Current Controversies in Philosophy of Film 9781138789517, 9781315764887, 1138789518

This volume advances the contemporary debate on five central issues in the philosophy of film. These issues concern the

247 11 1MB

English Pages 212 [222] Year 2016

Report DMCA / Copyright

DOWNLOAD PDF FILE

Table of contents :
Cover......Page 1
Title......Page 6
Copyright......Page 7
Contents......Page 8
Contributors......Page 10
Introduction......Page 12
Part I What Is the Relation between the Art and the Technology of Film?......Page 26
1 Cinematic Art and Technology......Page 28
2 Movie Appreciation and the Digital Medium......Page 47
Part II In What Ways Is Film a Realistic Medium?......Page 66
3 Imagined Seeing and Some Varieties of Cinematic Realism......Page 68
4 Realism in Film (and Other Representations)......Page 87
Part III How Do Films Work as Narrative Fictions?......Page 108
5 Fictional Indeterminacy, Imagined Seeing, and Cinematic Narration......Page 110
6 Motion Picture Narration......Page 126
Part IV How Do Films Engage Our Emotions?......Page 140
7 Putting Cognition in Its Place: Affect and the Experience of Narrative Film......Page 142
8 Mirror Neurons and Simulation Theory: A Neurophysiological Foundation for Cinematic Empathy......Page 159
Part V Can Films Philosophize?......Page 174
9 Film as Philosophy: The Pro Position......Page 176
10 Film, Philosophy, and the Varieties of Artistic Value......Page 193
Index......Page 214
Recommend Papers

Current Controversies in Philosophy of Film
 9781138789517, 9781315764887, 1138789518

  • 0 0 0
  • Like this paper and download? You can publish your own PDF file online for free in a few minutes! Sign Up
File loading please wait...
Citation preview

Current Controversies in Philosophy of Film

This volume advances the contemporary debate on five central issues in the philosophy of film. These issues concern the relation between the art and technology of film, the nature of film realism, how narrative fiction films narrate, how we engage emotionally with films, and whether films can philosophize. Two new essays by leading figures in the field present different views on each issue. The paired essays contain significant points of both agreement and disagreement; new theories and frameworks are proposed at the same time as authors review the current state of debate. Given their combination of richness and clarity, the essays in this volume can effectively engage both students, undergraduate or graduate, and academic researchers. Katherine Thomson-Jones is an associate professor of philosophy at Oberlin College. She is author of Aesthetics and Film (2008) and coeditor of New Waves in Aesthetics (2008), and she is currently writing a book on the philosophy of digital art.

Current Controversies in Philosophy Series Editor: John Turri University of Waterloo In venerable Socratic fashion, philosophy proceeds best through reasoned conversation. Current Controversies in Philosophy provides short, accessible volumes that cast a spotlight on ongoing central philosophical conversations. In each book, pairs of experts debate four or five key issues of contemporary concern, setting the stage for students, teachers, and researchers to join the discussion. Short chapter descriptions precede each chapter, and an annotated bibliography and suggestions for further reading conclude each controversy. Combining timely debates with useful pedagogical aids allows the volumes to serve as clear and detailed snapshots, for all levels of readers, of some the most exciting work happening in philosophy today. Volumes in the Series Published: Current Controversies in Philosophy of Film Edited by Katherine Thomson-Jones Current Controversies in Political Philosophy Edited by Thom Brooks Current Controversies in Virtue Theory Edited by Mark Alfano Current Controversies in Epistemology Edited by Ram Neta Current Controversies in Experimental Philosophy Edited by Edouard Machery and Elizabeth O’Neill Current Controversies in Philosophy of Mind Edited by Uriah Kriegel Forthcoming: Current Controversies in Cognitive Science Edited by Sarah-Jane Leslie and Simon Cullen Current Controversies in Bioethics Edited by S. Matthew Liao and Collin O’Neil

Current Controversies in Philosophy of Science Edited by Shamik Dasgupta and Brad Weslake Current Controversies in Philosophy of Perception Edited by Bence Nanay Current Controversies in Metaphysics Edited by Elizabeth Barnes Praise for the series “This series constitutes a wonderful addition to the literature. The volumes reflect the essentially dialectical nature of philosophy, and are edited by leading figures in the field. They will be an invaluable resource for students and faculty alike.” Duncan Pritchard, The University of Edinburgh

This page intentionally left blank

Current Controversies in Philosophy of Film

Edited by Katherine Thomson-Jones

First published 2016 by Routledge 711 Third Avenue, New York, NY 10017 and by Routledge 2 Park Square, Milton Park, Abingdon, Oxon, OX14 4RN Routledge is an imprint of the Taylor & Francis Group, an informa business © 2016 Taylor & Francis The right of the editor to be identified as the author of the editorial material, and of the authors for their individual chapters, has been asserted in accordance with sections 77 and 78 of the Copyright, Designs and Patents Act 1988. All rights reserved. No part of this book may be reprinted or reproduced or utilized in any form or by any electronic, mechanical, or other means, now known or hereafter invented, including photocopying and recording, or in any information storage or retrieval system, without permission in writing from the publishers. Trademark notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data A catalog record for this title has been requested ISBN: 978-1-138-78951-7 (hbk) ISBN: 978-1-315-76488-7 (ebk) Typeset in Minion by Apex CoVantage, LLC

Contents

Contributorsix Introduction1 Katherine Thomson-Jones

Part I What Is the Relation between the Art and the Technology of Film?

15

  1 Cinematic Art and Technology

17

Berys Gaut

  2 Movie Appreciation and the Digital Medium

36

Katherine Thomson-Jones

Part II In What Ways Is Film a Realistic Medium?

55

  3 Imagined Seeing and Some Varieties of Cinematic Realism

57

George M. Wilson

  4 Realism in Film (and Other Representations) Robert Hopkins

vii

76

viii  • Contents

Part III How Do Films Work as Narrative Fictions?

97

  5 Fictional Indeterminacy, Imagined Seeing, and Cinematic Narration

99

Angela Curran

  6 Motion Picture Narration

115

Noël Carroll

Part IV How Do Films Engage Our Emotions?

129

  7 Putting Cognition in Its Place: Affect and the Experience of Narrative Film

131

Carl Plantinga

  8 Mirror Neurons and Simulation Theory: A Neurophysiological Foundation for Cinematic Empathy

148

Dan Shaw

Part V Can Films Philosophize?

163

  9 Film as Philosophy: The Pro Position

165

Thomas E. Wartenberg

10 Film, Philosophy, and the Varieties of Artistic Value

182

Murray Smith

Index203

Contributors

Noël Carroll is a distinguished professor of philosophy at the Graduate Center of the City University of New York. His recent books include Living in an Artworld, Art in Three Dimensions, Minerva’s Night Out: Philosophy, Pop Culture and Motion Pictures, and Humour: A Very Short Introduction. Angela Curran is a faculty fellow at Colby College in Maine, USA. She is the author of The Routledge Philosophy Guidebook to Aristotle and the Poetics (2015). Berys Gaut is a professor of philosophy at the University of St Andrews and president of the British Society of Aesthetics. He is the author of Art, Emotion and Ethics (2007), A Philosophy of Cinematic Art (2010), and coeditor of The Routledge Companion to Aesthetics (3rd ed., 2013), as well as author of numerous articles in aesthetics, the philosophy of film, and the philosophy of creativity. Robert Hopkins is a professor of philosophy at New York University. His recent publications include “The Real Challenge to Photography (as Communicative Representational Art)” in the Journal of the American Philosophical Association, and “ ‘Remember Leonard Shelby’: Memento and the Double Life of Memory” in a memorial volume for Peter Goldie, edited by Julian Dodd. Carl Plantinga is a professor of film and media at Calvin College and former president of the Society for Cognitive Studies of the Moving Image. Among his books are Moving Viewers: American Film and the Spectator’s Experience (2009) and (as coeditor) Passionate Views: Film, Cognition, and Emotion (1999). ix

x  • Contributors

Dan Shaw is a professor of philosophy and film at Lock Haven University. He is the author of numerous articles on film and philosophy, and several books (including Film and Philosophy, Taking Movies Seriously, Morality and the Movies: Reading Ethics through Film, and the forthcoming Movies with Meaning: Existentialism through Film), and is managing editor of the print journal Film and Philosophy. Murray Smith is a professor of film studies at the University of Kent, Canterbury, UK, codirector of the Aesthetics Research Centre at Kent, and president of the Society for Cognitive Studies of the Moving Image. His publications include Engaging Characters: Fiction, Emotion, and the Cinema (Clarendon); Trainspotting (BFI); Film Theory and Philosophy (coedited with Richard Allen) (Clarendon); Thinking through Cinema (coedited with Tom Wartenberg) (Blackwell); and the forthcoming Film, Art, and the Third Culture (Oxford). Katherine Thomson-Jones is an associate professor of philosophy at Oberlin College. She is author of Aesthetics and Film (2008) and coeditor of New Waves in Aesthetics (2008), and she is currently writing a book on the philosophy of digital art. Thomas E. Wartenberg is a professor of philosophy at Mount Holyoke College, where he also teaches in the Film Studies program. He is the author of Thinking on Screen: Film as Philosophy and Unlikely Couples: Movie Romance as Social Criticism, and has edited or coedited four books on the philosophy of film. George M. Wilson is a professor emeritus of philosophy and cinematic arts at USC. He also taught at the University of California at Davis and at Johns Hopkins University. He is the author of two books on film, Narration in Light (1986) and Seeing Fictions in Film (2011), and of one book on theory of action, The Intentionality of Human Action (1990). He has also written extensively on the philosophy of language, the philosophy of mind, and the interpretation of the later Wittgenstein.

Introduction KATHERINE THOMSON-JONES

This volume takes up five central issues in the philosophy of film. Briefly, these issues concern the relation between the art and technology of film, realism in film, narration in film, our emotional engagement with films, and whether films can philosophize. In keeping with the title of the volume, these five issues are the subject of current controversy or debate even while they have important connections to the history of philosophical thinking about film. This history is practically coextensive with the history of film as an art form. When film first emerged as a technology for mechanical recording, it was assumed by many to be antithetical to artistic creativity. Thus the art status of film had to be won through the efforts of early filmmakers—for example, Georges Méliès and D. W. Griffith—and explicitly defended with arguments for the possibility of artistic expression in spite of or even because of mechanical recording. These early arguments for the independent art status of film appealed to artistic effects dependent on the technology of film, distinctive or expanded forms of realism and expression achieved in film, new modes of audience engagement, and forms of insight made possible by the new art form. The connections between these early arguments and many of the current controversies are clear and direct, and yet unlike the early arguments, the current controversies are pursued systematically in a recognized subfield of philosophical aesthetics. Philosophy of film did not emerge as a unified field of study until the 1990s, when analytic aesthetics in general began paying more attention to individual art forms. Since the 1990s, which issues have emerged as central in philosophy of film has depended on at least two main factors.

1

2  •  Katherine Thomson-Jones

On the one hand, issues that were first identified by classical film theorists, such as Rudolph Arnheim and André Bazin, continue to intrigue contemporary philosophers of film. This is the case with the issue of realism and, to some extent, with the relation between the art and technology of film. On the other hand, some issues are considered central given the influence of work in other areas of philosophy and the sciences. This is the case with the issue of emotional engagement: Contemporary philosophers of film draw on work in the cognitive sciences to explain the interplay between cinematic devices and audience response. It is also (at least partly) the case with the issue of narration, where authors draw on philosophical accounts of the imagination and depiction, as well as on narratology, in order to understand how films convey stories. The issue concerning whether films philosophize is perhaps more unusual in terms of its history of significance. Prior to the consolidation of the subfield of philosophy of film, Stanley Cavell developed a sustained argument for the philosophical self-consciousness of the film medium (see, e.g., Cavell 1981). Partly under Cavell’s influence, the question of film as philosophy has engaged philosophers in both the continental and the analytic traditions. Each issue in this volume is introduced by a question to which two authors offer independent but mutually informed answers. It would be inaccurate to say that the authors provide straightforwardly opposing positions on an issue. As we shall see, there tend to be as many points of agreement as disagreement, even though authors will frame their concerns in quite different ways. As a small and relatively new subfield, philosophy of film encourages particularly close engagement between its contributors. Close engagement in turn encourages collaboration and attention to detail: Rather than the elaboration of systems or the staunch defense of positions, the focus is on noting particular areas of contention and compatibility, largely for the sake of establishing the complexity and significance of the issues involved. Perhaps the debate in which the clearest lines of opposition are drawn is the debate over whether film can do philosophy. But even here, as our two authors on the issue reveal, reasonable opposition need not be so sharp and must proceed with a shared methodology. The tendency toward cooperation in the philosophy of film, and the rejection of fixed positions, also reflects an awareness of how much work remains to be done in the field. With many of the key issues, including the ones confronted in this volume, there is still a question of how to frame them correctly just so as to make their resolution a possibility. Closely related is the issue of how to define the object of study: The term “film” is sometimes used in a restrictive sense to apply to moving-image works made in a certain way and/or belonging to a certain genre—most typically, mainstream works of narrative fiction shot on traditional celluloid film. But the term “film” is also used interchangeably with “movie” and “cinema” to refer to all works in moving-image media, including feature-length narrative fiction films, television serials, documentaries, video

Introduction  •  3

art, and video games. Ongoing developments in the technology of filmmaking, as well as in the scientific study of our perceptual, imaginative, and emotional responses to films, cannot be ignored by any adequate philosophical account of the nature of film art and engagement. Under Noël Carroll’s well-known characterization (1998), cinema is a mass art and arguably the mass art form of the twentieth century. In the twenty-first century, cinema continues to dominate, with digital technology allowing for increased access to movies on portable devices as well as on TV screens and in movie theaters. It is fair to say that we all consider ourselves experienced consumers of moving imagery. We likely have an opinion on how movies are meant to engage us, and we have a grounding sense, even if it is not one we can easily articulate, of what a movie is. The accessibility and mass appeal of film can be part of what attracts philosophers to study its nature and value. But it also introduces particular challenges for philosophical analysis. For instance, philosophers of film may have to resist a stronger tendency than with some other art forms to take certain claims about film as obvious or beyond question. Most importantly, philosophers of film have to combat the assumption that, since watching movies is somehow easy— easier, for example, than engaging with many kinds of traditional fine art— philosophy of film is also going to be easy and perhaps not really worth doing. As this volume strongly demonstrates, philosophy of film is far from easy. But it is certainly easy to become fascinated by its puzzles.

