‘Role of Theory’, 32. 28. These are more interesting for two reasons: (1) they appear to correspond to lexical (i.e., monomorphemic) items in natural language (see n. 34); and (2) they do not have fixed extensions. 29. Weitz, ‘Role of Theory’, 31. 30. Ibid., 32. 31. The classic statement of this criticism can be found in Mandelbaum, ‘Family Resemblances’. 32. Anderson, ‘Aesthetic Concepts’; Zangwill, ‘Are there Counterexamples?’. 33. Weitz, ‘Role of Theory’, 31. 34. Roughly speaking, lexical concepts are those concepts that correspond to monomorphemic lexical items in natural language. See Laurence and Margolis, ‘Concepts and Cognitive science’, 4. ART and SCULPTURE appear to be lexical concepts. LARGE SCULPTURE MADE OF STEEL and SUPERHERO COMIC WRITTEN BY GRANT MORRISION are not lexical.

June 21, 2008 10:57 MAC/NWA Page-145 9780230_220478_08_cha07 146 From Defining Art to Defining the Individual Arts

35. Stich, ‘What is a Theory?’, 250; Ramsey, ‘Conceptual Analysis’, 74; Williamson, Knowledge and its Limits, 31. 36. Stich, ‘What is a Theory?’, 250. 37. Ramsey, ‘Conceptual Analysis’, 174. 38. Williamson, Knowledge and its Limits, 31. 39. Stecker, ‘Is it Reasonable?’, 54. 40. The very fact that some concepts (e.g., some mathematical and logical concepts as well as the other examples mentioned above) look to admit of fairly straightforward definition suggests that not all concepts are alike. See Machery, ‘Concepts are Not’, for additional reasons. See also Margolis and Laurence, ‘Concepts’, for the suggestion that ‘different types of concepts may well have correspondingly different types of structure’. 41. Note that this would leave open the possibility of appealing to a special ‘narrow’ inductive argument against the possibility of defining art (i.e., past instances of attempts to define the concept have failed so it is likely that future instances will fail too). I put this argument aside. For discussion of the narrow inductive argument, see Stecker, ‘Is it Reasonable?’, and Gaut, ‘Cluster Account Defended’. 42. Dean, ‘Nature of Concepts’. See Stich, ‘What is a Theory?’, and Ramsey, ‘Conceptual Analysis’, for general versions of the arguments. 43. See Ramsey, ‘Prototypes’, for relevant discussion. 44. Carroll, PhilosophyofArt, 251. 45. Adajian, ‘Prototype Theory’; Stecker, Aesthetics. 46. Adajian, ‘Prototype Theory’, 234; Stecker, Aesthetics, 103–104. 47. For robust criticism of prototype theory see Fodor, Concepts. One fundamental problem with prototypes is that they do not appear to respect compositionality— the idea that ‘complex concepts inherit their contents from those of their constituents’. Concepts, 100. 48. See Fodor, Concepts, for a defence of atomism. 49. Adajian, ‘Prototype Theory’, 234. 50. Ibid., 232–233; Rey, ‘Concepts and Conceptions’. 51. Adajian, ‘Prototype Theory’, 233. 52. Levinson, ‘Irreducible Historicality’, 367, and ‘Concept of Music’, 268. 53. Stecker, ‘Is it Reasonable?’, 60. 54. Ramsey, ‘Prototypes’, 174. 55. See Putnam, ‘Meaning of ‘‘Meaning’’ ’ , for the canonical account. 56. See Davies, ‘Essential Distinctions’, 3–6, for relevant discussion. 57. This thought is inspired by some of the claims made by Dominic McIver Lopes in ‘Nobody Needs a Theory’. 58. See Leddy, ‘Rigid Designation’, for relevant discussion. 59. Stecker, ‘Is it Reasonable?’, 60. 60. Adajian, ‘Prototype Theory’, 233. 61. See Laurence and Margolis, ‘Concepts and Cognitive science’, 16. 62. See Putnam, ‘Analytic and Synthetic’. 63. Dutton, ‘A Naturalistic Definition’, 368. 64. Jackson, ‘Critical Notice’, 518–519. See also Matthias Steup’s essay on the ‘The Analysis of Knowledge’ in the Stanford Encyclopedia of Philosophy: ‘It might be a mistake to expect that there is a decisive argument that settles the dispute between internalists and externalists one way or another. One way to respond to the intractability of the debate is to acknowledge that there simply is not one concept of knowledge for which there is an analysis that has any chance of

June 21, 2008 10:57 MAC/NWA Page-146 9780230_220478_08_cha07 Aaron Meskin 147

meeting with broad assent. Rather, we might conclude that, when we use the word ‘‘knowledge’’, we have sometimes one concept and at other times another concept in mind’. Steup, ‘Analysis of Knowledge’. 65. Cf. Zangwill, ‘Are there Counterexamples?’. 66. Dickie, Art and the Aesthetic, 34. 67. Zangwill, ‘Are there Counterexamples?’, 116. 68. Ibid., 117. 69. Carrier, Aesthetics of Comics, 74. 70. Hayman and Pratt, ‘What are Comics?’, 423. 71. Ibid., 422. 72. Meskin, ‘Defining Comics?’, 372. 73. Ibid., 374. See Carroll, ‘Moving Image’, 65–66, for discussion of the film case. 74. And so it turns out that there is fragmentation at the level of the art form concepts too! For another example, consider Levinson’s focuses on the contemporary West- ern concept of music. See Levinson, ‘Concept of Music’. Surely, there are other concepts of music. 75. Walton, ‘Categories of Art’. 76. Ibid., 349. 77. This is only a pro tanto reason. It might be outweighed by other considerations. Perhaps this is how to understand the case of the closed poetic forms mentioned above. The pro tanto value of using an ‘open’ concept is outweighed by the valuable payoff (e.g., with respect to encouraging creativity and the exhibition of skill) that comes from treating such forms as straightforwardly definable. 78. See Carroll, ‘Specificity of Media’, for devastating criticisms of the view that each art form has a different function or set of functions. 79. Williamson, Knowledge and its Limits, 32–33.

Bibliography

Adajian, T. 2005. ‘On the Prototype Theory of Concepts and the Definition of Art’, Journal of Aesthetics and Art Criticism, 63: 231–236. Anderson, J. 2000. ‘Aesthetic Concepts of Art’ in N. Carroll (ed.), Theories of Art Today, Madison: University of Wisconsin Press, 65–92. Beardsley, M. 1982. ‘What is Going on in a Dance?’ Dance Research Journal, 15: 31–37. Beardsley, M. 1983. ‘An Aesthetic Definition of Art’ in H. Curtler (ed.), What Is Art? New York: Haven, 15–29. Budd, M. 1995. Values of Art: Pictures, Poetry, and Music, Harmondsworth: Penguin. Carrier, D. 2000. The Aesthetics of Comics, University State Press: Pennsylvania University Park. Carroll, N. 1985. ‘The Specificity of Media in the Arts’, Journal of Aesthetic Education, 19: 5–20. Carroll, N. 1988. ‘Art, Practice, and Narrative’, Monist, 71: 140–156. Carroll, N. 1993. ‘Historical Narratives and the Philosophy of Art’, Journal of Aesthetics and Art Criticism, 51: 313–326. Carroll, N. 1994. ‘Identifying Art’ in R.J. Yanal (ed.), Institutions of Art, University Park: Pennsylvania University State Press, 3–38. Carroll, N. 1996. ‘Defining the Moving Image’, Theorizing the Moving Image, Cambridge: Cambridge University Press, 49–74. Carroll, N. 1999. Philosophy of Art: A Contemporary Introduction, London and New York: Routledge.

June 21, 2008 10:57 MAC/NWA Page-147 9780230_220478_08_cha07 148 From Defining Art to Defining the Individual Arts

Carroll, N. (ed.) 2000. Theories of Art Today, Madison: The University of Wisconsin Press. Currie, G. 1989. An Ontology of Art, New York: St. Martin’s Press. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Davies, D. 2004. Art as Performance, Malden, MA: Blackwell. Davies, S. 2003. ‘Essential Distinctions for Art Theorists’ in S. Davies and A.C. Sukla (eds), Art and Essence, Westport, CT:Praeger, 3–16. Davies, S. and Sukla, A.C. (eds) 2003. Art and Essence, Westport, CT: Praeger. Dean, J. 2003. ‘The Nature of Concepts and the Definition of Art’, Journal of Aesthetics and Art Criticism, 61: 29–35. DePaul, M. R. and Ramsey, W. (eds) 1998. Rethinking Intuition: The Psychology of Intuition and its Role in Philosophical Inquiry, Lanham, MD: Rowman and Littlefield. Dickie, G. 1974. Art and the Aesthetic: An Institutional Analysis, Ithaca, NY: Cornell University Press. Dickie, G. 1984. The Art Circle: A Theory of Art, New York: Haven. Dutton, D. 2006. ‘A Naturalistic Definition of Art’, Journal of Aesthetics and Art Criticism, 64: 367–377. Eaton, M. 2000. ‘A Sustainable Definition of ‘‘Art’’ ’ in N. Carroll (ed.), Theories of Art Today, 141–159. Fodor, J. 1998. Concepts: Where Cognitive Science Went Wrong, New York: Oxford University Press. Gaut, B. 2000. ‘ ‘‘Art’’ as a Cluster Concept’ in N. Carroll (ed.), Theories of Art Today, Madison: The University of Wisconsin Press, 25–44. Gaut, B. 2005. ‘The Cluster Account of Art Defended’, British Journal of Aesthetics, 45: 273–288. Hayman, G. and Pratt, H.J. 2005. ‘What Are Comics?’ in D. Goldblatt and L. Brown, (eds), Aesthetics: A Reader in Philosophy of the Arts, Upper Saddle River, NJ: Pearson Education Inc., 419–424. Jackson, Frank. 2002. ‘Critical Notice of Knowledge and its Limits by Timothy Williamson’, Australasian Journal of Philosophy, 80: 516–521. Kivy, P. 1993. ‘Differences’, Journal of Aesthetics and Art Criticism, 51: 123–132. Lamarque, P. and Olsen, S. 1994. Truth, Fiction, and Literature, Oxford: Clarendon Press. Laurence, S. and Margolis, E. (eds) 1999. ‘Concepts, and Cognitive Science’, Concepts: Core Readings, Cambridge, MA: MIT Press. Leddy, T. 1987. ‘Rigid Designation in Defining Art’, Journal of Aesthetics and Art Criticism, 45: 263–272. Levinson, J. 1979. ‘Defining Art Historically’, British Journal of Aesthetics, 19: 232–250. Levinson, J. 1989. ‘Refining Art Historically’, Journal of Aesthetics and Art Criticism, 47: 21–33. Levinson, J. 1990. ‘The Concept of Music’, Music, Art, and Metaphysics: Essays in Philosophical Aesthetics, Ithaca and London: Cornell University Press. Levinson, J. 2002. ‘The Irreducible Historicality of the Concept of Art’, British Journal of Aesthetics, 42: 367–379. Lopes, D.M. forthcoming. ‘Nobody Needs a Theory of Art’, Journal of Philosophy. Machery, E. 2005. ‘Concepts are not a Natural Kind’, Philosophy of Science, 72: 444–467. Mandelbaum, M. 1965. ‘Family Resemblances and Generalizations Concerning the Arts’, American Philosophical Quarterly, 2: 219–228. Margolis, E. and Laurence, S. 1999. Concepts: Core Readings, Cambridge, MA and London: Bradford Books/MIT Press.

June 21, 2008 10:57 MAC/NWA Page-148 9780230_220478_08_cha07 Aaron Meskin 149

Margolis, E. and Laurence, S. 2006. ‘Concepts’ in E.N. Zalta (ed.), The Stanford Encyclo- pedia of Philosophy (Spring 2006 Edition), URL = . Meskin, A. 2007. ‘The Cluster Account of Art Reconsidered’, British Journal of Aesthetics, 47: 388–400. Meskin, A. 2007. ‘Defining Comics?’ Journal of Aesthetics and Art Criticism, 65: 369–379. Passmore, J. 1951. ‘The Dreariness of Aesthetics’, Mind, 60: 318–335. Ponech, T. 2006. ‘The Substance of Cinema’, Journal of Aesthetics and Art Criticism, 64: 187–198. Putnam, H. 1962. ‘The Analytic and the Synthetic’ in H. Feigl and G. Maxwell (eds), Minnesota Studies in the Philosophy of Science III, Minneapolis: University of Minnesota Press, 358–397. Putnam, H. 1975. ‘The Meaning of Meaning’ in K. Gunderson (ed.), Language Mind and Knowledge, Minnesota Studies in the Philosophy of Science VII, Minneapolis: University of Minnesota Press, 131–193. Ramsey, W. 1998. ‘Prototypes and Conceptual Analysis’ in M.R. DePaul and W. Ramsey (eds), Rethinking Intuition, 161–177. Originally published in 1992. Topoi, 11: 59–70. Rey, G. 1985. ‘Concepts and Conceptions: A Reply to Smith, Medin and Rips’, Cognition, 19: 297–303. Ribeiro, A. 2007. ‘Intending to Repeat: A Definition of Poetry’, Journal of Aesthetics and Art Criticism, 65: 189–201. Scholz, B.C. 1994. ‘Rescuing the Institutional Theory of Art: Implicit Definitions and Folk Aesthetics’, Journal of Aesthetics and Art Criticism, 52: 309–325. Stecker, R. 1996. ‘What is Literature?’ Revue Internationale De Philosophie, 50: 681–694. Stecker, R. 1997. Artworks: Definition, Meaning, Value, University Park: Pennsylvania State University Press. Stecker, R. 2000. ‘Is it Reasonable to Attempt to Define Art?’ in N. Carroll (ed.), Theories of Art Today, Madison: University of Wisconsin Press, 45–64. Stecker, R. 2005. Aesthetics and the Philosophy of Art: An Introduction, Lanham, MD: Rowman and Littlefield. Steup, M. 2006. ‘The Analysis of Knowledge’ in E.N. Zalta (ed.) Spring 2006. The Stanford Encyclopedia of Philosophy, URL = . Stich, S. 1992. ‘What is a Theory of Mental Representation?’ Mind, 101: 243–261. Walton, K. 1970. ‘Categories of Art’, The Philosophical Review, 79: 334–367. Walton, K. 2007. ‘Aesthetics—What? Why? And Wherefore?’ Journal of Aesthetics and Art Criticism, 65: 147–250. Weitz, M. 1956. ‘The Role of Theory in Aesthetics’, Journal of Aesthetics and Art Criticism, 15: 27–35. Williamson, T. 2000. Knowledge and Its Limits, Oxford: Oxford University Press. Zangwill, N. 2002. ‘Are There Counterexamples to Aesthetic Theories of Art?’ Journal of Aesthetics and Art Criticism, 60: 111–118.

June 21, 2008 10:57 MAC/NWA Page-149 9780230_220478_08_cha07 8 Imagining Fact and Fiction Stacie Friend

Why does it matter whether a work is fiction or non-fiction? Gregory Currie claims that the distinction is crucial:

There can hardly be a more important question about a piece of writing or speech than this: Is it fiction or nonfiction? If the question seems not especially important, that’s because we rarely need to ask it. Most often we know, in advance of reading or hearing, that the discourse before us is one or the other. But imagine we did not know whether The Origin of Species is sober science or Borgesian fantasy on a grand scale. We would not know whether, or in what proportions, to be instructed or delighted by it. No coherent reading of it would be possible.1

Whether or not we agree with Currie’s view of the general importance of the distinction, the classification of a text as fiction or non-fiction can certainly make a difference to how we respond to it.2 We read history, not romance, to learn about the past. We worry about the serial killer identified in the newspapers, not the fictional psychopath in a popular novel. We praise Shakespeare for the artistically justified distortions in his history plays, whereas we would be critical of such blatant falsehoods in a history of England. Furthermore, there are serious debates about the classification of works as fiction or non-fiction outside philosophy. Consider the controversies over Edmund Morris’s Dutch: A Memoir of Ronald Reagan (1999) and John Berendt’s Midnight in the Garden of Good and Evil (1994). Morris (in)famously inserted himself into the story as a fictional narrator so that he could be on hand to witness many of the events of Reagan’s life. Berendt made up a scene and changed the chronology of some events in order to be present when a murder is committed. It is because the books were published as non- fiction that critics objected to such devices. Although we cannot simply

150

June 21, 2008 11:1 MAC/NWA Page-150 9780230_220478_09_cha08 Stacie Friend 151 assume that such debates hinge on a philosophically interesting distinction, we should expect a theory of fiction to shed light on what is at stake. That is, we should expect philosophical definitions of fiction and non- fiction to clarify the significance of the distinction for understanding and appreciation. Several recent theories define fiction at least partly in terms of its function, in particular its role in prompting a certain sort of response in audiences. These theories have the following two components: (1) a specification of what it means to treat a work as fiction, and (2) a further condition, or a set of conditions, to distinguish those works that are fiction from works that might merely be treated as such. While accounts of (2) diverge, there is considerable consensus that (1) involves make-believe or (in this debate, equivalently) imagining. Thus Kendall Walton, Gregory Currie, David Davies, and Peter Lamarque and Stein Haugom Olsen agree that treating a work as fiction means adopting a certain attitude to the content of the work, an attitude characterized by make-believe.3 This way of approaching the distinction reflects the widespread rejection of theories of fiction that rely on semantic notions such as reference and truth.4 Many works of fiction refer to real people, places, things and events, and many works of non-fiction make false claims, whether through ignorance or deception. A newspaper article that contains false information does not thereby become fiction. The difference, for the theories under consideration, is that newspaper articles ask us to believe rather than to make believe their content.5 This distinction is often (though not always) drawn in terms of intention. For instance, Davies writes: ‘Whereas it is a condition for assertion that the speaker intends her audience to believe what she states, in fictive utterance the author intends that her audience make-believe what is narrated.’6 The claim is that fiction can be defined in terms of an intended response of make-believe. I do not think, however, that this way of understanding the nature of fiction is promising. I argue that there is no interpretation of imagining or make-believe that designates a response distinctive to fiction as opposed to non-fiction. The class of works that invite make-believe, however it is determined, is substantially broader than our ordinary concept of fiction would allow. The question is whether there is a way of understanding the sort of imagining involved in our engagement with fictions that would carve out a narrower category. I consider various possible interpretations and argue in each case that works of non-fiction may invite the same imaginative responses as fiction, just as works of fiction may invite the same cognitive responses as non-fiction. These considerations cast doubt on definitions of fiction that appeal to make-believe, and the attempt to save the theory by restricting it to individual statements rather than whole works is unsatisfactory. A different approach to classification is required if we wish to understand the significance of the distinction.

June 21, 2008 11:1 MAC/NWA Page-151 9780230_220478_09_cha08 152 Imagining Fact and Fiction

I. Walt-fiction

I begin with Kendall Walton’s theory, since he introduced the notion of make-believe into discussions of fiction. Walton defines a work of fiction as a work whose function is to act as a prop in certain games of make- believe. The connection with children’s playing, where the props may be dolls and toy trucks, is intentional; for Walton there is continuity between the two types of games. The idea is that works of fiction are designed to prescribe imaginings about their content, and imagining what is prescribed is participating in the game of make-believe authorized by the work. The notion of games makes the connection between correctness on the one hand and the imagination on the other: some moves in a game are licensed, others are not. Other theories of fiction retain the appeal to make-believe while dropping the reference to games.7 Even so, they agree that works of fiction guide or constrain our imaginative responses. While I am not prevented from imagining that Oliver Twist is eight-feet tall, there is a clear sense in which this would be an inappropriate response to Dickens’s novel. It is the function of prompting imaginings that distinguishes fiction from non-fiction on Walton’s account:

It is not the function of biographies, textbooks, and newspaper articles, as such, to serve as props in games of make-believe. They are used to claim truth for certain propositions rather than to make propositions fictional. Instead of establishing fictional worlds, they purport to describe the real world.8

What determines the function of a work? To answer this question, Walton refers to a number of factors that may be relevant, either individually or in combination: these include the author’s intentions, the typical uses of the work, views about how the work is appropriately used, and so on.9 Walton is often criticized for failing to distinguish between a work’s being fiction and its being treated as fiction.10 For his criteria of fictionality allow the possibility of classifying a work according to the way it is typically treated. In this approach some works, such as the Greek myths, may be non-fiction for their original audience but fictional for us; but one might instead want to claim that the myths remain non-fiction even though we treat them as fiction. I prefer to construe Walton as relativizing the distinction. His notion of function is loosely defined to allow for differences in practice. In particular, he acknowledges that in our own practice we have a tradition of respect for the origins of works, for how they were produced and understood. This appeal to origin sustains the distinction for us, though perhaps not in every case. By contrast, Walton’s critics build reference to intention into their theories: a text does not count as a genuine fiction unless it was intended to prompt make-believe. The crux of Gregory Currie’s definition of fiction, for example,

June 21, 2008 11:1 MAC/NWA Page-152 9780230_220478_09_cha08 Stacie Friend 153 is a Gricean explanation of fictive utterance, which is the communicative act that characterizes fiction-making. In fictive utterance, the speaker wants the audience to make believe a certain proposition P, and to do so at least partly as a result of their recognizing the speaker’s intention that they make believe P.11 David Davies and Lamarque and Olsen also adopt this part of Currie’s theory. From their perspective, even if we treated the Greek myths as fiction they would not be, on the assumption that they were intended to prompt a response of belief rather than what we may call ‘make-belief’. For the moment it does not matter how precisely the function of the work is determined, though I argue below that the intention to produce make- believe should not be taken as relevant to classification. Instead, I want to focus on Walton’s claim that any work that has the function of prescribing imagining counts as fiction. Because there is a great deal of fuzziness in making this determination, Walton stipulates that if a work has as just one of its functions to prescribe imaginings, it is fiction in this sense.12 As a result, Walton’s category of fiction turns out to be much broader than our ordinary conception: many works we would normally label as non-fiction are included. Any work that has, to the smallest degree, the function of serving as a prop in a game of make-believe will count as fiction for Walton. Berkeley’s Dia- logues, because they prescribe imaginings about the conversation between two fictional persons, so qualify. Does every work that involves intentional uses of non-referring names count as fiction? At certain points Walton wants to resist that conclusion. He says in MimesisasMake-Believe, ‘William Hazlitt’s Characters of Shakespeare’s Plays (1817) is largely about mere fictions, yet nothing is more unambiguously a work of nonfiction.’13 But since according to his theory all pretended reference to fictional characters calls for us to imagine that there are such characters, Hazlitt’s book must qualify as fiction. Fur- thermore, any works of history, biography, and autobiography, newspaper articles, journal entries, and so forth, insofar as they aim in any of their features to prescribe imaginings, will count as fiction for Walton. Yet, numerous works of history are designed to get the reader to imagine, say, what it was like to live in a different time and place, and most interesting works have this as at least one of their goals. And the category of fiction for Walton is not limited to narrative texts, but includes any texts containing metaphors as well as all pictures.14 Wearelikelytothink,bycontrast,thattheuseofnon- literal language does not by itself make a work fictional. And to the extent we are willing to recognize a fiction/non-fiction distinction for pictures, we would presumably say that journalistic photographs, courtroom drawings, and the like count as non-fiction. Pointing out these differences between our ordinary notion of fiction and Walton’s conception is no criticism of Walton, however; he agrees that the category he delineates is broader than our ordinary notion. Moreover, he does not take himself to be analysing that notion and even denies that it

June 21, 2008 11:1 MAC/NWA Page-153 9780230_220478_09_cha08 154 Imagining Fact and Fiction is desirable to do so.15 Walton is interested, not in our everyday concept of fiction, but in those works that prompt make-believe, however we pre- theoretically classify them. Such works he labels as fiction or, equivalently, representational art. I shall call them works of walt-fiction. There are various theoretical reasons to be interested in the wider category of walt-fictions, most obviously with respect to the role of imagination in artistic representation generally. Still, to the extent that we want to understand our ordinary practice of distinguishing between fiction and non- fiction—the practice that explains controversies over works like Dutch and Midnight, for example—we should not be satisfied with the category of walt- fiction. Consideration of walt-fictions does, however, highlight the problem I wish to address. Make-believe is an appropriate response to walt-fiction by definition, but the category of walt-fiction is broader than the category of fiction. Is there a kind of make-believe that is special to the narrower category?

II. Imaginative constructions

There is one form of imagining we can quickly dismiss as special to fiction rather than walt-fiction: the experience of mental imagery, traditionally construed as definitive of imagination.16 Both fiction and non-fiction elicit imagery, a fact illustrated by the following passages. Here is part of Eric Schlosser’s decidedly off-putting description of his first visit to a slaughterhouse in Fast Food Nation:

On the kill floor, what I see no longer unfolds in a logical manner. It’s one strange image after another. A worker with a power saw slices cattle into two halves as though they were two-by-fours, and then the halves swing by me into the cooler. It feels like a slaughterhouse now. Dozens of cattle, stripped of their skins, dangle on chains from their hind legs. ...The kill floor is hot and humid. It stinks of manure. Cattle have a body temperature of about 101 degrees, and there are a lot of them in the room. Carcasses swing so fast along the rail that you have to keep an eye on them constantly, dodge them, watch your step, or one will slam you and throw you onto the bloody concrete floor. It happens to workers all the time.17

Or consider this harrowing true account of attempted abortion by a woman who became pregnant in a Soviet camp, quoted in Anne Applebaum’s non- fiction tome Gulag:

Imagine the picture. It is night. It is dark ...Andrei Andreevich [the camp doctor] is trying to cause me to abort, using his hands, covered in iodine, without instruments. But he is so nervous that nothing comes of it.

June 21, 2008 11:1 MAC/NWA Page-154 9780230_220478_09_cha08 Stacie Friend 155

I can’t breathe from the pain, but I endure it without a sound, so that no one will hear. ‘Stop!’ I finally shout from unbearable pain, and the whole procedure is stopped for two days. In the end, everything came out—the fetus, with a great deal of blood. That is why I never became a mother.18

These passages cause readers to imagine the scenarios described, perhaps too plainly for comfort. I take it that for most readers, the images involved are created rather than recalled from memory.19 They are therefore clear cases of imaginative construction. Of course, many works of non-fiction do not prompt much imagery: demographic studies or political essays, for instance. Even works such as Fast Food Nation and Gulag contain more exposition and argument than they do narrative or descriptive elements. I suspect that such observations are responsible for the tendency to associate imagination with fictions, which are paradigmatically narrative, rather than with non-fictions. Yet many non-fictions are largely narrative, and not every section of a work of fiction need prompt mental images. Once we limit ourselves to the narrative element of a work, we should recognize that vivid storytelling, whether fiction or non-fiction, has the capacity to evoke imagery. Perhaps we should look to a different form of imaginative construction as distinctive of fiction. Walton suggests that one difference between fiction and non-fiction as we ordinarily classify them is that works of fiction generate their own fictional worlds.20 While some identify fictional worlds with possible worlds,21 in Walton’s view a fictional world is just the content we are supposed to imagine when we correctly interpret a work—that is, whatever is ‘true in the fiction’. The metaphor of ‘worlds’ emphasizes the sense in which fictional truth goes well beyond what is explicitly stated: we are supposed to imagine the characters and events represented in a narrative in far more detail than the text provides. A possible distinction between fiction and non- fiction could be drawn from these observations. We could say that whereas non-fiction aims to instil beliefs about the real world, fiction prompts us to construct fictional worlds in our imaginations. It turns out, however, that we engage in just that sort of imaginative activity when we read non-fiction narratives. Our ability to make inferences from a finite number of sentences constitutes a basic aspect of our comprehension of any text, fiction or non-fiction, no matter how short or simple, most of it on a subconscious level. Thus we know that Candide has blood in his veins rather than oatmeal, even though the text does not say this explicitly. As readers, we are required mentally to construct or reconstruct the ‘world of the work’, to keep track, for instance, of the main characters and settings, of the chronology of events, and so on. These inferences must draw upon an enormous amount of background knowledge, including the kinds of entities in the world, relations of cause and effect, commonly known facts about history, psychology, science, and geography, and familiarity with discourse

June 21, 2008 11:1 MAC/NWA Page-155 9780230_220478_09_cha08 156 Imagining Fact and Fiction conventions.22 The resulting mental representations of characters and events are not merely propositional in form: ‘For readers to carry out complex reasoning with respect to a text ...they typically must construct more elaborate models, situation models, which integrate information from the text with broader real-world knowledge.’23 A situation model is the reader’s representation, not of the text itself, but of the situation the text is about. It may be described as a complex representation of a world, created in a reader’s imagination. Because this creative process is common to fiction and non-fiction, we should conclude that neither mental images nor situation models, though both arguably products of imagination, constitute a type of make-believe special to fiction. Reading non-fiction often prompts the creation of imagery and of imagined worlds in much the same way as fiction. So far we do not have a way to demarcate the narrower domain of fiction within the class of walt-fictions.

III. Belief and imagination

The reason for the conclusion of the previous section should be obvious: the kinds of imaginative constructions discussed there are consistent with our believing the content that we also imagine. This may sound paradoxical, but the compatibility between belief and make-belief is straightforwardly demonstrated. It is basic to accounts of fictional truth that we import various beliefs about the real world in filling out a fictional world.24 On one plausible interpretation of how we construct fictional worlds, we start with a mental representation of the real world and modify it as required by the story.25 The resulting representation contains all the relevant beliefs about the real world that remain consistent with what is fictionally the case. If we imagine what is fictionally true, and what is fictionally true includes what we believe, then we imagine what we believe. Moreover, it is standard in psychological theories of the imagination to assume that one can believe and make believe the same content, as Shaun Nichols makes clear:

This is nicely illustrated in an experiment of Alan Leslie’s. Leslie had young children watch as he pretended to pour tea into two (empty) cups. Then he picked up one of the cups, turned it over and shook it, turned it back right side up and placed it next to the other cup. The children were then asked to point at the ‘full cup’ and at the ‘empty cup’. Both cups were really empty throughout the entire procedure, but two-year-olds reliably indicated that the ‘empty cup’ was the one that had been turned upside down and the ‘full cup’ was the other one .... On the most natural interpretation of this, the child is imagining that the cup is empty. But the child also, of course, believes that the cup is empty.26

June 21, 2008 11:1 MAC/NWA Page-156 9780230_220478_09_cha08 Stacie Friend 157

If this is right, there is no bar to our imagining and believing precisely the same thing, in fiction or in non-fiction. The obvious rejoinder to this line of thinking is to say that when theorists of fiction invoke make-believe, they mean by this something that contrasts with belief. Nichols’s point is just that the contrast rests not on a difference in the contents of the imagined or believed representations, but instead on their functions.27 One way of construing this suggestion is to say that although the imagination is involved in forming both beliefs and make-beliefs, we do different things with the outputs of these processes. In the case of non- fiction, we could say, we are supposed to integrate our representations with our beliefs about the actual world. By contrast, we need not make this move in response to fiction: we simply enjoy the representations of the fictional worlds for their own sakes. Studies of the way we store information from fiction and non-fiction in memory support this idea up to a point. A series of experiments investigated the ways in which readers incorporate some information into our long-term beliefs while compartmentalizing other information into a distinct situation model.28 The results suggested that incorporation was more likely when subjects believed the material to be non-fiction (‘real’) while compartmentalization was more likely when they believed it to be made up, either because it was fictional or because it was created for experimental purposes. We might think of compartmentalized representations as ones contained within the imagination, restricted from our beliefs about the world. While there is room here for a distinction between belief and imagination, though, the same studies showed that there is no sharp distinction in responses to fiction and non-fiction. When subjects thought they were reading a true account and therefore incorporated more information, they also retained a compartmentalized, coherent story representation in memory; and when subjects thought they were reading a fictional story, they incorporated many claims into their beliefs. In both cases information from a text does not necessarily go directly into a reader’s long-term beliefs; whether or not the text is fiction, information usually passes through a situation model from which readers selectively incorporate. A variety of factors influence the degrees of incorporation and compartmentalization. It is possible that some readers might incorporate more information from, say, a roman aclef` than they do from a notoriously unreliable autobiography or political treatise. Our responses to fiction and non-fiction involve a mix of belief and imagination. Even if this is so, it seems to miss an important point. Surely (one might object) the purpose of a work of non-fiction is to cause belief, even if it also prompts imagining; and the purpose of a work of fiction is to prompt imagining, even if it also causes belief. It is not so clear, however, that fiction and non-fiction do have these different purposes. Such ‘nonfiction novels’ as Truman Capote’s In Cold Blood (1965) use real events as an excuse to create an entertaining narrative. Capote, who invented the genre (or at least its

June 21, 2008 11:1 MAC/NWA Page-157 9780230_220478_09_cha08 158 Imagining Fact and Fiction name), chose a brief newspaper article about the murder of a family in Kansas as the subject for a new kind of non-fiction narrative. He insisted that he did not make anything up but his motivation, like that of many recent ‘true crime’ writers and non-fiction novelists, was more literary than journalistic. At the same time, authors of fiction frequently intend their works to induce belief, and in many cases this intention constitutes the primary aim of the work. Think of Upton Sinclair’s The Jungle (1906), an expose´ of the US meat packing industry that helped lead to legislation prohibiting unsafe food han- dling practices; Aleksandr Solzhenitsyn’s One Day in the Life of Ivan Denisovich (1962), an account of a typical day for a prisoner in a Soviet labour camp; or Joel Klein’s Primary Colors (1996), the journalist’s roman aclef` about Bill Clinton’s 1992 presidential campaign. In each of these cases, the fictionalized presentation serves mainly to promote certain beliefs. Fiction and non-fiction do not seem to be distinguished by the overall purposes of their authors.29 Suppose we agree that works of fiction might have the same ends as works of non-fiction (and vice versa). Still, there does seem to be a difference in the means. Non-fictions such as In Cold Blood prompt belief directly, by explicitly asserting the content we are to believe.30 By contrast, fictions such as The Jungle do so only indirectly, by getting us to infer conclusions from content that we merely imagine. Consider this passage from Sinclair’s novel:

Jonas had told them how the meat that was taken out of pickle would often be found sour, and how they would rub it up with soda to take away the smell, and sell it to be eaten on free-lunch counters; also of all the miracles of chemistry which they performed, giving to any sort of meat, fresh or salted, whole or chopped, any colour and any flavour, and any odour they chose.31

Given the background facts about the novel, the reader knows that he/she can infer from this sentence that everything following ‘Jonas had told them’ was true of meat packing industry practice in early twentieth-century Chicago. Nonetheless, what is said by the sentence—the proposition literally expressed—is not true, since Jonas is a fictional character and thus did not tell anyone of the conditions reported.32 Although we are supposed to imagine the contents of sentences about fictional characters, we are not supposed to believe them. Call the attitude that consists in imagining-but-not-believing, mere-make-believe. Perhaps there is some sense in which mere-make-believe is special to fiction.

IV. Mere-make-believe

The hypothesis suggested by the previous section is this: with respect to the propositions expressed by the sentences in the text, fiction (as opposed to non-fiction) invites mere-make-believe, whereas non-fiction (as opposed to

June 21, 2008 11:1 MAC/NWA Page-158 9780230_220478_09_cha08 Stacie Friend 159 fiction) invites belief.33 This proposal may seem plausible given that mere- make-believe is appropriate to those features of a work that are made up (and known to be so),34 and it is common to associate fiction with such features. Moreover, appeal to mere-make-believe would carve out a category narrower than walt-fiction, which need not contain any made-up elements. For instance, Walton accepts the possibility of a ‘genre of historical novel in which authors are allowed no liberties with the facts and in which they are understood to be asserting as fact whatever they write’, of which New Journalism, ordinarily classified as non-fiction, provides an example.35 Other philosophers offer a further condition on fictionality designed to rule out such genres. They claim that it is not enough that a work be the product of a fictive utterance intended to invite make-believe; it must also contain fictive content, that is, content that is the product of the author’s imagination. According to Currie, ‘a work is fiction iff (a) it is the product of a fictive intent and (b) if the work is true, then it is at most accidentally true’.36 The condition (b) is designed to rule out cases where an author invites audiences to make-believe a true story. Davies also seems to take the combination of fictive intent and a restriction on content to supply necessary and sufficient conditions for fictionality, though he thinks that a non-accidentally true narrative could still be fiction.37 In his view fictionality requires that (1) the author intends that readers make believe the narrated events, and (2) it is not the case that ‘correspondence with the manner in which events actually transpired was taken, by the utterer, to be a constraint that the ordering of events in [the text] must satisfy’.38 If the author is guided by this fidelity constraint, the work counts as non-fiction.39 Though they do not offer necessary and sufficient conditions in these terms, Lamarque and Olsen agree that fictionality requires fictive content. They gloss this as the requirement that ‘how things are (in the fiction) is determined by how they are described to be in a fictive utterance’, rather than the other way around.40 Leaving aside the details, each of these proposals is motivated by the assumption that the content of non-fiction narratives is determined by what really happened—as Currie puts it, non-fiction works ‘display counterfactual dependence on the facts’41—while the content of fictional stories is determined by their authors. In this sense fictional content is ‘made up’, and thus invites mere-make-believe. The connection between non-fiction and belief on the one hand, and fiction and mere-make-believe on the other, is intuitive. Nonetheless, these considerations fail to identify a response distinctive to works in each category. First, we should reject outright the claim that fiction invites no belief, but only mere-make-believe, in its explicit content. We read this in Ian Flem- ing’s Thunderball (1961): ‘New Providence, the island containing Nassau, the capital of the Bahamas, is a drab sandy slab of land fringed with some of the most beautiful beaches in the world.’42 This statement is not only true, it was intended to be true, and any informed reader of Fleming will believe

June 21, 2008 11:1 MAC/NWA Page-159 9780230_220478_09_cha08 160 Imagining Fact and Fiction it. It meets all the standard requirements on (sincere) assertion. It might seem tempting to hold that Fleming is not making an assertion because the statement occurs within a work of fiction; however, this begs the question.43 Perhaps the concern is rather that in fiction we cannot ever be sure whether or not a statement is asserted; but non-fiction need not differ in this respect. Our needing to know something about the work or author in order to recognize the relevant speech act does not differentiate the fiction case from any other situation of interpretation. It would be wrong to think that we are supposed to believe, or even take seriously, the proposition expressed by every declarative sentence used in a work of non-fiction, simply because it is a work of non-fiction.44 We might need the background information correctly to interpret irony or hyperbole, for example. Leaving aside the common use of non-literal language, many works of non-fiction contain story elements that are simply made up. These include such controversial cases as Edmund Morris’s Dutch and John Berendt’s Mid- night, mentioned in the introduction. Despite the debate, Dutch is still the only authorized biography of Reagan. Berendt’s methods have been subject to more controversy, but even so Midnight spent a record-breaking four years on the New York Times non-fiction bestseller list. Of course, we should not accept the classifications of bookstores and bestseller lists as author- itative. Dutch and Midnight are borderline cases, and thus should not be taken as definitive of non-fiction. There are, however, many other works of non-fiction that prompt mere-make-believe. Discussions of scientific models posit frictionless planes, point particles and other idealized entities that do not exist. Philosophical dialogues in the tradition of Plato, Berkeley, and Hume cause belief only indirectly, by representing the conversations of (fictional) characters. The authors do not make assertions in their own voices, so whatever beliefs we may form are inferred from the mere-make-believe of invented conversations.45 To take a yet clearer case, anyone familiar with the conventions of classical history knows that speeches were invented, constrained by plausibility but otherwise more or less the creative output of the author. Both Thucydides and Tacitus, for instance, included invented speeches in their works. A practical reason for this convention is that although one could discover the main points made by a speaker, transcripts or notes from the original speeches were unlikely to be available. There were also stylistic considerations, including ‘the ancient writer’s conception of the unity of style required for a literary work’ and the consensus that frequent ‘use of reported speech ... makes for tedious listening; and history should not, if possible, be tedious’.46 Similar conventions attached to battle descriptions, which were largely made up. Tacitus’s Annals and Histories are replete with vivid battles and strikingly eloquent speeches, the contents of which readers are not supposed to believe. Of course, sometimes the reader is supposed to believe that something like this was said, or that something like that happened; but once we get to such

June 21, 2008 11:1 MAC/NWA Page-160 9780230_220478_09_cha08 Stacie Friend 161 a general level it is hard to see the distinction from historical fictions that are faithful to the general facts while inventing only the specifics. Significantly, Tacitus does not restrict himself to making up what people said and did; he also dramatizes their thinking, both through inner speech and indirect discourse. Here is a typical passage from The Annals (i.69), relating the Emperor Tiberius’s train of thought after witnessing the heroism of Agrippina, wife of Tiberius’s nephew Germanicus:

This made a deep impression on the mind of Tiberius. ‘Such zeal,’ he thought, ‘could not be guiltless; it was not against a foreign foe that she was thus courting the soldiers. Generals had nothing left them when a woman went among the companies, attended the standards, ventured on bribery, as though it showed but slight ambition to parade her son in a common soldier’s uniform, and wish him to be called Caesar Caligula. Agrippina had now more power with the armies than officers, than generals. A woman had quelled a mutiny which the sovereign’s name could not check.’47

Such passages are noteworthy given that ‘inside views’ of people’s minds from the third person are often considered to be a clear, and even defining, mark of fictionality.48 As far as I know there is no controversy over the classification of Thucy- dides’ History of the Peloponnesian Wars or Tacitus’s Annals and Histories.These are not merely historical examples of non-fiction; they remain important works of history today. Yet they prompt mere-make-believe because some elements of the story are made up. So mere-make-believe cannot be specific to fiction. Even if we accept the authorial intention to invite mere-make-believe as a necessary condition for fictionality, it cannot be sufficient.