Part I: What Is the Relation between the Art and the Technology of Film? Although film began as a mechanical recording technology and had to earn its art status amidst considerable skepticism, that status is now largely taken for granted. But this is not to say that, when we watch and think about films, we have simply forgotten or always ignore the underlying technology of filmmaking. Quite the contrary, in fact: Our understanding of the history of the art form seems to depend, at least in part, on important technological developments—for example, the introduction of synchronous sound in the 1920s. Film critics will frequently refer to the technology used to achieve a certain effect in order to explain or justify their assessment of a film in terms of the filmmakers’ achievement. And ordinary film viewers can be drawn into the movie theater by being told that a film has been shot in a new way—for example, at an unusually high rate of fps (frames-per-second; cf. The Hobbit); that it incorporates a new kind of special effect (e.g., the use of real-time motion-capture technology to create the character of Gollum); or, that it is to be projected in 3-D. Interestingly, we seem to have a sense of the unity and continuity of the cinematic art form even though the technology for making films is always changing. On the one hand, there is the sense that film as an art is inextricably linked to technology. On the other hand, there is the sense

4  •  Katherine Thomson-Jones

that we go on appreciating films as films, despite technological changes, able to meaningfully compare recent digital works with traditional analog ones. There seems to be a puzzle here, one with which both chapters in this section are concerned. In chapter 1, Berys Gaut argues that the art of cinema is dependent on technology even though cinematic technology is not dependent on the art of cinema. A proper appreciation of a film partly depends on understanding the technology deployed in its production. This is because a proper appreciation of a film partly depends on understanding the difficulties involved in making the film, and those difficulties are conditioned by the technology used in making the film. On Gaut’s analysis of the concept of achievement, it is plausible that the more difficult an achievement is, the greater it is, other things equal. Alternatively, the easier an achievement is, the less valuable it is, other things equal. Gaut shows how certain cinematic effects are easier to achieve by digital means than by traditional means. This seems to lead to the “digital dystopia argument” on which digital films, other things equal, are less of an achievement than traditional films. The possibility of a digital dystopia is unlikely, however, for at least two reasons: Gaut thinks that digital moviemakers can take on greater difficulties than traditional moviemakers and they can sometimes achieve even more valuable results. The “digital dystopia” that Gaut describes toward the end of his chapter is my primary focus in chapter 2. Like Gaut, I want to resist general skepticism about the value of digital cinema. My grounds for resistance are different but compatible with Gaut’s: Whereas he points to greater difficulties had with familiar kinds of filmmaking challenges, I suggest that digital filmmaking comes with its own distinct set of challenges. I begin with a detailed analysis of the notion of an artistic medium and show that the digital medium is properly understood as a symbolic medium of artistic appreciation. Then I argue for the importance of recognizing that the digital medium sets terms of appreciation for digital movies that are distinct from those of the traditional cinematic medium. In particular, the use of computer applications in moviemaking introduces new creative risks involved in having to choose from a vast array of equally and instantly available options. A principal creative problem for digital moviemakers is thus one of massive proliferation—in terms of options for the precise look and sound of a movie. This is a distinctively digital problem, since it arises from the use of binary computer processing, and it is one that requires our appreciative attention. Diverse solutions to it are being explored—actively, self-consciously, but only sometimes successfully.

Part II: In What Ways Is Film a Realistic Medium? The debate about film realism began in the 1940s when André Bazin took what had been considered an obstacle to cinematic artistry—namely, a reliance on

Introduction  •  5

mechanical recording—and made it the basis for his account of film as the supremely realistic art form (see Bazin 1967). Philosophers of film have tended to follow Bazin in trying to understand film realism in terms of the filmmaking process, at least the process involved in making live-action films rather than animated movies. The special causal connection between a mechanically recorded image and its subject is often treated as the key to understanding photographic and cinematic realism. On the basis of this causal connection Kendall Walton famously argues that photographs are transparent—we see their objects through them (1984, 1997). This is just one way to interpret the significance of the causal connection, however, and in fact, whatever its significance for the nature of photographic representation, a further argument is needed for a kind of realism guaranteed to images in virtue of the way they are made no matter how they look. The issue quickly becomes even more complicated when we consider not just the actual, recorded subject of a photograph but also the fictional subject matter of a narrative film. When a film is made by photographing actors and sets used to portray fictional characters and settings, what is the relevant sense in which the presentation of the fiction can strike us as realistic or not? Perhaps when we find the fiction realistic, we are under some kind of illusion about the fictional being actual. Perhaps we are engaged in a particular kind of imaginative activity, as prescribed by the film. Both these possibilities and others are thoroughly examined by our two authors. In chapter 3, George Wilson suggests that at least some of the ways in which films can be realistic depend on how and what we see in fiction films. Wilson is particularly struck by the way ordinary film viewers will talk about “seeing” both the actors and sets recorded for a film and the fictional characters and scenes represented by the recorded items. In both cases, Wilson thinks film viewers should be understood as speaking literally, although in the latter case, some “contextual amplification” is required to grasp the kind of seeing involved. In the former case, assuming the truth of photographic transparency, film viewers literally albeit indirectly see actors and movie sets through recorded film images. In the latter case, Wilson argues that film viewers imagine seeing the fictional characters and scenes, in the sense that they imagine of their actually seeing an actor that it is a seeing of a fictional character portrayed by the actor. Wilson is the first to recognize that the “Imagined Seeing Thesis” requires qualification and rigorous defense against some serious objections. In previous contexts, Wilson has defended the thesis as part of an epistemology of fiction film narration (e.g., 2011, 1986). In the next section of the volume, we see that the implications of imagined seeing are still a central concern in the debate about cinematic narration. In relation to realism, however, the Imagined Seeing Thesis is useful, Wilson thinks, for uncovering the dual nature of a kind of perceptual realism in fiction film. It is often assumed—and indeed

6  •  Katherine Thomson-Jones

Wilson himself relies on the assumption—that photographic transparency yields a sense in which films are perceptually realistic; any recorded image is realistic insofar as it allows us to really see its subject. Strictly speaking, this notion of perceptual realism applies only to nonfiction film (to photographic representation and not to what is fictionally portrayed by photographic representation), and yet Wilson thinks we will often speak as though fiction films also exhibit the same kind of realism, at the level of the fictional representation. Wilson explains this tendency by extending the relevant notion of perceptual realism with reference to the Imagined Seeing Thesis: When a fiction film is made by using transparent movie images to represent fictional items and events, the role of a “structuring narrative imagination” is such that we have a pervasive experience “as if ” we were really seeing the fictional items and events on-screen, and thus through the images—it is as if the images are transparent to the fiction rather than to the movie set. In drawing out the implications of the Imagined Seeing Thesis, Wilson is ultimately concerned to ensure that distinct but intimately related forms of cinematic realism are recognized and properly understood. This concern is also paramount for Robert Hopkins as he confronts the “bewilderingly diverse” range of phenomena that have been identified with realism in cinema. In chapter 4, Hopkins takes a step back to consider what in general realism might be, if all these phenomena are to count as species of realism. Like Wilson, his focus is on photographic fiction film, which involves both the (photographic) representation of “Events Filmed” and the (theatrical) representation of the “Story Told.” Hopkins gives us a “whirlwind tour” of five important and highly varied kinds of realism. In order to explain what might make each of them a species of realism, Hopkins proposes that we appeal to a conception of the ideal case. On this conception, a maximally realistic representation fully captures how something is as a result of the object being the way it is. Interestingly, the candidate realisms that most easily align with this conception are the two broadest—realism in terms of accuracy and precision, and realism in terms of recessive form—and the narrowest, a kind that may be limited to photographic fiction film and involves a “collapse” in our perceptual experience of Events Filmed as the Story Told. The two remaining candidates for realism are more difficult to explain as species of the same phenomenon. These candidates involve an illusion of the real and photographic transparency. As we have seen, Wilson relies heavily on the connection between transparency and realism in order to explicate a sense of realism connected with imagined seeing. But Hopkins questions the essential realism of a transparent image and thereby threatens Wilson’s argument for complex perceptual realism. This potential threat is particularly striking given that Hopkins’s notion of realism as “collapse” and Wilson’s notion of extended perceptual realism account for the same central aspect of film viewing experience.

Introduction  •  7

Part III: How Do Films Work as Narrative Fictions? In response to the foregoing question, both our authors, Noël Carroll and Angela Curran, are interested in the way narrative fiction films work on us, the viewers, in order to convey a story, specifically by prescribing us to imagine certain things in certain ways. Perhaps most basically, narration would seem to involve a prescription to imagine as having taken place the events depicted in or described by a work. But providing a full analysis of this prescription is surprisingly difficult—something of which our authors are only too aware. There are two main issues to be decided for the sake of understanding the imaginative workings of narration: One concerns the kind or kinds of imagining being prescribed in the conveyance of a movie’s narrative. The other issue concerns the extent of the imaginative prescription—specifically, whether not only the fictional content of the story but also features of its fictional conveyance are to be imagined by viewers for the sake of narrative understanding and proper appreciation. In chapter 5, Curran addresses both these issues head-on: The first part of her chapter is concerned with the question of whether movie viewers are prescribed to imagine the means by which they are shown the fictional events depicted in the movie. The second part of the chapter is concerned with the question of whether movie viewers typically imagine seeing the fictional events depicted in the movie or, alternatively, simply imagine those events occurring without imagining experiencing them. Curran’s approach to resolving these issues is to investigate the implications for each of adopting independently a general view about the indeterminacy of our means of access to fictions. On this view, many forms of narrative fiction operate on the principle that, unless something about the audience’s access to story events is made explicit in the story, the audience is not meant to imagine anything about their means of access. In the case of narrative fiction film, this would mean that we are not meant to imagine how we came to have a certain view on fictional events, as though it were provided by a ubiquitous and implicit visual narrator. On the other hand, the principle of indeterminacy permits a role for imagined seeing in our engagement with fiction films. Given fictional indeterminacy, it would not be the case that in imagining seeing depicted events, our engagement with the fiction would be automatically compromised by our also having to imagine the means by which we were seeing the fiction. At the same time, Curran does not think that imagined seeing is the default mode of imaginative engagement with the visual fiction in films; rather it is a mode that can be activated for particular effect with the use of particular cinematic techniques. In chapter 6, Carroll’s concern with imagined seeing is in relation to an influential argument, largely due to Wilson, for the ubiquity of implicit visual narrators in narrative fiction films. On the “seeing/showing” argument, given that we

8  •  Katherine Thomson-Jones

imagine seeing fictional events in a mediated fashion, we must posit a minimal agency that is responsible for providing our view on those events. Carroll offers a series of considerations for “resisting” the ubiquity of implicit cinematic narrators, both on the basis of the seeing/showing argument and on the basis of the “ontological gap” argument, which relies on the assertoric power of a fictional narrator to make things true in the fiction. Interestingly, Carroll’s main objection to the seeing/showing argument seems to invoke the reverse of Curran’s principle of indeterminacy in fiction: Precisely because the implicit cinematic narrator is implicit, questions reasonably arise about the precise nature and mode of operation of such an agency. Only when the means of access is explicitly given in the fiction (e.g., with Flash Gordon’s “Space-o-scope”) can it remain indeterminate as to exactly how that means of access operates. On the ontological gap argument, it is thought that to have a film in which it is fictional that anything is the case, there must be someone in a position to assert that something is the case in the fiction. The actual makers of a fiction film are not in the right position, since they are outside the fiction and separated from it by an “ontological gap.” The gap is closed by a narrator who is part of the fiction and for whom the fictional is actual and assertable. In response to this argument, Carroll tests our intuitions about the possibility of actual filmmakers narrating what is fictionally the case. In addition, and most fundamentally, Carroll challenges the account of truth in fiction presupposed by the ontological gap argument. Rather than relying on the assertions of a fictional narrator to make things true in the fiction, Carroll proposes that we rely on the actual filmmakers’ mandate to imagine. Carroll is the first to admit that his rival “imagination account” of truth in fiction may seem incomplete, given that we are mandated not only to imagine but also to believe many truths in a fiction film. But he suggests there are ways to overcome this deficiency as well as a serious motivation to do so. The motivation in question is that, unlike the cinematic-narrator account, the imagination account of truth in fiction is at least consistent with the “phenomenology of the plain viewer.”

Part IV: How Do Films Engage Our Emotions? The issues of narration and realism in film are intimately bound up with the fourth issue in our volume, concerning emotional engagement. Films are often thought to be particularly affecting because of their realism. And choices over how to convey story events are often explained by the filmmakers’ attempt to create certain emotional effects—for example, suspense. Compared with other art forms, particularly with other representational art forms, it is frequently observed just how easily and how powerfully we are moved by films. The ways in which we can be moved by different films are rich and varied, but responses to any particular movie tend to converge among audience members.

Introduction  •  9

Such convergence suggests an interplay between cinematic techniques and the general mechanisms of human emotion. Filmmakers find ways to reliably trigger the same kinds of emotions in many different people, and understanding how they do this will not only deepen our appreciation of certain films but also enable us to understand better the workings of human emotion. Of course it is not as though filmmakers necessarily set out to implement routine strategies for triggering different emotions. Through practice and experience, filmmakers may develop an intuitive grasp of ways to elicit different responses from their audiences. Despite its intuitiveness, however, the process whereby filmmakers work to emotionally engage audiences generates some interesting philosophical puzzles. Even though it seems clear that we are meant and are able to engage emotionally with a film’s characters or narrative, in the case of fiction films, there is a puzzle about the rationality of our doing so. As long as we remain aware of the fictionality of Ellen Ripley (Aliens), say, how does it make sense to fear for her safety? There may also be moments in a film when my responses to a character feel empathetic rather than sympathetic—perhaps I myself experience a surge of Ripley’s righteous rage against her alien attackers. Such shared feeling is often linked to our tendency to identify with certain characters, but it is surprisingly difficult to give a coherent account of identification as a process that supports our having another’s—including a fictional other’s—feelings. Addressing these puzzles and others requires getting clear on the nature of emotion, particularly on its relation to cognition and the body. Fortunately there has been a recent surge of activity in the cognitive sciences on the neural and psychological mechanisms underlying emotion. Philosophers of emotion then analyze the key assumptions and concepts mobilized in the scientific studies. All of this can then be brought to bear on film engagement, but with an additional sensitivity to and awareness of film-theoretical traditions involving psychoanalytic and politically inflected analyses of film spectatorship. The issue of emotional engagement is given an important theoretical context by Carl Plantinga, who begins chapter 7 by charting some major shifts in film studies when it comes to the prioritizing of different aspects of film viewers’ experiences and the popularity of different models and metaphors for understanding viewers’ experiences. Given this context, the difficulty of finding a simple or single answer to the question of how films engage us emotionally is due not only to the complexity of the film viewing experience but also to considerable unclarity and confusion over the terms of analysis. Recently there has been a strong tendency in film and media studies to assume that emphasis on bodily and affective experience of films precludes emphasis on cognitive engagement with films. Plantinga aims to show that this tendency rests on a misunderstanding about the nature of cognition, and a failure to recognize that non-conscious, automatic, and affective responses to films are aspects of,

10  •  Katherine Thomson-Jones

rather than alternatives to, cognitive engagement. The key to overcoming the false dichotomy between cognition and affect, Plantinga suggests, is recognition of the ways in which affect and emotion play a role in the defining cognitive activities of understanding and knowledge formation. Plantinga is at great pains to show that the possession of such a role is compatible with emotions having an automatic, non-conscious component, and with their being neither rational nor irrational in themselves when experienced toward fictions. The case of cinematic experience also challenges the cognition-affect dichotomy by its reliance on and integration of the full range of lower- and higher-level cognitive responses. Our sympathetic response to a character onscreen, for example, could be partly automatic, triggered by the mimicking of facial expressions, while also coming to involve thoughts or imaginings about the character’s suffering, and at the same time, being conditioned by our awareness of the fictional (or represented) status of the character. Ultimately, the “two-folded” nature of our engagement with film fictions, whereby we respond to characters and events even while retaining awareness of their fictionality, leads Plantinga to express serious doubts about Amy Coplan’s account of empathy in fiction film (see, e.g., Coplan 2011). In chapter 8, Dan Shaw also refers to Coplan’s account of empathy, but whereas Plantinga is concerned with a more complex phenomenon than that accommodated by Coplan, Shaw is concerned with something more basic and bodily. “Basic empathy,” Shaw argues, is an automatic and visceral response to a film character’s display of emotion prior to and independent of any assessment of the character’s situation. The primary aim of Shaw’s chapter is to draw on scientific research about the emotive function of mirror neurons to support the central role of basic empathy in our engagement with film characters. Shaw begins with a brief history and interpretation of the research on mirror neurons. He draws particularly on the work of Marco Iacoboni, who describes the emotive function of mirror neurons in terms of the “inner imitation” of another’s expression. Studies show that the same neurons fire in response to another’s emotional expression as would fire if one were making the expression oneself. The mirror neurons then interact with the emotional centers of the brain such that the emotion whose expression is imitated is then actually felt. This shared feeling, which occurs automatically as a direct result of mirror neuron activity, Shaw considers the most basic form of empathy. As the art of moving images, film is particularly reliant on this form of empathy for engaging its audiences. We see this reliance in the use of the facial close-up, and the emphasis in film acting on genuine, and thus contagious, facial expression.

Part V: Can Films Philosophize? The fifth and final issue taken up in our volume has a controversial status in the philosophy of film. On the one hand, there are those who seem to assume

Introduction  •  11

that the whole of philosophy of film is properly taken up with the question of whether films can philosophize. On the other hand, there are those who contend that the issue is not really part of the philosophy of film at all. The first way of thinking is most closely associated with the French philosopher Gilles Deleuze, who was deeply interested in what cinema could reveal about the nature of space and time (1983, 1985). The extensive writings of Stanley Cavell, mentioned earlier, have also contributed to the sense that film requires philosophical attention as philosophy in itself. In reaction to this way of thinking, some authors point out that such attention could neglect what is most interesting about films as films. The worry is that the debate about whether films can philosophize cannot advance our understanding of the nature and value of film art, for the debate requires our treating film as a philosophical tool and failing to appreciate film on its own terms. This worry may be overstated, however. Indeed there seems to be a way to bridge the two sides of the controversy: The issue of whether films can philosophize is part of the philosophy of film but only a part. As we shall see in the work of our two authors, productive debate on whether films can philosophize cannot proceed without simultaneous attention to the nature of the art form and the nature of philosophical activity. As a long-standing contributor to the film-as-philosophy debate, Thomas Wartenberg is in an ideal position to identify common misunderstandings about and objections to the claim that film as an art form can do philosophy. In chapter 9, Wartenberg points out that defenders of the film-as-­philosophy, or FAP, thesis are not confused about films acquiring philosophical agency, nor do they count as proof of FAP the case of a filmed philosophy lecture. Rather they seek evidence, usually on a case-by-case basis, for the film medium being successfully employed in philosophical activity—for example, in the development of a thought experiment, or in the provision of new insight on a complex philosophical problem. In order to further clarify what is at stake in the FAP debate and provide a moderate defense of the FAP thesis, Wartenberg rehearses and rejects some general objections to the FAP thesis. These objections concern worries about incompatibilities between the aims of philosophy and film as an art, as well as questions about whether an appeal to filmmakers’ intentions or to the collaborative nature of filmmaking counts against philosophical activity in a film. Wartenberg takes the objections seriously and yet responds to them decisively. In the process, he is able to advance some positive claims about the ways in which particular films do philosophy, thereby helping to address the “curious asymmetry” in the FAP debate, which tends to involve supporters of the FAP thesis in defensive action against their critics. Despite Wartenberg’s systematic treatment of objections, Murray Smith uses chapter 10 to show that a broader critical perspective on the relation between art and philosophy yields new and deeper worries about the FAP

12  •  Katherine Thomson-Jones

thesis. Smith interprets the FAP thesis as a particularly strong version of the thesis that artworks—in particular, films—can have cognitive value. On the one hand, Smith does not doubt that films can have cognitive value—they can provide insight and understanding of various kinds and in various ways. But their cognitive value is intimately tied up with the other kinds of value they possess—aesthetic, ethical, and perhaps also political and historical. To support this point, Smith draws on Arthur Danto’s Hegelian view about artworks providing a sensuous embodiment of ideas: As such, artworks support a wide array of perceptual, sensuous, imaginative, emotional, and cognitive responses that resist both hierarchization and conflation. In combination with an analysis of the basic epistemic norms guiding philosophical activity, Smith concludes that films, even when they offer philosophical insight, do not do so “in the manner of philosophy.” With this conclusion, one might well conceive of Wartenberg and Smith as fundamentally opposed not just on particular points of argument within the FAP debate but also on whether the debate is even one worth having. Nevertheless, there are ways in which each author’s essay opens up new possibilities for the other’s position: Wartenberg’s essay involves some close analysis of individual films doing philosophy; these case studies challenge Smith’s claim that to focus on the philosophical in a film is necessarily to neglect its other artistically significant features. Smith’s essay reminds us of the complexity of our engagement with films; the challenge is then to give an account of the ways in which films do philosophy that embraces this complexity.

How to Read This Book Each section of the volume stands on its own, without presupposing familiarity with issues in other sections. This means that one can choose to read the sections selectively or in any order, depending on which issues most interest the reader. It is certainly recommended that one read both chapters in any given section, in order to gain the fullest and most critically nuanced understanding of an issue that this volume can provide. The authors have suggested additional reading should one wish to pursue an issue further. Since all five issues are central to contemporary philosophy of film, for teaching purposes it is recommended that all five sections of the volume be covered. The order of the sections reflects the order in which I tend to address the issues when I am teaching the philosophy of film as an advanced undergraduate course. My students in this course tend to find most difficult the arguments concerning cinematic narration. The chapters on realism in this volume would help prepare students for the discussion of narration, since there is considerable overlap between parts II and III, particularly to do with film viewers’ imaginative access to a fiction.

Introduction  •  13

References Bazin, A. (1967) What Is Cinema? H. Gray (ed. and tr.), Berkeley and Los Angeles: University of California Press. Carroll, N. (1998) A Philosophy of Mass Art, Oxford and New York: Clarendon Press. Cavell, S. (1981) Pursuits of Happiness: The Hollywood Comedy of Remarriage, Cambridge, MA: Harvard University Press. Coplan, A. (2011) “Understanding Empathy: Its Features and Effects,” in A. Coplan and P. Goldie (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Clarendon Press, 218–229. Deleuze, G. (1983) Cinéma I: l’Image-Mouvement, Paris: Minuit. (H. Tomlinson and B. Habberjam [tr.], Cinema I: The Movement-Image, Minneapolis: University of Minnesota Press, 1986). ——— (1985) Cinéma II: l’Image-temps, Paris: Minuit. (H. Tomlinson and B. Habberjam [tr.], Cinema II: The Time-Image, Minneapolis: University of Minnesota Press, 1989.) Walton, K. (1984) “Transparent Pictures: On the Nature of Photographic Realism,” Critical Inquiry 11, 246–277. ——— (1997) “On Pictures and Photographs: Objections Answered.” In R. Allen and M. Smith (ed.), Film Theory and Philosophy, Oxford: Clarendon Press, 60–75. Wilson, G. M. (1986) Narration in Light: Studies in Cinematic Point of View, Baltimore: Johns Hopkins University Press. ——— (2011) Seeing Fictions in Film: The Epistemology of Movies, Oxford: Oxford University Press.

This page intentionally left blank

PART 

I

What Is the Relation between the Art and the Technology of Film?

This page intentionally left blank

CHAPTER 

1

Cinematic Art and Technology BERYS GAUT

Chapter Overview In this chapter I argue that cinematic art is dependent on cinematic technology, and that the proper appreciation of cinematic artworks is partly dependent on understanding their technological features. In support of the latter claim, I develop the achievement argument, which holds that appreciating a cinematic artwork is in part a matter of appreciating the achievement that it is, which depends in part on understanding the kinds of difficulties that the filmmakers overcame in making it, and that understanding this partly depends on understanding the technology that the film incorporates. I then consider the prospect of a digital dystopia, the view that digital films are, other things equal, lesser achievements than non-digital ones. This seems to follow from the difficulty principle (other things equal, the more difficult an achievement is, the greater it is), together with the observation that it is easier to create certain effects in digital cinema than in non-digital cinema. I argue against the inevitability of a digital dystopia by showing that both what is achieved and the difficulties overcome can be greater in digital than in non-digital cinema.

Cinema, Media, Art, and Technology Cinema was born as a technology, but rapidly grew into an art. What is the relationship between its artistic features and its technological ones? One kind of relationship concerns the historical influence of technological developments on artistic ones and vice versa. In the history of cinema there have been 17

18  •  Berys Gaut

episodes when technology led art, others when art led technology, and yet other times when there were positive feedback loops in which technological change drove artistic experimentation, which in turn led to greater demands for technological development. There are also more systematic ways in which art and technology in cinema are linked. I will argue for two dependence claims: Cinematic art is dependent on cinematic technology (ontological dependence), and appreciating cinematic artworks is in part dependent on understanding their technological construction (appreciative dependence); I will develop what I call the achievement argument for the latter claim. To defend these two claims, something needs to be said about the concepts figuring in them. Cinema I will understand as the medium of the moving image: that is, it is the medium that has the capacity to generate apparently moving images (Gaut 2010: 1; 2012: 202). So construed, cinema consists of a wide variety of different media, including traditional celluloid-based film; hand-drawn cinema, such as that developed by Émile Reynaud in the 1880s; shadow plays; television; and digital cinema, including its interactive varieties, of which the most common type is video games. In most of these kinds of cinema the images are only apparently moving because their motion is illusory (they really consist of a series of static frames projected very rapidly). However, in the case of shadow plays, the motion of images is real, so what appears to be real movement is real movement. So in the definition of cinema “apparently” is not identical to “illusory.” The cinematic medium’s capacity to generate apparently moving images is almost invariably realized, but need not be: Derek Jarman’s Blue (1994) is a cinematic work whose visual aspect consists of a constantly blue screen, but what makes it a cinematic work is that it is in a medium that could have produced moving images. Finally, in cinema the moving images must be publicly accessible—that is, ones that many people can simultaneously see. If someone reading a novel visually imagines the events portrayed therein, that does not make novels a cinematic medium, since the mental images formed are private; even if many people read the novel simultaneously and happened to form exactly the same mental images, that would still not show that the novel is a cinematic work, due to the privacy of the images: One person cannot access another’s mental image.1 A medium I will understand as a set of practices governing the use of a material for communicative purposes (Gaut 2010: 287–290). Cinema is the medium of moving images, but its material varies. It may be physical (as with celluloid-based film) or symbolic (as in the bitmaps used in digital cinema). The material does not uniquely establish the medium: Celluloid-based film could be used to make films or to make sculptures. The medium is the set of practices that governs the use of these materials: The cinematic medium employs materials including celluloid-based film and bitmaps in order to generate moving images. Media nest—that is, some media are contained within

Cinematic Art and Technology  •  19

others (Gaut 2010: 290–291). For instance, the medium of cinema contains the specific media of traditional film and digital cinema. Media can also overlap: Digital films belong to both the digital medium and the cinematic medium, in parallel fashion to the way that operas belong to both musical and theatrical media.2 “Art” I will understand as a cluster concept (Gaut 2000). Cluster concepts are ones that lack a definition in terms of individually necessary and jointly sufficient conditions, but that satisfy criteria—that is, properties possession of which as a matter of conceptual necessity counts toward an object falling under the concept. None of these properties is individually required, but possession of certain subsets of them qualifies something as falling under the relevant concept. An artwork is an artifact that satisfies criteria such as possessing positive aesthetic properties, being expressive, intellectually challenging, formally complex, and coherent, having a capacity to convey complex meanings, exhibiting an individual point of view, being an exercise of creative imagination, being the product of a high degree of skill, belonging to an established artistic form, and being the product of an intention to make a work of art (Gaut 2000: 28). On these criteria there are many cinematic artworks, which include many popular films as well as those that are branded as “art films.”3 Finally, what is technology? As Peter Kroes (1998) notes, there is no consensus on what is meant by the term. But the following account suffices for our purposes: Technology is knowledge of how to make physical artifacts. Technology is a kind of knowledge. When we talk of the advance of technology during, say, the nineteenth century, or more specifically of the advance of loom technology during that period, we talk of the increase in knowledge about how to make various artifacts, such as looms. We can also use the term to refer to a collection of things, as in “there’s a lot of technology in Steve’s office,” and this is to say that there are a lot of technological artifacts present, artifacts produced by a kind of knowledge. So the construal of technology as a kind of knowledge is the more fundamental usage. Knowledge-how, of which technology is a species, is to be distinguished from knowledge-that. Technology is often confused or identified with applied science, a kind of knowledgethat, but the two are distinct. As George Basalla, a historian of technology, notes, technology is vastly older than science (Basalla 1998: 27). The most ancient technology known to us consists of stone hand axes, the oldest, Oldowan hand axes, dating from about two and a half million years ago; these may have been made by australopiths rather than members of our own genus, Homo (Boyd and Silk 2015: 269). They far predate scientific knowledge; so science-based technology is only one kind of technology, and until the latter part of the nineteenth century most technology was a kind of craft-knowledge. Even cinema grew partly from nonscientific knowledge, such as the development of the Maltese cross gear that draws the filmstrip in stop-start motion

20  •  Berys Gaut

through the camera and projector, and the development of the projector out of earlier magic lantern technologies. The term “artifact” figures in our definition of technology. An artifact is a kind of object made for some purpose. Not all artifacts are physical: Some are abstract, such as laws, constitutions, and some kinds of artworks, as Amie Thomasson (1999) has argued. It is tempting to hold that technology is knowledge about how to make any kind of artifact, including abstract ones. The problem in doing so, however, is that this would, for instance, count writing the U.S. constitution as a technological achievement, since it involves knowledge about making an abstract artifact. It is better to require that technological artifacts are physical: Oldowan hand axes are physical objects, and when one thinks of technological artifacts, one has physical objects in mind. But there seems to be a problem with this proposal: If all technological artifacts are physical objects, it appears to follow that digital cinema, which consists of bitmaps used to generate moving images, is not a kind of technology, for bitmaps are not physical objects but abstract, mathematical structures that specify the colors and locations of a set of pixels (picture elements), and therefore would not count as artifacts if all artifacts are physical objects. However, the difficulty is illusory, for digital cinema is the medium of moving images generated by bitmaps, and the process of moving-image generation requires physical machinery in order to create images from abstract structures. So if we restrict technological artifacts to physical ones, we can still maintain that digital cinema is a kind of technology.