V. Defining fiction

The considerations of the previous section pose the most acute problem for Davies’s and Currie’s proposals for classifying works, since they seek to offer necessary and sufficient conditions in terms of mere-make-believe. One way to avoid the problem is to focus on fictionality as a property of individual statements or phrases. I discuss this approach below, after considering Lamarque and Olsen’s more complex theory of fiction. Lamarque and Olsen agree with Davies and Currie that fictionality requires a fictive utterance, characterized by a Gricean intention that the audience engage in make-believe, as well as content that is determined by the author rather than by the way events actually transpired. To this extent, their theory is subject to the same criticisms as Davies’s and Currie’s. Yet Lamarque and Olsen also say that the author’s expectation that audiences will make believe what they read, and the audience’s compliance with that expectation, hinge

June 21, 2008 11:1 MAC/NWA Page-161 9780230_220478_09_cha08 162 Imagining Fact and Fiction at least in part on mutual awareness of the storytelling conventions that define the social practice of fiction. It is in virtue of this awareness that audiences adopt the ‘fictive stance’, which involves the ‘disengagement from certain standard speech act commitments, blocking [many] inferences from a fictive utterance back to the speaker or writer, in particular inferences about beliefs’.49 Their acknowledgement of the role of conventions suggests that Lamar- que and Olsen’s proposal might be more sensitive to variations in fictional and non-fictional practices, such as those exemplified by the conventions of classical history; yet their characterization of the fictive stance remains static. This is not because they fail to recognize that the practice of fiction evolves; they acknowledge that there are many different sub-practices reflecting different genres and the various purposes that fiction may serve, and that the line between fiction and non-fiction changes over time. Nonetheless, their ‘enquiry is synchronic not diachronic’ because they ‘are concerned to describe the practice as it now exists, in paradigmatic forms’.50 As a result, they count as ‘transitional’ in the development of our current paradigm those ‘troublesome border cases—saints’ lives, chronicles, sagas, (early) histories of the world—which involve ...deliberate imaginative embellishment, of a kind we now associate with fiction, though without the attitudes of make-believe characteristic of a more recognizable practice of fiction’.51 It is certainly true that it can be valuable to understand the practices of writing, reading, and criticizing genres of fiction and non-fiction as they now exist. But interpreting past practices from our present perspective fails to do justice to the evolution of storytelling conventions that Lamarque and Olsen themselves acknowledge. There is no reason to think that genres of the past which exhibit some but not all of the features we typically associate with fiction constitute a transition to our current practice, as if that practice were the epitome or goal of ‘fictionality’ (compare the assumption that Egyptian art was a primitive, failed attempt at perspective). Even if we accepted that works produced by a ‘fictive utterance’ and which invite the ‘fictive stance’ constitute paradigmatic fictions for us, nothing follows about fictionality in general. A theory that has nothing to say about why Tacitus counts as non- fiction despite inviting mere-make-believe is at best incomplete. Lamarque and Olsen could reply that though conventions of accuracy in non-fiction have changed over time, the core of fiction—to invite imagining of made- up content—has remained the same.52 But if the conventions of non-fiction shift, so too must the scope of fiction. We would require further argument to think that the conventions central to our own practice should constitute the relevant core. Furthermore, because the conventions of fiction and non-fiction continue to change, it is unclear what exactly constitutes the current paradigm. If we count Dutch, Midnight, and many other contemporary works as non- fiction that trace their conventions back to the New Journalism of the 1960s,

June 21, 2008 11:1 MAC/NWA Page-162 9780230_220478_09_cha08 Stacie Friend 163 then the invitation to imagine content that is made up is not exclusive to fiction even now. Of course, Lamarque and Olsen are interested in paradigm cases, and we are still inclined to treat these works as peripheral. We do not, however, treat Thucydides or Tacitus as peripheral. It is part of our current practice to classify works from the past as fiction or non-fiction, and in so doing we do not base our judgements exclusively on current conventions. Substantial chunks of Tacitus’s Annals are products of ‘fictive utterance’ and invite the ‘fictive stance’, at least as these terms are understood by Lamarque and Olsen; parts of historical novels are asserted and invite belief. If we define the paradigm in terms of mere-make-believe, we ignore the complexities of actual cases. Perhaps, though, actual cases are too complex to be amenable to philosophical theory. If the works we classify as fiction and non-fiction display a mixture of different features, perhaps we would do better to concentrate on those features than on the vagaries of bookstore classifications. Currie is most explicit in taking this route. He argues that a philosophical theory of fiction should restrict itself to an analysis of fictional statements, rather than fictional works.53 Currie dismisses the question of how many statements must be fictional for the work to be fictional:

If we wanted to, we could define a numerical degree of fictionality, but it would be artificial and unilluminating. What is illuminating is a precise account of the fictionality of statements. For in some perhaps irremedi- ably vague way, the fictionality of works is going to depend upon the fictionality of the statements they contain. As long as we are clear about what water molecules are, it hardly matters for purposes of definition that most things we call ‘water’ actually contain much else besides.54

The analogy seems to turn on the fact that there is nothing general to say about the proportion of H2O molecules required for a given stuff to be called

‘water’ in unscientific settings. While possession of H2O molecules is probably necessary for a substance to count as water in our ordinary sense, it is not sufficient; a cup of tea may contain a greater proportion of H2O molecules than a bathtub of dirty water, but we count the latter and not the former as ‘water’. Similarly, while containing made-up elements is plausibly necessary for a work to count as fiction, it is not sufficient; many works of non-fiction also contain such elements, sometimes more than in a work of fiction. Cur- rie’s point is that just as scientists should focus on the chemical composition of molecules rather than worry about how the ‘folk’ classify tea, philosophers should focus on the statements that invite mere-make-believe rather than worry about how the folk classify texts. Yet the analogy is questionable. The reason for the scientists’ restricted focus is that, unlike tea or the liquid that comes out of bathtub taps, H2Ois

June 21, 2008 11:1 MAC/NWA Page-163 9780230_220478_09_cha08 164 Imagining Fact and Fiction a natural kind. There are scientific reasons to identify water with a particular molecular structure, and these are not motivated by worries about how to classify liquids in cups and baths. What is the parallel justification for identifying a ‘molecular essence’ of fiction, other than the role of fictional statements in somehow determining the classification of works? Why should we assume that fiction has such an essence, let alone that it consists in statements that are made up or invite mere-make-believe? Currie claims that the fictionality of a work depends on the fictionality of its parts. But a plausible alternative view is that the fictionality of works has explanatory priority over the fictionality of statements: in short, that a statement, whether made up or asserted, counts as ‘fictional’ because it is contained in a work of fiction, that it invites a certain sort of response or set of responses because of its context rather than its intrinsic features. Identifying what is ‘made up’ with what is ‘fictional’ begs the question against this alternative. Our concern with classification is first and foremost a concern with works. Thus, while we should not expect philosophers to resolve debates over the classification of such books as Dutch and Midnight, we should expect them to shed light on what is at issue. If we take Dutch to be non-fiction, what does this tell us about how to read it or criticize it? If we take it as fiction, what should change in our engagement with the text? Focus on individual statements is unlikely to assist us in answering these questions. After all, there is no controversy over which parts of Dutch are supposed to be believed and which merely imagined; it is precisely because some parts are to be believed, and others not, that controversy over this text has arisen. None of this is to deny the importance of determining which statements in a work are to be believed and which merely-made-believe. If we wish to avoid forming false beliefs, we need to know that Morris invented characters, just as we need to know that Tacitus invented speeches. But picking out the asserted and unasserted claims in a text is distinct from knowing how to approach the text as a whole. There seems to be a difference between reading or appreciating a work as fiction and reading or appreciating it as non-fiction.IfwereadDutch as non- fiction, for instance, Morris’s device of a fictionalized narrator will strike us as ground-breaking, noteworthy, provocative. If we read it as fiction, the device will not have the same effect; we might wonder why Morris chose to narrate from a first-person rather than a third-person point of view, but we are unlikely to find the technique of special interest. A good explanation of this difference is that we treat fictional narrators as standard for works in the category of fiction, but as contra-standard for works in the category of non-fiction biography, in the sense of those terms familiar from Walton’s ‘Categories of Art’.55 Since the identification of standard, contra-standard, and variable features makes sense only with respect to whole works that possess a variety of features, the categorization of a text as fiction or non- fiction should make a difference to how we are supposed to read or appreciate

June 21, 2008 11:1 MAC/NWA Page-164 9780230_220478_09_cha08 Stacie Friend 165 it. Prompting mere-make-believe is a standard feature of works of fiction; prompting belief is a standard feature of non-fiction. Yet given the variations in practices and conventions, we should not take these standard features as definitive.

VI. Fiction and imagination

Does this mean that there is no special connection between fiction and imagination, that there is nothing distinctive about fiction in this respect? That would be a rather counterintuitive conclusion, and it is not the one that I wish to draw. I suggest that theories of fiction have located the connection in the wrong place: instead of focusing on works or individual statements, we should look at the practices that underpin our interaction with various texts. What is right about Lamarque and Olsen’s approach is that the works we classify as fiction and non-fiction are part of practices of writing, reading, and criticizing texts, guided by various genre conventions. The mistake is to assume that there is a single, unified practice of fiction defined by the features of the works within it. A more promising approach is to say that we count certain works as fiction because they are embedded in the relevant practices—practices associated with particular genres —just as the statements that count as ‘fictional’ are those embedded in works of fiction. What constitutes ‘embedding’ in a practice? Our classifications are probably too complex to answer this question straightforwardly, though the sorts of factors Wal- ton identifies in ‘Categories of Art’—including features of the text as well as facts about the origins of the work—would surely be relevant.56 For instance, I take for granted that the author’s intention is a significant factor in classification. It is important to recognize, however, that the relevant intention is not the intention to invite belief or imagining, but rather to produce a work in a certain category. The fact that Tacitus intends readers to engage in mere-make-believe is not in conflict with his intention to write non-fiction history, and it is the latter, not the former, that matters for classification.57 A different question is what makes a given practice a practice of fiction. Again I doubt that we will find a general answer; but this question highlights a different way to construe the relationship between fiction and imagination. I am attracted to the idea that the existence of those practices we associate with fiction can be explained at least partly by the purpose of allowing authors to use their creative imaginations to engage audiences in mere-make- believe, even though authors exploiting the conventions of various genres also invite other responses. Anthony Savile has argued persuasively that a developed practice of literature requires traffic in invented characters and situations.58 By contrast, the practices associated with non-fiction genres are typically justified by other purposes, even if specific genre conventions permit or require authors to make things up. Of course, authors of both fiction and non-fiction write for a variety of different reasons: to entertain,

June 21, 2008 11:1 MAC/NWA Page-165 9780230_220478_09_cha08 166 Imagining Fact and Fiction to instruct, to persuade, and so on. Those who choose to write fiction rather than non-fiction reject at least some conventions of accuracy in achieving their goals. Because these conventions change over time, what matters to the decision to write fiction or non-fiction also changes. The invitation to merely-make-believe Tiberius’s thoughts was acceptable for non-fiction in Tacitus’s time; today an author who wanted to imagine the contents of a person’s mind is likely to be writing fiction. A definition of fiction might seem to be lurking behind these comments. Should we say that a work of fiction is one that breaks conventions of contemporary non-fiction practice, specifically those that prohibit making certain things up? This is the wrong conclusion to draw. Conventions change over time, and one catalyst for change is that writers break the rules. There should be a difference between an author’s challenging the boundaries of non-fiction and his writing fiction (and vice versa). When Morris puts forward Dutch as a work of non-fiction biography, he is saying that the text should be interpreted and criticized with the standards of non-fiction biography in mind. It is from that perspective that his narrative counts as surprising and inno- vative. Of course, classification is complicated, and there are a number of considerations other than intention. The important point is that however it is determined, categorizing a work as fiction or non-fiction should give us some idea of which genre conventions, which standards, we are supposed to apply in interpretation and criticism. I suggest that we give up the quest for necessary and sufficient conditions for fictionality, and focus instead on examining these features of the practices of fiction and non-fiction.

Acknowledgements

I would like to thank the audiences at the Conference on the Imagination and Thought Experiments (Bristol, May 2006), the British Society of Aesthetics Annual Meeting (Oxford, September 2006), and the Aesthetics Forum of the Institute of Philosophy (London, November 2006), for helpful discussion of previous versions of this chapter. I would also like to thank Katherine Thomson-Jones for insightful comments on a previous draft.

Notes

1. Currie, Nature of Fiction,1. 2. I use ‘work’ and ‘text’ interchangeably, as well as ‘reader’ and ‘audience’. In this paper I focus exclusively on verbal fictions. 3. Walton, Mimesis; Currie, Nature of Fiction; Davies, ‘Fictional Truth’; Davies, ‘Fiction’; Lamarque and Olsen, Truth, Fiction, and Literature. 4. Theorists who invoke make-believe also typically criticize the view that authors of fiction engage in pretend speech acts, but for my purposes there is little difference between the pretence and make-believe theories since both take fiction to invite imagining rather than belief. For the pretence view, see Searle, ‘Logical Status’.

June 21, 2008 11:1 MAC/NWA Page-166 9780230_220478_09_cha08 Stacie Friend 167

5. Of course, belief is not the only attitude other than imagining invited by texts. The debate is typically framed in terms of the contrast between believing and making believe, but the issues arise equally for any other ‘serious’ attitude we might contrast with imagining. The focus on belief should be treated as a convenient simplification. 6. Davies, ‘Fiction’, 265. 7. See, for example, Lamarque and Olsen, Truth, Fiction, and Literature, 47. 8. Walton, Mimesis, 70. 9. Ibid., 91. 10. Currie, Nature of Fiction, 36; Lamarque and Olsen, Truth, Fiction, and Literature, 48; Davies, ‘Fiction’, 264. 11. Currie, Nature of Fiction, 31. 12. Walton, Mimesis, 70. 13. Ibid., 74. 14. Walton’s theory of pictorial representation entails that all pictures ‘are fiction by definition’ (Mimesis, 351). His analysis of metaphors also involves recourse to make-believe, in ‘Metaphor’. 15. Walton, Mimesis, 70. 16. Arguably, visual fictions such as films do not invite (much) mental imagery. As Katherine Thomson-Jones has pointed out to me, this is another reason to deny that imagery is definitive of fiction. 17. Schlosser, Fast Food Nation, 170. 18. Applebaum, Gulag, 319. 19. Thanks to Chad Grine for emphasizing the contrast with memory images. 20. In conversation. 21. For example, Lewis, ‘Truth in Fiction’. 22. See Gerrig, Experiencing Narrative Worlds; Emmott, Narrative Comprehension; Goldman et al., Narrative Comprehension. 23. Gerrig, Experiencing Narrative Worlds,6. 24. See Lewis, ‘Truth in Fiction’; Walton, Mimesis, Chapter 4. 25. Skolnick and Bloom, ‘Intuitive Cosmology’. 26. Nichols, ‘Just the Imagination’, 460. The experiment is reported in Leslie, ‘Pretending and Believing’. 27. Nichols proposes a distinction in terms of the relation to desires about real and imaginary situations, but I will not pursue this idea here. 28. Potts et al., ‘Incorporating New Information’; Gerrig and Prentice, ‘Representa- tion of Fictional Information’; Prentice et al., ‘What Readers Bring’; Wheeler et al., ‘Fictional Narratives’. 29. Of course, we could deny that an author’s purpose or aim in writing a work is the same as the work’s purpose or function. But it will not do to say that a work’s purpose is to invite make-believe because it is fiction, if the goal is to define fiction in terms of make-believe. Moreover, the denial is incompatible with those theories of fiction that invoke the author’s intention to invite make-believe. (Thanks to Katherine Thomson-Jones for pressing me on this issue.) 30. See Walton, Mimesis, 78–79, on this difference. 31. Sinclair, The Jungle, 162. 32. The relevant contrast is between what is explicit and what is implied. 33. See n. 5. 34. The qualification, designed to rule out cases of deception, should be assumed in what follows.

June 21, 2008 11:1 MAC/NWA Page-167 9780230_220478_09_cha08 168 Imagining Fact and Fiction

35. Walton, Mimesis, 79. Currie claims that his theory is also consistent with an entirely asserted fiction (Nature of Fiction, 35). Yet this claim appears to be in some conflict with his view, discussed in the next paragraph, that fiction can be no more than ‘accidentally true’. If an assertion is true, this is unlikely to be by accident. I think we would count a work made up of statements either false or merely accidentally true as the product of deception or ignorance, not as a work of fiction. 36. Currie, Nature of Fiction, 46. 37. Though Davies does not use the term ‘necessary and sufficient conditions’, he does claim that works failing to meet his criteria of fictionality would fail to be fiction. See ‘Fiction’, 266. 38. Davies, ‘Fictional Truth’, 52. 39. Davies, ‘Fiction’, 266. 40. Lamarque and Olsen, Truth, Fiction, and Literature, 51 (italics removed). 41. Currie, Nature of Fiction, 47 (italics removed). 42. Fleming, Thunderball, 116. 43. This is not to deny that we might treat assertions in fiction differently from assertions in non-fiction. 44. Sentences that are not used but only mentioned are excluded from consideration here. 45. If we agree that philosophical dialogues are non-fiction, then they seem to constitute cases of non-fiction that provoke only mere-make-believe in the propositions expressed by the sentences in the text. 46. Miller, ‘Dramatic Speech’, 46. 47. Tacitus, Annals, 39–40. 48. See, for example, Cohn, ‘Fictional versus Historical Lives’. 49. Lamarque and Olsen, Truth, Fiction, and Literature, 46. The bracketed qualifier derives from a clarification on p. 44. 50. Ibid., 38. 51. Ibid. 52. A similar reply was suggested by Peter Lamarque in response to a presentation of a previous version of this paper. 53. Neither Davies nor Lamarque and Olsen consider this alternative explicitly. In any case they seem clearly to be discussing the classification of works. 54. Currie, Nature of Fiction, 49; Searle (‘Logical Status’) similarly claims that works of fiction are made up of a mix of genuine and pretend speech acts. 55. Walton, ‘Categories of Art’. 56. In ‘Categories of Art’, Walton considers only perceptual art; but with refinements the basic argument should apply equally well to literature. 57. This conflict would seem to arise on any theory that takes an intention that audiences engage in (mere-)make-believe to be definitive of fiction. 58. Savile, ‘Imagination’.

Bibliography

Applebaum, A. 2003. Gulag: A History, New York: Doubleday. Cohn, D. 1989. ‘Fictional versus Historical Lives: Borderlines and Borderline Cases’, Journal of Narrative Technique, 19: 3–24. Currie, G. 1990. TheNatureofFiction, New York: Cambridge University Press.

June 21, 2008 11:1 MAC/NWA Page-168 9780230_220478_09_cha08 Stacie Friend 169

Davies, D. 1996. ‘Fictional Truth and Fictional Authors’, British Journal of Aesthetics, 36: 43–55. Davies, D. 2001. ‘Fiction’, in B. Gaut and D.McIver Lopes (eds), The Routledge Companion to Aesthetics, London: Routledge, 263–274. Emmott, C. 1997. Narrative Comprehension: A Discourse Perspective, Oxford: Oxford University Press. Fleming, I. 1963. Thunderball, London: Pan Books. Gerrig, R. 1993. Experiencing Narrative Worlds: On the Psychological Activities of Reading, New Haven: Westview Press. Gerrig, R. and Prentice, D. 1991. ‘The Representation of Fictional Information’, Psychological Science, 2: 336–340. Goldman, S. R., Graesser, A.C., and van den Broek, P. (eds) 1999. Narrative Comprehen- sion, Causality, and Coherence, Mahwah, NJ: Lawrence Erlbaum. Lamarque, P. and Olsen, S. 1994. Truth, Fiction, and Literature, Oxford: Clarendon Press. Leslie, A. 1994. ‘Pretending and Believing: Issues in the Theory of ToMM’, Cognition, 50: 211–238. Lewis, D. 1983. ‘Truth in Fiction’, Philosophical Papers, Volume 1, New York: Oxford University Press, 261–280. Miller, N. P. 1975. ‘Dramatic Speech in the Roman Historians’, Greece & Rome,Second Series 22: 46. Nichols, S. 2006. ‘Just the Imagination: Why Imagining Doesn’t Behave like Believing’, Mind and Language, 21: 459–474. Potts, G., St. John, M., and Kirson, D. 1989. ‘Incorporating New Information into Existing World Knowledge’, Cognitive Psychology, 21: 303–333. Prentice, D., Gerrig, R., and Bailis, D. 1997. ‘What Readers Bring to the Processing of Fictional Texts’, Psychonomic Bulletin and Review, 4: 416–420. Savile, A. 1998. ‘Imagination and the Content of Fiction’, British Journal of Aesthetics, 38: 136–149. Schlosser, E. 2002. Fast Food Nation: The Dark Side of the All-American Meal, New York: Perennial. Searle, J. 1975. ‘The Logical Status of Fictional Discourse’, New Literary History,6: 319–332. Sinclair, U. 1985. The Jungle, New York: Penguin. Skolnick, D. and Bloom, P. 2006. ‘The Intuitive Cosmology of Fictional Worlds’ in S. Nichols (ed.), The Architecture of the Imagination: New Essays on Pretence, Possibility, and Fiction, Oxford: Clarendon Press, 73–86. Tacitus. 2003. The Annals and the Histories, A.J. Church and W.J. Brodribb (trans.), Hadas, M. (ed.), New York: Modern Library. Walton, K. 1970. ‘Categories of Art’, The Philosophical Review, 79: 334–367. Walton, K. 1990. Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Walton, K. 1993. ‘Metaphor and Prop Oriented Make-Believe’, European Journal of Philosophy, 1: 39–56. Wheeler, S.C., Green, M., and Brock, T. 1999. ‘Fictional Narratives Change Beliefs’, Psychonomic Bulletin & Review, 6: 136–141.

June 21, 2008 11:1 MAC/NWA Page-169 9780230_220478_09_cha08 9 Three Debates in Meta-Aesthetics Elisabeth Schellekens

Few philosophical debates seem to allow for as little theoretical disparity as that on the subject of Realism or Anti-Realism. That the two antithetical positions uphold the broad structure of a dichotomy may come as no surprise: the question under scrutiny is, after all, one about whether the world and its contents are autonomous of our minds, or whether the world and its contents simply cannot be said to exist independently of our perception and understanding of them. There does not, in other words, seem to be much leeway between the two stances, at least partly because what they capture is a deeply entrenched conceptual divide over what does and does not exist.1 How, one may ask, could some thing exist but a little? In the philosophy of value, Realism has traditionally stood not only for the strictly metaphysical view that value qualities exist independently of our minds, but also more broadly for a philosophical programme centred on the object to which value is ascribed and the correctness of value judgements. Correspondingly, Anti-Realism has represented the denial of the Realist programme. In addition, however, two debates—in important respects derivative of the wider Realism/Anti-Realism debates—have emerged and acquired something of a life of their own. First, the Objectivism/Subjectivism debate specifically concerns whether qualities are to be ascribed to the object or the subject of experience, and so whether it is the object or the subject that makes some thing appear as it does. Second, the Cognitivism/Non-Cognitivism debate concerns whether our judgements about an object’s qualities can yield knowledge and be assessed according to some generally available normative standards. This second debate takes two distinct (albeit intimately related) expressions. On the one hand, it concerns the possibility of correctness or incorrectness for a certain kind of judgement, while, on the other, it concerns whether we can gain knowledge from a certain kind of experience and judgement or not. How are we to understand the relations between the Realism/Anti- Realism debate, the Objectivism/Subjectivism debate, and the Cognitivism/ Non-Cognitivism debate? In this chapter I will argue that despite their close

170

June 21, 2008 11:5 MAC/NWA Page-170 9780230_220478_10_cha09 Elisabeth Schellekens 171 and numerous connections, it is important to conceive of these debates as distinct. For although there is a sense in which the broader, metaphysical understanding of the Realism/Anti-Realism (hereafter R/AR) debate encapsu- lates the other two, there is also a strong sense in which the difference in focus between all three debates is substantial enough to merit separate consideration. Thus the debate between Objectivism and Subjectivism (hereafter O/S) can be seen to address distinctively ontological issues directly related to the origin of our experience of aesthetic and moral value, whereas the question of Cognitivism versus Non-Cognitivism (hereafter C/NC) can be seen to address epistemological issues about the judgements we make concerning that experience. Indeed, part of my purpose in this chapter is to show why maintaining a properly conceptual distinction between the R/AR, O/S, and C/NC debates is crucial to any progress in the wider sphere of the philosophy of value. My aim in this is not to demonstrate the irrelevance of the metaphysics of aesthetic and moral value. Rather, it is to show that a more constructive and fruitful engagement with these philosophical issues can be had by taking our lead from concerns that fall more properly under the domain of Objectivism and Subjectivism, and Cognitivism and Non-Cognitivism. My intention is to weaken the philosophical stranglehold that the question of the metaphysical reality or otherwise of aesthetic and moral qualities has over value enquiry in general. In fact, I shall claim not only that it is largely due to discussions about Objectivism, Subjectivism, Cognitivism, and Non- Cognitivism that progress has been made in the general area of enquiry, but also that these developments have led to a partial reconciliation between the original, strictly opposed positions in the R/AR debate. It seems, in other words, that there is more complexity to the distinction between Realism and Anti-Realism than the disjunctive structure we began by sketching would appear to allow for.

Few philosophers would disagree with the claim that with regards to aesthetic and moral value, the burden of philosophical explanation has tended to rest more on the Realist than on the Anti-Realist. That is to say, built into much of the debate over the last 50 years or so seems to have been the idea that the most reasonable place to begin philosophical examination is to proceed from the view that, at bottom, aesthetic and moral qualities do not really exist. For the Realist to gain ground, very convincing philosophical cases would need to be made for the existence of qualities such as beauty, justice, cruelty, and ugliness—qualities which on a daily basis we nonetheless all refer to, think in terms of, and to all intents and purposes treat as if they exist.2 For the orthodoxy to prevail, however, the Anti-Realist has simply had to weather the Realist’s occasional thrust for change.

June 21, 2008 11:5 MAC/NWA Page-171 9780230_220478_10_cha09 172 Three Debates in Meta-Aesthetics

In spite of reflecting the very opposite of many of the common-sense assumptions implicit in our aesthetic and moral practices, Anti-Realism’s seeming advantage may nonetheless be well founded. After all, the Anti- Realist does seem to have the upper hand in dealing with the notoriously difficult questions famously expressed by John L. Mackie in his Ethics: Invent- ing Right and Wrong,3 concerning the regular and widespread disagreements about value on the one hand, and what may be described as the peculiar ontology of evaluative qualities4 on the other. For we often do have divergent opinions on matters related to aesthetic and moral value, and the task of pinning down the exact kind of existence that aesthetic and moral qualities enjoy is a challenge, to say the least. At best, and as we shall soon see in greater detail, the imbalance in favour of Anti-Realism has its roots in such difficulties. This being the case, the challenge Realists face is a daunting one. Nevertheless, in the case of the more specialised concerns of the O/S and C/NC debates, it seems that some progress towards a less intractable position has been achieved. Two developments in particular, common to both debates, stand out. First, there is the widely accepted response-dependence of (at least most) aesthetic and moral qualities. Following the lead of work on the ontological—and often dispositional—status of secondary qualities, it is generally held that qualities such as elegance, harmony, and callousness require some form of human response in order to be genuinely instantiated or manifested.5 To be beautiful or cruel, say, is thus to be such as to give rise to certain emotional responses in normal perceivers under normal circumstances. Second, there is the idea, taken more and more seriously, that there can be a standard of normative assessment valid for a community of judgement- makers even in cases where the subject-matter of those judgements is not capable of satisfying all the criteria for truth-assertability strictly conceived. No doubt, encouraged by theories of emotion that allow for the notion of appropriateness to serve as standard of correctness for emotional responses, it seems to have become less and less controversial to hold that targeted assess- ments or aesthetic and moral quality-ascriptions can allow for some form of accuracy.6 By emphasising their cognitive aspect, emotional responses may be understood to pick out qualities and characteristics that are not entirely agent-relative but available to general scrutiny, and so in that sense suitable, apt, or fitting. In the light of these developments in the O/S and C/NC debates, several philosophers of value have tried to forge an alternative approach, capable of unlocking the disjunctive structure of the traditional R/AR debate. This approach recognises that there is truth to the claim that, say, the graceful or the cruel appears graceful or cruel because (there is a sense in which) it is graceful or cruel. However, the approach also holds on to elements of the opposing claim that what it is for something to be graceful or cruel consists, fundamentally, in its appearing graceful or cruel to us.7

June 21, 2008 11:5 MAC/NWA Page-172 9780230_220478_10_cha09 Elisabeth Schellekens 173

As we shall see, the development of this alternative approach has, perhaps rather disconcertingly, led to two related but distinct interpretations of the terms ‘Realism’ and ‘Anti-Realism’. It will be important to distinguish between the narrower—strictly metaphysical—usage of the terms, and the broader, more general, usage which also covers, on the Realist side, Objec- tivism and Cognitivism, and, on the Anti-Realist side, Subjectivism and Non-Cognitivism. On the second usage, ‘Realism’ and ‘Anti-Realism’ refer to two general kinds of approach to questions in the philosophy of aesthetic and moral value rather than to specific theories about the metaphysical status of aesthetic and moral qualities.8 It is beyond the scope of this chapter to examine each and every contribution to these debates. Nor is it my intention to develop a comprehensive account capable of overcoming the main difficulties they bring to our attention. Rather, through an examination of some of the more interesting current accounts in Meta-Aesthetics, I aim to show that there is currently a lot more common ground between Realist and Anti-Realist positions than one might assume. As already suggested, it will be my contention that the two sides of the traditional R/AR debate have moved closer together in virtue of the progress made in the O/S and C/NC debates. What is more, I will suggest that at present, the more substantial philosophical discussions are actually to be found at the level of these two, originally derivative, debates. Whilst it is certainly true that Objectivism and Cognitivism are natural and comfortable philosophical partners, and to a considerable extent do have overlapping aspirations and methods, the two kinds of theories do nonetheless have different ambitions and pursue somewhat divergent goals. I shall begin by briefly outlining the principal tenets of the most compelling Objectivist and Cognitivist theories that have recently been developed. I will then examine two accounts that proclaim to belong firmly within the Anti- Realist camp. My analysis will show that these accounts do, to all intents and purposes, share their principal tenets with the Objectivist and Cognitivist theories under consideration.

Recent attempts to argue for Aesthetic Objectivism have tended to take one of two main forms. The first9 is rooted in an analogy with secondary qualities, and may be summarised in the following terms:

(O1) Aesthetic qualities can be said to be objective because ontologically they are very similar to secondary qualities such as colour properties (which are objective).

June 21, 2008 11:5 MAC/NWA Page-173 9780230_220478_10_cha09 174 Three Debates in Meta-Aesthetics

For this kind of Objectivist account, the analogy is grounded in the inher- ently perceptual character of both kinds of qualities. Roughly speaking, the argument has it that since we seem uncontroversially to adhere to Objectivism about colour properties, and aesthetic qualities are like colours in several significant ontological respects, there is no need to be wary of upholding Objectivism for aesthetic qualities too. The second form of the argument for Objectivism not only takes the analogy between aesthetic qualities and secondary qualities seriously, but also emphasises the non-inferential character of the way in which we make aesthetic judgements.10 Genuine aesthetic judgements, it is argued, must be based on direct first-hand perceptual experience; ascribing aesthetic features merely by drawing inferences from the presence of certain other (non-aesthetic) features cannot result in genuine aesthetic judgements. The theory’s central claim may be formulated thus

(O2) Aesthetic qualities can be said to be objective because ontologically they are very similar to colour properties, and even though there might be elements of aesthetic qualities the experience of which are individual to each perceiver, the phenomenal impression afforded by them is available to and sufficiently similar for all perceivers.

Strengthening the claim to objectivity here is the descriptive element of the phenomenal impression made by aesthetic qualities, defined as perceptual properties. The main thought here, roughly, is that a sufficiently similar experience of a certain aesthetic quality is accessible to anyone who perceives that quality first-hand.11 Cognitivist theories in Meta-Aesthetics divide into two broad kinds of account, which we may satisfactorily characterise in terms of Justification- Cognitivism (C) on the one hand and Knowledge-Cognitivism (C) on the other.12 As this terminology suggests, the former kind of Cognitivism is mainly concerned with the task of justifying aesthetic judgements, thereby endowing them with epistemic authority, whereas the latter kind is mainly concerned with the possibility of aesthetic judgements having significant cognitive content. Of the Cognitivist strategies centred on the notion of justification, the one advanced by the broader Realist theory which also promotes O2 has earned considerable attention.13 On this theory there can be a normative standard of assessment for aesthetic judgements, in the sense of their being grounded in a shareable aesthetic experience:

(C1) Aesthetic judgements can be (in)correct (in virtue of the generally available and shareable phenomenal impression afforded by aesthetic qualities).

June 21, 2008 11:5 MAC/NWA Page-174 9780230_220478_10_cha09 Elisabeth Schellekens 175

The possibility of correctness for aesthetic judgements thus derives from the objective character of the phenomenal impression afforded by aesthetic qualities, grounded in the descriptive element of that impression. A second kind of Justification-Cognitivism places less emphasis on the aesthetic experience and more on those features that seem to be responsible for the manifestation of the aesthetic property.14 To express the matter more formally:

(C2) Aesthetic judgements can be (in)correct (in virtue of the generally available non-aesthetic features on which aesthetic judgements should be based and which are appealed to in the process of justifying those judgements).

This kind of theory emphasises those features of the object of aesthetic contemplation upon which the aesthetic property is said to depend ontologically. Like O1, O2, and C1, C2 adheres to the non-inferential schema of the making of aesthetic judgements and emphasises the philosophical importance of first-hand perceptual experience. Central to both C1andC2, then, is a commitment to assertoric form for aesthetic judgements. This is because of the notion of correctness that is held to represent a rational and significant standard of normative assessment applicable to aesthetic judgements, a correctness which is grounded in features that are in principle generally available and shareable. Underlying Knowledge-Cognitivism in Aesthetics is the claim, not only that aesthetic experience (of art in particular) can yield knowledge, but that that knowledge contributes to the overall value of art. The question then arises as to what kind of knowledge this might be. In response, a multitude of interesting theories have been developed of late.15 In particular, two kinds of account have gained significant attention in recent philosophy.16 According to the first of these accounts, aesthetic experiences of art are said to be capable of yielding knowledge and insight into situations that we may not (yet) have undergone ourselves.17 For example, contemplating an artwork aesthetically may enable us to grasp what it is like to feel parental love or to be denied education because of gender. On this kind of Aesthetic Cognitivism,

(C1) Aesthetic experiences (of art) can yield understanding of what something (i.e. being in a certain situation or being a certain kind of person at a certain time and place) is like.

The claim is not merely that through art we can come to experience things as they will or could be were they to happen to us. Rather, the claim is also that art may help us understand how a certain situation or event could or

June 21, 2008 11:5 MAC/NWA Page-175 9780230_220478_10_cha09 176 Three Debates in Meta-Aesthetics should be interpreted, and enable us to see what is important or meaningful in a particular situation or event.18 On the second prominent account of Knowledge-Cognitivism in Aesthet- ics, emphasis is placed on the way in which aesthetic experience can help us develop our personalities and enhance our emotional lives and sensibilities.19 This view has it that

(C2) Aesthetic experiences (of art) can yield knowledge and understanding of how to do or perceive certain things in the world by educating and developing our emotions, imagination, and perceptions.

In other words, by experiencing art aesthetically we can gain knowledge not merely of the emotions and sensibilities expressed by the art as such, but also of our own and others’ personalities and perceptions. Engaging with art can enrich our lives substantially insofar as it may make us more mature, considerate, and creative human beings. The common aim of the Cognitivist strategies captured by C1and C2 is thus to establish that aesthetic experience can give us insight and understanding above and beyond our particular perception of the aesthetic character of an object. The principal idea underlying both C1andC2isthat the (assertoric) content of aesthetic judgements, in virtue of the knowledge and understanding that may be afforded by the experiences they report, can be important, meaningful, and open to rational scrutiny.