The Asymmetry between Cinematic Art and Technology Given these understandings of the terms, it follows that not all media are technological ones. Language is a medium, since it employs symbols according to linguistic practices in order to communicate. But language is not a kind of technology, because spoken words are not physical objects, so one does not need to know how to make physical artifacts in order to speak. Further, the novel is not a technological medium, since novels are abstract entities whose instantiation need not be by physical objects (novels can be read out); in contrast, book production is a technological medium, since it involves making physical artifacts and so requires knowledge of how to make them. Music is a medium, but is not in itself a technological medium: Singing requires no knowledge of physical-artifact production. But instrumental music is a technological medium since it requires knowledge about how to produce musical instruments. Cinema is a technological medium. Without knowledge of how to make physical artifacts, such as cameras and projectors, cinema could not exist. This is true of all kinds of cinema: Even shadow plays require knowledge about how to produce cut-out figures, lighting sources, screens, and so on.

Cinematic Art and Technology  •  21

Art forms are different from media. Not all media are art forms: Communication by telephone is a medium, but is not an art form. An art form is the use of a medium for artistic ends (Gaut 2010: 289). I earlier defined a medium as a set of practices governing the use of a material for communicative purposes. Does not that therefore mean that an art form is a use of something that is itself a use? That sounds odd. To clarify, consider an example. Graphite is a material that can be used in various ways—for instance, as a lubricant or to make drawings. What constitutes a graphite pattern as a drawing is not the pattern itself, for a lubricant pattern might be identical to a drawing’s pattern, but rather the fact that the pattern is used in the medium of drawing. Not all drawings are artworks: Most technical drawings and doodles are not. But one can use drawings for artistic purposes—that is, to realize artistic values. When I talk of the use of a medium for artistic ends, I mean the use for artistic ends of the entities (e.g., drawings) that are constituted as such entities by their use in a medium (e.g., of drawing). So the use of a medium for artistic ends is not the use of a use, but the use of the entities constituted by their use in a medium. Not all art forms are technological: For instance, novels and poems are not technological achievements. These art forms require genre knowledge and various skills, but these are not technological knowledge and skills, since novels and poems are abstract entities. Likewise, purely vocal music is an art form that is not technological. But the cinematic art form is technological, for the same reason as the cinematic medium is technological: It cannot exist without knowledge of how to make physical artifacts. So there is no cinematic art without cinematic technology. This ontological dependence claim may seem false. Could there not be kinds of cinema that do not rely on physical artifacts for their production? Suppose that there were a kind of creature, the cinematons, who had a natural ability to communicate with each other by projecting publicly observable moving images simply by thinking, employing no technology. There would be cinematic art without technology, and the ontological claim would be false. However, the ontological claim is not meant to apply to all logically possible kinds of cinema. Rather, it concerns cinema as it has been, is, and is likely to be, and it should be understood in this sense. The point of theorizing about cinema is to explain and evaluate actual cinema and cinematic practices, not to construct theories about merely logically possible types of cinema. This may raise a fresh worry: Perhaps someday humans will produce cinema as the cinematons do, simply by taking thought, because a neural implant will be invented that allows humans to do this. We cannot dismiss this as something that will never occur. However, the imagined neural implant is a piece of technology, so there is no counterexample here to the ontological dependence claim. So there is no cinematic art without cinematic technology. What of the converse claim: Is there cinematic technology without cinematic art? Cinema is

22  •  Berys Gaut

used for documentaries, news reports, industrial training films, weather forecasts, and so on, which are not artworks. And cinema was a technology before it was an art: Many of the pioneers of cinema, such as the Lumière brothers, and W.K.L. Dickson (working for Thomas Edison), had no interest in the new medium as art, and it took figures such as Georges Méliès to draw out the artistic possibilities of distinctively cinematic techniques (Ezra 2000). So there is cinematic technology without cinematic art. Hence there is an asymmetry between technology and art in cinema. There is no cinematic art without technology, but there is cinematic technology without art. So there is an ontological dependence of cinematic art on cinematic technology, but no converse dependence.

The Achievement Argument Distinct from the issue of ontological dependence is a question about appreciative dependence: Is the proper appreciation of cinematic artworks at least partly dependent on understanding their technological features? I am going to argue that this is indeed the case. Before doing so, it is helpful to consider an example. In Breaking Bad, season 3, episode 12, entitled “Half Measures” (2010), Jesse decides to kill the two drug dealers who murdered his girlfriend’s ten-year-old brother. To jangling, discordant music, he walks into the frame, pulling out a gun, his face tense with fear, as the dealers nonchalantly draw their guns. Shown in a series of alternating close-ups, Jesse continues walking toward the dealers, the camera lingering on their unconcerned expressions, and it seems certain, as Jesse draws ever closer to them, that they will kill him. Suddenly, in a single, continuous long shot, a car veers into the scene from the right, its wheels spinning and screaming, and crashes into one dealer, throwing him over the roof of the car, and runs over the other, bucking up as it drags and crushes him; the first dealer, in the same continuing shot, crashes to the ground behind the car. In the subsequent shots the driver, Walt, dashes out of the vehicle, grabs the dealer’s gun, shoots him in the head, and in extreme close-up says to Jesse the single word “run.” The scene is brutal, terrifying, and shocking. It is pivotal in many ways to the season, marking Jesse’s increasing terror and disorientation, and Walt’s slide into ever-greater brutality. Central to its power is the car hit; its powerful realism partly depends on the fact that it is shown in a single, continuous shot; so the options of cutting away to allow the stunt performers to be removed to safety and other tricks to be performed are all unavailable. How did the filmmakers do this? “Hit and Run,” one of the extras on the DVD, reveals how: The two actors playing the dealers were filmed without the car being present, reacting to its arrival, one of them bending forward as if hit; then a separate shot of the car entering the scene was created and sandbags were placed under

Cinematic Art and Technology  •  23

the car, so that it bounced up, as if running over a body, without the actors present. Digital doubles were created from the point at which the car apparently reached the actors so that the actor apparently being thrown over the car was a digital replica, as was the actor apparently being driven over by it. Finally, the live-action and computer-generated elements were composited together to create the final shot. It is a great special effects shot (not least because it does not look like one) and a considerable achievement. But suppose that one had to do it in a continuous shot without using digital effects. It would be extremely difficult or nearly impossible to do so without killing or seriously injuring the performers. Perhaps a trapdoor could be created to drop the stunt performer below the car down to safety and a dummy sprung up in his place so that the car would buck up; and perhaps a rubberized car hood and roof could be created, together with a rubberized part of the tarmac behind the car so that a stunt performer could be thrown over the car and land behind it without injury. But doing it this way would be extraordinarily dangerous, and it would also require considerable ingenuity to work out how to accomplish. The sequence illustrates the general point that technology helps people to do things that would otherwise be very hard or impossible to do. It also shows something else: While the actual digital effects shot is a considerable achievement, doing it successfully nondigitally and completely in-camera would have been an even greater achievement. So appreciating the shot correctly requires one to understand the kind of technology that was used in making it: As a digital product it is impressive, but if it had been done using traditional (i.e., non-digital) film techniques it would be truly extraordinary.4 The point well illustrates the argument for the appreciative dependence of cinematic art on technology. This argument, the achievement argument, is as follows: 1. Appreciating a cinematic artwork depends in part on appreciating the achievement that it is. 2. Appreciating the achievement that a cinematic artwork is depends in part on understanding the difficulties the filmmakers confronted in making it and how they overcame them. 3. Understanding the difficulties the filmmakers confronted in making a cinematic artwork and how they overcame them is in part dependent on understanding the technology that the cinematic artwork incorporates. Hence: 4. Appreciating a cinematic artwork depends in part on understanding the technology that the cinematic artwork incorporates. The argument is valid: But are its premises true? Let us consider each of them and the conclusion in turn.

24  •  Berys Gaut

Appreciation and Achievement The first premise is: Appreciating a cinematic artwork depends in part on appreciating the achievement that it is. Artworks are not merely objects: They are artifacts, things that someone has made for some purpose. As such, appreciating them involves an awareness of not only their properties they have independently of the manner of their making, but also those properties that they have by virtue of being made in a certain way. For instance, a film may be beautiful by virtue of its visual properties, and grasping its beauty need not involve understanding how those properties were produced. Likewise, we may correctly say of a film that it has a good plot and rounded characters and looks spectacular without implying anything about how the film was given those properties. So some artistic properties of cinematic artworks can be correctly ascribed to them without implying anything about how the artworks came to acquire those properties. This class of properties I call achievement-independent ones. Another class of properties, achievement-dependent ones, are ones such that ascribing them does imply something about how the artwork was made to possess them. For instance, if I say that a film is very skillful, I imply something about how it was produced. I cannot determine just by looking at it that is a skillful film, since those marks that I take to be ones of skill may be the result of mere luck. In the Breaking Bad example, to know just how skillful the car-crash sequence is and in what way it is skillful, we need to know whether it was made using digital doubles or filmed in-camera using stuntmen. Likewise, whether a film is creative requires reference to how it was made: If apparently creative features are the result purely of luck, then they are not really creative; and if apparently derivative features of a film were produced by someone who did not know the film from which we supposed those features were derived, then those features are really creative after all, in Margaret Boden’s (2004: 2) sense of being psychologically creative—that is, new to their maker, rather than being the first time that they have ever been produced. Digital doubles had become fairly standard in special effects by 2010, so their use was not very creative; but assuming that the combined use of trapdoors, rubberized car parts, and stuntmen was not standard in 2010, then filming the sequence in this way would have been more creative, even though visual inspection might not show which method was used. Even features to do with viewer reception of films, as well as films’ creation, may be achievement-dependent. The Breaking Bad sequence would be even more shocking and powerful if it turned out that it had been filmed in-camera using stuntmen, rather than digital doubles. In the former case we would have been

Cinematic Art and Technology  •  25

watching a photographic recording of real people putting their lives in danger, and that fact, when we are informed of it, conditions our perception of what we see, as well as making its effects far more visceral. Viewer interest in achievement-dependent properties is evidenced by the widespread inclusion of “Making Of ” sequences on DVDs of films. So appreciation is of both achievement-independent and achievementdependent properties. Hence we should maintain, as the first premise states, that appreciation of a cinematic artwork is partly dependent on appreciating the achievement; it is not entirely so dependent, since not all artistic properties of films are achievement-dependent. Achievement and Difficulty The second premise of the achievement argument is: Appreciating the achievement that a cinematic artwork is depends in part on understanding the difficulties the filmmakers confronted in making it and how they overcame them. I characterized an artwork as an achievement in terms of how the artwork was made. But the notion of an achievement is richer than this. To say that someone has achieved something is to say that she has succeeded in her aims. But not all fulfilled aims are achievements. If I want to turn on the light, and cross the room to do so, I have fulfilled my aim, but it would be odd to say that turning on the light was an achievement. But if I were disabled, so that moving were very difficult for me, or if it were an extremely complex light switch that required a great deal of figuring out to operate, then turning on the light would be an achievement. In both of these scenarios something that before was extremely easy to do has now become difficult. So the notion of an achievement requires that there be some difficulty to overcome. An achievement has a value-neutral sense in which something may be an achievement merely because it is extremely difficult to do. For instance, if I manage to cross my fingers ten thousand times without pause, there is a sense in which that is an achievement. But in another sense it is no achievement at all, since it is not worth doing. (Imagine the deathbed dialog: “What did you achieve in your life?” “Well, I crossed my fingers tens of thousands of times.”) In the sense in which appreciating a cinematic artwork involves appreciating the achievement it is, it must be the evaluative sense of “achievement” that is in play, since appreciating something requires recognizing and responding to its evaluative properties. Hence in the evaluative sense, an achievement is success in doing something difficult that is worth doing. (Another way to put this point is that achievements are solutions to problems worth solving, where a problem is a difficult task.) In the case of producing an artwork the doing is