III

Of the many Aesthetic Anti-Realist accounts that have been developed and defended in the last 10–15 years, two in particular break new philosophical ground in the R/AR debate whilst remaining, in principle at least, committed to an Anti-Realist framework. These are Sensible Subjectivism, based on David Hume’s Subjectivism about moral value,20 and Aesthetic Quasi-Realism, mod- elled on Simon Blackburn’s Quasi-Realism in Ethics.21 Myaiminthissection is to outline these two accounts in order to show that despite their alleged loyalty to the Anti-Realist approach, there are, nonetheless, good grounds for describing each theory as sharing the main tenets of Objectivism and Cognitivism, as outlined above. As was suggested in Section I, the occurrence and seeming irreconcilability of contradictory aesthetic opinions is one of Anti-Realism’s main corner- stones. Generally speaking, the concern relates to the regularity with which aesthetic disagreement occurs and the magnitude it often takes even amongst those who share a cultural background and are experienced and educated in matters of aesthetic taste. However, whilst this concern is still seen to cast a shadow on the prospects of the Realist approach, neither Sensible Subjectivism nor Aesthetic Quasi-Realism assume that such disagreement

June 21, 2008 11:5 MAC/NWA Page-176 9780230_220478_10_cha09 Elisabeth Schellekens 177 necessarily reflects a discrepancy in aesthetic taste and response that cannot, at least in some sense, be overcome.22 Let us begin by examining Sensible Subjectivism, and the way in which it may be seen to represent a rather more constructive account of aesthetic disagreement than that of previous theories developed within the framework of Anti-Realism. Like many other Anti-Realist accounts of aesthetic and moral value, Sen- sible Subjectivism has Hume’s theory of value as its starting-point.23 For Hume, aesthetic judgements are based on the emotional responses elicited in the subject by the object of perception. A statement such as ‘O is ϕ’ (where ␸ is a property such as ‘beautiful’) means that O is the kind of thing to arouse in certain subjects a certain sentiment of approbation associated with ϕ. The subjects in question here are ‘ideal critics’24 or well-educated and unbiased perceivers, sharing aesthetic tastes and sensibilities. Sensible Subjectivism develops the Humean claim that O is ϕ if and only if O is such as to arouse the sentiment of approbation, not by holding that value is merely a ‘gilding or staining’ of ‘natural objects with the colours borrowed from internal sentiment’,25 but rather by adopting a so-called ‘genealogical account’.26 The idea underlying the theory is that our primitive responses to objects led us to group together the objects that elicit responses of a particular kind. With time, the condition for belonging to a certain category evolved from the criterion that had, originally, given rise to the primitive response so that we begin to think of a particular response as something that is made appropriate by an object’s having a certain property. Eventually, and if the so-called ‘property, response’ pair takes hold, we reach a point where the appropriateness of finding that some object (O) has a particular property (ϕ) and the appropriateness of having the relevant response (R) are held answerable to each other. This is, then, a ‘subjectivism of subjects and properties mutually adjusted’.27 According to Sensible Subjectivism, the normative standard of assessment for judgements such as ‘O is beautiful’ is based on the way in which last- ing ‘property, response’ pairs are accountable to one another. To be ϕ is to be such as to give rise to R, and, correspondingly, to give rise to R is to be ϕ.(Or,forOtohaveϕ is for O to give rise to R and for O to give rise to R is for O to have ϕ.) What we have, then, is not the standard of correctness described by Hume, which ‘appears more and more implausible the further we move from the gustatory and olfactory cases’,28 but, rather, a normative standard of correctness for judgements such as ‘O is beautiful’ that is both ‘subject-involving and property-involving’.29 As Wiggins explains,

[a]n object’s or person’s or event’s being ϕ...consists in its being such as to evoke in the right way or such as to make appropriate some response, call it A ...where A is ...the response that we owetoitifitreallyisϕ.30

June 21, 2008 11:5 MAC/NWA Page-177 9780230_220478_10_cha09 178 Three Debates in Meta-Aesthetics

Interestingly, Wiggins’s account thus claims to remain committed to Subjectivism—at least insofar as ‘the sort of agreement that is in question’ for this normative standard of correctness ‘is only agreement in susceptibility to respond thus and so to ϕ things’31—despite openly adhering to Cognitivism.32 According to Sensible Subjectivism, our responses are correct ‘when and only when they are occasioned by what has the corresponding property ␸ and are occasioned by it because it is ϕ’. 33 The similarities between Sensible Subjectivism, on the one hand, and C1 and C2, on the other hand, need hardly be rendered more explicitly. Judge- ments of the form ‘O is ϕ’ (where ϕ is an aesthetic quality) uncontroversially allow for correctness and incorrectness on Wiggins’s account, and in this respect at least Sensible Subjectivism is openly committed to a view traditionally associated with the Realist approach. For with regards to aesthetic judgements, Sensible Subjectivism—just like C1andC2—lays claim to assertoric form, normative force, a normative standard of correctness, and the possibility of meaningful justification. What is more, we have already seen that O1 and O2 also draw heavily on the analogy between secondary qualities and evaluative qualities in terms of their response-dependence, and that an aesthetic quality is said to be objective precisely insofar as the ‘impressions or appearances’ it affords are ‘relativized’34 to perceivers who have the aesthetic sensibility with which to view the work correctly. Rather like the Objectivisms examined above, then, according to which what it is for O to have an objective aesthetic quality is for O to give rise to a phenomenal impression in the subject of experience, Sensible Subjectivism holds not only that we ‘perceive an object as red/graceful simply because it is red/graceful’, but also that it is the case that ‘an object is red/graceful in virtue of looking red/graceful to perceivers endowed with the appropriate visual apparatus or aesthetic sensibility’.35 That is to say, we perceive an object as, say, graceful because it is graceful, but what it is for something to be graceful is for that thing to give rise to a certain response in us. What about Aesthetic Quasi-Realism—does it manage to retain its distinctively Anti-Realist character? In Ethics, Quasi-Realism’s principal aim is to retain some of the advantages of Anti-Realism while making room for some of the most attractive features of Realism. In the words of Cain Todd, ‘[t]he ethical quasi-realist argues that by casting our moral attitudes in propositional style, we secure the advantages of thinking and arguing about our commitments as we think and argue about matters of fact’.36 Effectively, what Quasi-Realism sets out to uphold is

the expressive, projective origins of our evaluative discourse while at the same time arguing that our framing of such commitments in assertoric form, using evaluative predicates, and hence talking as if they are genuine judgements, with truth-conditions, is justified.37

June 21, 2008 11:5 MAC/NWA Page-178 9780230_220478_10_cha09 Elisabeth Schellekens 179

Aesthetic Quasi-Realism thus defines aesthetic judgements in terms of ‘projection[s] of our aesthetic sentiments onto the world’ that are nonetheless formulated as genuine assertions because of the ‘normative demands’ inherent to them.38 Interestingly, then, even though the theory is committed to the view that aesthetic agreement and disagreement are located ‘in the attitudes/experiences lying behind aesthetic judgements’, it simultaneously insists that some form ‘of demand for agreement is warranted’.39 The agreement that is called for with regards to aesthetic judgements is thus genuine and can be well grounded. Aesthetic Quasi-Realism is committed to assertoric form for aesthetic judgements even if that commitment is not purely formal. For Quasi-Realists ‘do not demand agreement because we happen to utilise assertoric descriptive discourse; rather, we use such discourse because of the normative nature of the attitudes we are expressing’.40 That is to say, aesthetic judgements have a normative force that must be taken seriously, especially since there is a standard of assessment available to all. This may not be truth of the kind applicable in the natural sciences, but

the quasi-realist sense of ‘truth’ ...[can] be supplied by appealing to a notion such as appropriateness; a notion which allows for the development of normative standards ...and concerning which we could justifiably have developed an assertoric discourse in which to express it.41

Despite there clearly being room for the genuine justification of aesthetic judgements, a Quasi-Realist theory in Aesthetics, as Rob Hopkins has pointed out, may nevertheless seem hard pressed to account for the possibility of resolving aesthetic disagreement. For it has been held that when someone finds himself/herself disagreeing with several others who share a view, a Quasi-Realist cannot genuinely commit to the claim that being in a minority of opinion may be ‘reason enough for [the person in the minority] to adopt the [majority] view’ not only in ‘ordinary empirical matters’, but also with regards to beauty.42 In reply to this challenge, Todd argues that Aesthetic Quasi-Realism is fully capable of overcoming this concern. On his view, ‘[u]pholding the kind of projectivist, expressivist picture supported by quasi-realism requires some sort of conception of reality independent of our attitudes and values and as amenable to the type of description expressed in genuine judgements’. What is required, Todd continues, is ‘a way of demonstrating how aesthetic judgements can be at once descriptive and expressive’. The Aesthetic Quasi- Realist need not ‘as Hopkins thinks the quasi-realist must’ defend the view ‘that the sole role of an aesthetic statement is to express some non-cognitive response’.43 Rather, there is room for a cognitive assessment which justifies the assertoric form and possibility of correctness for aesthetic judgements.

June 21, 2008 11:5 MAC/NWA Page-179 9780230_220478_10_cha09 180 Three Debates in Meta-Aesthetics

There is little doubt, then, that Aesthetic Quasi-Realism thus openly insists on allowing for significant assertoric form, substantial normative force, a genuine normative standard of correctness, and the possibility of meaningful justification. Aesthetic judgements, we are told, warrant a ‘demand for agreement’ and may be described as ‘true’ or ‘false’. This truth may not be verifiable through scientific procedure, but then neither is the notion of truth employed by C1andC2. Be it from the Realist or Anti-Realist perspective, the truth of aesthetic judgements is not held to allow for the kind of proofs we find in mathematics and the natural sciences.44 Instead, it is our actual experience of a particular aesthetic quality or character that serves as evidence for the truth of a certain aesthetic judgement. And this idea, as we have already seen, lies at the heart of both O1 and O2, since on those accounts, aesthetic judgements must be non- inferential and grounded in nothing other than the subject’s own perceptual experience. In a nutshell, Sensible Subjectivism and Aesthetic Quasi-Realism are committed to several tenets traditionally associated with the Realist approach to aesthetic value. Even though supporters of these views claim that their commitment to what we have described as Cognitivist and Objectivist tenets is superficial, it is far from obvious that this proviso is in actual fact available to them. For given the philosophical weight placed on the possibility of assertoric form, normative assent and truth, it seems reasonable to assume that Sensible Subjectivism and Aesthetic Quasi-Realism rely on these notions in more fundamental ways than they are willing to admit.45

What the two previous sections have shown is that whereas Aesthetic Anti- Realists have become more philosophically amenable to the Objectivist and Cognitivist demands implicit in much of our folk psychology about aesthetic value, Aesthetic Realists have, in turn, become more aware of the way in which human responses seem to play important ontological and epistemological roles in aesthetic experience. Having said this, contemplating the results of our investigation may lead an analysis of the matter further still. For on closer inspection it becomes possible to interpret the R/AR rapprochement as also, in one sense, a dissolution of the distinctively Realist and Anti-Realist character of particular theories. That is to say, if Anti-Realism has lost its long-established association with Non- Cognitivism, and the Objectivism that Realists are willing to commit to is one that even their traditional opponents require and rely on, it seems to be high time to revise our conception of the alleged divide in Meta-Aesthetics.

June 21, 2008 11:5 MAC/NWA Page-180 9780230_220478_10_cha09 Elisabeth Schellekens 181

Clearly, Anti-Realism no longer necessarily entails a straightforward Subjec- tivism and Non-Cognitivism, and Realism no longer necessarily implies what may uncontroversially be described as Objectivism. One might be inclined to infer from the above that what is called for is not so much a revised conception as an actual disbanding of the traditional R/AR dichotomy. After all, as we have seen, a brief study of some of the most interesting and successful theories associated with both the Realist and Anti- Realist approaches reveals that the more significant philosophical questions have passed beyond the narrowly metaphysical R/AR debate. This suggests that there is at least one sense in which the labels ‘Realism’ and ‘Anti-Realism’ are redundant. Whatever one’s response to this issue, it is important to note that retaining the status quo comes at a high price. Resisting any substantial revision of the philosophical landscape on this question imposes considerable limitations and presents serious difficulties to do with how aesthetic agreement is to be explained and how we should make room for the normative standards embedded in our aesthetic language and thought. The reason why the theories developed by Wiggins and Todd are so promising is precisely that they rest on the insight that if progress is to be made in Meta-Aesthetics, theories originating in Anti-Realism simply have to acknowledge the issues that have traditionally been favoured by Realists. Similarly, one may think that Realist theories that seek to incorporate ideas such as response-dependence and an analogy with secondary qualities have abandoned their distinctive identity, and that what is required here is a more isolated and solid Realist take on aesthetic value. Again, however, failing to engage with certain questions that have traditionally been more easily addressed by Anti-Realism most probably is a direct cause of what has hitherto been the relatively weak position of Objectivism and Cognitivism within Philosophical Aesthetics.

The philosophical developments discussed in this chapter clearly have ramifications beyond the R/AR debate. Contingent upon the kind of ontological status one is ready to confer upon aesthetic qualities and the form of epistemological authority one is willing to attribute to aesthetic judgements are a manifold of issues of considerable independent importance. These include the interpretation of artworks (e.g., whether such interpretations can be correct or not), the scope of our imaginative activity in aesthetic experience (e.g., the extent to which our imagination is constrained by the aesthetic character of the object of aesthetic contemplation), and the nature of artistic expression (e.g., exactly what it is that art expresses). The wide span of these ramifications is hardly surprising given that, fundamentally, these

June 21, 2008 11:5 MAC/NWA Page-181 9780230_220478_10_cha09 182 Three Debates in Meta-Aesthetics developments in Meta-Aesthetics urge us to reflect upon our understanding of aesthetic experience and aesthetic judgement. By way of concluding my discussion, then, let us address the question of whether the shift we have been examining may alter this understanding. One important aspect of this, and one that is certainly pivotal to any distinction between Realist and Anti- Realist approaches, concerns the precise role played by emotional responses in aesthetic judgement. The first point to note is that accounts that have traditionally emphasised emotions and emotional responses with regards to aesthetic value have now moved from a principally affect-based model to one more apt to recognise the cognitive element of emotions. According to theories such as Sensible Sub- jectivism and Aesthetic Quasi-Realism, emotional responses actually reveal something about the way in which the world and its contents are. For even if what it is for an object to have an aesthetic quality is, fundamentally, for that object to be able to give rise to an appropriate emotional response, there is an important sense in which being able to give rise to such a response is to reveal a property available to everyone, which therefore can be said to be a quality of the object itself. Interestingly, Realist theories of aesthetic judgement have, if anything, seemed less quick to recognise the significance of cognitive theories of emotion. For accounts based on O1, O2, C1, and C2 operate on the assumption that whilst there is an emotional element to most aesthetic experience, the claim to correctness and objectivity is not grounded in that element but, rather, in the perceptual response that aesthetic qualities call for. On this issue, aesthetic theories claiming to belong to the Anti-Realist camp seem to be ahead of the game. It appears, then, that the principal difference between accounts such as Sensible Subjectivism and Aesthetic Quasi-Realism on the one hand and those captured by O1, O2, C1, and C2 on the other lies in the philosophical task they expect emotional responses to perform in aesthetic judgement. According to the former kind of theory, it is by a process of understanding the emotional responses we have to aesthetic qualities as being less agent- relative and more cognitive in content that it becomes possible to suppose that a standard of normative assessment can be had for aesthetic judgements. However, on the latter kind of account, it is the perception, the phenomenal impression, that legitimises the correctness and incorrectness of aesthetic judgements. Perhaps, then, the doctrine of response-dependence in the R/AR debate has two relatively different expressions. Unlike theories such as Sensible Subjec- tivism and Aesthetic Quasi-Realism, Objectivist and Cognitivist theories rest on the idea that the response which lies at the heart of aesthetic experience and aesthetic judgement-making is not so much emotional but perceptual. That is to say, and as discussed above, theories based on the tenets formulated in O1 and O2 state that aesthetic qualities are response-dependent insofar as

June 21, 2008 11:5 MAC/NWA Page-182 9780230_220478_10_cha09 Elisabeth Schellekens 183 they give rise to perceptual responses or phenomenal impressions. What is more, C1andC2 also claim that it is a perception, rather than an emotion, that grounds our attribution of aesthetic qualities to some thing, person, or event. So, although in the previous section we established that one of the main reasons why aesthetic theories on both sides of the R/AR debate share more ground nowadays is that both kinds of theor recognise the response- dependent character of aesthetic qualities, it is also here, it seems, that the principal difference between them lies. On the first kind of account, the response in question is fundamentally perceptual, whereas on the second kind of account it is undoubtedly emotional. This observation also raises the further question of whether, on theories such as Sensible Subjectivism and Aesthetic Quasi-Realism, emotions themselves may, at least at times, be understood to be rather like perceptions or perceptual judgements. To all intents and purposes, this is a separate philosophical question in its own right. Whilst it is one thing to defend the view that emotions are to be cast in relatively cognitive terms, it is another thing altogether to hold that the precise way in which emotions are said to be cognitive is rather like the way in which our ordinary perceptions may be so described. For, far from all theories that emphasise the cognitive aspect of emotions also defend the stance that emotions are a kind of perception, and so it is important not to draw hasty conclusions on this important matter. Nevertheless, this move is at the very least readily available to Sensible Sub- jectivism and Aesthetic Quasi-Realism. Several philosophers have recently developed strong and interesting suggestions along these lines,46 and doing so may help the aforementioned theories to consolidate the standard of normative assessment upon which they rely: if emotions are a kind of perception or perceptual judgement, a Sensible Subjectivist or Aesthetic Quasi-Realist may have found a way to account for the cognitive content of emotions that validates the commitment to a standard of normative assessment. For as it stands, this standard—centred on the notion of appropriateness—requires more attention than Sensible Subjectivism and Aesthetic Quasi-Realism have given it so far.47 Clearly, one consequence of the suggestion that emotions, made to do so much work in theories such as Sensible Subjectivism and Aesthetic Quasi- Realism, are to be defined in terms of perceptions is that there will most probably only be a nominal difference between the two main philosophical approaches we have been examining. The alternatives available to philosophers working in the area may thus have become less obviously distinct and metaphysically marked. As a result, Realism and Anti-Realism with regards to questions of aesthetic value may no longer be immediately recognisable. But, as this chapter has suggested, rather than being cause for concern, this thought should be seen to strengthen the hope that a viable solution may soon be found to one of the most pressing set of questions in the philosophy of value.

June 21, 2008 11:5 MAC/NWA Page-183 9780230_220478_10_cha09 184 Three Debates in Meta-Aesthetics

Notes

1. According to one recent definition, ‘Realism is the thesis that the objects, properties, and relations the world contains exist independently of our thoughts about them or our perceptions of them ...Antirealists either doubt or deny the existence of the entities the realist believes in or else doubt or deny their independence from our conceptions of them.’ Khlentzos, Naturalistic Realism,2. 2. This tendency is present in works from both kinds of approaches. See, for example, Bender, ‘Realism and Aesthetic Disputes’, and Levinson, ‘Being Realistic’. 3. Mackie, Ethics. 4. For the purposes of this chapter, I shall use the phrase ‘evaluative qualities’ to refer to moral and aesthetic properties. There are, no doubt, evaluative qualities that are neither aesthetic nor moral. Nor is it necessary to assume that all aesthetic and moral properties are evaluative in the sense of an overall assessment or evaluation. 5. See, for example, Johnston, ‘Dispositional Theories of Value’, 139–174; Lewis, ‘Dispositional Theories of Value’, 113–137; Pettit, ‘Realism and Response- Dependence’, 587–626. 6. See, for example, de Sousa and Morton, ‘Emotional Truth’, 265–275. 7. See Blackburn, ‘Errors’, 1–22. See also McDowell, ‘Aesthetic Value’, 1–16; ‘Val- ues and Secondary Qualities’, 110–129; Wiggins, ‘Truth’, 139–184; ‘A Sensible Subjectivism?’, 185–214. See also my ‘Reasonable Objectivism’, 163–177. 8. Unless specified otherwise, the terms will hereafter be used in this more inclusive way. 9. Sibley, ‘Objectivity and Aesthetics’, 31–54. See also ‘Aesthetic and Non-Aesthetic’, 135–159. 10. Levinson, ‘Being Realistic’, 351–354. See also ‘Aesthetic Properties’, 61–80. For an account based on similar claims, see Pettit, ‘Aesthetic Realism’. 11. As aesthetic qualities are defined as perceptual, to say that there are elements of the phenomenal impression afforded by an aesthetic quality that are purely descriptive and so objective is not fundamentally distinct from the claim that there are elements of aesthetic qualities that are purely descriptive and therefore objective. 12. Please note that ‘Justification-Cognitivism’ and ‘Knowledge-Cognitivism’ are not terms in general usage. 13. See Levinson, ‘Aesthetic Properties’. 14. See Schellekens, ‘Reasonable Objectivism’. 15. For a very clear and comprehensive account of the different kinds of knowledge that may be gained from aesthetic experience and on the topic of art and knowledge in general, see Gaut, ‘Art and Knowledge’, 436–450. 16. For other theories promoting Knowledge-Cognitivism, see John, ‘Reading Fic- tion’, 331–348. See also Kieran, ‘Art, Imagination’, 337–351; Nussbaum, Love’s Knowledge. 17. Putnam, Meaning; Beardsmore, ‘Learning from a Novel’. 18. The claim is that the aesthetic appreciation of art yields knowledge (of certain kinds of experience), not just experience itself. 19. Robinson, ‘L’Education Sentimentale’. See also Currie, ‘Realism’. 20. Wiggins, ‘A Sensible Subjectivism’. 21. Blackburn, Essays in Quasi-Realism. 22. Unlike, for example, Goldman, ‘Realism’, 31–38, and Aesthetic Value. 23. Wiggins, ‘A Sensible Subjectivism’.

June 21, 2008 11:5 MAC/NWA Page-184 9780230_220478_10_cha09 Elisabeth Schellekens 185

24. Hume, ‘Of the Standard of Taste’. 25. Hume, Treatise of Human Nature, 294. 26. Wiggins, ‘A Sensible Subjectivism’, 195. 27. Ibid., 199. This kind of account may be said to belong to the category of ‘neo- sentimentalist’ theories. For references see note 8. 28. Ibid., 200. 29. Ibid. 30. Ibid., 202–203. 31. Ibid., 205. 32. Ibid., 209. 33. Ibid., 204–205. 34. Levinson, ‘Aesthetic Properties’, 62. 35. Schellekens, ‘Reasonable Objectivism’, 169. 36. Todd, ‘Quasi-Realism’, 281. 37. Ibid., 277. 38. Ibid., 282. 39. Ibid., 288. 40. Ibid., 282. 41. Ibid., 284. 42. Hopkins, ‘Kant, Quasi-Realism’, 169. 43. Todd, ‘Quasi-Realism’, 284. 44. One exception here is Zemach, Real Beauty. 45. Although the relations that Sensible Subjectivism and Aesthetic Quasi-Realism harbour with regards to C1 and C2 should be clear from our previous discussion, it is worth noting that there is nothing about the two accounts that necessarily excludes or undermines the claims central to the two forms of Knowledge- Cognitivism outlined in Section II. For more on some Anti-Realist arguments directed against Knowledge-Cognitivism, see Gaut, ‘Art and Knowledge’. 46. See, for example, Prinz, Gut Reactions. See also Nussbaum, Love’s Knowledge. 47. For more on this point, see, for example, D’Arms and Jacobson, ‘The Moralistic Fallacy’.

Bibliography

Beardsmore, R. 1973. ‘Learning from a Novel’ in G. Vesey (ed.), Philosophy and the Arts: Royal Institute of Philosophy Lectures 6, London: MacMillan. Bender, J. 1996. ‘Realism, Supervenience, and Irresolvable Aesthetic Disputes’, Journal of Aesthetics and Art Criticism, 54.4: 371–381. Blackburn, S. 1985. ‘Errors and the Phenomenology of Value’ in T. Honderich (ed.), Morality and Objectivity, London: Routledge and Kegan Paul, 1–22. Blackburn, S. 1993. Essays in Quasi-Realism, Oxford: Oxford University Press. Currie, G. 1998. ‘Realism of Character and the Value of Fiction’ in J. Levinson (ed.), Aesthetics and Ethics, Cambridge: Cambridge University Press. D’Arms, J. and Jacobson, J. 2000. ‘The Moralistic Fallacy: On the ‘‘Appropriateness’’ of Emotions’, Philosophy and Phenomenological Research, 71: 65–90. de Sousa, R. and Morton, A. 2002. ‘Emotional Truth: Emotional Accuracy’, Proceedings of the Aristotelian Society Supplementary Volume, 76.1: 265–275. de Sousa, R. 2007. ‘Truth, Authenticity and Rationality of Emotion’, Dialectica, 61. 3: 323–345.

June 21, 2008 11:5 MAC/NWA Page-185 9780230_220478_10_cha09 186 Three Debates in Meta-Aesthetics

Gaut, B. 2003. ‘Art and Knowledge’ in J. Levinson (ed.), The Oxford Handbook of Aesthetics, Oxford: Oxford University Press, 436–450. Goldman, A. 1993. ‘Realism About Aesthetic Properties’, Journal of Aesthetics and Art Criticism, 51: 31–38. Goldman, A. 1995. Aesthetic Value, Boulder, CO: Westview Press. Hopkins, R. 2001. ‘Kant, Quasi-Realism and the Autonomy of Aesthetic Judgement’, European Journal of Philosophy, 9.2: 169. Hume, D. 1888. A Treatise of Human Nature, L.A. Selby-Bigge (ed.), Oxford: Clarendon Press. Hume, D. 1965. ‘Of the Standard of Taste’ in J. Lenz (ed.), Of the Standard of Taste and Other Essays, Indianapolis: Bobbs-Merrill. John, E. 1998. ‘Reading Fiction and Conceptual Knowledge: Philosophical Thought in Literary Context’, Journal of Aesthetics and Art Criticism, 56: 331–348. Johnston, M. 1989. ‘Dispositional theories of value’, Proceedings of the Aristotelian Society Supplementary Volume, 63: 139–174. Khlentzos, D. 2004. Naturalistic Realism and the Antirealist Challenge, Cambridge, MA and London: MIT Press. Kieran, M. 1996. ‘Art, Imagination, and the Cultivation of Morals’, Journal of Aesthetics and Art Criticism, 54: 337–351. Levinson, J. 1994. ‘Being Realistic About Aesthetic Properties’, Journal of Aesthetics and Art Criticism, 52: 351–354. Levinson, J. 2001. ‘Aesthetic Properties, Evaluative Force, and Differences of Sensibility’ in E. Brady and J. Levinson (ed.), Aesthetic Concepts: Essays After Sibley, Oxford: Clarendon Press, 61–80. Lewis, D. 1989. ‘Dispositional Theories of Value’, Proceedings of the Aristotelian Society Supplementary Volume, 63: 113–137. Mackie, J. L. 1977. Ethics: Inventing Right and Wrong, London: Penguin Books. McDowell, J. 1983. ‘Aesthetic Value, Objectivity and the Fabric of the World’ in E. Schaper (ed.), Pleasure, Preference and Value, Cambridge: Cambridge University Press, 1–16. Reprinted in McDowell, J. 1998, Mind, Value and Reality, Cambridge, MA & London: Harvard University Press. MacDowell, J. 1985. ‘Values, and Secondary Qualities’ in T. Honderich (ed.), Moral- ity and Objectivity, Routledge and Kegan Paul, 110–129. Reprinted in J. McDowell, 1998. Mind, Value and Reality, Cambridge, MA & London: Harvard University Press. Nussbaum, M. 1990. Love’s Knowledge: Essays on Philosophy and Literature, New York: Oxford University Press. Pettit, P. 1983. ‘The Possibility of Aesthetic Realism’ in E. Schaper (ed.), Pleasure, Preference, and Value, Cambridge: Cambridge University Press, 17–38. Pettit, P. 1991. ‘Realism and Response-Dependence’ in Response-Dependent Concepts, Mind, 100: 587–626. Plato. 1883. Euthyphro in B. Jowett (ed. and trans.), The Dialogues of Plato, Volume One, New York: Scribner, 277–302. Prinz, J.J. 2004. Gut Reactions: A Perceptual Theory of Emotion, New York: Oxford University Press. Putnam, H. 1978. Meaning and the Moral Sciences, London: Routledge & Kegan Paul. Robinson, J. 1995. ‘L’Education Sentimentale’, Australasian Journal of Philosophy, 73.2: 212–226. Reprinted in S. Davies (ed.), 1997. Art and Its Messages, University Park: Pennsylvania State University Press.

June 21, 2008 11:5 MAC/NWA Page-186 9780230_220478_10_cha09 Elisabeth Schellekens 187

Schellekens, E. 2006. ‘Towards a Reasonable Objectivism for Aesthetic Judgements’, British Journal of Aesthetics, 46.2: 163–177. Sibley, F. 1959. ‘Aesthetic Concepts’, Philosophical Review, 68: 421–450. Reprinted in J. Benson, B. Redfern and J. Roxbee Cox (eds), 2001. Frank Sibley: Approach to Aesthetics – Collected Papers, Oxford: Clarendon Press. Sibley, F. 1965. ‘Aesthetic and Non-Aesthetic’, Philosophical Review, 74: 135–159. Reprinted in J. Benson, B. Redfern and J. Roxbee Cox (eds), 2001. Frank Sibley: Approach to Aesthetics – Collected Papers, Oxford: Clarendon Press. Sibley, F. 1968. ‘Objectivity and Aesthetics’, Proceedings of the Aristotelian Society,Sup- plementary Volume 42: 31–54. Reprinted in J. Benson, B. Redfern and J. Roxbee Cox (eds), 2001, Frank Sibley: Approach to Aesthetics – Collected Papers, Oxford: Clarendon Press. Todd, C. 2004. ‘Quasi-Realism, Acquaintance, and the Normative Claims of Aesthetic Judgements’, British Journal of Aesthetics, 44.3: 281. Wiggins, D. 1998. ‘Truth, and Truth as Predicated of Moral Judgments’, Needs, Values, Truth, Oxford: Clarendon Press, 139–184. Wiggins, D. 1998. ‘A Sensible Subjectivism?’ Needs, Values, Truth, Oxford: Clarendon Press, 185–214. Zemach, E. 1997. Real Beauty, University Park: Pennsylvania State University Press.

June 21, 2008 11:5 MAC/NWA Page-187 9780230_220478_10_cha09 10 Aesthetic Ideals Rafa¨el De Clercq

My point of departure in this chapter is a claim about aesthetic properties that seems hard to deny in the light of twentieth-century post-formalist aesthetics (as represented by, for example, Walton’s ‘Categories of Art’). The claim is this: what aesthetic properties an object has depends not just on what non-aesthetic, accidental (e.g., perceptual) properties it has but also on what kind of object it is, that is under what sortal it falls (e.g., ‘man’, ‘animal’).1 Using the concept of supervenience to single out the relevant sense of dependence, this claim can also be put as follows: any adequate supervenience base for aesthetic properties minimally includes a number of essential properties (let us assume that if a conjunctive property P∧Q belongs to the supervenience base, then so do P and Q). In other words, if

∀x∀F ∈ A Fx →∃G ∈ B(Gx ∧ ∀y(Gy → Fy))

is to be a correct, formal representation of the fact that aesthetic properties supervene strongly on non-aesthetic properties, then it will have to be assumed that the set B includes at least some essential properties.2 Essential properties tend to occur in clusters, and some of these clusters are designated by common kind-terms or sortals such as ‘man’, ‘animal’, and the like.3 Now, it has appeared to some philosophers that, although there is a general sense in which such clusters are aesthetically relevant—they occur in any supervenience base that is adequate for aesthetic properties—there is also a more special sense, distinguishing their contribution to the aesthetic character of objects from the contribution made by ordinary, accidental properties such as looking greenish or casting a certain shadow. However, the literature does not offer a clear, uniform analysis of what that special sense is. In this chapter, my focus will be on one proposal, advanced in posthumous work by Frank Sibley.4 On this (Kantian) proposal, the special aesthetic relevance

188

June 21, 2008 11:11 MAC/NWA Page-188 9780230_220478_11_cha10 Rafa¨el De Clercq 189 of sortals (i.e., sortal nouns or, metaphysically speaking, clusters of essential properties) consists in the fact that they provide standards or ideals of beauty.5 Here is one passage in which Sibley expressed the idea:

[T]o judge that x is a beautiful face, beautiful woman, beautiful body, beautiful horse, beautiful cat (non-comparatively ...), reference must be made via the noun to some ideal (...) of respectively facial, female, bodily, equine, and feline beauty.6

Whether an object is beautiful seems to depend, then, on two factors: first, on what aesthetic ideal is associated with the object through its membership of a particular kind; second, on whether the object sufficiently approximates that ideal by virtue of its non-aesthetic properties (in particular, by virtue of the properties included in the supervenience base). Unfortunately, Sibley is rather silent on how these aesthetic ideals should be conceived of. Little more is said than that they are ‘norms or ideals incorporating notions of appropriateness, etc.’ (pp. 182–3). In what follows, four interpretations of the notion of an aesthetic ideal (as it figures in Sibley’s writings) will be considered. Whichever interpretation is chosen, the notion turns out to be problematic: either it is a trivial, uninteresting notion; or it is an unclear, though possibly interesting, notion; or it is a clear, non-trivial notion that is empty because nothing corresponds to it. This negative conclusion regarding the notion of an aesthetic ideal is not intended to cast doubt on what probably motivated its introduction: the sense that sortals are aesthetically relevant in a special way. In fact, it will be admitted that the special aesthetic relevance of sortals resides in the ideals associated with them! What is objectionable, because it is misleading, is calling these ideals ‘ideals of beauty’ or, as in this chapter, ‘aesthetic ideals’. By way of ending, a notion that was introduced, or at least re-interpreted, in a previous paper—sortal-dependence—will be evaluated in the light of the announced conclusions.

Sibley seems to hesitate between an objective and a subjective (or subject- relative) conception of aesthetic ideals. On the one hand, he writes that nouns or concepts set ideals of beauty, suggesting a conceptual connection between non-aesthetic sortals and aesthetic ideals.7 A connection of this sort seems more or less independent of the circumstances in which the nouns or concepts are used. On the other hand, Sibley writes that the ideals in question are ‘no doubt multiple, varying, and shifting’ (p. 181) and that ‘such ideals might ...change or vary with fashion and across cultures’ (p. 194). Here the idea of a conceptual connection seems to be absent. The objective and the subjective conception can be reconciled, however, by the following

June 21, 2008 11:11 MAC/NWA Page-189 9780230_220478_11_cha10 190 Aesthetic Ideals thought: individual sortal concepts have associated with them, as a matter of conceptual necessity, a range of ideals, and to be able to make warranted aesthetic judgements using the concepts (e.g., ‘Isabelle Adjani is a beautiful woman’) it is necessary to know at least one such ideal. As a result, people in different cultures or at different times in history may grasp the same concept but associate a different ideal with it, thereby putting themselves in a position of faultless disagreement with one another. In any case, in what follows it will be assumed that there is something objective to aesthetic ideals. Otherwise, it would become hard to understand how they can determine what aesthetic properties an object has, or how knowledge of them enables us to arrive at warranted aesthetic judgements. (Of course, this is not to deny that in some contexts it may be legitimate to use the term ‘aesthetic ideal’ in a different way, for example, as applying to something believed to be an aesthetic ideal by a certain group or individual. A similar ambiguity has actually infected the use of the term ‘value’: sometimes it is used to say that something is valuable, sometimes it is used to say that something is believed to be valuable.) There is a trivial sense in which there are objective aesthetic ideals. After all, it seems hard to deny that (at least) for some sortal S there are non-aesthetic properties P1, ..., Pn such that, necessarily, for any object x, if x instantiates

P1, ..., Pn, then x is a maximally beautiful S, where a ‘maximally beautiful S’ is one whose beauty cannot be topped by any other S.8 For example, it is probably true that if a woman looks and moves like Isabelle Adjani, 9 then she is a maximally beautiful woman. The set of properties P1, ...,

Pn could thus be taken to define an aesthetic ideal with respect to S (e.g., ‘woman’). That aesthetic ideals in this sense exist follows from two seemingly trivial premises. First, the premise that, possibly, some S is maximally beautiful; second, the premise that whether or not something is a maximally beautiful S depends on what non-aesthetic properties it has. The second premise is a trivial supervenience thesis. The first premise may not be entirely trivial, but it is utterly plausible. At least some kinds of objects do not admit of endless variation in aesthetically relevant respects (e.g., elementary particles or certain abstract objects). As a result, there must be a limit to how beautiful they can be. Moreover, non-comparative, non-relational (i.e., absolute) judgements of beauty seem to be possible, at least in some cases, and this presupposes that there is a maximum for the cases in question. If this last claim is not intuitively clear, then my best argument would build on the following, bold premise: with properties that can be instantiated to various degrees (richness, redness, baldness, tallness, beauty, and so on), relational truths of the form ‘x is at least as ϕ as y’ are primary. Together with some general, non-relational truths such as ‘(possibly) some x is ϕ’, they represent all that is given.10 Non-relational truths concerning particular cases follow from them either logically or by convention. (For present purposes we can ignore negative non-relational truths, which are of the form ‘x is not ϕ’.

June 21, 2008 11:11 MAC/NWA Page-190 9780230_220478_11_cha10 Rafa¨el De Clercq 191

Moreover, it will be assumed that x and y can be in different worlds.) They follow by convention if it is stipulated—implicitly or explicitly—that if x is at least as ϕ as c, for some chosen c, then x is ϕ. This is what happens when a comparison class is involved. Or they follow logically if there is an x which is at least as ϕ as any y.11 This is what happens when a maximum is involved.12 Obviously, the x instantiating ϕ to the maximum degree will be ϕ, given the general truth that some x is ϕ. (For this latter truth to follow logically from the stated premises a meaning postulate would have to be assumed; for example, if x is ϕ, then no y is such that it is both not-ϕ and at least as ϕ as x. This can be seen as a convention, but it is constant throughout different contexts because it reflects our understanding of gradual differences in ϕ. Conventions introducing a comparison class are not so constant.) Here is a simple example. Let it be given that Alice is at least as beautiful as Beatrice, Beatrice is at least as beautiful as Clarisse, and Clarisse is at least as beautiful as Dorothea. Moreover, let it be given that the relation holds in the other direction just in one case: Beatrice is also at least as beautiful as Alice. Alice, Beatrice, Clarisse and Dorothea together exhaust the domain of the (reflexive and transitive) relation ‘is at least as beautiful as’. Now the question is: who is beautiful? If what has just been suggested is correct, then there are only two ways in which the question can be answered on the basis of what is given. One is to simply agree on the beauty of one of the girls, say Clarisse, and to reckon everyone beautiful who is at least as beautiful as she is (in the example, Alice, Beatrice and Clarisse herself). This introduces a conventional element that may be appropriate in some contexts but not in others. For example, if new girls are introduced to the domain, then we may agree to locate the ‘threshold of beauty’ elsewhere. The other way is to look for a maximum, a person at least as beautiful as anyone else, in this case: Alice and Beatrice, who are both equally beautiful. Given that at least one among Alice, Beatrice, Clarisse and Dorothea is beautiful, it follows that Alice and Beatrice are beautiful. Here no variable, conventional element is involved. Now, if this schematic account of the relation between relational and non- relational truths regarding particular cases of a gradable property such as beauty is roughly correct, then the following conclusion seems to follow: if non-relational judgements can be made in the case of beauty (e.g., ‘Isabelle Adjani is beautiful’) without involving, implicitly or explicitly, a comparison class (and a corresponding notion of what is average or mean), then there must be a maximum degree of beauty. Recall, however, that there was also another reason for believing that there is such a maximum, at least for certain kinds of objects. Whatever the case may be, the thesis that aesthetic ideals exist as non- aesthetic, sufficient conditions for being a maximally beautiful S does not strike me as highly significant, or indeed, as a highly plausible interpretation of what Sibley was after. For example, Sibley did not believe that there

June 21, 2008 11:11 MAC/NWA Page-191 9780230_220478_11_cha10 192 Aesthetic Ideals is an aesthetic ideal for just any kind of entity,13 whereas the trivialising interpretation seems to imply that there is. The thesis that aesthetic ideals exist becomes at least more significant when they are taken to be non-aesthetic necessary and sufficient conditions for being a maximally beautiful S so that there is a non-aesthetic property P, or a finite conjunction of such properties, which can be said to constitute maximal beauty in the case of objects that are S. However, there seems to be no good reason to believe that aesthetic ideals in this sense exist. Examples of principles that capture the essence of beauty of some kind or other do not readily come to mind. Perhaps, in some domains, we can arrive at some fairly reliable generalisations. However, if these generalisations are arrived at within a controlled, scientific setting—as, presumably, they should be—then chances are that they will merely record what certain subjects tend to find beautiful. Furthermore, there is no reason to expect that the generalisations will ever become more than contingent, ceteris paribus laws.Insum,aposte- riori insight into the essence of beauty, including S-beauty, does not seem to be a realisable goal. There is a general explanatory defect affecting all aesthetic ideals in the sense we are now considering: such ideals are capable of explaining only why certain objects are maximally beautiful. They cannot explain why objects exemplifying properties that do not exactly match the ideal may also appropriately be judged beautiful. For such an explanation to be possible, two conditions would have to be fulfilled. First, there would have to be a non- aesthetic way of measuring how close an object approximates the aesthetic ideal. In other words, it would have to be possible to judge such closeness on the basis of non-aesthetic criteria. Otherwise, there is nothing that aesthetic ideals might contribute to our understanding of why an item is beautiful. Sec- ond, the degree of closeness to the ideal, as measured by the non-aesthetic standard, would have to correspond to the degree of beauty exemplified by the object. Suffice it to say that these conditions seem unlikely to be fulfilled. (The reason why this lack of explanatory power constitutes a defect is that Sibley introduced aesthetic ideals to explain something, namely why warranted judgements of beauty—not just maximal beauty!—often presuppose sortal knowledge.) It may also be objected that the interpretation of ‘aesthetic ideal’ we are currently considering does not leave room for multiplicity or cultural-historical variation, a possibility that Sibley clearly wanted to allow for. After all, if aesthetic ideals are (non-aesthetic) necessary and sufficient conditions for S-beauty, then how could there be more than just one ideal for S-objects? However, there is a way of accommodating multiplicity in this conception. As long as the necessary and sufficient conditions that make up the aesthetically ideal S are not too determinate, it is possible to meet them in different ways—just as it is possible to meet the necessary and sufficient conditions for being rectangular or being married in different ways. Because a culture

June 21, 2008 11:11 MAC/NWA Page-192 9780230_220478_11_cha10 Rafa¨el De Clercq 193 may be aware of, or simply prefer, only one or two ways of meeting the conditions, it may come to identify the ideal S with them. This does not preclude the possibility of other cultures trying to realise the same ideal in a completely different manner. So even though there is, for each S, only one aesthetic ideal on this conception, there is room for multiplicity and cultural-historical variation since the conditions for being an aesthetically ideal S may be met in different ways (ways that could even be classified into types). Sibley sometimes suggests that the existence of ‘standards’ or ‘ideals’ or ‘norms’ of beauty merely consists in the existence of certain restrictions on what can truly be judged beautiful. Consider, for example:

By standards incorporating notions of rightness or appropriateness I mean this. The grace of a line lies in, say, its particular flowing curvature, the beauty of a pebble in its line, shape, and colouring; there are no particular restrictions on how these should be except that they should be beautiful ( ...) But upon those lines, colours, and shapes, on which, together with the interrelation of these, and no doubt much else, facial, equine, or female beauty depends, there are clear, if somewhat loose and variable, restrictions. However beautifully luminous the crimson eyes, however delicately green the mottling of the cheeks, however richly brown and however graceful the outward curve of the teeth, they are not beautiful eyes, cheeks,orteeth.14

The above passage suggests that a standard or ideal of beauty for a kind K

(e.g., ‘horse’, ‘man’) is nothing more than a set of properties P1, ..., Pn such that, necessarily, for any x, if x is a beautiful K, then x has none of P1, ...,

Pn. In other words, an ideal of beauty for a kind K would be no more than anegativenecessary condition for K-beauty. To avoid making the suggestion blatantly trivial, one might specify that the set of properties includes neither essential properties (e.g., not being K) nor negative aesthetic properties (e.g., being an ugly K). Paradoxical as it may sound, the problem with this suggestion is that it presupposes too little. Specifically, it presupposes no more than the thesis that aesthetic properties and their complements (e.g., not-being-beautiful) supervene on other, non-aesthetic properties. In colloquial language, this is the thesis that the aesthetic character of an object is determined by how it is in certain non-aesthetic respects. Now, as the logical formula quoted in the beginning of the chapter makes clear, if A-properties supervene on B-properties, then for every A-property there is a B-property (necessarily) guaranteeing its instantiation. So, if aesthetic properties and their complements supervene on certain non-aesthetic properties, then for any property of the first sort there is a property of the latter sort (necessarily) guaranteeing its instantiation. Consider, then, the property of not-being-a-beautiful-K.

June 21, 2008 11:11 MAC/NWA Page-193 9780230_220478_11_cha10 194 Aesthetic Ideals

This is a complement of an aesthetic property, if not itself an aesthetic property. So, according to the above reasoning, there must be a non-aesthetic property P, different from not-being-a-beautiful-K, which guarantees its instantiation (among K-objects). As a result, there is guaranteed to be a set of properties, {P}, such that, necessarily, no beautiful K exemplifies any member of the set. In other words, there is guaranteed to be an aesthetic ideal—in the sense defined above—for the kind K. Because it requires no more than the aforementioned supervenience thesis, the present proposal is unable to explain why sortal knowledge often is a precondition of making warranted aesthetic judgements. After all, the world abounds in supervening properties but only some of these properties require sortal knowledge for their attribution. Moreover, when sortal knowledge is a requirement, the reason usually is that the sortal in question defines the contextually relevant comparison class. However, this is not the (only) reason why Sibley believes that sortal knowledge is required for the attribution of aesthetic properties.

So far we have not made much progress with respect to finding a suitable conception of aesthetic ideals. Perhaps a more promising lead can be found in an observation made by Roger Scruton:

Features that we would regard as beautiful in a horse – developed haunches, a curved back, and so on – we would regard as ugly in a man, and this aesthetic judgement would be determined by our conception of what men are, how they move, and what they achieve through their movements.15

This passage is quoted approvingly by Sibley (on p. 180) just before he introduces the notion of an ideal of beauty. Later in the article (pp. 188–189) he returns to the passage, supplements it with other examples, and makes explicit the link with the notion of a sortal-specific aesthetic ideal:

To echo [Scruton’s] words with a different example, features that we would regard as beautiful in an English elm – the majestic sturdiness, solidity, and strength, which is part of what makes it a beautiful English elm – might be regarded (though still making it perhaps a beautiful tree) as defects vis-a-vis` the American elm (which has a vase-shaped form, a spreading, forking, open habit); to be judged beautiful as an American elm, a tree needs these characteristics, sources of its characteristic beauty, spreading elegance, graceful branching, lightness rather than solidity or sturdiness. (This is still ‘strong’ attributiveness, with respect to a norm or ideal of a species, not involving comparison within a species ...)16

June 21, 2008 11:11 MAC/NWA Page-194 9780230_220478_11_cha10 Rafa¨el De Clercq 195

In a nutshell, the point of the quoted passages is this: different sortals have different beauty-making properties associated with them. (Falling under a specific sortal is, presumably, not itself a beauty-making property.) So perhaps the thesis that sortals set standards or ideals of beauty could be interpreted as saying that sortals determine what properties of an object are beauty-making (‘sources of beauty’), or at least possibly beauty-making. More formally, one might express this thesis in at least one of the following two ways (for some sortals S and S):

(1) There is a property that is necessarily beauty-making in S and which is not necessarily beauty-making in S (even though it can be exemplified in both S and S); (2) There is a property that is possibly beauty-making in S and which is not possibly beauty-making in S (even though it can be exemplified in both SandS).

Theses (1) and (2) do not imply one another but they can both be true (for some S and S), namely if there is property P that is necessarily beauty- making in S while it is not necessarily beauty-making in S and if there is also a property P that is possibly beauty-making in S while it is not possibly beauty- making in S. P and P could be the same property but they need not be. Presumably, if there are beauty-making properties, then there are also ugly- making ones. So (1) and (2) may correspond to:

(3) There is a property that is necessarily ugly-making in S and which is not necessarily ugly-making in S (even though it can be exemplified in both SandS); (4) There is a property that is possibly ugly-making in S and which is not possibly ugly-making in S (even though it can be exemplified in both S and S).

Again, (3) and (4) can both be true for some S and S. In fact, theses (1)–(4) can be combined at will. However, some combinations are possible only if more than one property is involved because one and the same property cannot be both necessarily beauty-making and necessarily/possibly ugly-making in S, although it can be both possibly beauty-making and possibly ugly-making in S (for some S). An attractive feature of any combination of (1)–(4) is that it explains why sortal knowledge often is a precondition of making warranted aesthetic judgements. After all, it is plausible to suppose that a warranted judgement of beauty cannot be made without knowing what is contributing to, or detract- ing from, the beauty of an object—that is, without knowing what features are beauty/ugly-making. Since any combination of (1)–(4) entails that the beauty/ugly-making status of a feature is sortal-dependent, it follows that

June 21, 2008 11:11 MAC/NWA Page-195 9780230_220478_11_cha10 196 Aesthetic Ideals at least some aesthetic judgements will be insufficiently warranted without knowledge of what kind of object is being judged. Theses (1)–(4) do not look substantial from a metaphysical point of view. If they express the sense in which there are aesthetic ideals (either individually or in combination), then aesthetic ideals are not some special kind of entity. In other words, (1)–(4) do not reify aesthetic ideals. This, too, is an attractive feature of the present proposal but it does not make it preferable to the previous ones. The first three proposals committed one only to ordinary aesthetic and non-aesthetic properties; the current proposal involves the additional commitment to a second-order property of being beauty/ugly- making, though perhaps this property can be made to dissolve under analysis. Unfortunately, it is not clear at all how the property of being beauty- making is to be analysed. (Henceforth, all that is claimed to hold true of this property will be assumed to hold true of the property of being ugly-making as well—simply substitute ‘ugly’ for ‘beauty’ to get the corresponding claims.) Presumably, it is a property of properties that stand in a special kind of determination relation to beauty, one that is not reducible to the relation of supervenience.17 After all, far more non-aesthetic properties enter into the second relation than into the first. The property of falling under a certain sortal, for example, will never be reckoned a beauty-making feature although it is, by assumption, part of the supervenience base (‘He is beautiful because he is a man’ is something only extreme sexists would say). Similarly, by the transitivity of the supervenience relation, fundamental physical properties are part of the supervenience base but they are unlikely to be considered beauty-making. Beauty-making properties are somehow more closely tied to the beauty of an object than are other aesthetically relevant properties. But by what measure of closeness? Is it the likelihood that a certain property is mentioned in an explanation of the item’s beauty? If so, then in determining this likelihood one should certainly discard certain explanations: some will be wrong, some will be incomplete, some will be unintelligible (to us). However, one naturally wonders whether sorting out the relevant explanations from the irrelevant ones can succeed without some idea or other of which properties are beauty-making. Moreover, it seems that any complete explanation will have to mention aesthetically relevant properties that are not beauty-makers. One might suggest that the peculiarity of beauty-making properties lies in the fact that their presence guarantees beauty to some degree. On this suggestion, P is a beauty-making property of x if, and only if, [P is non- aesthetic and] necessarily, for all y, if x and y are of the same kind, then y exemplifies P only if y is at least minimally beautiful. As it stands, however, this suggestion faces two problems. First of all, although there is no doubt that it distinguishes beauty-making properties from some aesthetically relevant properties that are not beauty-making, it is questionable whether it distinguishes beauty-making properties from all such aesthetically relevant

June 21, 2008 11:11 MAC/NWA Page-196 9780230_220478_11_cha10 Rafa¨el De Clercq 197 properties. In particular, the complex, conjunctive property formed out of all properties that are aesthetically relevant in a particular case might here provide a counterexample (because it satisfies the condition that is supposed to uniquely characterise beauty-making properties). Furthermore, the suggestion rules out properties that are merely possibly beauty-making for a particular sortal S, whereas, at least on my understanding of the term, it makes perfect sense to say that a property is beauty-making in one S-object but not in another one. This is why (2) can be seen as supplementing (1). Yet another suggestion might be that P is a beauty-making property of x in w1 if, and only if, [P is non-aesthetic and] for all y and for all w2, if y exists in w2 and is indiscernible from x in w1, then it is known in w2 that y is beautiful only if it is known in w2 that y is P. Simplifying, and using only colloquial language, beauty-making properties are those we cannot ignore if we are to discern the beauty of an object. The nice thing about this suggestion is that it leaves room for merely possible beauty-makers: P could be a property of both x and y, x and y being of the same kind, and yet be a beauty-maker in only one of them. Still, the suggestion faces two prima facie problems. One is that it seems possible to know that something is beautiful without knowing everything that makes it beautiful. If this is true, then there are beauty-makers that we do not have to take notice of. The other problem is that some properties might not be beauty-makers in an object even though we have to notice them. Possible examples include having a spatial location, having a shape, having a colour, having parts, being a man, being an artefact and being striped. In any case, the foregoing considerations are not supposed to attack the notion of a beauty-making feature. Clearly, some sense can be made of the notion even if it cannot be analysed. The problem to which the foregoing considerations are supposed to point is merely that not much philosophical clarification is gained by translating claims about aesthetic ideals into claims about beauty-making features, since the latter is a notion that is itself in need of clarification. So what we would be left with, if this last interpretation were adopted, is a notion of aesthetic ideal that is non-trivial but also unclear. In particular, it seems to be too unclear to do philosophically interesting work.

III

Recall what philosophical work the notion of an aesthetic ideal was supposed to do: it was supposed to account for the special aesthetic relevance of sortals. Given that we have not found an appropriate interpretation of the notion, it may be worthwhile to look for alternative accounts, ones in which the notion of an aesthetic ideal plays no role. In what follows, the rough outlines of one such account will be sketched. Readers familiar with the aesthetic literature may see in it a generalisation of some of the central ideas of Walton’s ‘Categories of Art’.

June 21, 2008 11:11 MAC/NWA Page-197 9780230_220478_11_cha10 198 Aesthetic Ideals

According to David Wiggins, sortal concepts of spatio-temporal particulars (‘man’, ‘tree’, ‘lung’, ‘ship’ and other substance-concepts) have associated with them either a principle of activity or a mode of operation or a function.18 From this association derives an idea of what are ‘normal’ or ‘functional’ specimens of the kind in question. Let us call this, for the sake of contrast, a non-aesthetic ideal. A blossoming tree, a seaworthy ship and a lung capable of extracting oxygen from inhaled air may all count as ideal in the intended non-aesthetic sense. Now, it is clear that such non-aesthetic ideals are aesthetically relevant. A complete and reliable aesthetic evaluation of an artefact’s design, for example, will have to take into account what the artefact was designed for. Similarly, a complete and reliable aesthetic evaluation of a human being will have to take into account at least some facts concerning how human beings (typically) function, for example, ‘how they move, and what they achieve through their movements’.19 More generally, an object’s aesthetic appeal often lies in how it approximates, or deviates from, the non-aesthetic ideal associated with its sortal.20 In short, the relation to the non-aesthetic ideal is a matter of aesthetic importance. The question is how to bring out this importance in such a way that the special aesthetic relevance of sortals is captured. For example, it would not suffice to say that the non-aesthetic ideal is part of the supervenience base: this is also true of most accidental properties, including some very uninteresting ones. My suspicion is that the special aesthetic relevance of sortals is best captured in epistemological terms and, more specifically, in terms of why we need to know what sortal concept applies (in a particular case). In light of the foregoing, it is plausible to suggest that we need to know what sortal concept applies because we need to know what non-aesthetic ideal applies, for example what function the object is supposed to perform. The reason why we need to know what accidental properties are instantiated is, of course, different. Accidental properties have no ideals associated with them.21 For example, there is no particular way in which a red object can be expected to behave (except that, in appropriate conditions, it will reflect light of a certain wavelength). Nonetheless, we need to know an object’s accidental properties in order to know how it is related to the non-aesthetic ideal characteristic of its sortal. Moreover, such knowledge is required by anyone who wishes to determine an object’s overall aesthetic value. Perhaps this latter fact can even be explained in terms of beauty-making properties but, for obvious reasons, no attempt will be made here. Summarising, my suggestion is that the special aesthetic relevance of sortals consists in the fact that they are associated with non-aesthetic ideals— that is, roughly speaking, ideas of normality and functionality—knowledge of which is at least sometimes necessary for determining the aesthetic value of objects falling under the sortals. If this suggestion is correct, then it seems that we do not need the notion of an aesthetic ideal or a standard of beauty in order to account for the special aesthetic relevance of sortals.

June 21, 2008 11:11 MAC/NWA Page-198 9780230_220478_11_cha10 Rafa¨el De Clercq 199

It might be objected that we still need aesthetic ideals to explain how someone could be a good (reliable) aesthetic judge of Fs and a bad (unreliable) aesthetic judge of Gs while knowing both what Fs are and what Gs are, including what non-aesthetic ideals are associated with these sortals. (In fact, this is a situation one frequently encounters. Few people have universal taste.) In reply, it should be pointed that any explanation in terms of aesthetic ideals will suffer from the defects exhibited by this notion (at least if ‘aesthetic ideal’ is interpreted in any of the four ways considered in this chapter). Otherwise said, if the notion of an aesthetic ideal is either unclear, empty or trivial, then an explanation invoking this notion will not be fully satisfactory. Moreover, at least in some instances of the situation we are considering, an alternative explanation is possible which does not appeal to aesthetic ideals. For example, such an explanation might refer to the fact that the aesthetic value of Gs (e.g., musical pieces) is mainly determined by properties of a kind not frequently encountered in Fs (e.g., paintings). Obviously, someone who is insensitive to properties of this kind, who has difficulties detecting and remembering them and so on, is destined to be a poor aesthetic judge of Gs.

To end, let me indicate how the foregoing bears on a previous paper of mine. In ‘The Aesthetic Peculiarity of Multifunctional Artefacts’, I suggested that ‘beautiful’ may be a sortal-dependent adjective in the following sense: ‘[Necessarily] for all x, x is beautiful only if there is a sortal noun ‘‘S’’ such that x is a beautiful S, where ‘‘S’’ may be said to provide a standard or criterion or ideal of beauty for the entities falling under it’.22 Translating the linguistic idiom into a metaphysical one, this can be read as a suggestion concerning the property of being beautiful. More specifically, the suggestion would be that, necessarily, for any x that is beautiful, there is a sortal concept S such that x is S and x is beautiful by the standard of beauty associated with S. It might be asked whether this suggestion is still worth considering, given the scepticism towards standards or ideals of beauty that was vented in the previous sections of this chapter. One might answer this question in the affirmative because, strictly speaking, the notion of a standard or ideal of beauty has not been rejected. For all that has been said, there may still exist aesthetic ideals in some (objective) sense or other. In particular, the unclear though graspable sense, expressed by means of the term ‘beauty-making property’ (cf. Section III), might be thought appropriate here. For example, one might suggest reading the right- hand side of the definition of sortal-dependence as saying that, necessarily, for any x that is beautiful, there is sortal concept S such that x is S and x is beautiful (at least in part) in virtue of the beauty-making properties associated with S. Or simpler: necessarily, for any x that is beautiful, there is a sortal concept S such that x is S and x has at least some of the beauty-making properties

June 21, 2008 11:11 MAC/NWA Page-199 9780230_220478_11_cha10 200 Aesthetic Ideals associated with S. Unfortunately, the thesis that beauty is a sortal-dependent property becomes trivially true on this reading of the definition, at least if beauty-making properties can be merely possibly beauty-making. All the thesis requires now is that (i) each beautiful object falls under some sortal, and that (ii) each beautiful object is ‘made’ beautiful by (i.e., is beautiful ‘in virtue of’) some property it exemplifies. Conditions (i) and (ii) together imply that each beautiful object has a beauty-making property that is associated with its sortal (at least as a possible beauty-maker for that sortal). Perhaps the question whether beauty is a sortal-dependent property becomes more interesting when it is understood in terms that do not involve (implicit or explicit) reference to the notion of an ideal of beauty. For example, in view of how the special aesthetic relevance of sortals was explained in the previous section, one might suggest giving the definition of sortal- dependence an epistemological twist. On one version of this proposal, being beautiful is a sortal-dependent property if, and only if, necessarily, for any x that is beautiful, there is a sortal concept S such that knowing that x is beautiful requires knowing that x is S.23 No unclear concepts are used in this definition. In particular, the notion of an ideal of beauty is altogether absent. More- over, the definition does not trivialise the issue of whether being beautiful is sortal-dependent.

Acknowledgements

Thanks to Hans Maes and Kathleen Stock for helpful comments on a previous version of this chapter.

Notes

1. In the literature, the term ‘aesthetic property’ is not always applied to the same items. Nonetheless, there are properties that, on pain of displaying a serious misunderstanding of term, one cannot but classify as ‘aesthetic’. Examples include being beautiful, being handsome, being pretty, being graceful and being elegant. 2. In words, ‘necessarily, for each x and each property F in A, if x has F, then there is a property G in B such that x has G, and necessarily if any y has G, it has F’ (Kim, ‘Concepts of Supervenience’, 65; italics in original). This is a definition of ‘strong supervenience’. To get the corresponding definition of ‘weak supervenience’, it suffices to leave out the second (italicised) occurrence of ‘necessarily’. In what follows, the difference between strong and weak supervenience will matter only sporadically. 3. On the notion of a ‘sortal’, see Wiggins, Sameness and Substance Renewed,8–9.My use of the term is more or less guided by the principle that two objects fall under the same sortal if and only if they share their qualitative essential properties (cf. De Clercq, ‘A Criterion of Diachronic Identity’, 29, although ‘qualitative’ is used there in a slightly looser sense). 4. Sibley, Approach to Aesthetics. The relevant papers in the collection are ‘Aesthetic Judgements: Pebbles, Faces and Fields of Litter’ (Chapter 13) and ‘Some Notes on Ugliness’ (Chapter 14).

June 21, 2008 11:11 MAC/NWA Page-200 9780230_220478_11_cha10 Rafa¨el De Clercq 201

5. ‘Kantian’ because the notion of a sortal-specific ideal of beauty already figures in his aesthetics. However, according to Kant, there is just one sortal that has an ideal associated with it, namely ‘man’ or ‘human being’. 6. Sibley, Approach to Aesthetics, 194. 7. Cf. ‘Some things can legitimately be judged beautiful, etc., the adjectives being used predicatively; one need not know what these things are, because the nouns (or concepts) in question set no restrictive standards vis-a-vis` beauty or beauty-giving properties. Other nouns do set standards incorporating notions of appropriateness; with them aesthetic judgements are attributive’ (Sibley, Approach to Aesthetics, 183; italics in original). Also indicative of Sibley’s objectivist conception of aesthetic ideals is this: ‘Picasso’s distorted faces can be objects of considerable beauty; but no woman unfortunate enough to have such a face would have a beautiful face. Standards of facial, equine, or female beauty incorporate lines, shapes, colours and their ordering that, within limits, have to be right or appropriate for a face, horse, or woman’ (183). 8. The notion of a ‘maximally beautiful S’ is used by Malcolm Budd in his interpretation of Kant’s notion of an ideal of beauty. See Budd, The Aesthetic Appreciation of Nature, 43. Note that a non-beautiful object could be maximally beautiful in this sense. 9. This suggestion may seem to contradict Sibley’s well-known claim that aesthetic concepts lack positive (sufficient) conditions of application that can be stated in non-aesthetic terms. However, as is made very clear in Budd, ‘Aesthetic Judge- ments’, 301, Sibley denies the existence of sufficient conditions that are knowable a priori. 10. Roughly speaking, a set of truths regarding the instantiation of a property is primary if, and only if, it is the smallest set from which all the truths regarding the instantiation of that property can be derived. Obviously, if a set of truths is primary in this sense, then it may be said to ‘represent all that is given’ with respect to the instantiation of the property. 11. Given that x and y may be in different worlds, this has to be read as saying:

in some world w1 thereisanxwhich(inw1)isatleastasϕ as any y in any w2. Admittedly, this interpretation raises a problem of how to understand cross-modal relations (which is beyond the scope of this chapter). 12. If there is only a minimum, then positive absolute truths do not seem to follow (automatically, that is, without some stipulation or other). 13. See Sibley, Approach to Aesthetics, 181–183, on pebbles and crystals. 14. Ibid., 183; emphasis in original. 15. Scruton, The Aesthetics of Architecture, 10. 16. Sibley, Approach to Aesthetics, 188–189; emphasis in original. 17. Even when the supervenience relation is restricted to a particular case x so that only those properties are included in the base which are exemplified by x. 18. Wiggins, Sameness and Substance Renewed, 86–87. 19. Scruton, The Aesthetics of Architecture, 10. 20. Cf. Walton’s ‘Categories of Art’ on the aesthetic effect of perceiving certain features as ‘standard’, ‘variable’ or ‘contra-standard’. See also Davies, ‘Aesthetic Judgements’, 228–231. 21. For simplicity’s sake, let us suppose that accidental properties are purely accidental in the sense that they can be exemplified by objects of different kinds. For example, the property of being a man-sitting-in-the-sun-at-noon is accidental or contingent but is not ‘purely accidental’ because only men can possess it.

June 21, 2008 11:11 MAC/NWA Page-201 9780230_220478_11_cha10 202 Aesthetic Ideals

22. De Clercq, ‘Aesthetic Peculiarity’, 413. To be sure, the definition is rough and informal. 23. Apart from the fact that it mentions properties rather than adjectives, the definition is equivalent to the epistemological definition of ‘attributive’. See De Clercq, ‘Aesthetic Peculiarity’, 418.

Bibliography

Budd, M. 1999. ‘Aesthetic Judgements, Aesthetic Principles and Aesthetic Properties’, European Journal of Philosophy, 7: 295–311. Budd, M. 2002. The Aesthetic Appreciation of Nature: Essays on the Aesthetics of Nature, Oxford: Clarendon Press. Davies, S. 2006. ‘Aesthetic Judgements, Artworks and Functional Beauty’, Philosophical Quarterly, 56: 224–241. De Clercq, R. 2005. ‘A Criterion of Diachronic Identity based on Locke’s Principle’, Metaphysica, 6: 23–38. De Clercq, R. 2005. ‘The Aesthetic Peculiarity of Multifunctional Artefacts’, British Journal of Aesthetics, 45: 412–425. Kim, J. 1984. ‘Concepts of Supervenience’, Philosophy and Phenomenological Research, 65: 153–176. Reprinted in Kim, J. 1993. Supervenience and Mind, Cambridge: Cambridge University Press, 53–78. Kim, J. 1993. Supervenience and Mind, Cambridge: Cambridge University Press. Scruton, R. 1979. The Aesthetics of Architecture, London: Methuen. Sibley, F. 2001. Approach to Aesthetics: Collected Papers on Philosophical Aesthetics, Oxford: Clarendon Press. Walton, K. 1970. ‘Categories of Art’, The Philosophical Review, 79: 334–367. Wiggins, D. 2001. Sameness and Substance Renewed, Cambridge: Cambridge University Press.

June 21, 2008 11:11 MAC/NWA Page-202 9780230_220478_11_cha10 11 Configuring the Cognitive Imagination Jonathan M. Weinberg

I. The cognitive imagination: Analogies and mechanisms

My goal here is to advocate for a naturalistic theory of the imagination that concerns the ways that different parts of our mind can be configured to interact with the imagination. This account will serve as a framework for investigating a range of imaginative phenomena of interest to aestheticians, including genre, imaginative blockage, and the range of conative states we take on in engaging with various fictions. But to tell this story about how the imagination interacts with various parts of the mind, I first need to tell the tale of what those parts are, and indeed what the imagination is such that it makes sense to talk about such interactions. Let me start, as they say, at the beginning, with a standard-issue first-draft story of the history of the philosophy of art. For a very long time, everyone thought that the essence of art was its resemblance to the empirical world, for better or for worse (for Plato); then, growing out of romanticism, the essence of art was creative self-expression; then, most recently, the essence of art is not imitative nor expressive, but representational, even capable of representing its own philosophy (if Arthur Danto is to be believed; and here, as usual, he is). Given the intimate connection between art and imagination, it is perhaps unsurprising that thinking about the imagination has undergone a parallel set of developments. The ancients considered imagination primarily pictorially, as a locus of sensory images; Aristotle, in the De anima,makesit out to be a conduit of sorts between perception and cognition. The moderns were more appreciative of imagination’s creative and active capacities, such as in Hume, where imagination helps concoct a world of causes that our senses do not disclose to us, or in Kant, where judgments of taste involve a free play of the imagination and understanding together. And, most recently, some philosophers and psychologists have begun thinking of the imagination less in terms of pictures-in-the-head, and more as a symbolic representational system in the way that belief is a symbolic representational system. Because such a view posits a tight relation between imagining and believing, let us

203

June 21, 2008 11:13 MAC/NWA Page-203 9780230_220478_12_cha11 204 Configuring the Cognitive Imagination call this most recent, mental-symbol-oriented approach, the cognitive theory of the imagination. It merges a person-level account of the imagination that assimilates imagining to something very like believing, with a cognitive scientific account of how both these mental processes work, in terms of computations over tokened symbols, many of which are operated upon by mechanisms that work outside of our introspective awareness (and thus are not likely to show up in our person-level psychology). We can understand these developments in terms of a progression of analogies for how to think about what it is to imagine that p. The old perceptualist views construe imagining that p to be akin to seeing that p; followed by the simulationist line that imagining that p is similar to being a person who believes that p. My view follows the hybrid line in rejecting that imagining that p is like being a believer that p, though it does nonetheless share many important functional roles with having the belief that p. To imagine that p is not to redeploy one’s belief box (BB) off-line, but rather to entertain the content p in a separate and parallel representational system, which I will follow my earlier usage in calling the ‘imagination box’ (IB).1 That we have two separate systems working in tandem, and not just one system that gets redeployed in two functionally different ways, is the central difference between this theory and the simulation theory.2 But what makes this account very similar to a simulation account is that the contents in the IB possess the same format and structure of representations in the BB—they are, as Shaun Nichols has put it, written in the same code.3 Because of this BB/IB similarity, and because computational processes are sensitive to format and structure, representations in the imagination can interact with, serve as inputs to, and be added or removed by various other pieces of psychological machinery in the same way that beliefs do—with a few major and absolutely essential exceptions, as we will see shortly. For my purposes here, the most important systems that interact in the same way with the BB and the IB are (1) inference mechanisms, (2) affect systems, (3) the UpDater, (4) various domain specific processes, including moral evaluation systems, and (5) monitoring systems. The inference mechanisms and affect systems are the most intuitive. In the BB, if one believes IF P THEN Q and also believes that P, then ceteris paribus one’s inferential mechanisms (whichever one is responsible for modus ponens) will token the representation Q in the BB as well. (Of course, if one already has ∼Q in the BB, then matters get more complicated; see the UpDater below.) The core idea here is that when one imagines those contents, the exact same set of processes are engaged: one imagines IF P THEN Q, and imagines P, and the same modus-ponens-performing process gives rise to the presence of Q in the imagination as well. Very similarly, representations of the form SITUATION S IS LIKELY when we do not want S to be the case will produce affective responses like fear or apprehension when they are in our belief box, and I am claiming here that in the exact same way they can trigger those same affect-producing mechanisms when they occur in the IB as well. Mutatis

June 21, 2008 11:13 MAC/NWA Page-204 9780230_220478_12_cha11 Jonathan M. Weinberg 205 mutandis when we do want S to be the case, for happiness or anticipation. It is in this way that we can have affective responses to fictions, which even seem directed at the characters of those fictions. When the situation in question is the imagined plight of Anna Karenina, for example, those Anna representations in the imagination engage systems that produce a pity-type reaction and these are the same systems that produce such a reaction when we read about a real-world person in similar distress, and have formed our beliefs accordingly.4 So the IB and BB enjoy not just analogous inputs and outputs, but the exact same mechanisms are employed in producing them. Another mechanism that operates in the exact same way on both the BB and the IB is the UpDater. This system is posited by Nichols and Stich5 to handle the crucial task of updating our beliefs in the face of new information, in order to render the BB and IB each internally consistent. When we find ourselves in a situation like the one mentioned in the previous paragraph, where the inferential mechanisms are trying to infer Q when we already have a representation that ∼Q in the system, then the UpDater would manage the problem of either overriding the existing ∼Q belief or determining which of IF P THEN Q and P to reject. Such a mechanism is clearly involved when we engage with most fictions, for we regularly update the contents of the IB in the light of new input from a fiction, trying to maintain a coherent picture of the fictional world. The above-mentioned processes all correspond to parts of our mental lives that we have at least some introspective access to. We seem at least sometimes to know that we are making an inference with our beliefs or our imaginations, or that we are having an emotional response to a real or imagined scenario. The domain-specific processes, however, tend to operate below the threshold of consciousness. Although there is a great deal of debate as to whether our cognition is little more than a vast array of specialized ‘modules’, it is fairly uncontroversial to claim that there are at least some domain-specific processors that structure our understanding of the world. We all have ways of thinking about the world in terms of the ways that physical bodies interact, or the ways that biological entities operate in a fundamentally different way from non-biological ones, or for attributing beliefs, desires, and the like to psychological beings like ourselves in order to make sense of and predict their behavior. Moreover, we all seem to acquire such cognitive capacities from a very young age, without any formal schooling in them at all, and generally without ever thinking too much about them. These capacities are thus standardly referred to as our folk physics, folk biology, and folk psychology— ‘folk’ here in contrast to ‘scientific’, since they are not always consistent with what the corresponding scientists have to tell us. (Folk physics seems to be pretty close to the physics of Aristotle, and not at all like the physics of Einstein.) Because these systems tend to operate tacitly, below the threshold of consciousness, we do not usually think about or notice that we have these abilities to reckon with the world in physical, biological, and psychological

June 21, 2008 11:13 MAC/NWA Page-205 9780230_220478_12_cha11 206 Configuring the Cognitive Imagination terms, but they have been closely studied by psychologists.6 And we have good reason to think that, in addition to structuring the way we think about the real world, these domain-specific systems all get deployed in our interaction with fictions as well.7 The exact number and nature of domains for which we have such capacities is still a matter of significant debate, but I will be making heavy use later of the idea that at least some of our moral cognition is of this sort. The monitoring systems provide us with very reliable (though imperfectly and constrained by the occasional case of repression) self-knowledge as to the contents of the IB or BB. I follow Nichols and Stich in supposing them to operate quotationally: they take a representation p in the BB and insert a representation like I believe that p into the BB, or they can take a representation p in the IB and insert a representation like I am imagining that p into the BB. It will also be important to Section IV that the monitoring systems operate on the desire box in a very similar way. In addition to these systems that operate highly similarly on the BB and IB, there are also various systems that operate only on one or the other. (If this was not so, there would be no reason to consider the BB and IB to be different types of system!) For example, perceptual systems insert representations into the BB fairly automatically, whereas it is fairly rare for them to do so with the IB, and even then it is only in a limited way, that is, when we take the contents on a movie screen directly into the imagination.8 Another example of a system that works primarily with the BB is our action-production systems; it is important they are under the control of the BB but not the IB, lest we act out all our fantasies. There are also some systems that operate only with the IB, and I will mention the most important one here: the Inputter. Although we are not free to believe what we want just by choosing to do so, by and large we are free to imagine just about whatever we want.9 Some mechanism or set of mechanisms must be driven by our decision to imagine that p, for nearly any p, and to produce a token of p in the IB. Because doxastic voluntarism is basically false, there is no reason to think that the Inputter can produce representations in the BB as well. We may often, and unfortunately, find ourselves believing what we want to believe, but we do not seem generally able to believe what we decide to believe. We now have the basics of an architecture of the cognitive imagination: what the major pieces are, and how they can most basically interact with each other (including which ones cannot interact with each other). There are the two systems, the IB and BB, which can both interact in exactly parallel ways with inferential, affective, updating, monitoring, and a host of domain- specific systems. The BB interacts in proprietary ways with the perceptual and action-producing, while the IB has its own proprietary source of decision- based input. This architecture provides the framework within which the rest of the chapter will operate. This framework has already proved valuable in proposing accounts of various extant problems in aesthetics, such as fictional

June 21, 2008 11:13 MAC/NWA Page-206 9780230_220478_12_cha11 Jonathan M. Weinberg 207 emotions and imaginative resistance. Meskin and I have approached such problems architecturally, in that we have only made appeal to what can be explained by the functional distinction between the BB and the IB, in terms of the set of resources that is available to be recruited by both or by only one of those representational systems. But one might also consider under what particular circumstances those different mechanisms do or do not get recruited, and in what ways. For it turns out that this sort of naturalistic approach offers not just architectural, but also configurational explanatory resources—not just what the parts are and their functional roles, but the different ways in which they can fit together. My objective, in the rest of this chapter, will be to indicate some further ways in which taking this kind of naturalistic approach can bring further new resources to deploy in addressing various philosophical questions about how we engage with fictions. First, I will use this theory to suggest a framework for a theory of what it is for someone to master a genre. Second, I will consider a possible implication for the epistemology of thought experiments under the hypothesis that philosophical thought experiments are a kind of genre—thought experiments involve the contemplation of propositions in the imagination, after all, so it is not implausible that learning how to deploy thought experiments correctly involves the ability to reconfigure one’s imagination in much the same way that learning how to appreciate works in a given genre does. Third, I will address the question of whether we need to posit yet another box, with desire-like imaginings, in order to make sense of our imagination-involving conative states. Finally, I will use configurability to suggest a new understanding of the mixed ways in which we sometimes and in some ways can, and in others cannot, imagine morally deviant fictional worlds.