26  •  Berys Gaut

specifically the making of the artwork, so when an artwork is an achievement it is because it is the making of something worthwhile, which involves overcoming difficulties. It follows from this analysis of the concept that an achievement always involves some difficulties that are overcome. So to appreciate an achievement we have to understand the difficulties encountered and how they were overcome. For instance, in the Breaking Bad sequence we have to understand the difficulty of putting together a continuous shot of an apparently fatal car collision and how that difficulty was overcome by using digital doubles and compositing, and in our imagined in-camera variant we have to understand the difficulties of shooting the sequence using stuntmen. To put the same point in another way, to appreciate an achievement we have to understand the problem-­situation in which the filmmakers found themselves and how they went about solving the problem. This suffices to establish the main claim needed for the second premise, but we can also make a slightly stronger point about the degree of achievement, for given our analysis of the concept of achievement it is plausible that, other things equal, the more difficult an achievement is, the greater it is (call this the difficulty principle).5 For instance, it is an achievement to run a mile in four minutes, but it is a greater achievement, other things equal, to run it in this time if one is disabled than if one is not, since it is more difficult for the disabled person. “Other things equal” signals that all else remains the same in the comparison: For instance, that the runners are of a similar age, wear similarly good running shoes, the time is not a personal best for one but below personal average for the other, and so on. The difficulty principle also makes sense of our central example. The Breaking Bad collision sequence is a considerable achievement, since it was not easy to do, and it was worth doing, since it contributes importantly to the characterization of the key figures, the plot development, and thematic structure of the show. But if, as we imagined, it had been done by traditional in-camera techniques, it would have been a greater achievement, both because coming up with this idea of how to stage it would have been more difficult in the context than deploying standard digital techniques, and because implementing it would have been far more difficult and dangerous. One might object that the difficulty principle, absurdly, requires filmmakers to make things as difficult for themselves as possible in order to secure a greater cinematic achievement in overcoming those difficulties. For instance, the makers of the Breaking Bad sequence should have foresworn the digital route and spent their time instead making the crash sequence in-camera. However, that implication does not follow. The principle tells us, given the facts of the case, that the non-digital version of the sequence is a greater achievement and hence more valuable than the digital version. But what

Cinematic Art and Technology  •  27

someone ought to do depends not simply on the value (benefit) of one option available to her but on the benefits and costs of all of the options available to her, and those benefits and costs partly depend on available resources. Had the Breaking Bad filmmakers spent their time developing a non-digital version, given the tightness of their shooting schedule and budgetary limits, they would very likely never have finished the show. In contrast, when resources permit, it may make sense to take the non-digital route because of the greater achievement afforded by it. For instance, Christopher Nolan elected to shoot the spectacular main opening sequence of The Dark Knight Rises (2012) as a practical effect. The sequence involves a midair hijack with men descending from one plane into the fuselage of another, the second plane being cut into two, a hostage extracted by wire, and the sectioned plane crashed. The external shots were done using actual planes and stuntmen, with some model work, but minimal use of digital effects. Knowing it was done this way makes the sequence far more impressive than if it had been created using standard CGI techniques. Given the range of options available to him, Nolan made an artistically justifiable choice in filming this way. One may also object there is a better explanation for our intuitions than the difficulty principle. The reality-distance principle holds that an achievement in film is greater if the distance between reality (the pro-filmic event) and what appears on-screen is greater. To take another example from Breaking Bad, Walt’s son, Walt Jr., suffers from cerebral palsy, slurring his speech and needing crutches to walk. He is acted very well by R. J. Mitte, who really suffers from the condition, though in less severe form than does his character. According to the reality-distance principle, had Walt Jr. been played by an actor without the disability, the achievement would have been greater, since there would have been a larger gap between reality and the filmed fiction. In My Left Foot (1989) Daniel Day-Lewis, who is not disabled, plays Christy Brown, who had cerebral palsy of a more severe form than does Walt Jr. So Day-Lewis’s achievement is greater than Mitte’s—the principle holds—because the difference between his condition and that of the character he plays is greater than the difference between Mitte and his character. While the reality-distance principle looks plausible for such cases, it fails in others. In the case of the actual, digital version of the car-crash sequence the distance between the pro-filmic reality and the film scenario is large, since the actors were never in the same space as the incoming car, and there was nothing in pro-filmic reality corresponding to the drug dealers being thrown over the car and run over. But in the case of our imagined in-camera version there would be a smaller difference between what actually happened and what appears on-screen, since stuntmen would really have been thrown over and under the car. So according to the principle, the in-camera version would be a lesser achievement than the digital version, which is incorrect. Moreover, the

28  •  Berys Gaut

cases where the reality-distance principle appears to be correct, such as in the Mitte example, are explained by the role of difficulty in achievement. It would be, other things equal, more difficult for an actor without cerebral palsy to play Walt Jr. than it is for Mitte to play him, since Mitte has the foundation of a real disability that he can exaggerate for fictional purposes, whereas someone without the condition would have to build those propensities from scratch, as did Day-Lewis in his famously obsessive recreation of Brown’s condition. (Day-Lewis spent eight weeks in Dublin’s Sandymount Clinic for the disabled before starting filming, and stayed in character while on set, even being fed with a spoon and hardly getting out of his wheelchair.) So the difficulty principle explains the differences in achievement that the reality-distance principle fails to do. Finally, the second premise holds that appreciating a cinematic achievement depends in part on understanding the difficulties confronted and overcome. The “in part” formulation is employed because an achievement can vary not only in respect of the difficulties overcome, which is a matter of how one achieves something, but also in respect of what is achieved. So to appreciate an achievement we need to understand what was achieved, as well as understanding the difficulties overcome in securing that achievement. This point also applies to the difficulty principle, with its “other things equal” formulation, for the overall degree of achievement of a film depends not only on the difficulties that are overcome but also on what the achievement is, and these factors can vary independently to an extent. For instance, suppose two films, A and B, both need a lavish production design to be good films. Film A has the more lavish design, and so what is achieved by it is in this respect greater than what is achieved by B. But A also has a larger production budget, so its makers can easily produce these lavish effects by buying expensive materials, whereas the makers of film B have to find ingenious ways to make cheap materials look luxurious on-screen. So it is harder for the makers of B to reach the lesser level of what is achieved in their film than it is for the makers of A to attain the greater level of what is achieved in theirs. So difficulty of achievement and what is achieved can vary independently to an extent, and both matter to establishing the overall degree of achievement. Difficulty and Technology The third premise of the achievement argument is: Understanding the difficulties the filmmakers confronted in making a cinematic artwork and how they overcame them is in part dependent on understanding the technology that the cinematic artwork incorporates. Technology conditions both the difficulties filmmakers encounter and how they overcome them—that is, both the problems they face and the solutions to

Cinematic Art and Technology  •  29

them. In the Breaking Bad example there is an attempt to create a photorealistic depiction of a car crash: The difficulty of achieving that effect using digital technology is significant, but it is much greater if only non-digital technology is available, as we have seen. So the degree of difficulty of achieving the shot is dependent on the technology used. And technology also conditions how those difficulties are overcome—using digital doubles, or in-camera techniques with stuntmen. There are invariably some cinematic difficulties conditioned by technology, and that is because cinema is a technological medium. The kind of problem solving that cinema involves always has a technological aspect, since filmmakers have to deploy technology in making films: Editing, the use of lighting, camera movement, color correction, and very many other aspects of filmmaking require deploying technology to solve various problems. Even what is an addressable problem partly depends on this technology: Photorealistic animation was not in practice possible before the advent of digital cinema. Contrast cinema with a non-technological medium, such as novels. Here none of the difficulties and how they are overcome need be conditioned by technology: Inventing a good plot, coming up with convincing and engaging characters, mastering a prose style appropriate to the story, retaining readers’ interest, and so on are not technologically conditioned difficulties. Because the difficulties filmmakers confront and how they overcome them are in part dependent on cinematic technology, understanding those difficulties and how they are overcome is partly dependent on understanding cinematic technology. One cannot, for instance, understand the difficulties confronted by the makers of the Breaking Bad sequence unless one understands the technology they had available, nor can one understand how they overcame these difficulties without understanding that technology. The point generalizes because of the technological nature of the cinematic medium. Nevertheless, there are many cinematic difficulties that are not directly conditioned by cinematic technology. Creating a good plot and characters, finding and rehearsing actors to play those roles, scouting for locations, sorting out logistical problems, and many others are problems the solutions to which do not require the use of cinematic technology. So understanding these difficulties and how they were overcome does not require us to understand cinematic technology. And that is why the third premise is cast in terms of understanding the difficulties and how they are overcome being in part dependent on understanding cinematic technology. Appreciation and Technology The conclusion follows from the three premises that: Appreciating a cinematic artwork depends in part on understanding the technology that the cinematic artwork incorporates.

30  •  Berys Gaut

The “in part” formulation is because each of the premises incorporated that phrase: Not all appreciation is achievement-dependent, not all aspects of appreciating an achievement are a matter of grasping the difficulties encountered, and not all aspects of understanding cinematic difficulties involve understanding cinematic technology. This conclusion, the appreciative dependence claim, acts as a corrective to the otherwise tempting view that one can properly appreciate cinema as an art without troubling oneself to understand any of its technological aspects. A lot of work has to be put in to appreciate cinematic artworks, including understanding their technological construction. The conclusion is also important because it is a type of medium-specificity claim. Medium-specificity claims in cinema are controversial: For instance, one of the most prominent contemporary philosophers of film, Noël Carroll, has repeatedly argued against them (e.g., Carroll 2008: ch. 2). Some forms of medium-specificity claims hold, roughly, that we have to appeal to features of the cinematic medium if we are to understand and evaluate cinematic works correctly. In the versions that I have defended, the relevant cinematic features are not required to be unique to the cinematic medium, but may be only differential or distinctive ones—that is, features that distinguish one group of media, to which cinema belongs, from another group of media (Gaut 2010: 290–292). Being technological is a differential feature of cinema, since, as we have seen, some media are technological, such as cinema, and others are not, such as the novel. So the appreciative dependence claim is an example of a successful medium-specificity claim. It is in fact a version, applied to cinematic technology, of the MSV (Medium-Specificity Value) claim, which holds that some correct artistic evaluations of artworks refer to distinctive properties of the medium in which these artworks occur (Gaut 2010: 286). And the achievement argument is an instance and elaboration of a more general argument that I briefly developed in support of MSV (Gaut 2010: 293). So the arguments in the current chapter further support a perspective that stresses the indispensability of considering features of the cinematic medium if we are correctly to understand and evaluate cinematic artworks, and to give a full account of what makes the medium of cinema an art form.

A Digital Dystopia? I earlier argued for the difficulty principle, which holds that other things equal, the more difficult an achievement is, the greater it is. The principle could, equivalently, be formulated this way: Other things equal, the easier an achievement is, the lesser it is. We have seen how digital films allow their makers to produce certain effects far more easily than can be accomplished by nondigital means. It seems to follow that digital films must, other things equal, be

Cinematic Art and Technology  •  31

lesser achievements than non-digital ones. The digital dystopia argument runs thus: 1. Other things equal, the easier an achievement is, the lesser it is. 2. Digital films can produce certain effects far more easily than can nondigital films. Hence: 3. Other things equal, digital films are lesser achievements than are nondigital films. For example, the Breaking Bad car-hit sequence is a significant achievement, but it is a lesser achievement than had it been accomplished with traditional (i.e., non-digital) film techniques. Digital movies have witnessed an inflation of special effects—for instance, in each successive movie in the Transformers series—and audiences’ sense of wonder about how the filmmakers achieved those effects is rapidly eroding, for they were invariably done by using computer-generated images, rather than by finding clever one-off solutions to specific problems and confronting the dangers of grappling with reality. We seem to be heading, then, toward a digital dystopia, an evaporation of filmmaking achievement in an excess of digital filmmaking ease. The film theorist Rudolf Arnheim (1957) famously argued that the sound film is inherently aesthetically inferior to the silent film, and though his argument is flawed (Gaut 2010: 34–42), we appear to have a successful analog of it for digital cinema: The achievement of digital cinema is, other things equal, less than the achievement of traditional films. So are we entering a digital dystopia?6 The argument would not show that digital films are necessarily worse than non-digital ones, because of the “other things equal” clause. So, for instance, a digital film might be a greater achievement than a non-digital film, since it has a better plot, more engaging characters, better acting, and so on than the non-digital film. But if we hold all else equal, other than the fact that the one is a digital film and the other a non-digital film, then the argument maintains that digital film is the lesser achievement. However, the digital dystopia argument fails, for at least two reasons. First, in discussing the difficulty principle, we noted that the degree of difficulty overcome is only one ground for appreciating an achievement: The other is what is achieved. So in the difficulty principle “other things equal” signals that what is achieved is the same. But in certain respects, digital films can increase the level of what is achieved, so other things are not equal in those respects, and hence one cannot conclude that digital films are overall inferior to nondigital ones. Consider the 2005 film of King Kong, compared to the original 1933 version. In the latter, the two main models of Kong, designed by Willis O’Brien and

32  •  Berys Gaut

built by Marcel Delgado, that were used for the stop-motion animation were only eighteen inches tall. The animators could move Kong’s lips, eyebrows, and nose, but there was no possibility of producing subtle facial expressions, given the limited range of features that could be manipulated and the smallness of the models. In some close-ups a larger model of Kong’s head was used, which was powered by an air compressor and operated by three men; but there was still restricted control over facial features (Morton 2005: 33–36). In contrast, in the 2005 film, Kong has varied, subtle, and sometimes moving facial expressions. Since this Kong was built from a digital model, the filmmakers could construct the face in detail at any scale, and the digital facial-animation controls were far more complex and controlled more features than did the mechanical ones of 1933. So the digital filmmakers achieved subtle photorealistic facial expressions that were far beyond the grasp of the earlier filmmakers, and hence what they achieved went considerably beyond the level of their predecessors. And this would be true even if it were easier in 2005, given the state of technology, to produce the digital photorealistic effects than it was in 1933 to produce the physical-model effects. As we noted, degree of achievement depends on both the difficulty of securing a certain effect (how something was achieved) and how good that effect is (what was achieved). One cannot therefore conclude in general that since digital films generally make it easier to secure a particular result that they are lesser achievements, since they also make it possible to increase the level of what is achieved. The second reason that the dystopian argument fails concerns the second premise, for in certain respects digital films can be harder to make than can non-digital ones. For instance, developing the digital technology deployed to make a sequence can be extremely difficult and so a considerable achievement in itself. By the time of the Breaking Bad episode (2010) the use of digital doubles was well established in filmmaking: For instance, the technique had been deployed in The Lord of the Rings: The Fellowship of the Ring (2001) when the fellowship crosses the bridge of Khazad-dûm. Producing the software to make digital doubles and deploy them convincingly in film was a very difficult thing to do when first accomplished, even if it had ceased to be so by 2010. So to assess whether a digital effect counts as an achievement in respect of its difficulty we need to know the historical context in which it was created, and in certain contexts digital achievements can be considerable. After all, there is nothing about the nature of software development, or in its deployment, that means that it cannot count as a great achievement. There are also digital films where both what is achieved is greater and achieving it is more difficult than in traditional films. In Mission: Impossible—Ghost Protocol (2011) Tom Cruise’s character, Ethan Hunt, climbs up part of the outside of the Burj Khalifa in Dubai, the tallest building in the world. Hunt climbs up over ten floors from his hotel room, using special gloves that adhere to the