II. Toward a theory of genre mastery

Genre is a metacategory, a category of categories. I am not going to try to offer a substantive account here of what genre is, of what makes a given category of artworks a genre category as opposed to some other kind of artistic category; I am not even sure that such an account is possible, as there may be too diverse and disunified a set of categories that get lumped together under that term. However, I do think that there are phenomena involving subclasses of genres, and indeed specific genres, about which useful accounts can be offered. One such phenomenon is the mastery of literary genres. For a work that is a member of a genre G, in order to properly appreciate and evaluate that work, one must have mastery of G. I appeal to the notion of mastery here, because it is not enough simply to possess the concept of G. One can possess a concept in a thin way that allows one to do little more than use it to represent the referent of the concept. If someone watches Night of the Living Dead and offers as a criticism of the film, ‘that could never happen—there are

June 21, 2008 11:13 MAC/NWA Page-207 9780230_220478_12_cha11 208 Configuring the Cognitive Imagination no such things as zombies!’, then we know they simply do not have proper mastery of the horror genre—even if they possess the concept sufficiently to express thoughts like horror movies are so unrealistic. Moreover, the issue here is not one of whether one possesses a concept in only a partial or deferential understanding of the concept. Our clueless horror watcher may even be able to do a good job of distinguishing horror films from non-horror films—they may do so primarily to avoid accidentally watching the former, for example— and yet still not count as mastering the genre. Mastering a given genre is thus not even the same as having mastery of a concept of that genre. And yet we must also avoid requiring that appreciators and evaluators have anything like an explicit theory of G in order to engage successfully with the artwork. Such a theory may be part of one’s mastery, but it seems that a great many appreciators of a given genre will not have one. We do not expect most horror viewers to have an explicit theory of horror, for example, which is in part why books like Noel Carroll’s Paradoxes of the Heart can be informative, even surprising, to those of us who are veteran horror watchers. And, again, one may possess an explicit theory of a genre, even a very good one, and still not be a successful appreciator of the works in that genre. (Note that even if G is a genre of all bad artworks, then successful appreciation of works in G should be possible, namely in appreciating just how bad those works are.) This dissociation from explicit theorizing suggests that genre mastery may be analogous to a skill, just as mastery of driving cars is dissociable in both directions from having a very good explicit theory of driving. But can we say more than just an analogy-deploying gesture in the direction of the knowing how/knowing that distinction? The resources of folk psychology run aground here, but my hope is that the kind of psychological naturalism I am lobbying for here will allow us to do more. Here is my suggestion: mastering a literary genre G requires (and perhaps may be constituted by) learning how to appropriately configure one’s IB when engaging in works of G. At a minimum, this will require three different capacities. First, and perhaps rather obviously, one must be able to recognize works of G. Second, one must tacitly represent the appropriate import and export rules as to what the fictional ground rules are, and as to what real- world knowledge can be brought into the IB from the BB, and indeed what can be brought back to the BB from the fictional contents represented in the IB as well. Third, one must configure which particular subpersonal systems should be brought to bear on the IB while engaging in the work, and which should remain off-line. I will consider each in turn. In recognizing an instance of G, an explicit theory of G may be of use, but again it cannot be a necessary way of doing so. Gregory Currie has suggested in his Arts and Minds that our understanding of genres is at least partially constituted by having a well-trained neural network that maps which characteristics are likely to co-occur with each other or are unlikely to do so; for example, the work’s being set in the American desert badlands in the late

June 21, 2008 11:13 MAC/NWA Page-208 9780230_220478_12_cha11 Jonathan M. Weinberg 209 nineteenth century makes it likely to be a Western, which also makes it likely that there will be a protagonist operating under a certain kind of honor code, and that it will be likely that there will be significant gunplay in the work. And vice-versa: given such a protagonist and such gunplay, a certain kind of setting is made likely as well. None of these are entailments, as Serenity (a Western set in outer space) and Back to the Future III (a sci-fi movie set in the old west) demonstrate.10 Moreover, none of these likelihood relations need be represented consciously at all. The title and director’s bio of John Carpenter’s Vampires would indicate a paradigm member of the horror genre. But we soon see that the film is set in New Mexico, and that the vampire hunters pursue their quarry like a posse after a band of outlaws, and that the tone is one of raucous action more than horrific suspense and moody ambience. In short, we soon see that John Carpenter’s Vampires is not so much a horror film as a Western, albeit a Western with significant supernatural elements. Viewed as a horror film, it is not terribly satisfying: the vampires are fierce but not especially mysterious or seductive in the way that horror fans are used to, and the film has a fairly straightforward action plot, and nothing like the complex discovery plot that Carroll has shown to be typical of the horror genre.11 So a proper appreciation of the film requires that we export the characteristics of the film back to the outer perspective, where we can re-evaluate its generic identity, and thereupon import the axioms for its appropriate genre into the imagination. My suspicion is that the ultimate reconceptualization of the film from horror to Western is driven by, but does not drive, the subpersonal activation of the associated prototype of the relevant genres. The recognition of the genre must then drive at least two kinds of configurations of the IB: what import/export rules it is to have to regulate the passing of contents between the IB and the BB; and what further systems are to be allowed to operate on our imaginative representations and which to be prevented from doing so. To continue with our example of the horror genre, we expect to be able to import into the IB a great deal of our real-world knowledge (as represented for us in the BB). Not all of it, of course—we do not think werewolves really exist, but all bets as to their existence are off in a horror fiction, and once we see that there are, say, vampires, then we will not be scandalized should lycanthropes appear as well. But we do expect the everyday, daylit world of the story to work like our real everyday world, and if someone is not prepared for that, then they have not mastered the horror genre. Once he sees that the protagonists are being threatened by werewolves, if he expects that they might be rescued by a roving squadron of longsword-wielding Dunedain,´ then he has failed to view the work as horror, and is misapplying a fantasy genre instead. And, of course, if the protagonists do turn out to be rescued in such a way, then it is likely that this is a case of a genre switcheroo on the part of the artist,

June 21, 2008 11:13 MAC/NWA Page-209 9780230_220478_12_cha11 210 Configuring the Cognitive Imagination in which part of appreciating the work is to engage with it under the guidance of one’s mastery of one genre for an initial period, before switching at one point, usually involving a surprising plot twist. The Matrix is an excellent example of this phenomenon. It starts off keyed as a contemporary or near- future high-tech thriller along the lines of Enemy of the State, Swordfish,or Strange Days; something with impossibly stylishly-garbed hackers and cool gadgetry, perhaps, but basically operating within the world as we understand it. Of course, once Mr Anderson takes the red pill, we find out that the movie is really a work of far-future dystopian sci-fi, in a world that is nothing like the way ours is (or seems?). This has important consequences for how we understand what we see on the screen: although the ‘bullet time’ special effect initially seems merely a very cool but purely extradiagetic visual device, we come to see that in fact it has representational content. When we see the effect, it means that the rebels of Zion are using their powers as self-aware humans to bend the rules of the matrix.12 We must also understand what to bring back with us from the world of the fiction. Of course, a large number of metafictional contents are always appropriately retrievable—in general, finding p in our IB during a fiction licenses our coming also to have the real belief in the fiction, p (though that will, of course, be revisable if we later come to revise our take on the contents of the fiction). But we very often take ourselves to learn non-metafictional contents as well. For example, we typically expect to be able to export many beliefs about the general contours of world events and the texture of everyday life from works of historical fiction. Some works in the genre of alternative history aim, in part, in arguing for counterfactuals: what would the long-term consequences have been if, say, McClellan had not acquired Lee’s battle plans just prior to the battle at Antietam?13 One may even think of non-fiction as a genre itself, characterized in part by our having a more open setting on this parameter than in works of fiction.14 Third, and perhaps most interestingly for my purposes here, mastering G involves coming to be able to calibrate which of one’s subpersonal psychological systems will interact with the IB while entertaining the contents of works in G, and which will not. There is a recent trend in cognitive science to recognize the importance of the ‘soft assembly’ of complexes of processes and mechanisms that recruit from available resources to create a ‘task-specific device’, and need not be in place when we are not attempting that task.15 Although the focus of such work has been overwhelmingly in the domain of human motion, the idea carries over to less muscular aspects of cognition. I am suggesting that our imaginations have a set of recruitable resources that can be soft-assembled into different configurations in order to perform different tasks for us; and that to be a master a literary genre is, in part, to be able to assemble the right imaginative machine for the job. Another aspect of these soft-assembled task-specific devices is that they typically work by means of coordinating the organism with activity in its

June 21, 2008 11:13 MAC/NWA Page-210 9780230_220478_12_cha11 Jonathan M. Weinberg 211 environment. In a case of, say, running to catch a ball, our posture tilts forward as we unconsciously count on gravity to pull us in the direction we want to go. With fictions, we configure our imaginations in ways that (also often unconsciously) count on confronting appropriate contents and, if it is a film, play, radio drama or the like, appropriate stimuli that in particular will facilitate the activation or suppression of different affective responses. Horror, practically by definition, requires an affective response of a particular sort. Partly this is accomplished by the author’s having created a legitimately scary tale; partly this may be accomplished by stimuli like ominous music or images of strange shadows on a dark night; but partly this has to be accomplished by the reader or audience being appropriately receptive. If we have registered that the work is an instance of the romance genre instead of horror, though, we might view shadows in the darkness with a view towards prowlers more amorous than lycanthropic. Noel Carroll writes of the value of emotions as a source of criterial prefocusing, with different emotions directing our attention to different aspects of the fiction.16 However, we can also consider the other direction—recognizing what genre a work belongs to also directs our attention to the details that might facilitate certain affective reactions, as well as reducing the thresholds needed to trigger such emotions in the first place. Such observations about the role of emotions with genre works are not new. But I want to recontextualize such observations, considering them in tandem with other ways in which we calibrate our deployments of the imagination. In our previous papers, Meskin and I have treated the moral judgment systems as one ‘box’, not because we have thought it to be a unitary system but because our purposes did not yet require any further discrimination. But there are now good scientific reasons to posit at least five different psychological ‘foundations’ of moral cognition, which Jonathan Haidt and other social intuitionists have labeled: Harm/Care, Fairness/Reciprocity, Ingroup/Loyalty, Authority/Respect, and Purity/Sanctity.17 These different psychological systems do not present an inflexible monolith, but rather some are operative at times when others are not, with significant variability across cultures and even across individuals within a culture; for example, political liberals tend to be primarily active with regard to the first two systems, whereas political conservatives tend to use all the five systems more generally, which is why the two populations can talk past each other rather badly on issues such as sexual ethics.18 My conjecture here is that mastery of different genres also can require bringing some of these moral systems more fully online, while forcing others to recede into the background. To the extent that we are able to appreciate certain sorts of Christian religious fictions like the Grail legends, for example, we need to be able to engage them with the Purity/Sanctity system active, or we will be to those fictions what someone is to horror films who simply finds them boring and emotionally unengaging. People who can ‘get

June 21, 2008 11:13 MAC/NWA Page-211 9780230_220478_12_cha11 212 Configuring the Cognitive Imagination into’ ancient epics most likely can do so in part by being able to more fully deploy the Ingroup/Loyalty systems than those who find such stories too alien to be appreciated well. Revenge fictions call on us to deploy the Fair- ness/Reciprocity systems, which are strongly backed by affective responses of anger and contempt. Yet, at the same time, such fictions are constructed so as not to allow the Harm/Care system to make too salient to us the awful- ness of what the protagonist has wrought upon the wrongdoers. Perhaps one difference between works of straight ‘genre fiction’ and more complex non- or trans-generic works is just how restricted and stable is the set of psychological processes that are properly brought to bear on the IB. Neither volume of Kill Bill is a morally complex film, as their moral universe falls entirely along the dimension of Fairness/Reciprocity—the Bride must have her revenge, no matter who gets hurt or what society’s laws might dictate. Alternatively, one key achievement of Unforgiven lies in its getting us to use different parts of our moral psychology at different times, or even at the same time (as when Clint Eastwood’s Munny kills Gene Hackman’s Little Bill). In addition to fictions that call on us to deploy some, but not all, of our possible moral systems, still other fictions require us to suspend the activity of our moral judgment systems altogether. This has frequently been said of Patricia Highsmith’s Ripley novels, for example. In addition to affective and moral judgment systems, the epistemic machinery of our inferential systems and the UpDater are not always appropriately used in the same way across different genres. The UpDater is an essential piece of our epistemic machinery, detecting discrepancies among contents and trying to weed out inconsistencies. The great majority of fictions require its use, and it is especially relevant for obvious reasons to mystery fictions and fictions in which much early information is revealed to be misleading. (The vertiginous phenomenology of the very end of The Usual Suspects is the feeling of your poor, limited, human UpDater straining into the red zone and then melting down altogether.) But some fictions require us to suspend the UpDater’s operations, such as with works of surrealism or theater of the absurd. One is also ill-advised to allow one’s UpDater to work too hard on the broader sort of farces, such as Team America or the Scary Movie franchise. Cartoons, such as many classic Warner Brothers shorts, have resisted for a long time having too much consistency imposed on them, and one can see this aspect of some cartoon genres today in such fare as Aqua Teen Hunger Force or Family Guy. Interestingly, I am not aware of any fictions that call on us to suspend the operation of any of our intuitive physics, biology, or psychology. What one does find, of course, are fictions with alternative scientific laws, but they tend to be ones that are intelligible to the reader primarily by being described metaphorically in ways that engage these intuitive systems. Engaging in such fictions often has a literally ‘magical’ feel to it, which is one phenomenological product of contemplating our intuitive laws of nature being

June 21, 2008 11:13 MAC/NWA Page-212 9780230_220478_12_cha11 Jonathan M. Weinberg 213 violated. When confronted with apparently impossible physical events, such as traveling from one point in space to another without passing through any intervening points, our intuitive physics does not reject the event flat-out so much as flag it as highly unusual, a good candidate for further investigation— probably something with a trick to be discovered but maybe, just maybe, a real case of magic.19

III. Genre and the imagination: A metaphilosophical application

In addition to providing the theoretical machinery for an account of genre mastery, the configurable-imagination framework makes available new resources for debates about the epistemology of thought-experiments. One challenge that has been recently made against the philosophical practice of appeal to intuitions about hypothetical cases is that there is epistemically worrisome variation in such intuitions. For example, Weinberg, Nichols, and Stich20 surveyed Rutgers University undergraduates on a number of epistemological thought experiments, including this variation of the Gettier case:

Bob has a friend, Jill, who has driven a Buick for many years. Bob therefore thinks that Jill drives an American car. He is not aware, however, that her Buick has recently been stolen, and he is also not aware that Jill has replaced it with a Pontiac, which is a different kind of American car.

When asked whether Bob ‘really knows’ or ‘only believes’ that Jill drives an American car, students who were of European descent overwhelmingly (about 80%) said that he only believed. But slightly more than half of the students who were of East Asian descent answered that he really knows! Such studies suggest that, when philosophers appeal to ‘the’ intuition about a case, there may not be sufficient uniformity of intuition to license the definite article; or when they appeal to ‘our’ intuition, it is unclear just who is and is not covered by that first-person plural pronoun. Such unexpected variability in intuitions arguably renders them unsuitable for deployment as evidence for philosophical theses.21 But Ernest Sosa has recently suggested a way that proponents of intuitions may be able to deflect challenge of this sort of data, by appealing to the phenomena of import and export:

When we read fiction we import a great deal that is not explicit in the text. We import a lot that is normally presupposed about the physical and social structure of the situation as we follow the author’s lead in our own imaginative construction. And the same seems plausibly true about the hypothetical cases presented to [the] subjects. Given that these subjects are

June 21, 2008 11:13 MAC/NWA Page-213 9780230_220478_12_cha11 214 Configuring the Cognitive Imagination

sufficiently different culturally ...they may because of this import different assumptions as they follow in their own imaginative construction the lead of the author of the examples.... Perhaps, for example, subjects who differ enough culturally or socio-economically will import ...different background assumptions about how likely it is that an American who has long owned an American car will continue to own a car and indeed an American car.22

As it stands, though, this response can offer little succor to the would-be deployer of intuitions. For suppose that, as Sosa hypothesizes, that we cannot generally tell when two people are considering the same case in the same way, because of unstated differences in importation from the BB. It is true that these studies would no longer demonstrate real divergence in intuitions, since the different intuitions may be about what are, strictly speaking, different cases. But then the same would hold for convergence of intuitions, too! I would not be able to make use of your reported intuition about a case, because I would not know whether you have made relevantly different but unstated importations. Uncharted variation in filling-in is as almost as epistemically dangerous as uncharted variation in the intuitions themselves. Sosa appeals to the idea that the subjects in the experimental philosophy studies were ‘sufficiently different culturally’, but we do not even have the first inklings of a complete picture as to just what cultural differences do or do not produce variant intuitions, nor for that matter do we know what non- cultural differences (e.g., political, personality, or religious) may also produce such differences. We would thus seem to have some reason to expect, and no evidence at all to rule out, the possibility that it is an extremely rare event for two philosophers to consider the same thought experiment in the same way, and thus just as rare for them to ever really share evidence. Yet, what if philosophical thought experiments were a genre—at least in the sense that engaging in them successfully requires mastery along the same lines as I have sketched for the mastery of literary genres? There are rules to engaging properly with a hypothetical scenario, after all. To make just some of the more obvious generalizations about our imaginative practices with thought experiments: one should embellish as little as possible; generally it is a practice conducted in an affectively ‘cool’ manner; and our inferential systems must often be brought to bear in this particular sort of imaginative project as well. And there are surely other, and more subtly articulable, rules for the proper performance of thought experiments still to be detailed. If it is a plausible hypothesis that philosophical thought experimentation constitutes a genre (at least psychologically), then we could reasonably expect that philosophers who have mastered that genre might converge much more closely in their intuitive judgments than will those who lack such mastery. Sosa’s argument may thus be strengthened by appeal to a suitably psychological account of genre mastery.

June 21, 2008 11:13 MAC/NWA Page-214 9780230_220478_12_cha11 Jonathan M. Weinberg 215

Of course, the conceptual availability of this move does not render it thereby empirically available. It would remain to be seen that philosophers’ engagement with thought-experiments really are sufficiently uniform to obviate concerns about variation, and moreover it may well be that ordinary, non- philosopher subjects—or at least the ones who have had a bit of higher education—engage with thought-experiments in basically the same way that philosophers do. Someone looking to run a full-scale version of Sosa’s argument will thus need to do some of their own experimental work first. But my point here is just that, if someone is to do so, he/she would be well advised to consider the psychology of genre in the process of designing his/her study.

IV. Configurability and the variations of imaginative motivation

Although I have been writing here of the samenesses and differences between representations in the BB and those in the IB, I have not yet spoken to the question of desires, and whether or not there is a desire-like version of the IB that parallels the ‘desire box’ (DB) in much the same way that the IB does the BB. Several other theorists have posited the existence of such desire-like imaginings.23 There are a number of different arguments put forward for why we should posit such entities, but I will only engage with two of them here. First, it seems as if we must in some sense be able to come to care about fictional entities. We can, infamously, pity Anna Karenina, or at least do something which we typically describe as ‘pitying Anna Karenina’. Even putting aside the question of whether it is real pity or only a close ersatz like quasi-pity,24 this still seems to require either a real desire or a close ersatz desire that such bad outcomes not befall the poor Anna.25 Many of us also seem with perfect sincerity to express all sorts of attitudes about fictional characters, like, ‘Of course I like Han Solo better! Doesn’t everyone? Luke is such a whiner, after all.’ Such discourse—including sometimes very lengthy and passionate arguments about the comparative merits of, say, William T. Kirk and Jean-Luc Picard, or Eowyn and Arwen,26 or Jane Eyre and Elizabeth Bennet—strongly suggests that many consumers of fiction come to have what are to them meaningful conative attitudes towards persons that they know not to exist. Yet it can seem odd to attribute people real, full-fledged desires concerning persons of such questionable ontological status. One resource frequently appealed to here is the real, full-fledged desires about the artwork itself, which is, of course, a perfectly good real-world object about which to have propositional attitudes. Debates superficially about Kirk and Picard may really be debates about the acting chops of Shatner and Stewart; arguments about Jane versus Lizzy may be serving as a stand-in for

June 21, 2008 11:13 MAC/NWA Page-215 9780230_220478_12_cha11 216 Configuring the Cognitive Imagination what is really an argument about the merits of their respective novels. Yet, as Currie and Ravenscroft note,

[w]hen I am sorry and upset about the fate of Desdemona, I am not sorry that this fiction has it that an innocent and good-hearted girl suffers a cruel fate. One might be sorry about that, deploring that there are fictions with such unhappy outcomes. That is not what at least many of us are sorry about; we are glad that Shakespeare’s fiction has it this way.27

Our fiction-directed cognition can point off in different directions than our fictional-character-directed cognition.28 This is, of course, closely parallel to the relationship between the contents of the BB and the IB. I do not believe that Desdemona exists; but I imagine that she does. So perhaps similarly, I do not have a real desire that Desdemona not suffer a cruel fate, but I do in some imaginative sense desire that she not do so. These have been called ‘i-desires’ (Doggett and Egan) or ‘make-desires’ (on analogy with ‘make-believe’; Currie) or simply ‘desire-like imaginings’ (Currie and Ravenscroft). I will follow this last usage. Perhaps real desires that take entities known to be fictional as their objects would seem something of a metaphysical oddity. But, looking at the question from the point of view of naturalistic psychology, there is little reason not to allow that we have ordinary, real desires that take fictional characters as their objects. Human psychology is not particularly careful with regard to its implicit ontology, or about consistency across all of its commitments, which is why psychologists have been so interested in documenting the conflicts between intuitive, folk physics and explicit physical beliefs, even in trained scientists.29 So metaphysical queerness is not, by itself, enough of a reason to add desire-like imaginings to our psychology. There is another argument for positing a desire-like form of the imagination, however. Even if we do not need them to make sense of how we can care for fictional people at all, we may need them to make sense of how we can care for fictional people in ways that depart significantly from our ordinary patterns of care. Some of the fictional characters that we like are just like people that we would like in the real world, and some of the outcomes that we wish for in fictions are just like the ones we would wish for in the real world—but some, interestingly, are not. For example, a person may be a committed opponent to the death penalty in the real world, yet enjoy fictions that require us to want the serial-killing villain to get lethal just desserts. Currie has called this the ‘problem of personality’, and as he writes,

Just as we react to fictions in ways that seem to depend on beliefs that we do not really have ...so we react to fictions in ways that seem to depend on desires we do not have. This can manifest itself in interesting

June 21, 2008 11:13 MAC/NWA Page-216 9780230_220478_12_cha11 Jonathan M. Weinberg 217

ways: peaceful folk can seem, even to themselves, to desire violent revenge to be visited upon a fictional character, and lovers of justice can desire (so it seems) likeable rogues to escape the punishment they deserve. It is plausible that there are people who simulate, for the duration, desires they do not really possess.30

Our real desires are not good candidates for explaining such variation in our fiction-directed conative states. It is possible for extraordinary artworks to forge permanent alterations in what we desire and value—that is part of the life-changing power of great art. But the sort of phenomenon we are looking at here are transient shifts that even B-movie schlock can induce. But desire-like imaginings are problematic explanantia here as well. A key part of the functional profile of the IB is that the InPutter can insert contents into it at will (modulo cases of imaginative blockage). For almost all propositions p, we can imagine that p simply by choosing to do so. But the space of possible alternative conative states we can take on is much more restricted, and even for the ones that we can take on, we cannot do so anywhere near as effortlessly as with belief-like imagining. We cannot watch Rope through to the end and feel it an unfortunate tragedy that Rupert foiled the boys’ fun little plot; nor can we watch Othello in a way that involves our rejoicing in Desdemona’s death.31 Here is where the configurability of the IB comes in. We can adjust the desire-like aspects of our imaginative ‘persona’ not by tokening representations in a desire-like analogue of the IB, but rather by adjusting which affective, moral, or other conative systems interact with the IB at a particular time. As noted above, we are able to engage with works whose moral universe diverges from our own considered views by making a different use of the moral psychological hardware that is already there. The configurability theory can thus capture both explananda that the desire-like imagining theory could not. First, it predicts the comparatively narrow contours within which our imaginative conation can vary. Revenge tragedies, ancient epics, and so on each engage a potential human moral psychology that we can operate with, at least briefly, by using some combination of Haidt’s ‘foundation’ systems. But, because none of those systems produces an approval of the frivolous taking of innocent life32 to amount of reconfiguration will let us take on an appropriate moral psychology to sus- tain rooting for the murderers in Rope to get away with their crime. The rare fictions that do let us take morally odious characters at least sometimes as protagonists, such as with the first half of Rope or with, say, Lolita, typically do so by leading us to suspend moral engagement altogether, and not by taking on the character’s perverse morality. The desire-like imagining theory needs a special explanation for why we cannot stipulate such imaginings as unrestrictedly as with belief-like imaginings, but the configuration theory has no such need.

June 21, 2008 11:13 MAC/NWA Page-217 9780230_220478_12_cha11 218 Configuring the Cognitive Imagination

Second, the monitoring systems report on the contents of the BB, the IB, and also of the DB, and so we would expect that if there were a desire-like imaginings box, the monitoring systems would be able to report just as easily on its contents, too. This exception is particularly poorly motivated, because we do seem to have fairly good access to what our ordinary non-imagined desires are. (Don’t be faked out by interesting-but-rare Freudian cases here; there are such cases for beliefs too, after all. Think instead of common-but- philosophically-boring cases like your explicit desire to get home in time for supper, or to change channels during an annoying used car commercial, or to finish off the paper that you promised to the editor. This great bulk of our desires is very easy for us to monitor.) But the configuration of the affective and other systems is a fact about the IB at a given time, without in any way being implemented by any token in the IB. The configuration is not itself a belief, or a token of imagination. There is thus no reason to expect that we should have particularly good introspective access to these configurational states of the imagination, least of all from the monitoring system, since it only tracks the explicit contents of the BB and the IB. However, it is telling that proponents of desire-like imaginings find themselves needing to argue for the existence of the mental state whose existence they are arguing for. This may sound almost tautological, yet contrast desire- like imaginings with beliefs, desires, and belief-like imaginings: all of these other mental states are so vividly available to introspection that philosophers and psychologists have not seen their existence as standing in much need of being argued for.33 Relatedly, these other states all figure into our folk psychology, whereas desire-like imaginings do not. A likely explanation here is that the monitoring systems operate easily over much of both beliefs and belief-like imaginings, and also over desires, and yet for some strange reason fail to operate over desire-like imaginings. The desire-like imagining theorist is thus committed to an explanatory oddity that the configurability theorist is not. So imaginative configurations both give us a very good fit for both the range of conative states that we can take on in the imagination, and are not the sorts of states that we should expect to have good introspective access to. Neither is true of desire-like imaginings. This hardly proves the non-existence of desire-like imaginings, of course. Yet it does indicate that as yet we have no good reason to posit their existence.

V. Conclusion and directions for future research

There is much to be gained for aesthetics if it takes a naturalistic approach to the mind seriously. To do so does not at all mean outsourcing our problems to experimental psychologists, nor that all aestheticians must now develop a mastery of the scientific psychological literature. But it does mean getting to look for explanatory resources wherever we can find them—and often, they

June 21, 2008 11:13 MAC/NWA Page-218 9780230_220478_12_cha11 Jonathan M. Weinberg 219 will be outside the purview of folk psychology and introspection. This theory is not meant at all to displace person-level psychology, which works fine as far as it goes; and just because it is more or less restricted to the workings at the conscious surface manifold of our mental lives, there is still significant room for subtle and important nuances to be illuminated there, as indeed artists frequently do.34 But regarding the causal mechanisms that operate underneath that surface, these folk psychological explanatory resources are rather thin. And since it turns out that much of the imagination’s action takes place at that level, we must expand our explanatory resources beyond those of our folk heritage. I have suggested here, building on my work with Aaron Meskin (which built on the naturalistically minded work of Stephen Stich and Shaun Nichols, which itself had built on the naturalistically minded work of Gregory Currie and Kendall Walton), that taking such an approach toward the imagination not only allows for architectural explanations of various imaginative phenomena, but also for configurational explanations as well. I have suggested here that such psychological resources may be particularly relevant to developing a theory of genre mastery, and to explaining the particular sorts of conative variation we are—and are not—able to display in our engagements with many fictions. I will close briefly with three directions for further research. First, one aspect of our engagement with fictions that may seem tricky for a naturalistic approach is the normative. The ways we fill in, respond emotionally to, and morally evaluate the situations within stories are not just things that we do, but also frequently activities that can be done correctly or incorrectly, better or worse. I do not think that this normativity itself is something that should be ‘naturalized away’, and turned into a strictly descriptive psychological matter. However, psychology will be essential in helping us understand how we cognize those norms, how we learn them, and how we are guided by them.35 Such knowledge will be particularly valuable as part of the complete story about the complicated relationship between aesthetics and ethics. Second, the list of psychological systems canvassed here should by no means be expected to be exhaustive. Further exploration of sub- or unconsciously operating systems will extend the resources available to this approach. Tamar Szabo´ Gendler’s recent work on ‘aliefs’ is very suggestive in this regard. The promise of a psychologically naturalistic approach to the imagination is thus an open-ended one. Third, since this general architectural approach does not invoke anything particular about narrative, it leaves open the interesting possibility that at least some fictions are enjoyed not so much for their stories but for the fictional worlds themselves. Some fictions provide pleasure simply in allowing us to encounter other worlds, as a sort of metaphysical tourism.36 This is a not-infrequent component of evaluation of works in the fantasy genre, for example. This account can help give an account of world appreciation, in

June 21, 2008 11:13 MAC/NWA Page-219 9780230_220478_12_cha11 220 Configuring the Cognitive Imagination terms of the particular configuration of the imagination that particular fictional worlds call for from us and, typically with successful works, aid us in attaining. Part of the pleasure of a work like the Iliad may be its capacity to help us experience its world appropriately. So when we enjoy the distinctive flavor of a particular fictional world, it may in large part be the reconfiguration of our own imaginative capacities that we are particularly savoring.

Notes

1. Weinberg and Meskin, ‘Puzzling over the Imagination’. We do not now follow the usage of Nichols and Stich, ‘Cognitive Theory of Pretense’—upon which this two-box architecture is heavily dependent—and our own ‘Emotions, Fiction, and Cognitive Architecture’ in calling it the ‘possible worlds box’, because it is important to our overall story that it can hold impossible sets of contents. 2. See Weinberg and Meskin, ‘Imagine That’, for our arguments based on the aesthetics of fiction as to why one should prefer two boxes over one. 3. Nichols, ‘Imagining and Believing’. 4. Whether the results of having those mechanisms count as genuine emotions is a question that Aaron Meskin and I try to offer a principled-but-fence-sitting answer on in our ‘Emotions, Fiction, and Cognitive Architecture’. 5. Nichols and Stich, ‘Cognitive Theory of Pretense’. 6. For evidence for the existence of such systems, see, for example, Atran, ‘Folk Biol- ogy’; Baron-Cohen, Mindblindness; Leslie, ‘ToMM and ToBy’; McCloskey, ‘Intuitive Physics’; Premack, ‘Infant’s Theory’. All that is needed for our purposes is that these systems exist independently from any particular set of representations in the BB, and thus that they do not operate in the imagination as a set of representations in the IB. There are active ongoing debates as to whether such systems should be construed as either modular or innate, and this chapter is neutral with regard to such debates. Also, we should note that one of the standard methodologies for investigating such mechanisms is by having subjects make judgments about hypothetical cases; we can therefore be confident that these systems are used not just on the BB, but on the IB as well. 7. ‘The kinds of details that authors rely on audiences to supply come in all different shapes and sizes, ranging from facts about human biology to facts about geography, history, politics, religion, and so on. In many cases, the author relies upon what we know or believe about human psychology in order for her narrative to be intelligible’ (Carroll, Philosophy of Mass Art, 321). Our point is that it is not simply factual knowledge that is required—the work of various domain-specific systems that underwrite our ordinary capacities to judge in these domains is also required. 8. See Weinberg and Meskin, ‘Puzzling over the Imagination’, for a brief discussion of the possibility of ‘streaming’ inputs into the IB. 9. Modulo cases of imaginative blockage; see below. 10. Though it is the case that mastery of both the Western and sci-fi genres may be necessary for an appreciation of those films. 11. Carroll, Paradoxes of the Heart, especially ch.3. 12. Some artists are less adept at pulling this trick off. For example, Event Horizon starts off sci-fi and turns out to be horror, but somehow the transition does not quite succeed. The is at least in part due to just plain bad writing in the horror

June 21, 2008 11:13 MAC/NWA Page-220 9780230_220478_12_cha11 Jonathan M. Weinberg 221

half—at the point that Sam Neill says, ‘It was a dimension of pure chaos ...a dimension of pure evil’, all aesthetic hope perishes—but also the initial half of the film fails to prepare the audience for the kind of content and images that are to come, and especially for their affective consequences. (See below on the importance of affective calibration for genres). The Matrix, however, crosses from one sort of mystery to another, and the basic sci-fi premise is, in its way, just an intensified and metaphysicalized version of a premise of every mystery: that nothing is quite as it seems. The audience’s transition is also facilitated by our watching the character of Mr Anderson/Neo go through his parallel transition. In Event Horizon, the character that the audience initially expects to follow as the main character (Neill’s Dr Weir) becomes demonically possessed, ultimately transmogrified into something of a spokesman for the hellish spaceship, and the audience has a rather unfacilitated switchover from him to Laurence Fishburne’s character, Capt. Miller, as the protagonist. 13. As is indeed the premise of Harry Turtledove’s How Few Remain and its sequels. 14. This last idea is greatly indebted to conversations with Stacie Friend, as well as hearing her 2006 talk ‘Imagining Fictions’ at the Conference on the Imagination and Thought Experiments at the University of Bristol. 15. For example, Bingham ‘Task-Specific Devices’; Clark, Being There; Thelen and Smith, Dynamic Systems Approach. 16. Carroll, Philosophy of Mass Art. 17. Haidt, ‘New Synthesis’; Haidt and Joseph ‘The Moral Mind’; see also Blair et al., ‘Neuro-cognitive Systems’. 18. Haidt and Graham, ‘When Morality Opposes Justice’. 19. See Harris, Work of the Imagination, for discussion of how even very young children treat ‘magic’ as an explanation distinct from physical explanation, and indeed as an explanation only of last resort. 20. Weinberg et al., ‘Normativity and Epistemic Intuitions’. 21. See Alexander and Weinberg, ‘Experimental Philosophy’, for further discussion of the implications that experimental philosophy may have for philosophical intuitions. 22. Sosa, ‘Defense of Intuitions’. For an extended argument for the claim that philosophical thought-experiments might be properly treated as fictions, see Ichikawa and Jarvis, ‘Thought-experiment intuitions’. 23. Currie and Ravenscroft, Recreative Minds; Currie, ‘Imagination as Motivation’; Doggett and Egan, ‘Wanting Things You Don’t Want’. 24. Walton, ‘Fearing Fictions’, and Walton, Mimesis. 25. Currie and Ravenscroft, Recreative Minds, 20. 26. At the time of writing, there was at least one entire Web site dedicated entirely to the proposition that Aragorn should have chosen the former over the latter. So as not to prejudice the reader, the author will keep his own views on this question to himself. 27. Currie and Ravenscroft, Recreative Minds, 21. 28. Shaun Nichols has suggested in his ‘Review of Currie and Ravenscroft’ that we may simply have conflicting desires about the fiction. 29. Singh, ‘When Physical Intuition Fails’. 30. Currie, ‘Imagination and Make Believe’, 261. 31. See Carruthers, ‘Why Pretend?’; Nichols, ‘Review of Currie and Ravenscroft’. 32. At least, not in non-pathological minds!

June 21, 2008 11:13 MAC/NWA Page-221 9780230_220478_12_cha11 222 Configuring the Cognitive Imagination

33. Except in philosophically extreme circumstances, when fending off the predations of eliminative materialists, and even there the dialectical presumption is in favor of their existence. 34. See, for example, Stock, ‘Fiction and Psychological Insight’. 35. Sripada and Stich, ‘Framework’. 36. Harold, ‘Value of Fictional Worlds’.

Bibliography

Alexander, J. and Weinberg, J. 2007. ‘Analytic Epistemology and Experimental Philosophy’, Philosophy Compass, 2: 56–80. Atran, S. 1998. ‘Folk Biology and the Anthropology of Science: Cognitive Universals and Cultural Particulars’, Behavioral and Brain Sciences, 21: 547–609. Baron-Cohen, S. 1995. Mindblindness: An Essay on Autism and Theory of Mind, Cambridge, MA: MIT Press. Bingham, G. 1988. ‘Task-specific Devices and the Perceptual Bottleneck’, Human Movement. Science, 7: 225–264. Blair, J. Marsh, A.A. Finger, E. Blair, K.S. and Luo, J. 2006. ‘Neuro-cognitive Systems Involved in Morality’, Philosophical Explorations, 9.1: 13–27. Carroll, N. 1990. Paradoxes of the Heart, or, A Philosophy of Horror, London: Routledge. Carroll, N. 1998. Philosophy of Mass Art. Oxford: Oxford University Press. Carruthers, P. 2006. ‘Why Pretend?’ in S. Nichols (ed.), The Architecture of the Imagination, Oxford: Oxford University Press, 89–111. Cavell, S. 1979. The World Viewed, Cambridge: Harvard University Press. Clark, A. 1997. Being There: Putting Brain, Body and World Together Again, Cambridge, MA: MIT Press. Currie, G. 1995. ‘The Moral Psychology of Fiction’, Australasian Journal of Philosophy, 73: 250–259. Currie, G. 2001. ‘Imagination and Make-Believe’ in B. Gaut and D. McIver Lopes (eds), Routledge Companion to Aesthetics, London and New York: Routledge. Currie, G. 2002. ‘Imagination as Motivation’, Proceedings of the Aristotelian Society, 102.1: 201–216. Currie, G. and Ravenscroft, I. 2002. Recreative Minds: Imagination in Philosophy and Psychology, Oxford: Oxford University Press. Doggett, T. and Egan, A. 2007. ‘Wanting Things You Don’t Want’, Philosopher’s Imprint. Gendler, T.S. 2000. ‘The Puzzle of Imaginative Resistance’, Journal of Philosophy, 94: 55–81. Gendler, T.S. (forthcoming) 2008. ‘Alief and Belief’, Journal of Philosophy. Haidt, J. 2007. ‘The New Synthesis in Moral Psychology’, Science, 316: 998–1002. Haidt, J. and Graham, J. 2007. ‘When Morality Opposes Justice: Conservatives have Moral Intuitions that Liberals may not Recognize’, Social Justice Research, 20: 98–116. Haidt, J. and Joseph, C. 2007. ‘The Moral Mind: How Five Sets of Innate Intuitions Guide the Development of Many Culture-Specific Virtues, and Perhaps Even Mod- ules’ in P. Carruthers, Laurence, S. and Stich, S. (eds), The Innate Mind, Vol. 3: Foundations and the Future, Oxford: Oxford University Press, 367–391. Harold, J. Unpublished. ‘The Value of Fictional Worlds; Or, Why, Lord of the Rings is Worth Reading. Harris, P. 2000. The Work of the Imagination, Oxford: Blackwell. Ichikawa, J. and Jarvis, B. (forthcoming) ‘Thought-experiment intuitions and truth in fiction’, Philosophical Studies .

June 21, 2008 11:13 MAC/NWA Page-222 9780230_220478_12_cha11 Jonathan M. Weinberg 223

Leslie, A. 1994. ‘ToMM, ToBy, and Agency: Core Architecture and Domain Specificity’ in L. Hirschfeld and S. Gelman (eds), Mapping the Mind: Domain Specificity in Cognition and Culture, New York: Cambridge University Press, 119–148. McCloskey, M. 1983. ‘Intuitive Physics’, Scientific American, 248.4: 122–130. Meskin, A. and Weinberg, J.M. 2003. ‘Emotions, Fiction, and Cognitive Architecture’, British Journal of Aesthetics, 43: 18–34. Nichols, S. 2004. ‘Imagining and Believing: The Promise of a Single Code’, Journal of Aesthetics and Art Criticism, Special Issue on Art, Mind, and Cognitive Science, 62: 129–139. Nichols, S. 2004. ‘Review of G. Currie and I. Ravenscroft, Recreative Minds’, Mind, 113: 329–334. Nichols, S. and Stich, S. 2000. ‘A cognitive theory of pretense’, Cognition, 74: 115–147 Stich, S. 1992. ‘What is a Theory of Mental Representation?’ Mind, 101: 243–261. Perkins, V.F. 1972. Film as Film, New York: Da Capo Press. Premack, D. 1990. ‘The Infant’s Theory of Self-Propelled Objects’, Cognition, 36: 1–16. Scribner, S. 1977. ‘Modes of Thinking and Ways of Speaking: Culture and Logic Recon- sidered’ in P.N. Johnson-Laird and P.C. Wason (eds), Thinking: Readings in Cognitive Science, New York: Cambridge University Press, 483–500. Singh, C. 2002. ‘When Physical Intuition Fails’, American Journal of Physics, 70.11: 1103–1109. Sosa, E. forthcoming. ‘A Defense of Intuitions’ in M., Bishop and D. Murphy, (eds) Stich and His Critics, Oxford and Boston: Blackwell Publishers. Sripada, C. and Stich, S. 2007. ‘A Framework for the Psychology of Norms’ in P. Car- ruthers, S. Laurence and S. Stich (eds), Innateness and the Structure of the Mind,Volume 2, New York: Oxford University Press, 280–302. Stock, K. 2006. ‘Fiction and Psychological Insight’ in M. Kieran and D. McIver Lopes (eds), Knowing Art: Essays in Aesthetics and Epistemology, Dordrecht, The Netherlands: Springer, 51–66. Thelen, E. and Smith, L.B. 1994. A Dynamic Systems Approach to the Development of Cognition and Action, Cambridge, MA: MIT Press. Walton, K. 1978. ‘Fearing Fictions’, Journal of Philosophy, 75.1: 5–27. Walton, K. 1990. Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Warshow, R. 1964. The Immediate Experience, Cambridge: Harvard University Press. Weinberg, J.M. and Meskin, A. 2005. ‘Imagine That’ in M. Kieran (ed.), Contemporary Debates in Aesthetics and the Philosophy of Art, Malden, MA: Blackwell, 222–235. Weinberg, J.M. and Meskin, A. 2006. ‘Puzzling over the Imagination: Philosophical Problems and Architectural Solutions’ in S. Nichols (ed.), The Architecture of the Imagination, Oxford: Oxford University Press, 175–202. Weinberg, J., Nichols, S., and Stich, S. 2001. ‘Normativity and Epistemic Intuitions’, Philosophical Topics, 29: 429–460.