Cinematic Art and Technology  •  33

building’s glass surface, plummets several stories when he loses his grip, and then, having accessed the server room, rappels back down. But the line is too short, so he swings from it sideways and leaps into the hotel room, to be caught by Brandt, his fellow agent. Anyone with minimal knowledge of digital technology would assume that this was all accomplished safely in the studio using green screens, with the Burj Khalifa composited around Cruise. In fact, the actor performed the stunts on the real building: The chief use of digital techniques was simply to erase the cables attached to him. Cruise undertook very real risks to make this sequence: Not only could the cables have failed, but also hurling himself around at a height of more than two thousand feet could have led to very serious impact injuries. Planning and shooting the sequence also required significant logistical and practical problems to be solved: It was filmed in IMAX, which had to be frequently reloaded because the magazines hold a limited amount of film, with several shots taken from helicopters hovering next to the Burj Khalifa, and their position had to be exactly coordinated with Cruise’s. In traditional film, making this sequence in this way would have been impossible.7 Mattes can be employed in non-digital filmmaking to remove unwanted elements, and to composite in other elements, but matte lines are generated when the elements do not match up correctly. And mismatches are endemic, since traditional film is an analog medium, so getting an exact match between two elements is nearly impossible. Mismatches are particularly likely when the elements to be occluded are moving (requiring traveling mattes) and when they are very thin. So with non-digital techniques, trying to erase cables would lead to aesthetically unacceptable matte lines sprouting around Cruise, a problem exacerbated if filming were in the ultra-high resolution of the IMAX format. So in a traditional film Cruise would have had to be filmed in a studio, showing a mock-up of part of the Burj Khalifa, with safety equipment below him to arrest his fall, and the studio section composited into a shot of the real building, probably from a distance to reduce the matte mismatch problem. So by employing digital techniques, the filmmakers were able to set themselves a much more difficult problem than could have been set in a traditional film: to place an actor on a real building at an enormous height and for him to perform stunts on that building. When one learns of the real dangers and difficulties encountered, the sequence is rendered far more visceral, exciting, and impressive, and the artistic achievement is enhanced. Moreover, the resulting images are more detailed, realistic, and visually arresting than could have been achieved if the sequence had been shot using traditional techniques in a studio, where, for instance, it would have been impossible in practice to reproduce the myriad reflections of surrounding structures in the facets of the Burj Khalifa’s curtain wall. So sometimes digital films are greater achievements than traditional films both because the tasks set can be more difficult and because what is achieved—the results shown on-screen—is more impressive.

34  •  Berys Gaut

So, despite the apparent plausibility of a digital dystopia, there is no reason to think that we are confronting this situation. If digital special effect films seem increasingly tired and to be losing their ability to impress, this is because of how the new technology is being deployed, but not because its use will inevitably diminish artistic achievement.

Conclusion I have argued, then, for two kinds of dependency claims about the art and technology of cinema. The artistic properties of film are ontologically dependent on technology: There is no cinematic art without cinematic technology, since cinema is a technological medium. And there is also an appreciative dependence between the artistic and technological properties of films: Appreciating a cinematic artwork depends in part on understanding the technology it incorporates. I have also argued that this does not show that digital films are, other things equal, inferior to traditional films. The dependence claims also fit smoothly into the medium-specificity framework for understanding and evaluating cinematic artworks that I developed in A Philosophy of Cinematic Art. Cinema and the artistry that cinema evinces are through-and-through conditioned by technology.8

Notes   1 “Film” I will use to refer to types of cinema that are non-interactive (so video games are excluded) and involve illusory motion (so excluding shadow plays). On this understanding television also consists of films. The dependence claims apply to all types of cinema, though I will concentrate on films and so will henceforth use “cinematic work” and “film” interchangeably. One could narrow the usage of the term “film” further so that films are what are shown in the first instance in movie theatres, or that are high art, but I will not adopt these usages.   2 See Katherine Thomson-Jones’s chapter for further discussion of nesting.   3 This conclusion could be challenged, on the grounds that films are merely photographic recordings of artworks (they are in effect recorded theatre) rather than being genuinely cinematic artworks. To refute this objection we have to show both that artistry enters into how films record any pro-filmic events and that films represent in artistically distinctive ways. For the argument that cinematic artworks fulfill these conditions, see Gaut (2010: 21–50 and 300–306).   4 This point is about the relative difficulty of constructing the shot in 2010, the time of its making; it does not depend on the claim that digital shots or films are invariably easier to make than non-digital ones. I argue against the latter view in my discussion of digital dystopia.   5 The principle is about the difficulty of an achievement, in the evaluative sense in which an achievement is worth doing; so the principle does not maintain that merely overcoming difficulties, as in the finger-crossing example, suffices to make something an achievement. And since an achievement is a success, the difficulties must to some extent be overcome.   6 Thomson-Jones also discusses this issue in her chapter, offering a different solution from the one proposed here.

Cinematic Art and Technology  •  35   7 Historically, IMAX is an analog format, but in order to add digital effects, analog films are scanned into a digital format, creating a digital intermediate; the processed film can then, if desired, be scanned back to an analog format for projection.   8 I would like to thank Kate Thomson-Jones and audiences at the University of St. Andrews, the University of Groningen, and the Scottish Aesthetics Forum, particularly Patrick Greenough and Xinkan Zhao, for their helpful comments on earlier drafts of this chapter.

References Arnheim, R. (1957) Film as Art, Berkeley: University of California Press. Basalla, G. (1988) The Evolution of Technology, Cambridge: Cambridge University Press. Boden, M. (2004) The Creative Mind: Myths and Mechanisms, 2nd ed., London: Routledge. Boyd, R. and Silk, J. (2015) How Humans Evolved, 7th ed., New York: Norton. Carroll, N. (2008) The Philosophy of Motion Pictures, Malden, MA: Blackwell. Ezra, E. (2000) Georges Méliès: The Birth of the Auteur, Manchester: Manchester University Press. Gaut, B. (2000) “ ‘Art’ as a Cluster Concept,” in N. Carroll (ed.), Theories of Art Today, Madison: University of Wisconsin Press, 25–44. —— (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. —— (2012) “Replies to Ponech, Curran, and Allen,” British Journal of Aesthetics 52, 201–208. Kroes, P. (1998) “Technology, Philosophy Of,” in E. Craig (ed.), Routledge Encyclopedia of Philosophy, Vol. 9, London: Routledge, 284–288. Morton, R. (2005) King Kong: The History of a Movie Icon from Fay Wray to Peter Jackson, New York: Applause Theater and Cinema Books. Thomasson, A. (1999) Fiction and Metaphysics, Cambridge: Cambridge University Press.

Suggestions for Further Reading and Viewing Hugo (2011) directed by Martin Scorsese. (A fictionalized version, employing digital technology, of Méliès’s life, which incorporates shots from some of his films.) Price, D. (2008) The Pixar Touch: The Making of a Company, New York: Random House. (A detailed account of how a marriage of art and technology led to the first digitally animated movies.)

CHAPTER 

2

Movie Appreciation and the Digital Medium KATHERINE THOMSON-JONES

Chapter Overview This chapter considers the implications of the digital revolution in moviemaking for medium-based appreciation. One line of thinking suggests that the ephemerality and endless mutability of digital code, along with the computer automation of many traditional moviemaking tasks, count strongly against the artistic achievements of digital moviemakers. Insofar as we appreciate individual digital movies despite their being digital and thus easy to make, it must be the case that we appreciate them with no reference to the materials and methods of their making. A systemic response to such skepticism first requires an account of how the digital medium can be an artistic medium. Then we see that the digital medium sets different terms for the appreciation of movies than the traditional cinematic medium. This is due (at least in part) to the fact that the technology of computer automation fundamentally changes what counts as an achievement in moviemaking. Not everything is easy in digital moviemaking, and yet some of the most significant challenges arise precisely from the ways in which digital moviemaking appears easy.

Introduction Any artistically significant feature of a movie is made possible by the use of particular cinematic technologies. The groundbreaking use of deep focus in Citizen Kane required faster film stock than had previously been available and 36

Movie Appreciation and the Digital Medium  •  37

experimentation with new development techniques. The particularly sinister sequences in The Shining that involve a rapidly moving, floor-level point of view required the invention of the “Steadicam” camera-stabilizing system with an additional “low mode” bracket. And the presence in live-action movies of such fantastical creatures as Gollum and the Na’vi depended on the introduction of real-time motion-capture systems for computer imaging.1 In his related chapter in this volume, Berys Gaut provides a general argument in support of the dependence of cinematic art on cinematic technology, without the dependence of cinematic technology on cinematic art. The foregoing examples exhibit dependence relations of the kind relevant to Gaut’s argument. Such dependence relations could be specified in a more fine-grained manner: One could point to a particular application of motion-capture technology as making possible a momentary effect on-screen. Or the dependence relations could be specified very broadly, in terms of kinds of expressive, representational, or formal effects achievable with functionally defined categories of technology. At any level of specification, however, it is reasonable to ask just how relevant is awareness of these dependency relations for a proper and full appreciation of a movie.2 Detailed knowledge of, say, the mechanism of a movie camera or the coding of postproduction special effects clearly is not required for meaningful engagement with a movie. Some general, non-technical knowledge is required, however, of the kinds and capacities of available technology. Knowing that The Gold Rush was made in 1925, we expect to see black-and-white images on the screen; it would be unreasonable of us to complain about the lack of color in the movie, or the lack of a synchronized soundtrack, given the capacities of the technology available to the movie’s makers. (Not that sound and color were entirely unavailable to filmmakers in the 1920s, but reliable commercial systems for color printing and synchronous sound recording were not widely available until the 1930s.) On the other hand, if a contemporary movie is shot in black and white or without a dialog track, while we shouldn’t necessarily complain about it, we should certainly take an interest. The technology is standard for showing moving images in color and in sync with dialog; if a contemporary moviemaker chooses to eschew color or dialog, it is presumably for an artistically significant reason that we can hope to discover. From a technological perspective, moviemaking in the digital age appears, at first glance, strikingly different from moviemaking in the age of traditional, celluloid film. Different skills and a different distribution of those skills would seem to be required for moviemaking that involves, on the one hand, film-reel cameras, cut-and-splice editing, and in-camera, optical special effects, and, on the other hand, lightweight digital video cameras, computer editing software, and postproduction computer-generated special effects. Such a radical and pervasive shift in moviemaking practice due to

38  •  Katherine Thomson-Jones

technology is unprecedented: Even though there were many changes in moviemaking technology prior to the digital age—most notably, perhaps, with the introduction of synchronous sound in the late 1920s—none of these earlier changes simultaneously affected every stage of the moviemaking process, from production through distribution.3 By contrast, once you use a computer to convert dense image and sound information into digital code, you have the option of using digital tools at any stage of the moviemaking process, where these tools are simply computer applications, and the hardware for running those applications, enabling the indirect manipulation of the computer code for particular audiovisual results. Even if you shoot on traditional 35mm celluloid film, you can digitally scan the footage so that editing and mastering can be carried out on a computer. During this process, special effects, new backgrounds, and even new characters can be added to the footage. The final version of the movie can be streamed across the Internet, sent directly to movie theaters for digital projection. With such a marked technological divide in the history of moviemaking, it is interesting to consider whether it changes the relation between the art and technology of cinema. One line of thinking suggests that the ephemerality and endless mutability of digital code, along with the computer automation of many traditional moviemaking tasks, count strongly against the artistic achievements of digital moviemakers. Put another way, the speed and ease with which some previously onerous moviemaking tasks can be carried out digitally, along with the endless possibilities for modification built into these tasks, suggest that digital movies are just too easy to be artistically interesting. This line of thinking runs through a good deal of movie criticism and media theory, particularly in the early days of digital cinema (i.e., the mid1990s) when there was considerable skepticism about digital contributions to a mature art form already replete with masterpieces on celluloid. Essentially the same problem is raised by Gaut, in the final section of his chapter in this volume, with a particular emphasis on the use of digital effects: Digital movies have witnessed an inflation of special effects—for instance, in each successive movie in the Transformers series—and audiences’ sense of wonder about how the filmmakers achieved those effects is rapidly eroding, for they were invariably done by using computergenerated images, rather than by finding clever one-off solutions to specific problems and confronting the dangers of grappling with reality. We seem to be heading, then, toward a digital dystopia, an evaporation of filmmaking achievement in an excess of filmmaking ease. (Gaut, this volume: 31) As a first response to such skepticism, one might point to particular examples of digital movies that are artistically interesting, including ones that

Movie Appreciation and the Digital Medium  •  39

are heavily reliant on computer-generated special effects, such as Gravity. But the digital skeptic has a ready response, taking our appreciation of such movies to show that the medium is no longer relevant to appreciation in the digital age. As the means for conveying the artistic content of a work, an artistic medium comprises the technology employed in making the work as well as the ways in which the technology is employed.4 The skeptic argues that insofar as we appreciate individual digital movies despite their being digital and thus easy to make, it must be the case that we appreciate them with no reference to the materials and methods of their making. If we are able to appreciate at least some digital movies as significant artistic achievements, we must do so by ignoring the way in which they are made, for if we paid attention to the ease and automation with which they are made, our appreciation would not be what it is. In other words, digital moviemaking technology undermines the appreciative relevance of the art-making process because it removes the traditional challenges involved in that process. To say that the medium is no longer relevant to appreciation in the digital age is shorthand for saying that, although the dependency relations remain between moviemaking technology and the artistically relevant features of a digital movie, any kind of awareness of these relations is unnecessary and indeed best suppressed for the sake of the enjoyment of digital movies. This chapter will serve as a systematic response to this skeptical view concerning the relation of art and technology in digital cinema. The response I develop is different from but compatible with the response given by Gaut in this volume. Whereas Gaut considers how digital moviemakers can increase both the difficulty and the significance of traditional artistic challenges, I consider the possibility that new kinds of artistic challenge arise out of the very features that make digital moviemaking appear easy. More generally, we can acknowledge that a technological revolution occurred—is still occurring—in moviemaking with the introduction of digital tools, and yet show that the medium still matters for movie appreciation in the digital age. The first step is gaining a clearer understanding of what an artistic medium is, and thus how the digital medium, despite the ephemerality and mutability of computer code, can be an artistic medium. In thinking about how the medium properly informs appreciation, the notion of the “recalcitrance” of the medium is a particularly important one (Wollheim 1980). When it comes to digital cinema, we discover that the digital medium sets different terms for the appreciation of movies than the traditional cinematic medium. One way of understanding the difference is in terms of the digital medium having features of symbolic rather than material recalcitrance, to which digital moviemakers respond creatively and expressively. The recalcitrance of the digital medium is a function of the technology of computer automation where that technology fundamentally changes what counts as an achievement in moviemaking.