June 21, 2008 11:13 MAC/NWA Page-223 9780230_220478_12_cha11 12 Personifying Art Brian Soucek

Stanley Cavell once wrote that ‘The answer to the question ‘‘What is art?’’ will in part be an answer which explains why it is we treat certain objects, or how we can treat certain objects, in ways normally reserved for treating persons.’1 The claim is striking, not least because it treats the personification of art as a given, if one in need of explanation or justification. Though it is not always acknowledged, the treatment of artworks ‘in ways normally reserved for treating persons’ is a commonplace both in philosophy and criticism, in our everyday and more theoretical responses to art. It emerges in banalities like ‘the music is sad,’ or ‘that piece speaks to me.’ It finds deeper (or at least more mystical) expression in discussions of the artwork’s ‘soul’ and ‘body,’ whether they be in Kant, Hegel, Dewey, or Danto,2 as well as in descriptions of art’s life, death, and even afterlife.3 Benjamin’s ‘aura’—to perceive which is to invest an object ‘with the ability to look at us in return’4—provides a clear instance of personification in the visual arts, as, more recently, does Roger Scruton’s ‘indispensable metaphor’ about music that ‘the notes in music follow one another like bodily movements.’5 When Nelson Good- man asks ‘how buildings mean,’ or Robert Pippin asks how paintings can be authentic in the existentialist’s sense, not just the forger’s, both are importing into the art world categories native to personhood.6 The centrality of autonomy in post-Kantian aesthetics as well as the insistence on artworks as ends rather than means both point even more significantly to personification’s reach. Cavell is right, then, to highlight this and ask how it can be so. But he does still more: Cavell suggests that we should look to the metaphysics of art if we want to discover why art is personified. The very definition of art, he predicts, will reveal the source of art’s peculiar treatment and value. As far as I know, no one has ever made good on Cavell’s prediction; no one has explained what art is in a way that justifies its personification. Nor will I. In fact, I think the attempt to do so is misguided—though suggesting why that is will constitute much of the argument to follow. I believe Cavell is

224

June 21, 2008 11:20 MAC/NWA Page-224 9780230_220478_13_cha12 Brian Soucek 225 right to claim that we treat artworks ‘in ways normally reserved for treating persons.’ In fact, some of these ways of treating art have pervaded aesthetics for almost as long as aesthetics has been a subject for philosophy. But whereas Cavell seeks to ground such personification in the nature of art itself, I want to argue the reverse: that our understanding of art’s nature is more the result of our long-standing practices of personification. Doing so will require me first to identify just what these personifying practices are. I begin this in the following section by discussing what they are not. In the two sections after that I argue that accounts, first, of how we value art and, second, of how we understand its meaning both make frequent recourse to personifying language—language which is then either supported or resisted on the basis of some argument about the metaphysics of art. It is this latter strategy which I want to resist, and I do so by exploring an alternative: an argument about personification, suggested by Kant and extended by Schiller, which refuses to base our treatment of art on any metaphysical claim about artworks themselves. Understanding their argument and exploring its implications for more recent philosophies of art is the work of the final two sections of the chapter. Ultimately, my concern is to trace the way a justification of art’s personification might proceed—and to discover what would be at stake in those proceedings.

I. What personification is not

Three decades ago, an author writing in The Journal of Aesthetics and Art Criticism claimed that ‘aesthetic theory ...has given rise to an analogy which seems to be gaining acceptance: a work of art is like a person.’7 Judging certain recent books merely by their covers, it seems that acceptance has only grown—or at least spread beyond philosophy. The very titles of W.J.T. Mitchell’s What Do Pictures Want? and Nancy Armstrong’s How Novels Think demonstrate the trend, as does that of Lorraine Daston’s collection Things That Talk, especially its essay on Pollock entitled ‘Talking Pictures.’ The title of Michael P. Steinberg’s Listening to Reason is less explicit, but the jacket makes clear just who (or perhaps what?) is doing the listening: classical Germanic music’s ‘embodiment of subjectivity,’ we are told, ‘involved its apparent capacity to ‘‘listen’’ to itself, its past, its desires.’8 Symphonies listen, canvases talk, novels think, and pictures want. Some- one reading these books together might wonder whether these claims are interchangeable. In other words, we might ask not just what it means to ascribe thought, agency, or expression to a particular art, but whether such ascriptions can only be made to a particular art. Is personification unique, say, to images, to British novels, or to music of the ‘long nineteenth century’? In the books just mentioned, discussion of personification is limited to a single art or era; it is unclear whether the authors thereby mean to suggest that personification itself is limited to that art or era. I take the similarity

June 21, 2008 11:20 MAC/NWA Page-225 9780230_220478_13_cha12 226 Personifying Art of their claims across diverse fields as evidence that it should not be so limited. This suggests a first thing personification is not: it is not unique to any one of the arts. I mean this not as proof but as a working goal in what follows, where I will try to offer an explanation for personification across the arts. That this goal is a specifically philosophical one suggests why it is not shared by any of the books just mentioned, all of which come from outside of the field of philosophy. On the other hand, this goal has seldom been made explicit even within philosophy. One notable, if brief, exception comes from the work of Garry Hagberg, who uses Charles Taylor’s capacities of personhood as a standard against which to test the purported analogy between persons and artworks.9 For Hagberg, the analogy is strong: all of the claims already suggested—that artworks can speak, listen, think, and desire—find support in his argument. As he maintains, the work of art ‘is without question a bearer of rights’; it (often) exhibits an ‘internally generated coherence’ akin to Taylor’s ‘sense of self’; the artwork can, in his words, ‘frame representations of things,’ ‘hold values,’ and exhibit a ‘capacity to respond’ when addressed; more, the work, like a person, is an appropriate object of address, even of love—Hagberg quotes Eva Schaper’s belief that ‘the closest analogue to the regard for an object of aesthetic preference is that of the love in which one person can hold another.’10 Hagberg freely admits that artworks do not, say, respond or hold values quite as persons do. Yet the two ways, he writes, are ‘no farther than analogy is from identity.’11 This response is apt; the problem with Hagberg’s discussion is that he too sometimes forgets that analogy as opposed to identity is the goal. Discussing the possibility, for example, that works of art might adopt a ‘life-plan,’ Hagberg gives instances of artworks figuring in the devel- opmental trajectory—that is, the life-plan—of the artist. Similarly, to Taylor’s requirement that persons can revise their beliefs and values, Hagberg notes that artworks often prompt such revisions in us. This is clearly true; still, the point is not that artworks can figure in persons’ life-plans or cause revisions in their values. The question is whether artworks might have life-plans or values—or something importantly analogous. That is a question Hagberg leaves open. His work thus delineates a variety of ways in which art is personified. But it also inadvertently alerts us to further things that personification is not. Personification does not involve attributing personal characteristics to the artist—who, after all, is literally a person. Personification involves the attribution of such characteristics to the artwork itself. Questions surrounding personification might include whether artworks have something like desires, not just result from them, or whether it makes sense to speak of artworks not just as expressions of an artist but as themselves expressing something. It should be noted further that these questions are not about persons, often imaginary ones, within works of art. Personification involves Oliver Twist (the novel), not Oliver Twist (the character).12 The question here

June 21, 2008 11:20 MAC/NWA Page-226 9780230_220478_13_cha12 Brian Soucek 227 is what it might mean to attribute agency, say, to a novel—not to a fictional person within it. Thus circumscribed, the ways in which we treat artworks like persons begin to become clearer. I want now to suggest that they can be divided into two broad and related categories: our ways of valuing art and our ways of interpreting it.13

II. Valuing art

The distinctive ways artworks are valued have already begun to emerge. Eva Schaper’s comparison of aesthetic preference and interpersonal love, cited approvingly by Hagberg, resonates with Richard Wollheim’s suggestion that ‘A sense of what is special about art depends on our recognizing that art is, strangely enough, something that we humans can love.’14 The theme of love—of artworks and persons alike—arises more pointedly, if perhaps unexpectedly, in debates about forgeries in the visual arts. It is invoked, for example, by Mark Sagoff, whose argument for the value of ‘the particular, substantial, actual’ work over its copies includes this claim: ‘There is an analogy with love. Love attaches to individuals and not simply to their qualities or to the pleasures they give.’15 It recurs in Francis Sparshott’s memorable suggestion that ‘The frame of mind in which one asks ‘What is wrong with forgery?’ is the same in which one asks ‘What is wrong with free sex?’—for, after all, one cannot say how there could be anything wrong with pure pleasure, provided that no personal bond is implied or violated.’16 And this confluence of art and persons, love and sex, appears once more in a statement by Arthur Danto, that appreciating an unacknowledged copy of a painting is ‘like making love to a ghost who happens precisely to resemble the one we believe we are making love to.’17 Danto returns to this theme at greater length elsewhere, and the strategy his argument about forgeries employs is particularly relevant here. In an essay entitled ‘Moving Pictures,’ Danto asks whether a painting’s status as a copy should make any difference to our appreciation of it. By way of response, he immediately goes on to write:

Consider, for example, the possibility of duplicating persons. Suppose a man is killed in an automobile accident, but the widow is promised delivery in, say, three weeks of someone exactly like her husband in all obvious respects. Would it matter? Is she required to love, honor, and cherish the exact simulacrum of her husband, or what?18

What is striking in this argument is the abrupt shift from a discussion of artworks to one about cloned humans. We might wonder what a thought- experiment about persons can tell us about copied paintings—unless, of course, we assume some analogy between the two.

June 21, 2008 11:20 MAC/NWA Page-227 9780230_220478_13_cha12 228 Personifying Art

Danto himself does exactly that. ‘There is something like a parallel between what one might call the metaphysics of persons and the metaphysics of artworks,’ he claims. As he then explains the parallel,

Persons have bodies as parts of themselves, and works of art have physical objects as parts of themselves. ...An artwork is then a physical object with whatever in the philosophy of art corresponds to the soul in the philosophy of the person.19

Danto identifies the ‘soul’ of the artwork with its meaning, a suggestion with notable precedents. Hegel says nearly this in his Lectures on Aesthetics when he writes of the work of art having ‘an inner life, feeling, soul, a content and spirit, which is just what we call the significance of a work of art.’20 In the following century, John Dewey sounds a similar theme with his claim, in Art as Experience, that a painting’s meaning is ‘either, and equally, the soul of which colors are the body or the body of which colors are the soul.’21 The ways we personify when interpreting such meaning is the subject of the next section. Even within this discussion of art’s value, however, the analogies between artworks and persons continue to deepen. Valued as objects worthy of love, both are ascribed an individuality and uniqueness in need of our fidelity and defense. (The uniqueness invoked in the forgery debates often is less a statement of fact—that there is, say, only one Mona Lisa—than a normative claim: that we would somehow debase our love for the work were that love available to just any qualitatively similar or even identical canvas.) At times this desire to protect the work has been taken to require a still more drastic form of personification. Having given artworks their love, some want to give them rights as well. When Garry Hagberg writes that a work of art is ‘without question a bearer of rights’ since its integrity can be violated and its abuse elicits our shock, he is echoing a 1973 article by Alan Tormey and disregarding the dissent which it later provoked.22 Like Hagberg, Tormey argues not just that we should regard artworks as rights-bearers but that we already do. For Tormey, the ‘aesthetic pain’ that viewers or listeners feel when a painting is desecrated or a sonata is mangled in performance offers evidence that we regard such works of art as ‘invested with interests that are obligation-generating.’23 In other words, we regard artworks as ‘invested with rights of a kind analogous to those that we normally concede to persons.’24 The success of this argument has been sharply contested and my goal is not to defend it. What I want instead to pursue emerges in Tormey’s suggestion, made in a postscript, as to why we treat artworks this way in the first place:

Art works, no more than persons, are fully interchangeable entities, and it is their logical similarity with respect to their identity that may account

June 21, 2008 11:20 MAC/NWA Page-228 9780230_220478_13_cha12 Brian Soucek 229

for our treating art works as quasi-persons, and thus, in accordance with Kantian injunctions, for our apparent willingness to regard them as endowed with rights.25

An analogy between the identity of artworks and that of persons is offered here as support for their analogical treatment. The metaphysical analogy is used to ground our alleged ascription to artworks of rights analogous to those of persons. Of course, this is the same strategy Danto and others pursue when they argue against copied paintings by analogy with persons. There the analogical arguments make sense only if we assume (as Danto clearly does) that artworks and persons are relevantly similar. This is a point to which we will return, for arguments about interpretation, discussed in the next section, will share the same assumption, namely that the personification of art should be grounded upon some similarity between artworks and persons themselves. It is just this assumption that will be challenged in subsequent sections.

III. Interpreting art

That theorists from at least Hegel onward have (as we have seen) written of art’s meaning in terms of the soul should perhaps prepare us for a link between art’s interpretation and its personification. Still, this sort of personification has hardly gone unchallenged. Joseph Margolis’s criticism of Nelson Goodman for treating paintings ‘anthropomorphically’26 is an illustrative example. Goodman had written that a work of art exemplifies those properties that it ‘makes manifest, selects, focuses upon, exhibits, heightens in our consciousness—those that it shows forth.’ Margolis responded: ‘it is very strange to suppose that the work itself has reflexive power—akin to the power of human beings ...—to ‘‘select’’ or ‘‘focus’’ or ‘‘exhibit’’ anything.’27 Both the tendency to personify and the attendant fear of personification continually reappear in discussions of the ways artworks mean, particularly in the literature on one of those ways: expression, especially in music.28 Much work on musical expression is driven by a reluctance to attribute it either to the composer or to the audience. As many have observed, a musical passage may be recognized as ‘sad’ despite the fact that neither its composer nor its audience feel the slightest gloom.29 The attempt to avoid these alternatives has led to accounts that postulate some hypothetical persona in the work. In Jerrold Levinson’s widely discussed definition of musical expression, ‘A passage of music P is expressive of an emotion or other psychic condition E iff P, in context, is readily and aptly heard by an appropriately backgrounded listener as the expression of E, in a sui generis, ‘‘musical,’’ manner, by an indefinite agent, the music’s persona.’ 30 This indefinite source of agency within or behind the work allows Levinson to avoid personifying the work itself. But avoiding personification often proves

June 21, 2008 11:20 MAC/NWA Page-229 9780230_220478_13_cha12 230 Personifying Art harder than expected. Even Levinson himself writes of a case where, although we could not say that Khatchaturian alluded to Mussorgsky-Ravel’s Pictures at an Exhibition, ‘we could, I think, rightly say that that section of the piece [his Masquerade Suite] ...was doing so.’31 The emphasis here on ‘the piece’ is Levinson’s, and it is in keeping with similar claims made elsewhere in the same essay. Referring to what he calls an artwork’s ‘doings,’ Levinson writes that ‘where the ‘‘doings’’ in question are semantic, for example, symbolic or expressive ones, as opposed to ones of creating, framing, projecting, it is open to us to ascribe such ‘‘doings’’ to the work directly, and only indirectly to the artist ...’32 The scare quotes around ‘doings’ are telling: Levinson’s need to personify in these cases is matched by an aversion to doing so. He distances himself from the personifying language his theory seems to require. Other theorists of musical expression try to maintain a similar distance in different ways. Saam Trivedi, for example, discusses several ways of experiencing musical expression, but he dwells especially on cases where ‘we make-believedly animate the music itself so that the music itself is imaginatively experienced as the very being that is animate and whose emotions are being experienced musically.’33 Peter Kivy argues that we are ‘evolutionarily hard-wired to perceive things as animate, when they do not otherwise prevent that reading.’34 Musical works, along with weeping willows and the faces of Saint Bernard dogs, are often seen as expressive of certain emotions. These expressive properties are perceived features of the works (and trees and dogs) themselves. Kivy is careful to distinguish ‘expressive of emotion’ from ‘expresses emotion,’ for only the latter requires that someone possess the emotion in question.35 The candidates for possessor of that emotion seem limited to the composer—an option already rejected—or the work. Yet, as Kivy asks (on behalf of the so-called musical purist), if Beethoven’s symphony really were triumphant, ‘Shouldn’t we give it the Victoria Cross, or chide it for being boastful?’36 This is the kind of challenge both Kivy’s expressive/expression distinction and Trivedi’s emphasis on imagination try to avoid. It is nothing other than Margolis’s objection to Goodman, his insistence that artworks simply do not have sentience, or self-reflection, or agency. Insofar as personification is the attribution of such things to the work of art, personification is seen as misguided—a remnant, perhaps, of na¨ıve or superstitious approaches to art.37

IV. The alternate approach to personification

These debates have been traced in admittedly broad strokes. But that is only so certain similarities in their outlines might be made explicit. However varied they are in other respects, the arguments above all share the same form, insofar as the personification of art is rooted in a comparison between artworks and persons themselves. An ontological analogy between artworks

June 21, 2008 11:20 MAC/NWA Page-230 9780230_220478_13_cha12 Brian Soucek 231 and persons is used to justify the treatment of artworks (to quote Cavell again) ‘in ways normally reserved for treating persons.’ Put more simply, artworks and persons are treated alike because they are thought to be somehow alike. This, I believe, is the wrong approach to personification. There is a difference between the arguments of the two preceding sections: in the first, the ontological analogy between persons and artworks is said to justify their analogous treatment; in the second, an ontological disanalogy argues against such treatment. Because artworks, unlike persons, do not have intentions or emotions, theorists of the second sort attempt to ascribe expression elsewhere—say, to a hypothetical persona. Both these arguments, however, share Cavell’s strategy, quoted at the outset. To understand whether or not we should treat art in such a peculiar way—namely like persons—it is thought that we need to understand the peculiar sort of things that artworks are. This is precisely the approach I want to oppose—in fact, to reverse. The alternative to be considered is not found in recent aesthetic theory; it comes instead—significantly—from the years surrounding aesthetic theory’s emergence as a discipline. (I say ‘significantly’ because, as I will argue, personification was crucial both to the emergence of aesthetics as a philosophical discipline and to the conception of art it produced.) The alternative predates Kant, yet one of its best statements can be found in a paragraph, not often discussed, within The Critique of Judgment. There, near the end of the section ‘On Beauty as the Symbol of Morality,’ Kant writes,

[B]eautiful objects of nature or of art are often called by names that seem to presuppose that we are judging [these objects] morally. We call buildings or trees majestic and magnificent, or landscapes cheerful and gay; even colors are called innocent, humble, or tender, because they arouse sensations in us that are somehow analogous to the consciousness we have in a mental state produced by moral judgments.38

Two centuries old, the passage confronts the personification of art, the ascription of emotive, expressive, even moral attributes to artworks (and the beauties of nature as well, since Kant, typically, is concerned with both). Kant does not argue that such ascriptions are inappropriate, much less that they should be attributed elsewhere.39 Instead, he offers an explanation of personification that does not rest on any properties of the objects personified. This is the approach which deserves to be better understood, as a corrective to the contemporary insistence on personification’s metaphysical basis. Kant claims that our personification of ‘beautiful objects of nature or of art’ occurs because those objects arouse sensations in us ‘that are somehow analogous to the consciousness we have in a mental state produced by moral judgments.’ In other words, we personify art because our ways of judging or appreciating artworks and persons are analogous—not because there is any analogy between artworks and persons themselves. In the section from

June 21, 2008 11:20 MAC/NWA Page-231 9780230_220478_13_cha12 232 Personifying Art which this passage is taken, Kant traces in unusual detail the various parallels between aesthetic and moral experience—between taste and practical reason. Taste, he says, ‘legislates to itself, just as reason does regarding the power of desire.’40 Further, judgments of beauty and morality both involve an immediate, disinterested pleasure; both are universally valid; and both involve a harmony between freedom and lawfulness.41 Kant was not the first to compare moral and aesthetic judgment or to model the moral on the aesthetic. Francis Hutcheson did so explicitly in 1725 in his Inquiry into the Original of Our Ideas of Beauty and Virtue.42 The sense of beauty, which was seen as innate and disinterested, provided Hutcheson his model for the still higher sense of virtue. This analogy between aesthetic and moral judgment was not entirely original even in Hutcheson; its roots were in Shaftesbury’s work, though the latter’s neo-Platonism often made the aesthetic and the moral more identical than analogous. Both, reacting against Hobbesian egoism, used the purported disinterestedness of aesthetic judgment as a model for moral judgment. Their interest was in moral theory, but the analogy with taste—and the way that analogy underscored taste’s alleged disinterest—had at least as much effect on the shape of their aesthetic theories. Disinterest, which later became the autonomy of aesthetic judgment in Kant’s work, emerged as a central notion within aesthetics.43 The point is that this analogy between aesthetic and moral judgment had, by the end of the eighteenth century, become so established that that Kant could claim ‘common understanding’ habitually bears it in mind.44 And this, he said further, is what leads to, and explains, our widespread personification of art and the beautiful. We do not treat artworks and persons alike because they share some ontological similarity. Rather, characteristics which seem shared —a common capacity, say, for majesty, cheerfulness, or innocence— are merely byproducts of an affinity between our mental faculties, an analogy between our aesthetic and moral modes of judgment. Kant might have been overly optimistic in claiming that common understanding intuitively bears in mind the analogy which had taken him 59 sections of the Third Critique (if not the entire Critical system) to reach. Still, whatever we think of the specifics of his account, I want to maintain that its general shape and direction are correct. Broadly, Kant argues that familiarity with analogies between judgments of beauty and of goodness leads to the personification of beautiful objects themselves. I want to say that this is true of history, not necessarily of ‘common understanding.’ In other words, the analogy, increasingly familiar over the course of the eighteenth century, eventually brought about the personification of the beautiful (art, of course, included). The first half of this story has just been outlined. What remains to be told is the second half of the story: how this conception of aesthetic judgment, shaped through the analogy with moral judgment and seen (by Kant’s time) as autonomous, led historically to the personification of works of art.

June 21, 2008 11:20 MAC/NWA Page-232 9780230_220478_13_cha12 Brian Soucek 233

The decisive step, I would like to suggest, was taken by Friedrich Schiller. Though sympathetic to much of the Third Critique, Schiller was quick to complain that Kant’s aesthetics neglected ‘the objective constitution of the objects held to be beautiful.’45 In a series of letters on the subject,46 Schiller took seriously Kant’s insistence that taste remain autonomous and beauty be independent of concepts; as such, he associated beauty with practical reason instead of theoretical reason. (Here again is the eighteenth-century link between aesthetics and ethics, even if under another guise.) Schiller, however, hardly makes beauty the same thing as morality. Whereas the moral is the product of a pure will, a will determined by the form of practical reason and not some external force, the beautiful imitates the form of practical reason. Objects of beauty give the appearance of being determined autonomously; in Schiller’s slogan, beauty becomes ‘freedom in appearance [Freiheit in der Erscheinung].’47 In other words, beauty ‘necessitates us to produce the idea of freedom from within ourselves, and to apply it to the object.’48 ‘Reason lends the object ...a power to determine itself, a will, and then examines the object under the form of that will (not its will, since this would yield a moral judgment).’49 Seeing ‘freedom in appearance’—experiencing beauty—thus involves our treating the object of beauty as if it had an autonomous will. For Schiller, aesthetic experience and personification go hand-in-hand. Schiller struggled unsuccessfully to prove that freedom in appearance really is the same thing as beauty. As a result, he never published the dialogue that his letters on beauty were meant to prefigure. Still, the definition of beauty as freedom in appearance plays a key role in two of Schiller’s most influential published works, On Grace and Dignity and the Letters on the Aesthetic Education of Man. Both repeat an idea already broached in the Kallias Letters:that freedom is so important to us that we wish (though cannot command) its presence even in the realm of appearance. We wish, in other words, for everything, from our person to our clothes to our surroundings, to have the beauty of a work of art. Schiller’s ideal state thereby goes beyond even Plato’s republic or Kant’s kingdom of ends. As he writes in the Kallias Letters,ina passage echoed later in the Aesthetic Education,

Everyone is a free citizen and has the same rights as the most noble in the world of aesthetics, coercion may not take place even for the sake of the whole—everyone must consent. In this aesthetic world, which is quite different from the most perfect Platonic republic, even the gown I wear on my body demands respect for its freedom from me, much like a humble servant who demands that I never let on that he is serving me. In exchange, it promises to use its freedom in such a way that it will not curtail my own freedom; and if both keep their word, the world will say that I am well dressed.50

June 21, 2008 11:20 MAC/NWA Page-233 9780230_220478_13_cha12 234 Personifying Art

Schiller’s so-called aesthetic state is thus a better-dressed version of the kingdom of ends. Less glibly, his ideal is a morality made beautiful: a world free of even the appearance of coercion or domination.

V. Lessons learned

Having looked at Schiller’s theory, what remains is to see how elements of his theory explain, or might be used to justify, the personification of art which pervades later philosophies of art, those of our own time included. A first thing to note is how the autonomy of art, so central to Schiller’s view, also underlies the various arguments and debates canvassed earlier. Regardless of whether it is made explicit, this is the assumption on which most recent attempts at personification rest. Certainly, this is true of the debates surrounding meaning in art —debates which arise precisely because of the work’s alleged autonomy from its creator. As Jerrold Levinson claims of literary works (as opposed to ordinary writing), ‘the text is held to have a certain amount of autonomy, to be something we interpret, to some extent, for its own sake, and thus not jettisonable in principle if we could just get, more directly, at what the author had in mind to tell us.’51 Writing about music, Stephen Davies makes a similar point:

We react to the musical work as if it were autonomous and organic; we expect to find sufficient reason for the work’s progressing as it does within the context set solely by its style, form, and content. Thus, ‘its maker intended it to be so’ is incomplete as an explanation of why the music is as it is if no reason internal to the world of the work indicates why the composer had that intention.52

Davies goes on to make the now-familiar claim that understanding a musical work is like understanding a person. The fact that both are experienced as, or as if ‘autonomous and organic’ points significantly toward the connection. Beyond discussions of meaning, this insistence on the work’s autonomy reappears just as clearly in Alan Tormey’s arguments about aesthetic rights. What Tormey seeks to protect are not the interests of artists or of audiences. His argument turns on the interests and entitlements of artworks themselves. It is only because artworks are seen as having interests of their own—not reducible to those of the persons surrounding them—that the notion of aesthetic rights becomes an issue. The difference between Schiller’s account and these more recent ones is Schiller’s insistence on the autonomy of the artwork as something we ascribe to it, not something artworks simply possess or lack. We apply our idea of freedom to the work, he says; we lend the work of art a will. Works of art can then be treated as if they were self-determining.

June 21, 2008 11:20 MAC/NWA Page-234 9780230_220478_13_cha12 Brian Soucek 235

This ascription is possible, Schiller claims, whenever an object does not betray any outside determination, whether by laws of nature, the nature of its constitutive material, predetermined rules, or—most relevantly—the will or personality of its creator. In Schiller’s own words,

[A] form appears as free as soon as we are neither able nor inclined to search for its ground outside it. ...It is thus a tenable principle that an object presents itself as free in appearance, if its form does not compel reflective understanding to seek out a ground for it. A form is therefore beautiful only if it explains itself ...without the help of a concept.53

Artworks, of course, do not create themselves. In them, outside determination or external forces are necessary, but they have to be hidden, made to seem internal to the work. Schiller gives the example of a landscape painting in which a tree might cover some beauty in the background. The artist cannot, he claims, ‘require of the tree that it not do this’ without infringing on its freedom and revealing ‘dilettantism.’ Instead, ‘He allows that branch of the tree ...to sink down under its own weight and thus freely make place for the view behind it; thus the tree fulfils the will of the artist by following its own.’54 Though more vivid in his personification, Schiller is describing nothing other than Davies’s requirement that the work be explained on the basis of reasons internal to it, not ones like ‘the artist wanted it so.’ One might wonder how Schiller’s argument avoids the strategy being resisted here. That is to say, how is Schiller’s account different from those, already discussed, in which the personification of art is underpinned by some metaphysical view of artworks themselves? It is not sufficient to say that, for Schiller, autonomy is something ascribed by the viewer. For admittedly, such ascriptions are hardly arbitrary or unconnected to properties the work possesses. Schiller’s landscape makes this clear. Its material features are independent of our ascriptions: the branch’s appearance of sinking down under its own weight is a feature we perceive in the painting, not one we give it. That feature in turn makes appropriate our ascription of self-determination to the painting. A novel that reveals the hand of an author who manipulates its characters’ destinies, or a symphony with a modulation that makes no internal sense—these are cases where ascriptions of autonomy would prove unwarranted. Still, a crucial difference remains between Schiller’s account and those which sought to justify personification on the basis of some property in the artwork. For Schiller, properties of the artwork make some works more appropriate than others as candidates for personification. What they do not give us is any reason to personify art in the first place. Properties of the artwork do not motivate our practices of personification, even if they defeat some (or many) of our attempts to do so. Why, then, do we personify? Schiller remains true to—in fact, he embodies—the story Kant tells, tracing personification

June 21, 2008 11:20 MAC/NWA Page-235 9780230_220478_13_cha12 236 Personifying Art not to the metaphysics of the artwork, but to aesthetic judgment. Schiller makes the personified, self-determining work of art his ideal, his very notion of beauty. Yet he does so exactly because he accepts Kant’s notion of disinterested and conceptually unfettered aesthetic judgments. Put the reverse way: if we accept a broadly Kantian account of what aesthetic judgment is, we might wonder, as Schiller did, what kind of objects will provoke the free play of imagination and understanding that characterizes such judgments. Schiller’s answer is that the free play is occasioned by freedom in appearance, that is, by objects which can be seen as self-determining. These, we might say, are those objects that prove most readily ‘personifiable.’ An important aspect of this last point is that personifiability is an ideal for art—not part of its definition. This is signaled in part by the fact that Schiller’s aesthetic investigations concerned beauty, not the nature of art as such. They did so largely for contextual reasons; Schiller, after all, was responding to a tradition concerned at least as much with natural beauty as with the beauty of art. Still, there is a deeper lesson to be drawn from the difference between Schiller’s concerns and ours. An apparent slippage in my argument thus far—treating eighteenth-century discussions of beauty as if they were discussions of art—actually reflects a slippage in the history of aesthetics, wherein accounts of beauty became accounts of art. The personification once tied to a normative ideal thereby became a feature of a certain class of objects—works of art. The result is the kind of questions with which I began—questions which ask what is it about art that leads us to personify it. This approach, as I have been saying, gets things backward. It is not that art is a certain sort of thing, or that artworks share some particular trait, on the basis of which we feel justified in personifying them. Rather, our normative ideals involve personification; in fact, our ideal amounts to, for lack of a better word, personifiability. This is to say that we, no less than Schiller, seek and prize freedom in appearance in works of art. Works which are personifiable, that is those which can be seen as self-determining, are precisely those we value. This ideal was not discovered; it was, over time, developed and promoted, precisely because of the freedom which lies at its heart. Cognizant of this ideal, artists aim to make works which give the appearance of self-determination. They try to make artworks which support our ascriptions of autonomy. This is to say, then, that our personifying ideal has led to certain widespread features of works of art, not that some trait characteristic of art supports or justifies our practices of personification. Does this mean that I am reasserting beauty as the central term in our evaluation of art? Such a move would seem dubious given the direction of recent philosophy of art, and of art itself. Here, however, we find an unexpected advantage in one of Schiller’s shortfalls. Schiller’s theory is at its most convincing in showing how and why people value freedom in appearance in the objects (and persons) around us. What his theory lacks is a convincing

June 21, 2008 11:20 MAC/NWA Page-236 9780230_220478_13_cha12 Brian Soucek 237 reason to believe that freedom in appearance is the same thing as beauty.55 And yet, in the current context, this link is precisely what is not needed or even wanted. I would argue instead that it is freedom in appearance, self- determination—not necessarily beauty—which remains the mark of artistic excellence, in our time no less than in Schiller’s. Schiller might have failed in giving a theory of beauty, but that does not mean he failed to offer a normative ideal for art. His argument that freedom in appearance should constitute this aesthetic ideal continues to resonate. In practical terms, this involves a model of interpretation which begins with the hope that the artwork will appear self-determining. Our initial assumption is that the work can support our personification, our ascriptions of agency, our desire to treat it as if autonomous. But this is a goal, and one which many (perhaps most) works fail fully to reach. In those cases, our relation to the work shifts; our interpretive approach gets revised. Imagine, for example, that while interpreting a musical work, we find a note which fits oddly into its surroundings. We first struggle to explain the note as part of a harmony which the work ‘needs’ or ‘calls for’ at just that place. Should we fail in this attempt, however, we might shift gears and ask questions of a different sort. Perhaps instruments at the time left the composer few other choices, for example. (Why is the timpani playing an f at the end of Act Three of Verdi’s Ernani when all the other instruments are playing in A-flat minor? Because the rest of the finale is in F, Italian orchestras at the time only had two timpani, and they were constructed in a way which made quick retuning impossible.)56 Interpreters of a play or an opera, faced with a scene which seems to serve no dramatic or musical purpose, might similarly look for an explanation elsewhere—to a soprano who demanded a second showpiece, or a scenic effect which required extra time to prepare. These kinds of explanation are heteronomous. They shift our attention outside of the work, forcing us to recognize the hand of the artist, the constraints of his or her materials, or the intervention of a censor or some other external agent. My point is that interpretive shifts like these should be seen as fallback positions which suggest a certain failure on the part of the work. Notably, it is a failure in the work’s apparent self-determination, its personifiability. It is worth noting in passing that a similar variety of interpretive stances have been described in our dealings with other people. As Daniel Den- nett has argued, we make predictions of another’s behavior by considering what would be the most rational course of action given a goal and a set of constraints. (Alternately, given a particular behavior and a set of constraints, we might ascertain a person’s goal by considering what goal such behavior would be optimal for attaining.) This is the so-called ‘intentional stance.’57 But the optimality assumption on which it is based is only retained so long as it works. Once we have evidence of less than perfect rationality, we switch to other modes of interpretation—to another stance, in Dennett’s description. The intentional stance is replaced by something more

June 21, 2008 11:20 MAC/NWA Page-237 9780230_220478_13_cha12 238 Personifying Art like the design stance (which makes explanations or predictions on the basis of some outside programmer); ‘This,’ says Dennett, ‘is the fundamentally different attitude we occasionally adopt toward the insane. To watch an asylum attendant manipulate an obsessively countersuggestive patient, for instance, is to watch something radically unlike normal interpersonal relations.’58 Dennett’s example suggests that in persons, as in works of art, we stop ascribing rational self-determination only when doing so no longer works, when attributions of autonomy are no longer plausible. The argument just made holds that while the personification of any particular artwork might be justified by properties of the work, the practice of personifying art cannot be justified that way. That practice results instead from our particular notions of aesthetic judgment; the Kantian account of taste leads to the Schillerian ideal of personifiable art. The question of whether this ideal is justified turns, then, on the question of whether Kant’s notion of aesthetic judgment is justified. That is not an issue this chapter will settle. I can at best suggest what is at stake in rephrasing the question this way. One thing Kant and Schiller share, and share with their predecessors, is a concern with the link between taste and morality. As we have seen, both use aesthetic judgment as a stepping stone, whether to an understanding of morality (or moral reasoning) or to morality itself. Taste is said, for example, to make manifest the freedom which moral reasoning presupposes,59 or to offer (in Schiller’s description) a bridge from sense to form, physical drives to rational law. Given such stakes, the shape of one’s notion of aesthetic judgment matters for reasons beyond the art world. To put matters another way, having asked about art’s personification, we are led to questions about judgments of the beautiful; having pursued those, we are led to questions about judgments regarding moral action. Seen this way, it is tempting to suggest that the personification of art is not ultimately about art at all. It is rather about us—about persons. The question of why we treat artworks in ways normally reserved for treating persons does not produce the sort of answer initially expected. The answer does not reveal, as Cavell expected, something about what art is. If anything, it reveals more about our beliefs about what persons are, or perhaps what we want them to become. These are grand claims, and ones which do little but point forward to future research. What I have argued here is that questions about the personification of art are not merely questions about whether symphonies really have emotions, novels really have agency, or paintings really speak to us. Clearly, they do not. But that does not mean we lack reasons to treat them as if they did. As I have shown, discussions of these reasons permeated aesthetics in its early years. The personification of art has not waned in the years since; what we have lost—and what contemporary aesthetics can perhaps regain—is an awareness of what is at stake in that practice.

June 21, 2008 11:20 MAC/NWA Page-238 9780230_220478_13_cha12 Brian Soucek 239

Notes

1. Cavell, ‘Music Discomposed,’ 189. 2. See Kant’s discussion of art’s Geist, which needs constraint or else it would have ‘no body at all.’ Critique of Judgment, 171 (§43, Ak. 304). Cf. 193 (§51, Ak. 324) where Kant writes of works that are attributed a spirit and so ‘speak, as it were, by mime.’ Quotations and references from Hegel, Dewey, and Danto appear below, in the section ‘Valuing Art.’ 3. See Walter Benjamin’s claim that ‘The idea of life and afterlife in works of art should be regarded with an entirely unmetaphorical objectivity ...’ Benjamin, ‘The Task of the Translator,’ 71. 4. Benjamin, ‘On Some Motifs in Baudelaire,’ 188. 5. Scruton, The Aesthetics of Music, 76. Scruton adds that music’s movements exhibit ‘a causality that makes immediate sense to us’; this ‘causality that binds tone to tone in music is not the dead causality of a machine, but the causality of life, whose principal manifestation, for us, is in the world of human action’ (ibid.). For Scruton’s discussion of ‘indispensable metaphors,’ see ibid., 91–92. 6. See ‘How Buildings Mean,’ 31–48; and Pippin, ‘Authenticity in Painting,’ 575–598. 7. Binkley’s ‘Piece: Contra Aesthetics’, 265–277. Binkley himself objected to the analogy because of its emphasis on the artistic equivalent of a person’s body: the perceived art object. In other words, it turns art necessarily into an aesthetic phenomenon. 8. Mitchell, What Do Pictures Want?; Armstrong, How Novels Think;Jones,‘Talking Pictures’; Steinberg, Listening to Reason. A short but fascinating study which examines the way we personify objects of many kinds, not just artworks, is Tamen’s Friends of Interpretable Objects. 9. Hagberg, Art as Language, 180–185. For Taylor’s view, see ‘The Concept of a Person’,97–144. 10. Schaper, ‘The Pleasures of Taste,’ 51. 11. Hagberg, Art as Language, 182–183. 12. Armstrong often fails to make this distinction in How Novels Think;attimes(in the title, for instance), she attributes novels themselves a certain agency. But more often it is characters within the novels that are said to embody the notions of individuality and subjectivity which her book studies. 13. Consider Cavell’s claim that ‘We treat [works of art] in special ways, invest them with a value which normal people otherwise reserve only for other people—and with the same kind of scorn and outrage. They mean something to us, not just the way statements do, but the way people do.’ He adds about music: ‘We follow the progress of a piece the way we follow what someone is saying or doing. Not, however, to see how it will come out, nor to learn something specific, but to see what it says ...’ (Cavell, ‘Music Discomposed,’ 198). Though this is not Cavell’s intention, we might read this as suggesting that artworks mean as much as they do to us precisely because artworks mean—that is, because they (like persons) have something to say. 14. Wollheim, ‘Objects of Love,’ 553. 15. Sagoff, ‘On Restoring and Reproducing Art,’ 453. 16. Sparshott, ‘The Disappointed Art Lover,’ 253–254. 17. Danto, ‘Responses and Replies,’ 197–198. 18. Danto, ‘Moving Pictures,’ 212–213. To his question ‘Would it matter?’ Danto goes on to speculate that, yes, it would matter a great deal.

June 21, 2008 11:20 MAC/NWA Page-239 9780230_220478_13_cha12 240 Personifying Art

19. Danto, ‘Responses and Replies,’ 199–200, his emphasis. In two of his major works on the philosophy of art, Danto invokes Peter Strawson’s theory of the person to explain the distinction between artworks and the mere things from which they are constituted. See Danto, ‘The Artworld,’ 576, and The Transfiguration, 104. Cf. Strawson, Individuals, especially Chapter 3. 20. Hegel, Aesthetics, 20. 21. Dewey, Art as Experience, 118. 22. Tormey, ‘Aesthetic Rights,’ 163–170. See also Goldblatt, ‘Do Works of Art Have Rights?’ 69–77; Hein, ‘Aesthetic Rights: Vindication and Vilification,’ 169–176; and Sparshott, ‘Why Artworks Have No Right to Have Rights,’ 5–15. 23. Tormey, ‘Aesthetic Rights,’ 168. 24. Ibid., 163. 25. Ibid., 169. 26. Margolis, ‘What is When? When is What?’, 267. 27. Ibid. 28. The paraphrase of Goodman is in reference to his essay ‘How Buildings Mean.’ 29. Cf. the exhaustive account of this topic in Davies, Musical Meaning, Chapter 4. 30. Levinson, ‘Musical Expressiveness,’ 107, emphasis added. 31. Levinson, ‘Intention and Interpretation,’ 211. 32. Ibid., 202. 33. Trivedi, ‘Expressiveness as a Property,’ 414. The emphases are his. 34. Kivy, The Corded Shell, 172. 35. Ibid., Chapter 2. 36. Ibid., 6. 37. Cf. Kivy’s discussion of Maimonides and the anthropomorphizing of God in The Corded Shell (59–60). There might be reason ‘to cleanse our theological thinking of anthropomorphic overtones,’ since God does not possess human characteristics. But this is where God and music part ways, Kivy believes, for expressive qualities (akin to those of humans) are ‘genuine ‘‘objective’’ qualities of music.’ 38. Kant, Critique of Judgment, 229–230 (354 in the Akademie edition). In the notes that follow, page numbers from the Akademie edition will be used. 39. The notion of a hypothetical persona obviously does not figure into Kant’s aesthetics. His running together of art and nature raises a question, however, for theorists who do employ it in their treatment of art: are they willing to ascribe hypothetical personas to cheerful landscapes or humble colors just as they do to sad symphonies? 40. Kant, Critique, 353. 41. Ibid., 353–354. Paul Guyer emphasizes this last point in particular in Kant and the Experience of Freedom, 109. My argument here draws significantly from his. 42. Hutcheson’s stated concern there was ‘to show that there is some sense of beauty natural to men,’ for, as he wrote, ‘If the reader be convinced of this, it will be no difficult matter to apprehend another superior sense, natural also to men, determining them to be pleased with actions, characters, affections. This is the moral sense ...’Hutcheson,An Inquiry. 43. See Stolnitz, ‘On the Significance,’ 105–108; and ‘On the Origins of ‘Aesthetic Disinterestedness,’ 132–133. 44. Kant, Critique, 354. 45. Schiller, ‘The Aesthetical Lectures,’ 474. 46. Schiller, Kallias. Schiller’s concern with beauty, or beautiful objects, includes artworks but is not limited to them. The slippage from talk of the personification

June 21, 2008 11:20 MAC/NWA Page-240 9780230_220478_13_cha12 Brian Soucek 241

of art to the personification of the beautiful will be addressed in the following section. 47. It is worth emphasizing that Schiller does not mean by this that beautiful objects are produced by agents acting autonomously. They may or may not be. (Their creation thus may or may not be moral; this is not Schiller’s concern.) It is the beautiful object, not its maker, that appears autonomous; the beautiful object appears to determine itself. 48. Schiller, Kallias, 161. 49. Ibid., 151. 50. Ibid., 170. In the Aesthetic Education, Schiller claims that taste makes ‘the fetters of serfdom fall away from the lifeless and the living alike.’ He adds, ‘In the Aesthetic State everything—even the tool which serves—is a free citizen, having equal rights with the noblest ...’, 219. 51. Levinson, ‘Intention and Interpretation,’ 177. 52. Davies, Musical Meaning, 367. 53. Schiller, Kallias, 155. 54. Ibid., 172. 55. Schiller worried in his letter of February 23, 1793: ‘I have not yet shown that the quality of things which we call beauty is one and the same with this freedom in appearance; and this shall be my task from now on’ (Kallias, 160). Frederick Beiser claims convincingly that Schiller never succeeded in showing this (Schiller as Philosopher, 74–75). 56. Gossett gives this example in Divas and Scholars, 416–419. The f makes no sense within the context of the work, where it can only be described as the ninth of the dominant seventh chord. Recourse to an external, historical explanation is necessitated by this failure in our initial interpretive approach. 57. Dennett, ‘Intentional Systems,’ 87–106. 58. Ibid., 94. 59. This description is taken from Guyer, ‘Thomson’s Problems with Kant,’ 318.

Bibliography

Armstrong, N. 2005. How Novels Think: The Limits of Individualism from 1719–1900, New York: Columbia University Press. Beiser, F. 2005. Schiller as Philosopher: A Reexamination, Oxford: Clarendon Press. Benjamin, W. 1968. ‘On Some Motifs in Baudelaire’ in H. Arendt (ed.), H. Zohn (trans.), Illuminations, New York: Schocken Books. Benjamin, W. 1968. ‘The Task of the Translator’ in H. Arendt (ed.), H. Zohn (trans.), Illuminations, New York: Schocken Books. Binkley, T. 1977. ‘Piece: Contra Aesthetics,’ Journal of Aesthetics and Art Criticism, 35: 265–277. Reprinted in P. Alperson (ed.), The Philosophy of the Visual Arts, New York: Oxford University Press, 450–460. Cavell, S. 1969. ‘Music Discomposed’ in Must We Mean What We Say? Cambridge: Cambridge University Press. Danto, A. 1964. ‘The Artworld,’ Journal of Philosophy, 61: 19. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Danto, A. 1993. ‘Responses and Replies’ M. Rollins (ed.), Danto and His Critics, Cambridge, MA: Blackwell.

June 21, 2008 11:20 MAC/NWA Page-241 9780230_220478_13_cha12 242 Personifying Art

Danto, A. 1999. ‘Moving Pictures’, Philosophizing Art: Selected Essays, Berkeley: University of California Press. Davies, S. 1994. Musical Meaning and Expression, Ithaca, NY: Cornell University Press. Dennett, D.C. 1971. ‘Intentional Systems,’ Journal of Philosophy, 68.4: 87–106. Dewey, J. 1934. Art as Experience, New York: Perigee Books. Goldblatt, D.A. 1976. ‘Do Works of Art Have Rights,’ Journal of Aesthetics and Art Criticism, 35. Goodman, N. 1988. ‘How Buildings Mean’ in N. Goodman and C.Z. Elgin (eds), Reconceptions in Philosophy and Other Arts and Sciences, London: Routledge, 31–48. Gossett, P. 2006. Divas and Scholars: Performing Italian Opera,Chicago:Universityof Chicago Press. Guyer, P. 1992. ‘Thomson’s Problems with Kant: A Comment on ‘‘Kant’s Problems with Ugliness’’,’ Journal of Aesthetics and Art Criticism, 50.4, 317–319 Guyer, P. 1996. Kant and the Experience of Freedom, Cambridge: Cambridge University Press, 109. Hagberg, G.L. 1995. Art as Language, Ithaca, NY: Cornell University Press. Hegel, G.W.F. 1975. Aesthetics, T.M. Knox (trans.), Oxford: Clarendon Press. Hein, H. 1978. ‘Aesthetic Rights: Vindication and Vilification,’ Journal of Aesthetics and Art Criticism, 37. Hutcheson, F. 1973. An Inquiry concerning Beauty, Order, Harmony, Design, P. Kivy (ed.), The Hague: Martinus Nijhoff. Jones, C.A. 2004. ‘Talking Pictures: Clement Greenberg’s Pollack’ in L. Daston (ed.), Things That Talk: Object Lessons from Art and Science, New York: Zone Books. Kant, I. 1987. Critique of Judgment, W.S. Pluhar (trans.), Indianapolis: Hackett. Kivy, P. 1980. The Corded Shell: Reflections on Musical Expression, Princeton: Princeton University Press. Levinson, J. 1996. ‘Musical Expressiveness’, The Pleasures of Aesthetics, Ithaca, NY: Cornell University Press. Levinson, J. 1996. ‘Intention and Interpretation in Literature’, The Pleasures of Aesthetics, Ithaca, NY: Cornell University Press. Margolis, J. 1981. ‘What is When? When is What? Two Questions for Nelson Goodman,’ Journal of Aesthetics and Art Criticism, 39.3, 266–268. Mitchell, W.J.T. 2005. What Do Pictures Want?: The Lives and Loves of Images,Chicago: University of Chicago Press. Pippin, R. 2005. ‘Authenticity in Painting: Remarks on Michael Fried’s Art History,’ Critical Inquiry, 31: 575–598. Sagoff, M. 1978. ‘On Restoring and Reproducing Art,’ The Journal of Philosophy, 75. Schaper, E. 1983. ‘The Pleasures of Taste’ in E. Schaper (ed.), Pleasure, Preference, and Value, Cambridge: Cambridge University Press. Schiller, F. 1967. On the Aesthetic Education of Man in a Series of Letters, E.M. Wilkinson and L.A. Willoughby (trans.), Oxford: Clarendon Press. Schiller, F. 1988. The Aesthetical Lectures’ in W.F. Wertz, Jr. (trans.), Friedrich Schiller: Poet of Freedom, Volume 2, Washington, DC: Schiller Institute. Schiller, F. 2003. ‘Kallias or Concerning Beauty: Letters to Gottfried Korner’¨ in J.M. Bernstein (ed.) and S. Bird-Pollan (trans.), Classic and Romantic German Aesthetics, Cambridge: Cambridge University Press. Scruton, R. 1997. The Aesthetics of Music, Oxford: Oxford University Press. Sparshott, F. 1983. ‘The Disappointed Art Lover’ in D. Dutton (ed.), The Forger’s Art, Berkeley: University of California Press.

June 21, 2008 11:20 MAC/NWA Page-242 9780230_220478_13_cha12 Brian Soucek 243

Sparshott, F. 1983. ‘Why Artworks have no Right to Have Rights,’ Journal of Aesthetics and Art Criticism, 42. Steinberg, M.P. 2004. Listening to Reason: Culture, Subjectivity, and Nineteenth-Century Music, Princeton, NJ: Princeton University Press. Stolnitz, J. 1961. ‘On the Significance of Lord Shaftesbury in Modern Aesthetic Theory,’ Philosophical Quarterly, 11: 43. Stolnitz, J. 1961. ‘On the Origins of ‘‘Aesthetic Disinterestedness’’,’ Journal of Aesthetics and Art Criticism, 20. Strawson, P.F. 1959. Individuals: An Essay in Descriptive Metaphysics, London: Methuen. Tamen, M. 2001. Friends of Interpretable Objects, Cambridge, MA: Harvard University Press. Taylor, C. 1985. ‘The Concept of a Person’, Human Agency and Language: Philosophical Papers, Volume 1, Cambridge: Cambridge University Press, 97–144. Tormey, A. 1973. ‘Aesthetic Rights,’ Journal of Aesthetics and Art Criticism, 32: 163–170. Trivedi, S. 2001. ‘Expressiveness as a Property of the Music Itself,’ Journal of Aesthetics and Art Criticism, 59.4, 411–420. Wollheim, R. 1990. ‘Objects of Love,’ Times Literary Supplement, May 25–31.

June 21, 2008 11:20 MAC/NWA Page-243 9780230_220478_13_cha12 13 Danto and Kant, Together at Last? Diarmuid Costello

I. Danto and Kant

Danto and Kant: for anyone who has followed debates in either the theory or the philosophy of art over the last 40 odd years, this has to look like a very odd couple. Indeed, ‘The Odd Couple’ might have served equally well as a title for what follows: an attempt to show how much Danto’s and Kant’s aesthetics have in common, counter-intuitive as that may sound; and, within the context of this broad commonality, to offer a comparative analysis of the merits of their respective accounts of our relation—both cognitive and affective—to works of art. Given that art since the 1960s is widely thought to pose particular problems for aesthetic theories of art (such as Kant’s), to which various forms of cognitivism in the philosophy of art (such as Danto’s) have been offered as solutions, I intend to conduct this comparison on artistic terrain with which Danto (but not Kant) would be naturally associated—to see whether Kant’s aesthetics withstands the challenge. That Danto’s theory speaks to contemporary art hardly bears saying; his ontology was conceived in order to meet the challenges posed by art after modernism— but Kant and contemporary art? According to current consensus, the value of Kant’s aesthetics for the theory of art was tied to the fate of formalism, with which it is widely regarded (at least outside Kant scholarship) to have sunk. Contrary to this perception, I believe that Kant’s theory of art remains a valuable, if under-used, resource for understanding our relation—both cognitive and affective—to art after modernism. Until recently, one would have had little difficulty anticipating where Danto stood on this question—having tarred Kant’s aesthetics with the brush of Greenbergian formalism in After the End of Art.1 Despite this, the view that Kant’s aesthetics may prove amenable for contemporary theories of art received Danto’s surprise endorsement with the publication of ‘Embodied Meanings, Isotypes, and Aesthetical Ideas,’ in 2007. Here Danto asks whether his own most basic claim that works of art

244

June 21, 2008 11:25 MAC/NWA Page-244 9780230_220478_14_cha13 Diarmuid Costello 245 are ‘embodied meanings’—that is, entities that are, irreducibly, both about something, and embody what they are about—might be seen as an inher- itance (albeit unwitting) of Kant’s theory of fine art as ‘the expression of aesthetic ideas.’2 Whether this rather counter-intuitive sounding suggestion is plausible is what I want to examine here. My procedure will be straightforward: I shall look at Danto’s and Kant’s respective views of the embodiment of meaning in art in turn. Both Danto and Kant can be seen as proponents of expressionism in the philosophy of art, to the extent that both hold works of art to embody, and thereby express, the mental states (broadly construed to encompass beliefs, attitudes and feelings) of those that created them and, if successful, to dispose their viewers to a similar state. Moreover, both appeal—explicitly, in the case of Danto, implicitly in the case of Kant—to a conception of metaphor in their accountofwhataworkofartis. I shall begin with Danto’s most recent remarks about Kant, and track backwards, through the qualified ‘aesthetic turn’ of the Abuse of Beauty, to Danto’s most developed account of artworks as ‘embodied meanings’ in Transfigura- tion of the Commonplace. My goal in doing so is to show both that those aspects of Danto’s recent work, which might otherwise be taken as a fundamental reorientation of his earlier work, may be traced back to its foundational statement in Transfiguration of the Commonplace,andthat they can poses internal problems for Danto’s standard arguments against aesthetic theories to date. This raises the heretical thought that Danto’s own theory, contrary to both his own presentation of it and its orthodox reception, might commit Danto, despite himself, to an aesthetics after all. In this light, I juxtapose Danto’s proposed cognitivist alternative to aesthetic theories—that to understand a work of art (through interpretation) is, at root, ‘to grasp the metaphor that is always there’—with Kant’s theory of fine art as the ‘expression of aesthetic ideas,’ in order to bring out their common commitment to a metaphorical conception of artistic meaning. In doing so, I draw on a recent paper by Kirk Pillow that distinguishes two competing conceptions of metaphor in Kant’s theory of art: I suggest that Danto’s broadly Aristotelian view of metaphor has one significant feature in common with what Pillow calls Kant’s ‘weak’ conception of metaphor. I go on to argue that what Pillow calls the ‘strong’ conception of metaphor at work in Kant’s theory of aesthetic ideas offers something that neither Kant’s ‘weak’ conception nor Danto’s thorough-going cognitivism can provide, namely a basis for understanding the longevity of art—the widely held intuition that we never reach the end of a successful work of art, that if we do the work has failed in some crucial respect as art—by grounding it in the open-ended imaginative play that works of art elicit on Kant’s theory. I do this by bringing Kant’s account of aesthetic ideas to bear on Art & Lan- guage’s Index 01 (1972), the kind of work that would typically be thought to shipwreck Kant’s aesthetics in the face of recent art. My goal is to show otherwise.

June 21, 2008 11:25 MAC/NWA Page-245 9780230_220478_14_cha13 246 Danto and Kant, Together at Last?

II. Recent Danto on Kant’s aesthetics

Danto’s perception of Kant’s aesthetics has always been mediated, and in my view distorted, by the controversial use to which Clement Greenberg put a version of Kant’s theory of pure aesthetic judgement—for Kant, a judgement concerning ‘free,’ conceptually unconstrained, beauties—in underwriting his activity as a formalist art critic. This remains true of what Danto has to say about what he calls, in my view misleadingly, the ‘Kant-Greenberg aesthetics of form’ in his recent ‘Embodied Meanings’ papers. I have taken issue with Danto’s tendency to take his Kant at Greenberg’s word before; but as several of the worries I have raised elsewhere still pertain to Danto’s presentation of Kant, I will briefly enumerate them for the record.3 To begin with an old saw, and the basis for Danto’s antipathy for the general tenor of Kant’s aesthetics over the years: the thought that for Kant ‘artistic excellence is at one with aesthetic excellence.’4 Although there is no doubt some prima facie plausibility to this charge—unlike Hegel, Kant does not set natural beauty aside in his opening remarks—it nonetheless ignores the crucial distinction Kant makes between ‘free’ and ‘dependent’ beauty (CJ §16). According to Kant the latter, but not the former, puts conceptual constraints on the beauty of works of art, in line with ‘what they are meant to be’—thereby building intention, and some minimal cultural milieu, into the account at the ground level.5 To cite Kant’s own example: a beautiful church must not only be beautiful, its beauty must be fitting to its purpose as a place of worship; much that might otherwise please freely in aesthetic judgement would fall foul of this constraint. As such, the notion of dependent beauty puts a restriction on the scope of free beauty: it requires, to put it in Danto’s terms, that the beauty of a church must be conceptually constrained by (and hence ‘internal to’) its meaning or purpose. This idea of conceptually constrained beauty, as anyone who has followed Danto’s recent work will recognise, is remarkably similar to Danto’s own account of ‘internal beauty.’6 Indeed, it is axiomatic to both Danto’s philosophy of art and philosophy of criticism that works of art should be judged for the appropriateness or ‘fit’ of their form of presentation to the content thereby presented. If this is correct, there is nothing that need trouble Danto in Kant’s notion of dependent beauty—on the contrary, it is Kant’s notion of dependent (rather than free) beauty to which he should look. Of course, not all works of art are dependently beautiful for Kant—there are free artistic beauties, such as Kant’s notorious ‘designs alagrecque` , the foliage on borders or on wall paper, etc.’—though the vast majority will be: if a work has any semantic content at all, it will be dependently beautiful (or ‘beautiful as ...’) for Kant.7 But this need not trouble Danto either, given that not all art is internally beautiful on his own account. Similarly, Danto routinely conflates Kant’s central claim that aesthetic pleasure is undetermined by concepts with the more contentious claim that

June 21, 2008 11:25 MAC/NWA Page-246 9780230_220478_14_cha13 Diarmuid Costello 247 it is independent of concepts. This would make Kant’s conception of pure aesthetic judgement (judgements of free rather than dependent beauty) ‘non- conceptual’ in the strong sense of conceptually empty as opposed to the weaker sense of conceptually unconstrained.8 But Kant’s commitment to safeguarding the freedom of aesthetic judgement only commits him to the latter, namely that beauty cannot be directly inferred from the fulfilment of any concept or rule—a point since widely associated with Frank Sibley’s work on aesthetic concepts (the thought that no array of non-aesthetic properties ever suffices, logically, to license the inference of an aesthetic one).9 Then there is the contentious equation of what Greenberg means by ‘formal’ (the design features of works of art) with what Kant means by this (generally, the aprioriconditions of aesthetic judgement, though also, in the notorious remarks in CJ §13–14, the spatio-temporal configuration of works of art); and local infidelities such as the claim that judgements of taste are universally valid for Kant when, as Kant believes we can never know that we have made such a judgement (given the stringent conditions such judgements must ful- fil and our opacity to our own motives and interests), the best we can do is lay claim, in full expectation of disagreement, to the assent of others. All these come together in Danto’s use of the generic ‘Kant-Greenberg aesthetics of form.’10 The foregoing examples show how problematic that equation is. That said, in fairness to Danto, his recent ‘Embodied Meanings’ paper is also more attentive than before to various disparities between Greenberg and Kant, distinguishing between Kant’s concern (across art and nature) with beauty and Greenberg’s concern (exclusive to art) with quality; and acknowledging that while for both aesthetic judgement is non-conceptual (albeit in the sense I have contested above, which is not true of Kant and may not even be true of Greenberg), it is not therefore non-cognitive.Moreover, Danto now grants that many of Greenberg’s views on criticism owe more to Hume than to Kant, and accepts that one ought to calibrate claims about what counts as ‘aesthetic,’ ‘anti-aesthetic’ or ‘non-aesthetic’ to the relative restrictiveness of the aesthetic theory in question: Though one will find Duchamp’s ready-mades anti-aesthetic on a narrowly formalist aesthetics that turns on the perceptual features of works of art, this leaves entirely undetermined whether the readymades might count as aesthetic according to some richer conception of aesthetic value. Hence it is no longer clear, even for Danto, that works such as Duchamp’s or Warhol’s shipwreck aesthetic theories tout court; it all depends what one means by ‘aesthetic.’11 All this is to be welcomed, but rather than dwell further on these refinements to Danto’s previous positions vis-a-vis` Greenberg and Kant here, I want to focus for the remainder of this chapter on what is new in Danto’s relation to Kant, namely his attempt to recoup Kant’s conception of aesthetic ideas for his own theory of art. I am sympathetic to this move to the extent that it looks like the right place to start in any discussion of the applicability of Kant’s aesthetics to recent art.12 About §49 Danto now claims: ‘the Kant of

June 21, 2008 11:25 MAC/NWA Page-247 9780230_220478_14_cha13 248 Danto and Kant, Together at Last?

§49 is not the Kant of Kantian aesthetics, which is based almost entirely on the Analytic of Taste. I owe it to Kant—and to myself—to show how close my views are to his in this section of his book.’13 The plausibility of this claim is what I want to consider. I shall briefly set out what Danto takes Kant to mean by an aesthetic idea, with a view to providing a more detailed account of my own later. In nuce, Danto regards aesthetic ideas as Kant’s recognition of the inadequacy of taste in the domain of art and, hence, of the fact that art requires more than taste, namely ‘spirit,’ the inner animation conferred by ideas. What most interests Danto about such ideas is that they are given sensory embodiment; they are aesthetic in the sense, that is, of being ‘given to sense’: ‘Kant has stumbled onto something that is both given to sense and intellectual—where we grasp a meaning through the senses.’14 Danto notes in passing how, on Kant’s theory, the way this works relies on indirect meaning of the sort that characterises both irony and metaphor—something I shall come back to, given the centrality of metaphor to both accounts. He also notes that presenting ideas in this way ‘expands’ the imagination for Kant, though he does not elaborate on what this might mean. Finally, he equates Kant’s conception of aesthetic ideas, shorn of the baggage of taste and aesthetic judgement, with his own notion of embodied meanings:

Aesthetical ideas have nothing much to do with the aesthetics of taste, and they are what is missing entirely from Greenberg’s agenda, who seldom spoke of meaning in his discussions of quality in art. [ ...] My own view is that the relationship of aesthetics to art was always external and contingent [ ...] But the theory of art as embodied meanings—or the ‘aesthetical presentation of ideas’—makes it clear how aesthetic qualities can contribute to the meaning of the work that possesses them.15

Danto’s attempt to hive off ‘the Kant of §49’ from the rest of the third Critique (‘the Kant of Kantian aesthetics’) is tailor-made to goad Kantians; and though I am neither sympathetic to it myself nor regard it as faithful to Kant’s own position, it would take me too far from my present purposes to address it here, given how complex and contentious the relation between formalism and expressionism in the third Critique still is.16 Instead, I want to examine whether Danto’s identification of Kant’s ‘aesthetic presentation of ideas’ with his own ‘embodied meanings’ is credible, taken on his own terms.

III. Late Danto’s ‘aesthetic turn’?

Giving an account of the relation, assuming there is one, between aesthetic ideas and ‘embodied meanings’ involves weighing the extent to which Danto may have been implicitly committed to an aesthetics of sorts all along.

June 21, 2008 11:25 MAC/NWA Page-248 9780230_220478_14_cha13 Diarmuid Costello 249

I believe that he has. But to show this requires two steps. In this section I consider what Danto himself says about the relation between aesthetics and the concept of art—specifically, whether aesthetic properties are necessary to the definition of art. Danto holds that they are not. In the next I argue that this conflicts with basic features of the ontology he has held since Transfigu- ration of the Commonplace. If this is correct, the latter has to win out, on pain of Danto having no way to differentiate works of art, properly so-called, from what he calls ‘mere representations,’ according to his own theory. Danto’s protestations to the contrary notwithstanding then, I conclude that his conception of what a work of art is has had an ineliminable, if implicit, aesthetic dimension from the outset. This lays the ground for a direct comparison of Kant and Danto on our cognitive and affective relation to works of art. Since The Abuse of Beauty appeared, in 2003, Danto has been concerned primarily with the legacy of his major work, Transfiguration of the Common- place, on the one hand, and a reconsideration of aesthetics, notably Kant’s aesthetics—largely ignored in that work—on the other. As a result, several commentators, myself included, have drawn attention to an apparent, if heavily qualified, ‘aesthetic turn’ in Danto’s late work, exemplified by his incorporation of the idea of ‘internal beauty’ into his theory of art. This is the concession that beauty may be a necessary feature of some works of art after all, but only insofar as it is ‘internal to’ their meaning. This would be the case when a work’s meaning requires a beautiful presentation, which must therefore figure in its interpretation as art.17 Failing that, a work’s beauty will remain adventitious to its meaning and hence, strictly speaking, irrelevant to it as art. On Danto’s account it would be incorrect to say that the work itself is beautiful in such cases; rather it is the ‘mere real thing’ from which the work cannot be visually discriminated, but to which it nonetheless cannot be reduced, that is beautiful in such instances. The mistake is to think the work itself must therefore be. For Danto, Duchamp’s Fountain and Warhol’s Brillo Boxes would be examples of the latter, were they to be admired—perhaps formally—for their beauty as objects; they are ‘externally’ beautiful. As Danto puts it, they would be merely ‘freely’ beautiful, that is beautiful in a way that is orthogonal to their meaning as works of art. Maya Lin’s Vietnam Veteran’s Memorial, Robert Motherwell’s Elegies to the Spanish Republic and Jacques-Louis David’s Death of Marat, by contrast, would be examples of the former; because their beauty is required by their meaning as works of art they are internally beautiful. In the case of Motherwell, for example, it is internal to the meaning of these paintings as works of mourning for the death of a political ideal (Motherwell himself described them as a ‘lamentation or funeral song,’ ‘his private insistence that a terrible death happened that should not be forgot’) that they be beautiful; the sombre nature of their beauty—their limited palette of forms and restricted use of colour—works, in Danto’s terms, to ‘colour’ or ‘inflect’

June 21, 2008 11:25 MAC/NWA Page-249 9780230_220478_14_cha13 250 Danto and Kant, Together at Last? our attitude to what has been lost in such a way as to dispose us, the work’s viewers, to see it as something to be mourned.18 An analogous case could be made for Gerhard Richter’s 18 Oktober 1977 (1988) cycle of paintings—based on police photographs of members of the Baader-Meinhof gang found dead in Stammheim prison—given Richter’s intention to transform the ‘horror’ of these documents into ‘something more like grief,’ and thereby open up a space for mourning.19 Given, however, that it is clearly not the case that all works of art need be internally beautiful in Danto’s sense, beauty remains external to art’s definition nonetheless. But what of aesthetic properties more generally—could it be that having some aesthetic property is a necessary feature of works of art after all? Danto thinks not. But that he does, I believe, is grounded on certain assumptions Danto seems never to have questioned. Only if one begins from the premises that (i) all aesthetic properties are perceptual, and (ii) it is the primary task of aesthetic theories to distinguish art from non-art on the basis of such properties, is there reason to think (as Danto does) that drawing attention to the non-manifest differences that must obtain between art and non-art, in the event that the two are indiscernible, suffices to demonstrate the inadequacy of aesthetic theories of art. Clearly, this would compromise such theories if one construes them as essentially sortal mechanisms for distinguishing art and non-art perceptually. But why suppose that? Many aesthetic theories are more concerned to define what, if anything, distinguishes the experience of art than they are to specify what is required to identify it. Indeed, there is no prima facie incompatibility between granting the force of those considerations Danto adduces for discriminating art, in the event of indiscernibility, whilst holding an aesthetic theory nonetheless: one can accept Danto’s strictures on identifying art—the necessity for a background knowledge of art history and theory that ‘the eye cannot descry,’ for example—while retaining an aesthetic theory of the experience of art once that identification is in place. Aesthetic theories need not, as formalist theories arguably do, turn on those perceptual features of objects, narrowly construed, that may or may not (as the case may be) be used to discriminate art from non-art, as Danto supposes. If this is correct, aesthetic theories are ill-characterised in Danto’s characteristic thought-experiments. Consider wit, understood as a piquant relation between ideas, as a non-perceptual aesthetic property that many contemporary works of art exhibit. One does not literally see a work’s wit‚ the way one sees that it is so-high, so-coloured or made of such and such materials; one appreciates it when one appreciates the work as a whole.20 But why should the fact that one does not directly perceive a work’s wit by means of the senses warrant the conclusion that it is not an aesthetic property, as opposed to the more modest, and more warranted, conclusion that it is not a perceptual property, narrowly construed? In sum, to begin by defining aesthetic properties perceptually looks contentious, and perhaps even stipulative.21 One might,

June 21, 2008 11:25 MAC/NWA Page-250 9780230_220478_14_cha13 Diarmuid Costello 251 with greater justification, define art’s aesthetic properties functionally, as any feature of a work that engages the imagination in some particular way. Be that as it may, Danto maintains that all aesthetic properties are perceptual; more specifically, he takes them to be those perceptual features of a work employed rhetorically to dispose its viewers to see its subject matter, the meaning embodied in the work, in a particular light. For Danto, then, an aesthetic quality is any quality that ‘inflects’ or ‘colours’ the attitude of a work’s recipient to its subject matter. The point of construing aesthetic properties in this way is to encourage a wider diet of aesthetic properties than has traditionally been entertained: on a rhetorical conception, the list is in principle open-ended. Hence Danto’s frequent invocation in recent years of J.L. Austin’s dictum, from ‘A Plea for Excuses’: ‘How much it is to be wished that [ ...] we could forget for a while about the beautiful and get down instead to the dainty and the dumpy.’22 Nonetheless, while Danto may now be prepared to grant a wider variety of aesthetic properties, so conceived, a greater rhetorical role than previously in inflecting viewers’ attitudes towards the meaning of works of art, such properties remain as irrelevant, ontologically, as ever to a work’s existence as art: Although aesthetic properties may be a necessary feature of some works of art, if they inflect their viewers’ attitudes in ways that are internal to the meaning those works embody, they are not a necessary feature of all works of art, and so remain external to art’s definition nonetheless. Now, it seems to me that this conclusion ought to present a serious problem for Danto, for reasons internal to his own theory; for it appears to entail that there can be artworks that express no point of view towards their own content (and hence have no recourse to aesthetic properties understood as ‘inflectors’ of said content). To show that this is so, and why it ought to be a problem for Danto, I now turn to the ontology of art Danto set out in Transfiguration of the Commonplace.

IV. Danto on works of art as ‘embodied meanings’

Noel¨ Carroll has provided the most elegant summary of the ontology set out in Transfiguration of the Commonplace, one that has been endorsed by Danto himself. Carroll claims that, for Danto,

something x is a work of art if and only if (a) x has a subject (i.e., x is about something); (b) about which x projects some attitude or point-of-view (this may also be described as a matter of x having a style); (c) by means of rhetorical ellipsis (generally metaphorical ellipsis); (d) which ellipsis, in turn, engages audience participation in filling-in what is missing (an operation which can also be called interpretation); (e) where the works

June 21, 2008 11:25 MAC/NWA Page-251 9780230_220478_14_cha13 252 Danto and Kant, Together at Last?

in question and the interpretations thereof require an art-historical context (which context is generally specified as a background of historically situated theory).23

In what follows, I shall focus on the second, third and fourth of Carroll’s five conditions (b–d): the second speaks to the problem at hand (namely why Danto’s view that aesthetic properties are not necessary to art’s definition presents a problem for the integrity of his own theory); the third and the fourth speak to the parallels with Kant. ‘Aboutness’ (a) is self-evidently definitional of works of art conceived as ‘embodied meanings’: for a work to possess meaning requires minimally that it be about something or other. Danto has a number of arguments to this conclusion, which seek to show that the difference between a ‘mere real thing’ and a work of art that exactly resembles it must consist in the meaning that the latter, but not the former, may be supposed to convey, in virtue of their different histories of production—the most notorious featuring a doppelganger¨ of Rembrant’s Polish Rider produced by a freak explosion in a paint shop. Only the original, and not its unlikely counterpart, supports an interpretation (d). I do not propose to examine these arguments, or the plausibility of the thought-experiments that underwrite them—for the sake of argument, I propose to simply grant aboutness here.24 Nonetheless, as Danto points out, there are many everyday artefacts (such as maps or diagrams) that are also about something, and hence possess semantic content, without that sufficing to make them art. Danto calls this sub-class of ‘mere real things’ ‘mere representations’ and asks, ‘What must we add to the concept of representationality which will make the difference between ordinary representations and works of art?’25 More is required, and Danto argues that expressing some attitude, or point of view, towards what they are about (b) provides what is necessary to distinguish mere representations, which are only about what they represent, but do not project any point of view towards it, from works of art, which do. Consider, in this light, the contrast between a computer print out of a curve generated by plotting co-ordinates against two axes, and a perceptually indiscernible counterpart, also produced by computer, presented as art. While there might be good mathematical, demographic or scientific reasons for asking why the former presents the precise curve that it does, depending on what it represents, it would make no sense to seek an interpretation for why the diagram itself takes the form that it does: that is just the way diagrams produced this way look. But the perceptually identical work of art might credibly be interpreted as ‘mechanical’: that is, as intentionally aping the artlessness of the former to make some artistic point—perhaps, in line with Danto’s penchant for angry young artists, to pour scorn on touch and virtuosity in la belle peinture. In Danto’s words, such an artwork would ‘use the way the non-artwork presents its content to make a point about how that content is presented.’26

June 21, 2008 11:25 MAC/NWA Page-252 9780230_220478_14_cha13 Diarmuid Costello 253

In Danto’s full-blown example, a diagram of a portrait of Mme. Cezanne in Erle Loran’s book analysing the structure of Cezanne’s paintings is appropriated by Roy Lichtenstein to make a point about how Cezanne sees the world: where Loran tries to make the scaffolding of a particular Cezanne perspicuous by means of a diagram, Lichtenstein’s painting uses the former to make a point about the way the world—including the artist’s own wife—was perceived by Cezanne as so many geometrical figures and planes. Thus, although Lichtenstein’s painting may resemble a diagram, it is not in fact one; instead, it uses a diagrammatic idiom rhetorically, to express an attitude towards what it presents, and thereby colour its viewers’ attitudes accordingly. So, where Loran produces a diagram of a painting, Lichtenstein employs a diagrammatic idiom within painting rhetorically, ‘transfiguring’ his source material in the process. From this, and other, examples Danto concludes:

[W]orks of art, in categorial contrast with mere representations, use the means of representation in a way that is not exhaustively specified when one has exhaustively specified what is being represented. This is a use that transcends semantic considerations (considerations of Sinn and Bedeutung). Whatever Lichtenstein’s work finally represents, it expresses something about that content.27

Danto’s argument here, on which his later work depends, makes expressing some attitude or point of view towards their own meaning in this way a necessary feature of all works of art; it does so because expressing a point of view is what constitutes the difference between works of art and ‘mere representations.’ Earlier, I remarked that Danto’s claim that aesthetic properties are not a necessary feature of art creates a serious problem for his own theory. The difficulty should now be evident: if Danto’s argument in Transfiguration of the Commonplace is correct, then aesthetic properties must be a necessary feature of works of art according to his own theory; such properties are what enable artworks to express an attitude towards whatever they are about. According to his own conception of what an aesthetic property is, such properties mark the difference between ‘mere representations’ and works of art, properly so- called. And this is at odds with his contention that such properties play no necessary part in the definition of art. Hence, where I once thought the qualified ‘aesthetic turn’ of Danto’s late work might represent some kind of return of the repressed—a reconsideration of precisely those features of art that the method of ‘indiscernible counterparts’ had been designed to dispatch—this a view that I no longer hold.28 That I do not is because it now seems to me that Danto’s ‘aesthetic turn’—if turn it is—would be better characterised as an acknowledgement or extrapolation of features of his ontology that have been present all along. The more interesting (and certainly more provocative) question, it therefore

June 21, 2008 11:25 MAC/NWA Page-253 9780230_220478_14_cha13 254 Danto and Kant, Together at Last? seems to me, is whether Danto may not have been implicitly committed to an aesthetics compatible with Kant’s all along. To assess this, the next two sections consider their respective views of metaphor, which is where the comparison can best be made.

V. Danto on metaphor

On Danto’s account, the kind of rhetorical structure in which artworks preeminently trade is metaphor: to understand the work is ‘to grasp the metaphor that is [ ...] always there.’29 Artworks invite us to see one thing as, or in the light of, another. In doing so, the work’s viewers must ‘fill in’ (c–d) what the work leaves unsaid, thereby making connections for themselves that the work itself leaves implicit. On Danto’s account, the metaphors embodied in a work are closely tied to how the artist perceives the world; when the work is successful it brings its viewers to see whatever it is about from the point of view of the artist. The extent to which Danto conceives artists and artworks on the model of rhetoricians and rhetoric cannot be overstated; as he puts it in ‘Metaphor and Cognition,’ an essay subsequent to Transfiguration of the Commonplace, ‘The rhetorician uses metaphor to drive the hearer’s mind where he wants it to go. [ ...] Metaphors belong to the theory of manipulation.’30 The same applies to Danto’s conception of the artist. It is because Danto takes works of art to express the artist’s point of view towards a given subject, and in so doing to dispose their audience to a similar view, that Danto’s theory is at root a variant of expression theories of art.31 Works of art succeed in ‘colouring’ their viewers’ attitudes in this way, according to Danto, by means of ‘rhetorical ellipsis’—leaving something unsaid for their viewers to fill in, in much the same way that a joke’s recipient must make the connections that illuminate its punch-line for herself. Danto draws on Aristotle’s understanding of enthymemes to illuminate his case, and that he does reveal something fundamental about how he understands metaphor. An enthymeme is a syllogism with a missing premise (or conclusion) that yields a valid syllogism when what is missing is supplied. What is missing is taken to be an obvious truth, one that anyone within the target community could be counted on to supply—that regarding one’s own wife as just so many cones and pyramids, for example, leaves something to be desired. Thus, in ‘Metaphor and Cognition’ Danto professes to be doubtful that metaphors ‘ever, as metaphors, tell us something we do not [already] know.’32 On this account, grasping a metaphor requires that one already possess the background knowledge it depends on, if one is to be able to supply what it leaves unsaid. The important pragmatic consideration for Danto is that metaphors activate the reader/viewer/listener nonetheless. They must supply what is missing themselves, the rhetorical gain being that whatever they supply for themselves will carry far more weight, and hence persuade more readily, than anything provided for them would be capable of doing.