40  •  Katherine Thomson-Jones

Understanding the Medium In general terms, defending the appreciative importance of the medium involves an appeal to the so-called recalcitrance of the medium. The notion of recalcitrance is most often associated with Wollheim (1980) (e.g., by Davies 2003: 182; Lopes 2014: 135), who distinguishes the materials of artistic media by the way that they “present difficulties that can be dealt with only in the actual working of them” (Wollheim 1980: 42). Here Wollheim is responding to a view, attributed to Croce and Collingwood, that compels “indifference” to the making and materials of artworks in our true appreciation of the artist’s imaginative achievement. The trouble with this view, Wollheim suggests, is that it fails to acknowledge the inseparability of the “how” and the “what” of an artist’s imaginative achievement, that the artist’s achievement lies precisely in how he or she works with certain materials to overcome those materials’ inherent “difficulties.” When we appreciate artworks for what they are as works, we acknowledge, implicitly or explicitly, what must have gone into making them. A work’s history of making is always a history of making in a certain medium, in terms of how the artist worked with certain materials in certain ways. The medium presents particular challenges and possibilities for artistic creativity, and the artwork makes manifest the artist’s response to these challenges and possibilities. Although the ordinary art consumer may lack detailed, technical knowledge of art-making methods and materials, a general sense of them is built into our appreciation of artworks. Moreover, art critics will often elucidate the kinds of problems that arise in a certain artistic medium for the sake of explaining distinctive features of particular artworks. This practice is perhaps most common in criticism of the visual and plastic arts, comparing what can or cannot be done, more or less easily, with different kinds of paint and painting surfaces, different film stocks and types of cameras, and different sculpting materials and tools. An artist who achieves certain effects despite their difficulty in his or her medium, or an artist who discovers new (and interesting) effects in his or her medium, is duly praised; his or her work appreciated as a particular kind of, and a notable, achievement.5, 6 Discussion of the artist’s achievement often equivocates between distinct senses of the “medium.” This reflects the fact that a creative achievement need not only be based on a technical or artisanal achievement, involving the skillful and successful manipulation of materials and tools. A creative achievement can also be conceptual or institutional, providing the impetus for a new way of thinking about art or the kinds of things that can be artistically significant. In the most general terms, a medium is the means by which something is conveyed, transferred, or accomplished. When we describe water and air as physical media, it is in virtue of their serving to carry acoustic signals, say, or heat

Movie Appreciation and the Digital Medium  •  41

energy. When we describe newspapers and the Internet as news media, it is in virtue of their serving to deliver information to us about current events. This suggests that when we describe painting, poetry, or sculpture as art media, it is in virtue of their serving to convey artistic content, which most typically comprises the formal, expressive, and representational features of artworks. In the case of the art of painting, the medium is the means of conveying the aesthetically significant appearance of a marked surface. For any particular artwork, its artistic content is conveyed by a particular structure that we apprehend in order to access the work. The content of de Kooning’s Woman I (1950–52), for example, is conveyed by just its particular configuration of (particular) painted marks. But that configuration, referred to either as the “display” of the work or the “vehicle” of the work’s artistic content,7 is an instance of a kind, one common way of conveying artistic content—namely, the medium of painting. It is crucial to note that artistic content, including the content of Woman I, cannot be conveyed just by the painter’s tools and physical materials, nor even by the ranges of colors, textures, and shapes that can be realized with those tools and materials. Paint is used for purposes other than making art, and particular configurations of colors, textures, and shapes need not have artistic content. When the materials of painting are used to make art, there must be some further means by which the painted canvas represents a certain subject, say, or has certain expressive properties. This also reflects the fact that the same paint and even the same painterly configuration could be used to convey different artistic contents. Painting as an artistic medium is thus the combination of the perceptible properties of the painted canvas and the artistic intentions behind those properties, or the perceptible properties of the painted canvas accurately conceived as the result of an artist’s purposive activity. David Davies appropriately calls this the “agential” sense of the medium (2003), since it is a sense of the medium that moves beyond the materials of art to include certain of their uses, specifically for artistic ends. To speak of the brushstrokes and design of a painting, rather than merely of its marks and pattern (see Margolis 1980: 41–42), is to identify the purposively manipulated materials of an artwork, the work’s artistic medium, in just this sense. Understood agentially, the cinematic medium is not to be confused either with the raw materials of moviemaking or with the bare results of using those materials. Moving images are the result of using moviemaking technologies just as painted configurations of color, shape, and texture are the result of applying pigment to a surface. But neither moving images nor painted images need have artistic content. Used only in certain ways, moving or painted images serve specifically as artistic media. When movies are works of art, therefore, they are in the cinematic medium in so far as moving images (and sound)8 have been used in such a way as to mediate specifically artistic content. If an artist is trying to tell a story specifically with moving images rather

42  •  Katherine Thomson-Jones

than with words, with painted images, or with any other artistic materials, he or she will face certain challenges and opportunities that are specific to her chosen medium. The final work is then properly understood and evaluated as a particular kind of achievement on the part of an artist (or team of artistic collaborators) who faced just those medium-specific challenges and opportunities (as well, perhaps, as other challenges and opportunities common to several media). Interestingly, however, acknowledgment of a movie’s agential medium can seemingly become automatic and merely tacit. As veteran movie viewers, and given our image-saturated daily lives, we tend to take for granted the general artistic capacities of moving images, losing the ability to recognize as limits, possibilities, and challenges features of the medium that we treat instead as in some sense inevitable. Thus our engagement with movies is not informed, either implicitly or explicitly, by comparisons between the artistic potentialities of moving images versus still images, text, or theatrical performances. If our appreciation still relies on any comparisons, these are more likely between different kinds of movies. So we think about what can be done artistically with moving images of one kind compared with moving images of another kind— say, with live-action moving images compared with animated moving images. If this is indeed the case, must we conclude that awareness of the medium is not in fact crucial to appreciation? Not if we take seriously Berys Gaut’s suggestion that media can “nest.” This suggestion follows upon Gaut’s observation that “ordinary usage allows us to speak of one medium containing several others” (2010: 19). Print is a medium, for example, that contains other media, like etching, woodcuts, engravings, and lithographs. Print, moreover, is itself contained within the medium of visual imagery, along with painting and cinema. In this and in all his examples, Gaut consistently uses “medium” in the agential sense and avoids any reduction of the medium to the artist’s materials. Thus he notes that “the material does not invariably determine the medium. . . . Rather, the medium is how one uses the material.” At the same time, however, “when we finely individuate media, the materials employed sometimes play a role” (2010: 288)—as with the distinction between the print media of etching and lithography, or with the distinction between the media of oil painting and watercolor. Perhaps, then, any particular movie is best appreciated as being in a neat series of nesting media: images, then moving images, then moving images of a particular kind—say, animated moving images—then animated images made in a certain way—say, photorealistic computer animation.

Defending the Digital Medium The nesting idea is very important for Gaut, since he thinks it “lays out some of the methodology for the study of cinema” (2010: 19). We want to be able to

Movie Appreciation and the Digital Medium  •  43

appeal to the medium in explaining many of the distinctive features of a work and in evaluating the achievement that the work represents. We can do this for cinema as long as we are always careful to specify the relevant level of mediumspecific analysis. Thus our appreciation of a particular film can involve attention to the relative difficulty, ease, or exclusivity of certain effects in any and all of the nested media of the work as long as the terms of our comparison are at a single level of specificity. So we can compare effects achievable in pictorial and in lexical media, or effects achievable in live-action and in animated movingimage media, or effects achievable in hand-drawn and in computer-animated moving-image media. We can make all of these comparisons in relation to the same movie, and indeed this is just what is involved in its full and appropriate appreciation. When the movies to be appreciated are digital, however, one might worry about additional complications. A digital movie can be said to be in different media where the media in question do not nest. In light of my earlier remarks about the impact of digital moviemaking technology, it is natural to think of a digital movie as in the digital medium. But the digital medium cannot be said to nest in the medium of moving images. After all, the same digital technology mediates sounds and text, as well as images. Nor can the medium of moving images be said to nest in the digital medium. Gaut can easily acknowledge this fact and yet claim that the medium of digital cinema nests in both the cinematic medium and the digital medium. This is possible because (nested) media can overlap; digital cinema is found at one such point of overlap between two media. A practical worry might then arise about the feasibility of medium-based appreciation in the face of not only nested comparison classes but also multiple, overlapping series of (nested) comparison classes. Perhaps we must make a difficult decision about which medium matters, or matters the most, in our appreciation of a digital movie. Then again, such complications could simply encourage a preexisting tendency to regard traditional media distinctions as irrelevant in the digital age. In recent years, a host of film and media scholars have come to assume that, through so-called digital convergence, we have entered into a “post-medium condition.”9 One early and influential statement of this view is found in Friedrich Kittler’s 1999 work, Gramophone, Film, Typewriter: Before the end, something is coming to an end. The general digitization of channels and information erases the differences among individual media. Sound and image, voice and text are reduced to surface effects, known to consumers as interface. Sense and the senses turn into eyewash. Their media-produced glamour will survive for an interim as a byproduct of strategic programs. Inside the computers themselves everything becomes a number: quantity without image, sound or voice. And once optical fiber networks turn formerly distinct data flows into a

44  •  Katherine Thomson-Jones

standardized series of digitized numbers, any medium can be translated into any other. With numbers, everything goes. Modulation, transformation, synchronization; delay, storage, transposition; scrambling, scanning, mapping—a total media link on a digital base will erase the very concept of medium. (Kittler 1999: 1) Summing up this view, it supposedly follows from the fact that modern (digital) computers encode every kind of information in the same way—that is, as a sequence of binary digits—that an artwork that crucially relies in its production on computer processing is no longer defined by its mode of presentation— whether in images, moving images, sound patterns, or text. A work’s display is rendered merely contingent by the fact that it is generated from a common code. By adding a particular instruction to the code sequence specifying a work, imagery associated with that work could be instantaneously converted into sounds or text, or just into different imagery. Given the considerable effects of digital technology on artistic production, and particularly on movie production, it is perhaps understandable that commentators like Kittler are inclined toward a radical overhauling of art theoretical concepts. But their arguments in support of such an overhaul are, at best, incomplete. We see this once we cite some important continuities between ways of making and thinking about art in the analog age and in the digital age. It has always been the case, for example, that “any medium can be translated into any other”: Without using a computer, someone could manually devise a set of rules (an algorithm) for the translation of image values, say, into sounds or text.10 Moreover, a common storage and transmission means for (moving) imagery and sound is not unique to digital technology: As Doron Galili points out (2011), electronic image transmission going back to the late nineteenth century—in other words, TV and its precursors—relies on the conversion of both images and sound into electronic pulses. The electronic signal causes a membrane to vibrate in the receiving device, and these vibrations can determine either levels of illumination across the scanning disc of a visual display screen or a pattern of sound to be amplified. Apart from these important continuities, the media theorists’ inference from translatability to medium-free art simply does not hold. We could devise a set of rules for translating the imagery of 8½ into a symphony, but the original work remains a movie and thus in the medium of moving images. The new symphonic work that results from our translation is simply in a different medium, that of notated music for performance. The process of translation, in this case, is simply the process of creating a new work in a different artistic medium. But there is also the case of the same work being preserved and transmitted by different means. A literary work, for example, can be physically

Movie Appreciation and the Digital Medium  •  45

stored and transmitted by the printed page, in audio recordings, and by memory. This suggests a distinction between artistic media and the media in which works are recorded, stored, and transmitted. The same literary work could be accessed on the printed page or through an audio recording—the storage medium changes. But the artistic medium remains the same whatever the storage medium insofar as the work remains the same. Moreover, as long as the work is a work of literature, it is in the medium of literature. As an artistic medium, the medium of literature does not reduce to the materials by which the work is first recorded by the author or transmitted more generally to readers. It consists in letters and words used in certain ways. And as a symbolic medium, literature is, necessarily, a medium for translation. A literary work that is identified with a certain sequence of words, and not with any particular physical surface or object, can be stored successfully in any code or language for which we possess the rules of translation. Thus the possibility of translation does not so much erase media distinctions as confirm the role of symbolic artistic media. Written languages give us a way of preserving and transmitting literary works in many different physical formats. Similarly, binary numeric code, particularly when it is processed by high-speed computers, gives us a way of preserving and transmitting digital works in different formats. In this way, a plausible account of the digital medium can be given on the model of the literary medium: Instead of words and their artistic uses, binary code and its artistic uses mediate the artistic content of digital works. This allows for the possibility that the digital medium contains various sub-media, or “nested” media. Interestingly, digital sub-media seem to be distinguished in traditional terms. For instance, within the medium of digital art, the medium of digital visual art comprises artistic uses of computer code specifically to create images. In technical terms, such uses can be referred to as (artistic) “bitmapping,” given that a computer ultimately stores all images (2-D and 3-D vector) as bitmaps, code sequences that specify the integers assigned to light intensity measurements in a pixel grid. The medium of bitmapping is thus distinguished by a kind of digital technology, but the kind used to produce just those items belonging to the traditional medium of images. It is worth pausing to note that I have identified as the medium for a kind of visual image something (bitmapping) that is not in itself visible. One might consider this a problem, and indeed, Gaut argues that bitmaps in themselves cannot be the medium of digital images, since digital images must, by definition, be visible and thus in a visible medium. Bitmaps, according to Gaut, are just a constitutive material of the medium of digital imagery (2010: 14). I find this argument unconvincing, however. What is it for a medium to be “visible”? Assuming the agential account of the medium is correct, and a medium is more than (physical) materials but also essentially involves certain uses of those materials, it is

46  •  Katherine Thomson-Jones

hard to see how any medium—and not just the physical materials it involves— could be straightforwardly visible. At the same time, I think we can preserve the intuition about the essential visibility of a visual image while claiming that some visual images are conveyed by bitmaps. This is because the uses of binary code that comprise bitmaps incorporate the precise specifications for the visible realization of an image—bitmaps are just specifications for realizing an arrangement of light intensities. It is tempting to compare a nested digital medium, like bitmapping, to a nested print medium, like lithography, which is similarly distinguished by a particular kind of image-making technology. The nesting relations are different, however: Lithography and bitmapping are both technologically delimited ways of making images, but lithography neatly nests within the visual media— printed images and then images in general—whereas bitmapping does not; as a digital medium, it nests alongside other artistic uses of computer code to produce very different kinds of things, including sounds and text.11 In elaborating his idea of nesting media, Gaut cites the medium of digital cinema alongside other image media (Gaut 2010: 290), but he does not consider the relation of bitmapping to other digital media and how this relation might complicate medium-based appreciation of digital cinema. The complications are significant, however, in light of the fact that medium-based appreciation standardly appeals to the recalcitrance of the medium, the particular challenges and possibilities presented by a medium, to which a work in that medium manifests a particular response. When we consider the role of recalcitrance in appreciating digital cinema, it quickly becomes apparent that it is the recalcitrance of the digital medium, rather than of the traditional, cinematic medium, that matters most—matters most, that is, for the sake of confronting the digital skeptic who wishes to dismiss digital cinema as too easy to be interesting. As I shall go on to argue, the proper appreciation of digital movies depends in part on acknowledgment of a particular kind of challenge faced by digital moviemakers. This kind of challenge involves having to choose between a vast array of equally and instantly available options, and it arises specifically from the adoption of digital technology. Moreover, the challenge is confronted in making creative choices both about the look and about the sound (and about the relation of look and sound) of a movie. In the case of digital recording and editing, for example, many of the appreciatively relevant challenges are the same for the processing of images and sounds. Note that the presence of these common limits should not lead us to reconsider the popular but suspect “post-medium” view. Instead their presence should be seen as providing further support for the existence of the digital medium as a new artistic medium. Even though it is true and critically relevant that the medium of digital images nests inside the medium of moving images, we must pay attention to the fact that the medium of digital images also nests inside the digital

Movie Appreciation and the Digital Medium  •  47

medium. For it is acknowledgment of the recalcitrance of the digital medium, as manifest in both the look and sound of the movie, that allows us to successfully locate an important, and easily overlooked, kind of artistic achievement in digital moviemaking.