June 21, 2008 11:25 MAC/NWA Page-254 9780230_220478_14_cha13 Diarmuid Costello 255

Danto brings this Aristotelian discussion of rhetorical ellipsis in general to bear on metaphor in particular in the following way. On this conception, grasping a metaphor requires ‘finding a middle term t so that if a is metaphorically b, there must be some t such that a is to t what t is to b. Ametaphor would then be a kind of elliptical syllogism with a missing term and hence an enthymematic conclusion.’33 To grasp the metaphor is to complete the syllogism by ‘filling in’ what it leaves unsaid. What remains unclear, at least in Transfiguration of the Commonplace, is whether Danto himself endorses this conception of metaphor: :

The important observation [ ...] has less to do with whether Aristotle has successfully found the logical form of the metaphor than with the fact that he has pragmatically identified something crucial: the middle term has to be found, the gap has to be filled in, the mind moved to action.34

That said, some of Danto’s later remarks, in ‘Metaphor and Cognition,’ suggest that he does:

My sense is that metaphor must work in this way: the rhetorician demon- strates whatever it is that he wants the audience to believe whatever he is talking about is— and for this reason the connection must be as obvious as the suppressed premise of the enthymeme or the stifled answer to the rhetorical question.35

Now, assuming that Danto does in fact endorse this conception of metaphor as incomplete syllogism, he would be right to hold that metaphors require their hearers to make connections for themselves, but wrong to suggest this entails supplying a determinant missing premise. If that were true, metaphors would be finite, and hence amenable to paraphrase. Once more, however, Danto’s position on this remains hard to pin down. Danto seems to hold, in line with contemporary theories of metaphor, that if artworks are metaphorical, they cannot admit of paraphrase:

no paraphrase or summary of an artwork can engage the participatory mind in all the ways that it can: and no critical account of the internal metaphor of the work can substitute for the work inasmuch as a description of a metaphor simply does not have the power of the metaphor it describes [ ...] it is rather the power of the work which is implicated in the metaphor, and power is something that must be felt.36

Notice, however, that Danto does not say metaphors cannot be exhaustively paraphrased—what he says leaves that possibility open. What he says, in effect, is that even if they can, what cannot be captured by paraphrase, no

June 21, 2008 11:25 MAC/NWA Page-255 9780230_220478_14_cha13 256 Danto and Kant, Together at Last? matter how exhaustive, is the richness of the cognitive experience of grasping the metaphor, which may only be had—interestingly, for the comparison with Kant, Danto’s word is felt—first-hand. Thus, in Danto’s main example, depicting Napoleon as a Roman emperor is not only meant to elicit a particular view of Napoleon in the work’s viewer, namely to view Napoleon in the light of Caesar, but also meant to occasion a complex set of feelings about that possibility. So garbed, the figure is a metaphor of dignity, authority, grandeur and political power. A depicted as B, or under the attributes of B, has the same metaphorical structure as ‘Juliet is the Sun’ or ‘Man is a Wolf;’ we are encouraged to see Napoleon in the light of Caesar, as we are encouraged to see Juliet under certain attributes of the sun, man under certain attributes of a wolf and so on. If successful, by making the connections and associations the work invites, the audience will be brought to feel towards Napoleon, what they would have felt towards Caesar: amongst other things, presumably, awe—perhaps mingled with respect or fear. Hence, the aim is not solely, or even predominantly, to bring the viewer to entertain some simple proposition, namely to see Napoleon as Caesar, but to feel towards what is depicted, what would have been felt were what is depicted the case; in this way the audience is actively solicited to adopt the point of view of the work.37 Now, it seems to me that Danto’s own account of how this takes place in the metaphor of Napoleon-as-Caesar conflicts with his recourse to Aristotle on metaphor as enthymeme. His own description of what this involves suggests a far richer process than merely supplying some determinant missing premise; for this fails to capture the non-paraphraseable, non-transferable dimension of being moved by the work that he draws attention to. In this respect, Danto’s best descriptions of metaphors at work often exceed the theory into which he appears, intermittently, to be trying to shoehorn them. This being so, I now turn to the conception of metaphor at play in Kant’s theory of art, to see whether it is better able to accommodate the richness of metaphorical meaning, and the experience thereof, that Danto himself draws attention to.

VI. Metaphor in Kant

The most illuminating account of the conception(s) of metaphor implicit in Kant’s theory of art that I know of is Kirk Pillow’s ‘Jupiter’s Eagle and the Despot’s Hand Mill: Two Views of Metaphor in Kant,’ and I am indebted to Pillow’s analysis in what follows.Pillow’s central claim is that there are two competing conceptions of metaphor in Kant that need to be distinguished: one based on what Kant calls ‘symbolic hypotyposis’ (essentially a kind of analogy) as when beauty serves as a ‘symbol’ of morality (CJ §59) and a second, far richer, account that Pillow wants to show is compatible

June 21, 2008 11:25 MAC/NWA Page-256 9780230_220478_14_cha13 Diarmuid Costello 257 with contemporary ‘interactionist’ theories of metaphor, of the sort pio- neered by Max Black.38 It is the latter that is of most interest here, though I will briefly sketch the former, as it will help illuminate Danto’s account. On Pillow’s account, this takes the form of a (qualitative) proportional analogy, of the form A is to B as C is to D [A:B::C:D]; it consists, essentially, in drawing attention to an analogy between relations in two domains. In Kant’s example, a hand mill serves as a metaphor (in this limited, analogical sense) for a tyrannical state, not by virtue of exhibiting tyranny, obviously, but by virtue of their shared activity of grinding down resistance. In Pillow’s words, ‘the material fed through a hand mill is to its operator as the subjects under an absolute monarch are to the despot [ ...]whenyouseea state apparatus mangling its subjects’ freedoms the way a hand mill crushes through force, you know you are dealing with a tyrant.’39 It is this relational isomorphism between the two that allows the hand mill to symbolise the power of a despotic state—the problem being that this allows metaphors to be reduced to complex similes, the basis of which may be precisely specified. Now this is similar, in one crucial respect, to the theory of metaphor that Danto claims to find in Aristotle. Indeed, it is striking, given Danto’s recourse to Aristotle’s Rhetoric, that Pillow traces the notion of metaphor as symbolic exhibition, Kant’s first conception of metaphor, back to Aristotle’s notion of implicit analogy in the Poetics.40 Though ‘metaphor’ conceived analogi- cally has a crucial role in Kant’s critical philosophy of facilitating reflection on rational ideas, it runs the risk of reducing metaphors to determinant, and hence in principle finite analogies: grasp the analogy and you grasp the metaphor. On this conception of metaphor, the finite meaning it commu- nicates may be exhaustively specified, as in Kant’s own hand mill example, and this is the feature it shares with the theory Danto derives from Aristo- tle: both can be completed by providing the missing determinant premise or relational term. And this runs counter to contemporary views that metaphors resist paraphrase because their content cannot be exhaustively specified in either of these ways. Now contrast this with the second conception of metaphor that Pillow claims to find in Kant. On the foregoing account metaphor is not so much generative as revelatory; it reveals previously existing, if insufficiently remarked, similarities, rather than creating new affinities. Pillow contrasts what he calls the ‘weak’ creativity of implicit analogy with the ‘strong’ creativity of metaphor proper. On the latter conception, metaphors create new affinities between what they bring together, thereby bringing new thoughts into being. Paul Ricoeur has this stronger conception of metaphor in mind when he describes metaphor as ‘world-disclosive’: it affords the possibility of new ways of carving up the world, thereby allowing new patterns of salience to emerge.41 It is this stronger, productive sort of creativity that Pillow attributes to aesthetic ideas, and elucidates as a kind of

June 21, 2008 11:25 MAC/NWA Page-257 9780230_220478_14_cha13 258 Danto and Kant, Together at Last? proto-‘interactionism’. On the interactionist theory of metaphor, ‘Man is a Wolf’ encourages us to view the principal subject, man, in the light of our ‘system of associated commonplaces’ about the subsidiary subject, wolf, thereby creating novel associations between previously unrelated terms. We are invited, in Black’s terms, to see the principal subject through its metaphorical expression. The theory is ‘interactive’ in the sense that our perception of the nature of the principal subject has a bearing on what aspects of the subsidiary subject appear relevant (the fact that wolves hunt in packs, for example, rather than the fact that they have cold snouts), and those connotations then generate further reflection on both the principal and the subsidiary subjects in turn. Although, in common with the theory of metaphor as implicit analogy, the interactionist account requires a metaphor’s target audience to share broadly similar sets of connotations and associations about men and wolves—that is, their ‘associated commonplaces’ about wolves need to be broadly compatible, internally, with one another, and likewise for their associations about men—the theory does not require each member of the audience to have the same connotations for each term. On the contrary, because the theory assumes that individuals will have somewhat different, if not incompatible, overlapping sets of associations, the theory secures the openness and inexhaustibility of metaphorical meaning at the ground level; as such, the richness of a metaphorical meaning will be dependent to a significant degree on the knowledge, sophistication and interpretative elan´ of its audience—as with artistic meaning in general.

VII. Kant on aesthetic ideas

Armed with Pillow’s two conceptions of metaphor, I now want to consider Kant’s claim that artworks express ‘aesthetic ideas.’ To put this in the most straightforward terms possible, an aesthetic idea is Kant’s account of what is distinctive about either the content of a work of art, or the way in which it presents that content. What is distinctive about the content of works of art is either that they present concepts that may be encountered in experience, but with a completeness that experience never affords, or that they communicate ideas that cannot—in principle—be exhibited in experience.42 What is distinctive about the way in which works of art present such content is that they imaginatively ‘expand’ the ideas presented, by metaphorically embodying them in sensible form. For rather than seeking to present the idea itself directly to intuition, which would be impossible—ideas being by definition what cannot be exhibited in experience for Kant—aesthetic ideas present the ‘aesthetic attributes’ of their object, thereby expressing an idea’s ‘implications’ and ‘kinship with other concepts.’43 To take Kant’s own example: ‘Jupiter’s eagle with the lightning in its claws’ expands the idea of God’s majesty by presenting it aesthetically. What Kant

June 21, 2008 11:25 MAC/NWA Page-258 9780230_220478_14_cha13 Diarmuid Costello 259 calls the ‘logical’ attributes of an idea, in this case God, would be those in virtue of which it fulfils a concept, in this case majesty. Jupiter’s eagle with the lightning in its claws, by contrast, is a metaphorical expression of those same attributes, through which we are encouraged to view God’s majesty in the light of the wealth of thoughts provoked by Jupiter’s eagle, thereby opening up a rich—and in principle endless—seam of further associations. Through the interaction of the work’s idea or theme (God’s majesty), the image by which it is presented (Jupiter’s Eagle), and the specific aesthetic attributes through which the latter encourages us to view the former (the lightning bolt in the Eagle’s claws, etc.), the embodiment of ideas in art provokes ‘more thought’ than a conceptual elaboration of their content could hope to facilitate. Works of art ‘expand’ the ideas they embody in this way because, in Kant’s words, the aesthetic attributes through which they present ideas

prompt the imagination to spread over a multitude of kindred presentations that arouse more thought than can be expressed in a concept determined by words. These aesthetic attributes yield an aesthetic idea [ ...] its proper function is to quicken the mind by opening up for it a view into an immense realm of kindred presentations.44

In doing so, aesthetic ideas might be thought to achieve the impossible: they allow works of art to present rational ideas in sensuous form. Con- sider Delacroix’s Liberty Leading the People (1830) as a sensible embodiment of the idea of freedom: the aesthetic attributes through which freedom is embodied in the guise of ‘Liberty,’ and shown leading her people to victory—fearlessness, spontaneity, resoluteness, leadership, all attributes of an active self-determining will—while holding a flag, symbol of freedom from oppression, aloft in one hand, and clutching a musket in the other, serve to ‘aesthetically expand’ the idea of freedom itself. By presenting freedom in the guise of ‘Liberty,’ in this way, freedom is depicted concretely as something worth fighting for—indeed, as something requiring courage and fortitude to attain. This is what Kant means when he claims that the expression of ideas in art ‘quickens’ the mind, by freeing imagination from the mechanical task of schematising concepts of the understanding. No longer constrained to present concepts of the understanding in sensible form, as it is in determinate judgement, aesthetic ideas free the imagination to move swiftly over a multitude of related thoughts. By doing so, aesthetic ideas stimulate the mind, albeit in a less structured way than determinate thought, enabling us to think through the ideas presented in a new light. Now it might be objected that the foregoing account only works because it takes a representational painting as its object, and that this will be of little use to art in its contemporary context of non-traditional media and

June 21, 2008 11:25 MAC/NWA Page-259 9780230_220478_14_cha13 260 Danto and Kant, Together at Last? forms, which is where I need to demonstrate the purchase of Kant’s theory. To show that this is not the case, and that the argument runs irrespective of whether one takes a figurative painting as one’s example, my second example will be a work from the opposite end of the artistic spectrum. But before turning to that, I want to acknowledge the force of a question that Pillow has raised about Kant’s account, as I have presented it thus far: namely where—precisely—is the aesthetic idea to be located? In Kant’s example, is it in the image of the Jupiter’s Eagle with the lightning in its claws, or is it in the ‘aesthetically expanded idea’ to which reflection on that image gives rise? Pillow opts for the latter: ‘Kant indicates that the presentation of such attributes yields an aesthetic idea, which suggests [ ...] the aesthetic idea is the aesthetically expanded idea of majesty resulting from the reflection that the eagle image prompts.’45 The presentation itself obviously initiates this reflection, but what Pillow calls ‘the fulfilled aesthetic idea’ is the enriched conception of majesty that emerges. The same is true of the terms in which I presented Delacroix’s Liberty above; the aesthetic value of an ‘aesthetic idea’ resides in the richness and longevity of the imaginative play to which it gives rise. From Pillow’s perspective this is important because, in line with interactionist theories, it secures a role for both creator and recipient to play in the metaphorical exchange the work elicits. From my perspective it is important because it suggests a way of reading Kant that rings true of our imaginative engagement with art, indeed perhaps above all recent art, given its often bewildering array of non-traditional media and forms, by putting the experience of finding oneself imaginatively stimulated and cognitively stretched centre-stage. If this is true, it is a historical irony of the sort that Danto would be the first to appreciate, given that it was art after modernism (or what Danto calls art ‘after the end of art’) that his own theory was framed to accommodate. Kant’s account, as I have presented it, has this much in common with Danto’s: the creation of a work of art is, in part, the creation of an interpretative context—of something that gives rise to a rich interpretative play. On Kant’s account, however, this cannot be reduced to ‘filling in’ some determinant missing premise, on pain of ceasing to be aesthetic, but must involve an open-ended imaginative exploration of the enriched idea that emerges in reflection on the work, in which there is in principle always more to be discovered. Though not, broadly speaking, incompatible with Danto’s cognitivism, on Kant’s account aesthetic ideas are not governed by a syllogistic logic of the sort that Danto finds in Aristotle, and are not to be confused with the finite way in which symbols, such as the hand mill, function in Kant’s own account. This raises a question as to whether Kant’s conception of aesthetic ideas may not be better able to accommodate Danto’s richer examples than his own theory. To address this I turn, in closing, to my contemporary example.

June 21, 2008 11:25 MAC/NWA Page-260 9780230_220478_14_cha13 Diarmuid Costello 261

VIII. Aesthetic ideas, a contemporary example

My contemporary example is Art & Language’s Index 01 (1972). In addition to demonstrating the applicability of Kant’s theory to art after modernism, I would like this example to capture something of what it means, in practice, for a work to ‘yield’ an aesthetic idea. My choice of a work by Art & Lan- guage (and from just this period) is far from innocent, given that their work from this period might be thought to show, as well as any individual works of art might, the inapplicability of Kant’s aesthetics to recent art. Against this perception, I suggest that this work may be understood, in the terms advanced here, as a metaphorical embodiment of the idea of an exhaustive catalogue. Documenta Index consists of a cross-referenced index of the group’s writings on art to that date, and of the relations between them. Though it had various later incarnations, it originally took the form of eight small metal filing cabinets, displayed on four grey plinths, consisting of six tray-like drawers each, containing both published writings and unpublished writings, some of which raised the question of their own status as artworks. These were hinged one on top of the other in a series of nested sequences determined alphabetically, and sub-alphabetically, in terms of their order and degree of completion. The cabinets and their contents were displayed together with an index listing their contents in terms of three logical relations (of compatibility, incompatibility and incomparability) believed to obtain between them.46 The latter was papered directly onto the walls of the room in which the work was displayed, as if in an attempt to provide an ‘external’ vantage that would render the work’s internal relations perspicuous. At least in terms of its rhetoric of display and address, such a work seems to propose not only an exhaustive catalogue of the group’s writings to date, which is feasible, being finite, but to also aspire to map a set of logical relations between those writings. But the latter is something that can only exist as an idea, in Kant’s sense, because there are in principle always further relations to be mapped, were we acute enough to spot them, and had we sufficient time (and patience) at our disposal. Moreover, by embodying the idea of a self-reflexive catalogue, the project of the work makes internal reference to ideas of infinite relationality, discursivity, even socialibilty—another strata of ideas that cannot be directly presented—in that the production of the index itself creates a further layer of relations to be mapped, which would then have to be mapped in turn, and so on ad infinitum, in order to realise its goal. As such, undertaking the work itself makes its goal unrealisable. Nonetheless, by embodying this aspiration in sensible form, this seemingly austere—even ‘administrative’—work of art opens up a potentially limitless array of imaginative associations: to lists, taxonomies, and typologies; to attempts at self-documentation, self-reflexivity, and (ultimately) to ideals of complete self-knowledge or transparency; to machines for self-replication,

June 21, 2008 11:25 MAC/NWA Page-261 9780230_220478_14_cha13 262 Danto and Kant, Together at Last? self-generation, recursivity, and even perpetual motion; to conversation, col- laboration, interaction, study, and learning; and, of course, to various regimes of archiving, cataloguing, and the like.47 By doing so, the work imaginatively ‘expands’ the ideas it embodies in ways that Kant’s theory of aesthetic ideas is well placed to capture: it ‘prompts the imagination to spread over a multitude of kindred presentations that arouse more thought than can be expressed in a concept determined by words.’ It is just such imaginative engagement with sensibly embodied ideas—far removed from the astringent formalism typically attributed to the third Critique as a reception theory—that I would like to see retrieved for contemporary debates about art.48 Moreover, although Kant conceived of fine art in representational terms for historical reasons, there is nothing in his account of aesthetic ideas that requires art be representational, in a narrow sense, as my second example is intended to show.49 All Kant’s account requires is that artworks indirectly present ideas to sense, and in doing so engage their viewers in imaginatively complex ways—and there does not seem to be anything wrong with that thought in the light of more recent art that Kant could not have envisaged.50 Whether Kant’s account, schematic as it is, is ultimately better able to accommodate the richness of contemporary art than Danto’s is not something I claim to have demonstrated here. More would be required to establish that; though the fact that this possibility can be seriously raised at all, in virtue of his theory’s capacity to encompass both the cognitive and the affective dimensions of our engagement with art, is itself worth remarking. What I do take myself to have demonstrated, against the orthodox receptions of both Danto and Kant, is how much more their theories have in common than is typically acknowledged, and the costs that attach to too narrow, too exclusively cognitivist, a conception of what it is to fully engage with a work of art.

Acknowledgements

I would like to thank Charles Harrison for his expertise on Index 01, Arthur Danto for correspondence relating to this chapter, and Kathleen Stock for helpful comments on an earlier draft.

Notes

1. Danto, After the End of Art, Chapters 4–5. 2. Danto, ‘Embodied Meanings,’ 126–127 3. Costello, ‘Greenberg’s Kant,’ 223–224. 4. Danto, ‘Embodied Meanings,’ 123. 5. Kant, Critique of Judgement, §16, Ak. 229, 76. Citations from the third Critique give the section number, followed by the pagination of the Akademie Ausgabe, followed by the pagination of the Pluhar translation. 6. Danto, The Abuse of Beauty, Chapter 4, discussed below.

June 21, 2008 11:25 MAC/NWA Page-262 9780230_220478_14_cha13 Diarmuid Costello 263

7. That is, the judgement would take the form ‘This is a beautiful x’ and not ‘This x is beautiful.’ Only the former judges x to be beautiful as an x, thereby taking its beauty to be constrained by its being an x. 8. For more on this, see Janaway, ‘Kant’s Aesthetics.’ 9. ‘Aesthetic Concepts,’ in Sibley, Approach to Aesthetics. 10. Danto, ‘Embodied Meanings,’ 126. 11. On this possibility, see Shelley, ‘The Problem of Non-Perceptual Art.’ 12. See also Guyer, ‘From Jupiter’s Eagle to Warhol’s Boxes.’ Though I share Guyer’s intuition about the broad compatibility of Kant’s and Danto’s theories of art, I find several of the specific parallels he draws strained, notably that involving aesthetic pleasure—which Guyer is forced to claim Danto thinks so obvious as to not require comment. 13. Danto, ‘Embodied Meanings,’ 126–127. 14. Ibid., 127. 15. Ibid., 127–128. 16. The literature on the relation between formalism and expression in Kant’s third Critique is substantial, but Kenneth Rogerson’s work is particularly interesting to the point at hand: Rogerson argues that Kant’s formalism is not only compatible with his theory of aesthetic ideas but is only coherent in virtue of the theory of aesthetic ideas that completes it. See Rogerson, Kant’s Aesthetics, especially pp. 156–165. 17. On the notion of ‘internal beauty,’ see Gilmore, ‘Internal Beauty.’ 18. See Sarah Lucas, MoMA Highlights, 244. 19. Richter, The Daily Practice of Painting, 189. 20. For a view compatible with that put forward here, see Shelley’s ‘The Prob- lem of Non-Perceptual Art,’ 371–372. See also Carroll, ‘Non-Perceptual Aesthetic Properties,’ 418. 21. On this point, see Shelley’s critique of Carroll in ‘The Problem of Non-Perceptual Art,’ 377. For Caroll’s response, see ‘Non-Perceptual Aesthetic Properties,’ 422–423. 22. Danto cites this frequently, most recently in ‘Embodied Meanings,’ 125. See Austin, ‘A Plea for Excuses,’ 131. 23. Carroll, ‘Essence, Expression and History,’ 80. 24. On Danto’s use of thought-experiments, see Wollheim, ‘Danto’s Gallery of Indis- cernibles.’ I make use of Wollheim’s argument in Costello, ‘Whatever Happened to ‘‘Embodiment’’.’ 25. Danto, Transfiguration of the Commonplace, 139. 26. Ibid., 146. 27. Ibid., 147–148. Since Transfiguration, Danto has appealed to Frege’s notion of Farbung, or ‘colouration’ to capture this pragmatic aspect of meaning, pertaining to the way something it presented rather than what is presented. See Danto, Transfiguration, 162–164, and The Abuse of Beauty, 121–122. 28. I presented Danto’s turn to aesthetics as a feature of his recent work’s ‘lateness’ in ‘On Late Style.’ 29. Danto, Transfiguration, 172. 30. Danto, ‘Metaphor and Cognition,’ 74. 31. On this, see Carroll, ‘Essence, Expression and History,’ 85–87. 32. Danto, ‘Metaphor and Cognition,’ 86–87. 33. Danto, Transfiguration, 171. 34. Ibid. 35. Danto, ‘Metaphor and Cognition,’ 85.

June 21, 2008 11:25 MAC/NWA Page-263 9780230_220478_14_cha13 264 Danto and Kant, Together at Last?

36. Danto, Transfiguration, 173–174. 37. In his responses to the remarkable ‘Online Conference in Aesthetics’ (http://artmind.typepad.com/onlineconference/2007/02/arthur danto co.html), Danto cites this example as an aesthetic idea. See ‘Transfiguration Transfigured,’ 4–5: http://www.vanderbilt.edu/AnS/philosophy/events/.OCA/DantoDanto Conference.pdf. 38. See, for example, Black, ‘Metaphor,’ ‘More About Metaphor,’ and ‘How Metaphors Work.’ 39. Pillow, ‘Two Views of Metaphor,’ 194. 40. Ibid., 195; Aristotle, Poetics, §21, 1457, b7–29. 41. See Ricoeur, ‘The Metaphorical Process,’ 151–152, and The Rule of Metaphor, 197, as quoted in Pillow, ‘Two Views of Metaphor,’ 199. 42. Kant, Critique of Judgment, §49, Ak. 314, 182–183. 43. Ibid., Ak. 315, 183. 44. Ibid., Ak. 315, 183–184. 45. Pillow, ‘Two Views of Metaphor,’ 200. 46. On the index itself these relations were symbolised, respectively, by ‘+,’ ‘−,’ and ‘T.’ The latter stood for ‘Transformation,’ indicating that the documents in question did not occupy the same logical or ethical space and hence were incomparable. On Index 01, see Harrison, ‘The Index as Art-work,’ 65. 47. Harrison’s writings on Index 01 stress the degree to which it revolves around changed ideas of art’s production, display, address, and potential publics. As such it might be thought to embody a further series of ideas about community and communication, and even—in true avant-garde spirit—a sociability or society to come, in sensible form. See ‘The Index as Art-work.’ 48. One aesthetician who has consistently brought Kant’s aesthetics to bear on such debates is Paul Crowther, as has the art theorist Thierry de Duve. I take issue with de Duve’s appeal to Kant on pure aesthetic judgement in ‘Retrieving Kant’s Aesthetics.’ 49. For a detailed examination of whether Kant is committed to a representational concept of art, see Allison, Kant’s Theory of Taste, 290–298. 50. I am aware that more remains to be said about the question of form:specifically, whether the form of much contemporary art can be squared with the more restrictive aspects of Kant’s formalism. Unfortunately, the debates around Kant’s formalism are too complex to go into here. For a defence of Kant’s aesthetics against the charge of formalism, see Guyer, Kant and the Claims of Taste, 199–210. and Allison, Kant’s Theory of Taste, 131–143. For a defence of Kant’s formalism, see Crowther, ‘Kant’s Pure Aesthetic Judgement.’

Bibliography

Allison, H. 2001. Kant’s Theory of Taste, Cambridge: Cambridge University Press. Aristotle. 1970. Poetics, G. F. Else (trans.). Ann Arbor: University of Michigan Press. Austin, J.L. 1961. ‘A Plea for Excuses’ in J.O. Urmson and G.J. Warnock (eds), Philosophical Papers, Oxford: Oxford University Press. Black, M. 1962. ‘Metaphor’, Models and Metaphors, Ithaca: Cornell University Press. Black, M. 1979. ‘How Metaphors Work: A Reply to Donald Davidson’ in S. Sacks (ed.), On Metaphor, Chicago: Chicago University Press. Black, M. 1981. ‘More about Metaphor’ in A. Ortony (ed.), Metaphor and Thought, Cambridge: Cambridge University Press.

June 21, 2008 11:25 MAC/NWA Page-264 9780230_220478_14_cha13 Diarmuid Costello 265

Carroll, N. 1993. ‘Essence, Expression and History: Arthur Danto’s Philosophy of Art’ in M. Rollins (ed.), Danto and his Critics, Oxford: Blackwell. Carroll, N. 2004. ‘Non-Perceptual Aesthetic Properties: Comments for James Shelley’, British Journal of Aesthetics, 44.4: 413–423. Costello, D. 2004. ‘On Late Style: Arthur Danto’s The Abuse of Beauty’, British Journal of Aesthetics, 44.4: 424–439. Costello, D. 2007. ‘Greenberg’s Kant and the Fate of Aesthetics in Contemporary Art Theory’, Journal of Aesthetics and Art Criticism, 65.2: 217–228. Costello, D. 2007. ‘Whatever Happened to ‘‘Embodiment:’’ The Eclipse of Materiality in Danto’s Ontology of Art’, Angelaki, 12: 2. Costello, D. 2008. Retrieving Kant’s Aesthetics for Art Theory after Greenberg: Remarks on Arthur C. Danto and Thierry de Duve, in T. O’Connor, F. Halsall and J. Jansen (eds), Re-Discovering Aesthetics, New York: Standford University Press. Crowther, P. 1996. ‘The Significance of Kant’s Pure Aesthetic Judgement’, British Journal of Aesthetics, 36.2: 109–121. Danto, A. 1981. The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. Danto, A. 1992. ‘Metaphor and Cognition’, Beyond the Brillo Box: The Visual Arts in Post-Historical Perspective, New York: Farrar, Straus, Giroux. Danto, A. 1997. After the End of Art: Contemporary Art and the Pale of History,Princeton: Princeton University Press. Danto, A. 2003. The Abuse of Beauty: Aesthetics and the Concept of Art, Chicago: Chicago Open Court. Danto, A. 2007. ‘The Transfiguration Transfigured: Concluding Remarks’ in ‘Online Conference in Aesthetics: Arthur Danto’s Transfiguration of the Commonplace—25 Years Later’, URL = www. http://www.vanderbilt.edu/AnS/philosophy/events/.OCA/ DantoDantoConference.pdf. Danto, A. 2007. ‘Embodied Meanings, Isotypes and Aesthetical Ideas’, Journal of Aesthetics and Art Criticism, 65.1: 121–129. De Duve T. 1996. Kant after Duchamp, Cambridge, MA: MIT Press. Gilmore, J. 2005. ‘Internal Beauty’, Inquiry, 48.2: 145–154. Guyer, P. 1997. Kant and the Claims of Taste, Second Edition, Cambridge: Cambridge University Press. Guyer, P. 2005.‘From Jupiter’s Eagle to Warhol’s Boxes: The Concept of Art from Kant to Danto’, Values of Beauty: Historical Essays in Aesthetics, Cambridge: Cambridge University Press. Harrison, C. 2001. ‘The Index as Art-work’, Essays on Art & Language, Cambridge, MA: MIT Press. Janaway, C. 1997. ‘Kant’s Aesthetics and the Empty Cognitive Stock’, Philosophical Quarterly, 47.189: 459–476. Kant, I. 1987. Critique of Judgment, W.S. Pluhar (trans.), Indianapolis: Hackett. Lucas, S. 2004. MoMA Highlights, New York: Museum of Modern Art, revised edition. Pillow, K. 1994. ‘Form and Content in Kant’s Aesthetics: Locating Beauty and the Sublime in the Work of Art’, Journal of the History of Philosophy, 32.3: 443–459. Pillow, K. 2000. Sublime Understanding, Cambridge, MA: MIT Press. Pillow, K. 2001. ‘Jupiter’s Eagle and the Despot’s Hand Mill: Two Views on Metaphor in Kant’, Journal of Aesthetics and Art Criticism, 59.2: 193–209. Ricoeur, P. 1978. The Rule of Metaphor, London: Routledge. Ricoeur, P. 1979. ‘The Metaphorical Process as Cognition, Imagination and Feeling’ in S. Sacks (ed.), On Metaphor, Chicago: Chicago University Press.

June 21, 2008 11:25 MAC/NWA Page-265 9780230_220478_14_cha13 266 Danto and Kant, Together at Last?

Richter, G. 1995. The Daily Practice of Painting, London: Thames & Hudson. Rogerson, K.F. 1986. Kant’s Aesthetics: The Role of Form and Expression, Lanham, MD: University Press of America. Shelley J. ‘The Problem of Non-Perceptual Art’, British Journal of Aesthetics, 43.4: (October 2003) 363–378. Sibley, F. 1959. ‘Aesthetic Concepts’, Philosophical Review, 68: 421–450. Reprinted in J. Benson, B. Redfern, and J. Roxbee Cox (eds), 2001. Frank Sibley: Approach to Aesthetics—Collected Papers, Oxford: Clarendon Press. Wollheim, R. 1993. ‘Danto’s Gallery of Indiscernibles’ in M. Rollins (ed.), Danto and his Critics, Oxford: Basil Blackwell.

June 21, 2008 11:25 MAC/NWA Page-266 9780230_220478_14_cha13 Index

Adajian, Thomas, 132–4 Caplan, Ben, 21, 24–7 Addison, Joseph, 105, 106 Carrier, David, 140–1 aesthetic cognitivism, varieties of, 174–6 Carroll, Noel,¨ 126, 132, 135, 208, 211, aesthetic ideals, 188–202 251–2 aesthetic ideas, 258–62 Cavell, Stanley, 224–5, 231 aesthetic objectivism, 173–4 cognitivism in aesthetics, see aesthetic aesthetic supervenience, 188–9 cognitivism, varieties of analytic aesthetics, xi–xiv Costello, Diarmuid, xviii–xix anti-definitionalism about art, 125–49 creativity, 105–24 and closed concepts, 135–6 and anti-intentionalism, 109–12 inductive argument for, 131 common features of analyses of, normative argument for, 137–40 106–8 psychological argument for, 131–5 conditions upon, 116–20 Weitz’s argument for, 127–31 as process, 113–20 Anti-Realism in Meta-Aesthetics, see to what it should be attributed Meta-Aesthetics, Realism and basically, 108–16 Anti-Realism in Currie, Gregory, 2–3, 6, 110, 150, Aristotle, 203 152–3, 159, 161, 163–4, 208, Armstrong, David, 25 216–17, 219 Arnheim, Rudolf, 94 and Ravenscroft, Ian, 216 art defining, 125–49 Danto, Arthur, 203, 224, 227–9 ontology of, see ontology of art, the and Kant, 244–66 and personification, see the late ‘aesthetic turn’ of, personification in appreciation 248–51 of art on metaphor, 254–6 art forms, definitions of, 140–2 recent work on Kant’s aesthetics by, artifact expression, see expression 246–8 artworks on works of art as ‘embodied multiple instances of, 16n meanings’, 251–4 not exclusively grasped through Davidson, Donald, 41, 44–6, 58–83 seeing, 15–16 Davies, David, 2–3, 6, 21, 30–1, 108, 110, ontological diversity of, 6–14 151, 153, 159, 161 Austin, J.L., 251 Davies, Stephen, 32, 33, 89, 234 Dean, Jeffrey, 132 Beardsley, Monroe, 44–5, 58, 94, De Clercq, Rafael,¨ xvii–xviii 109–12, 126 Denham, Alison, 71–6, 78, 80 Benjamin, Walter, 224 Dennett, Daniel, 237–8 Blackburn, Simon, 176 Dewey, John, 224, 228 Black, Max, 257 Dickie, George, 99, 139 Boden, Margaret, 117–18 Dodd, Julian, 21, 23–9, 31 Budd, Malcolm, 127 Dutton, Dennis, 115, 137

267

June 21, 2008 11:29 MAC/NWA Page-267 9780230_220478_15_ind01 268 Index

Ekman, Paul, 86 Kania, Andrew, xiv, 21 expression Kant, Immanuel, 106, 108, 203, 224, artifact, 84–104; competing theories 225, 231–3, 244–66 of, 88–96 on aesthetic ideas, 258–60 human, 85–8 metaphor in, 256–8 musical, 41–57; against explanations see also Danto, Arthur, and Kant of in terms of metaphor, Kivy, Peter, 89, 105, 125–6, 230 41–57 Koestler, Arthur, 107 Kulvicki, John, xv–xvi fiction defining, 152–65 Lamarque, Peter and Olsen, Stein and imagination, 152–66 Haugom, 126, 151, 153, 159, Walton’s definition of, 152–8 161–3, 165 Fogelin, Robert, 71, 73 Leslie, Alan, 156 Friend, Stacie, xvii Levinson, Jerrold, 126, 133, 135, 229–30, 234 Gaut, Berys, 110, 126, 135 Livingston, Paisley, 110 Gendler, Tamar Szabo,´ 219 Longinus, 105, 106 genre mastery, towards a theory of, Lopes, Dominic, 18n, 86–7, 89, 90–1, 207–15 98–9 Gombrich, Ernst, 96, 100 Gonzalez-Torres, Felix, 6–11 McGonigal, Andrew, xv Goodman, Nelson, 20, 41, 43–4, Mackie, J.L., 172 84, 85, 89, 94–6, 97, 99, Margolis, Joseph, 229–30 224, 229–30 Matheson, Carl, 21, 24–7 Gracyk, Theodore, 32 Meskin, Aaron, xvi Greenberg, Clement, 246–8 see also Weinberg, Jonathan, and Meskin, Aaron Meta-Aesthetics Hadamard, Jacques, 107 Quasi-Realism in, 178–80, 182–3 Hagberg, Garry, 226, 228 Realism and Anti-Realism in, 170–87 Haidt, Jonathan, 211, 217 sensible subjectivism in, 176–8, 182–3 Hayman, Greg and Pratt, Henry John, metaphor 126, 140–1 Donald Davidson’s ‘causal theory’ of, Hegel, Georg Wilhelm Friedrich, 224, 58–83; complaints against, 59–62 229, 246 and the Frege-Geach problem, 63 Helmholtz, Hermann von, 107 and imaginative insight, 64–9 Hospers, John, 89 and other linguistic phenomena, Hume, David, 176, 203 62–76 Hutcheson, Francis, 232 and reduction to paraphrase, 44–8, 71–6 imagination, 152–66, 203–2 and resemblance and imagination, 49, and belief, 156–8, 204–7 51–4 cognitive accounts of, theories of, 49–54 203–7 whether they have a literal meaning, and desire, 215–18 44–6 Irvin, Sherri, xiv see also expression; metaphorical Isenberg, Arnold, xii exemplification; Kant, Immanuel, metaphor in; musical, against Jackson, Frank, 138 explanations in terms of metaphor

June 21, 2008 11:29 MAC/NWA Page-268 9780230_220478_15_ind01 Index 269 metaphorical exemplification, 43–4 Sagoff, Mark, 227 Millikan, Ruth, 86 Savile, Anthony, 165 musical expressiveness, see expression, Schaper, Eva, 227 musical Schellekens, Elisabeth, xvii music, ontology of, see ontology of Schelling, Friedrich von, 45 music Schiller, Friedrich, 225, 233–8 Scruton, Roger, 33, 41–2, 43, 51–4, Nichols, Shaun, 156–7, 204 194, 224 and Stich, Stephen, 205–6, 219, 220n Shaftesbury, 232 and Stich, Stephen, and Weinberg, Shusterman, Richard, xi–xii Jonathan, 213–15 Sibley, Frank, 188–94, 247 Silvers, Anita, xii objectivism in aesthetics, see aesthetic sortals, aesthetic relevance of, 189–200 objectivism Sosa, Ernest, 213 Olsen, Stein, see Lamarque, Peter and Soucek, Brian, xviii Olsen, Stein Haugom Sparshott, Francis, 227 ontology of art, the, 1–19 Stecker, Robert, 93, 126, 132, 133, 135 critical practice constraint upon, Stern, Josef, 61–2 5, 13–14 Stich, Stephen how the artist’s sanction affects, 4–16 see also Nichols, Shaun, and Stich, ontology of music, 20–40 Stephen; and Stich, Stephen, and platonism versus creationism in the, Weinberg, Jonathan 21–9 Stock, Kathleen, 21 relationship of fundamental to Stokes, Dustin, xvi higher-level, 32–3 supervenience, see aesthetic relationship of performance to work supervenience in the, 22 scepticism about the, 30–2, Taylor, Charles, 226 33–4 Thomasson, Amie, 12, 28, 30–1 Todd, Cain, 178–9, 181 Passmore, J.A., xiii Tormey, Alan, 228, 234 personification in appreciation of art, Trivedi, Saam, xv, 230 224–43 clarifying the nature of, 225–43 Urmson, J.O., xi Pillow, Kirk, 245, 256–7 Pippin, Robert, 224 Vendler, Zeno, 113–14 Plato, 105, 106, 203 Poincare,´ Henri, 107 Wallas, Graham, 107 Ponech, Trevor, 126 Walton, Kendall, 118, 142, 151, 152–8, Pratt, John Henry, see Hayman, 165, 188, 197, 219 Greg and Pratt, Henry John Weinberg, Jonathan, xviii Predelli, Stefano, 21 and Meskin, Aaron, 207, 211, 219 see also Nichols, Shaun, Stich, Stephen, Ramsey, William, 134 and Weinberg, Jonathan Realism in Meta-Aesthetics, see Weitz, Morris, 125, 127–31, 137–40 Meta-Aesthetics, Realism and Wiggins, David, 177–8, 181, 198 Anti-Realism in Williamson, Timothy, 143 Rey, George, 31 Wollheim, Richard, 115, 227 Ribeiro, Anna, 126 Wright, Crispin, 76–80 Ricoeur, Paul, 257 Ridley, Aaron, 33–4 Zangwill, Nick, 139

June 21, 2008 11:29 MAC/NWA Page-269 9780230_220478_15_ind01