Digital Recalcitrance At first glance, it might seem odd to speak of “challenges” or “limits” or, indeed, “recalcitrance” in relation to digital processes, which allow for instantaneous and endless modification with increasingly user-friendly applications and devices. The high degree of automation in the process of capturing an image with a digital camera, along with increasingly high image resolution and memory capacity, could make it seem as though digital images are too easily achieved to be artistically interesting. Then there are the practically endless possibilities for “correcting” the captured image with applications like Photoshop. Digital sound recording is likewise increasingly automated, increasingly fine-grained, and reliant on ever-greater computer memory capacities. Modifying and mastering recorded sound with digital editing software allow for an almost unlimited testing of options. In digital movie editing, sequence changes are instantaneous and entirely reversible—quite unlike when the editing process involved the physical cutting and splicing of a film (image or sound) strip. Although professional movie editing software is highly complex such that training and experience are required to master the full range of its capacities, editors are not required to be computer programmers: With the click of a mouse or the twist of a console knob, an editor can directly implement a choice about features of the image sequence or sound track, which is then translated automatically and instantaneously, first into a high-level programming language, and then into machine code. When an image is specified in binary code, its manipulation comes down to the selective exchange of 1s and 0s in the code sequence. A modern digital computer represents binary code with the physical arrangement of millions of tiny binary switches, electronic transistors etched into silicon on the computer’s circuit boards. Given their physical design, computers can process binary code at a tremendous rate such that a task like image manipulation is executed seamlessly, rapidly, and with no risk of degradation to the image. Digital tools thus allow moviemakers to focus (almost) purely on the making of choices about the look and sound of the movie without having to worry about the technical difficulty or finality of implementation. It is undeniable that digital technology makes it easier, in particular ways, to make movies. But what does this mean for medium-based movie appreciation? Perhaps it means that, compared to traditional movies, digital movies are less interesting, less of an achievement on the part of movie artists who

48  •  Katherine Thomson-Jones

increasingly rely on many forms of computer automation. But this seems too quick. If the digital medium is an artistic medium in the same agential sense as other, traditional artistic media, then it is also subject to recalcitrance of a kind relevant to appreciation. Remember, recalcitrance is not just that of physical materials and tools. When the appreciation of literature is an appreciation of the artist’s achievement, this is not in virtue of recognizing the difficulties involved in wielding a pen or learning to type. Insofar as the recalcitrance of the literary medium consists in the limits of using a written language in certain ways, perhaps the recalcitrance of the digital medium consists in the limits of using computer code in certain ways. Most of the ways in which moviemakers use computer code are of course highly indirect, and so the kinds of practice that are part of the digital movie medium might include the navigation of a computer imaging application by way of a computer’s graphical user interface and the use of a mouse with some awareness of the implications of each mouse click. Literature and digital art are both in symbolic media, and, in either case, medium recalcitrance is surely a function of the syntax of the symbol system in question. Then again, the analogy between the media of literature and digital art may extend only so far. One might question, for instance, whether medium recalcitrance in the case of digital art is always syntactically derived. Some of the creative challenges faced by digital moviemakers appear to derive simply from constraints on the manipulation of tools under different real-world conditions—for example, the challenge involved in using lightweight digital cameras attached to gliders in order to give a dynamic and up-close view of birds in flight. In taking on this particular challenge, the makers of Winged Migration did not have to think about the limits of the code written for the operation of the digital cameras. Rather, they had to think about such physical challenges as those involved in securely attaching a camera to a glider and making sure the glider kept up with the flying birds. In cases like this one, where the creative challenges taken on by digital moviemakers are those involved in using hardware in situ, it is more difficult to think of medium recalcitrance as a function of the syntax of a symbol system. Ultimately, however, hardware-related challenges are conditioned by the underlying digital code. Winged Migration could not have been made before the introduction of small, lightweight digital cameras that can be controlled remotely, via a computer. In fact, the breathtaking aerial imagery in the final movie is made possible by digital cameras in at least two ways. On the one hand, the aerial nature of the imagery is due to the portability of digital cameras. On the other hand, the breathtaking qualities of the final imagery are greatly heightened by the process of editing down a vast amount of original footage—in one case, two months of recording in a single location was edited down to less than one minute in the final movie. By using digital cameras, the moviemakers can count on having a great deal of footage to work with in editing. The reason they can have so much footage, however,

Movie Appreciation and the Digital Medium  •  49

is that the process of digital image-capture culminates in the encoding of light intensity values, and once images are registered as binary sequences, their storage and replication are extremely efficient and reliable. This suggests that, even when digital moviemakers are facing challenges involving the manipulation of hardware in situ, the significance of those challenges has to be understood with reference to the underlying symbolic medium. Interestingly, in the case of the digital medium, what looks at first glance like evidence for a lack of recalcitrance turns out to be evidence for a new and distinctive kind of recalcitrance. Digital recalcitrance is a form of symbolic recalcitrance akin to literary recalcitrance, even though moviemakers in the digital age rely on a great deal of hardware and the object of movie appreciation is a perceptible artifact, a sequence of moving images and sounds. The combination of physical and symbolic materials in digital moviemaking contributes to the kind of skepticism, described at the beginning of the chapter, about the art of digital cinema. Arguably, such skepticism is based on a misunderstanding about which materials, physical or symbolic, contribute to the medium most relevant in the appreciation of digital cinema. Those moviemaking tasks that have been automated—handed over to a computer application—are not the ones we need to know about for an appreciation of a digital movie. It is not worth reflecting on exactly how a certain hue was given to an image sequence, in order to determine the difficulty of achieving just that look for the sequence and thus its relevance to the achievement of the movie’s makers, if the hue in question was simply the result of a selection made with the click of a mouse. On the other hand, it is worth reflecting on why the editor selected just that hue for the image sequence when an almost unlimited range of alternative hues were instantaneously available at the click of a mouse. Moreover, reflection on the editor’s choice has to take account of the conditions under which she or he made that choice. In the most general terms, such conditions are those of working with symbolic rather than with physical materials, and specifically with symbolic materials that are digital—a digital code that permits endless replication and recombination. With digital tools, many moviemaking tasks can be completed with greater ease and flexibility. But with this increase in ease and flexibility comes an overwhelming proliferation of creative options for moviemakers. Choosing the right option under these conditions can be very difficult, with a great deal of room for mistakes. Of course moviemakers can learn to ignore many options that are deemed irrelevant or unimportant to the making of a certain choice, but this in itself is a new challenge. In addition, there will be cases in which having more options rather than fewer makes a particular choice easier to make. But in many areas of digital moviemaking, the proliferation of options represents a genuine creative challenge. Add to this a new set of pressures from movie audiences, who are at least broadly familiar with digital possibilities for

50  •  Katherine Thomson-Jones

image and sound synthesis. Such familiarity is largely based on the ubiquity of amateur digital imaging tools (e.g., the video camera and editing apps on my smart phone). It is a reasonable expectation on the part of viewers that even live-action movies incorporate some computer-synthesized effects. Increasingly sophisticated motion-capture technology also allows for the photorealistic presentation of fictional beings, like Gollum and the Na’vi, as mentioned earlier. Almost anything can thus be pictured in a (“live-action”) movie, which likely makes it harder to impress viewers or hold their attention with image content alone.12 Of course image content is not all that matters for movie engagement; the mode of presentation matters too, including the style of editing. Here, too, a proliferation of options indicates a new kind of challenge for digital moviemakers. It was expected that digital movie editing would be far more efficient than cut-and-splice editing. In fact, however, movie editing in the digital age is still an extremely time- and labor-intensive process. Digital editing technology makes the implementation of changes fast and easy, but it also presents far more variables concerning the precise look of an image sequence. An editor working with a traditional film negative is faced with a great many decisions about whether to reorder or remove particular frames. But apart from keeping an eye on color consistency between frames, she will not be faced with decisions about changes to the content of the frames—that is fixed by the filming and printing processes. By contrast, an editor working with digital film footage, along with making decisions about the order of frames, also faces decisions about the tweaking of brightness, hue, and saturation values within a frame or even within part of a frame. She can also remove depicted or recorded items from a frame so that the frame’s insertion in a particular sequence does not break the consistency of what’s being shown on-screen. If the footage with which the editor is working was shot on digital video cameras rather than on traditional movie cameras, there also tends to be a lot more of it. Digital video footage is far cheaper than film-reel footage as well as easier to store efficiently and transmit rapidly. So the digital editor arguably faces more creative risks than the analog editor: Even though committing to certain editorial decisions does not entail irreversible alterations to a filmstrip, arriving at those decisions involves sifting through and eliminating far more options, a process that can easily become overwhelming and therefore more error-ridden. When we properly appreciate a digital movie, therefore, part of what we need to appreciate is the significance of any scene or sequence looking and sounding just the way it does when it could have, all too easily, looked and sounded so many other ways. This is undeniably a form of medium-based appreciation, but the medium to which we appeal is the digital medium rather than the traditional cinematic medium. This is not to deny the general appreciative significance of the cinematic medium. After all, when we take into

Movie Appreciation and the Digital Medium  •  51

account the enormous process of selection involved in arriving at the final cut for a digital movie, we presuppose the significance of there being a sequence of moving images that mediates the artistic content of the work. My aim here is simply to draw attention to the distinct terms for appreciating a movie in the digital medium, terms that we must embrace in order to respond to the digital skeptic. The significance of the particular look of a digital movie sequence derives from the challenge of choosing just that look when so many others can be tried out. The computer processing of binary code simultaneously makes far easier the implementation of creative decisions and far more difficult the initial making of those decisions. If the digital medium is computer code used in certain ways—for example, to make images—then it is only when we think of a digital movie as in this medium that we can appreciate it as a particular response to the creative problem, introduced by coding, of finalizing selections from a vast array of equally and instantly available options. Note, however, that this problem of selection in the face of proliferating options need not be the exclusive emphasis of movie appreciation in the digital medium. In his chapter in this volume, Gaut discusses another aspect of digital movie appreciation that refers to the possibilities for doing things in digital movies that could not be done with traditional moviemaking technology. In terms of the proliferation of options, it is interesting to compare different digital moviemakers’ responses to the problem. Some digital movies betray an over-enthusiastic and insufficiently selective response to the broad availability of options. The image track might look overly processed; the use of special effects may be overwhelmingly glitzy or comically exaggerated. Think of Oz the Great and the Powerful, whose scenery is described by New York Times reviewer Manohla Dargis as “a garishly hued, digitally rendered land that brings to mind the (bad) [album] cover of a prog rock band,” and whose characters are all rendered “eerily gleaming” by “unnecessary digital facial smoothing” (New York Times, March 7, 2013). Relatedly, a movie may suggest a fetishistic fascination with the capacities of digital technology, sacrificing narrative interest or coherent mise en scène for the sake of lurid effects. As Anthony Lane quips about Clash of the Titans, its “magic tricks are the foe of [narrative or diegetic] logic” (New Yorker, April 10, 2010). Sometimes, however, digital moviemakers get it right, successfully navigating the teeming space of computer-generated creative options. Gravity is uniformly lauded for its stunning beauty, with its distinctive look understood as both a creative and a technical achievement. The Observer’s film critic, Mark Kermode, fully expects viewers to marvel . . . as director of photography Emmanuel Lubezki opens a Pandora’s paintbox of light, bouncing the brilliant reflections of celestial bodies around the virtual set with a clear, crisp precision that not even the darkening effect of 3D glasses can dampen.

52  •  Katherine Thomson-Jones

Kermode concludes his review with a pithy injunction: “Honestly, if you’re not mesmerised by the look of Gravity, then it may be time for you to stop going to the cinema” (Observer, November 9, 2013). This level of praise for a movie that relies heavily on digital technology (and particularly on computergenerated imagery) is, in fact, quite rare. But, in light of my remarks on the nature of the digital medium, the rarity of such praise should not surprise us. On my account, the peculiar recalcitrance of the digital medium is such that moviemakers are facing a new balance of technical and creative challenges. Being impressed by digital moviemaking technology is not the same thing, after all, as being impressed by a digital movie.13

Notes   1 I will not be offering explicit analysis of the concept of technology in what follows. For helpful analysis of this kind, see Berys Gaut’s chapter in this volume.   2 I use the term “movie” throughout the chapter to include cinematic works made with either digital or traditional tools and materials.   3 The introduction of sound did not affect every stage of moviemaking, but it did affect at least two stages in profound ways. (1) Production was affected insofar as the movie set had to be newly equipped with microphones; early sound-cameras were hard to move due to the added bulk of sound insulation; and acting performances for early sound films were constrained by the need to stay close to stationary sound equipment. (2) Distribution was also affected insofar as the introduction of synchronous sound meant the replacement of live, local musicians in the movie theater with loudspeakers.   4 A small qualification may be needed here, if Gaut is right that not all art media are technological (see his chapter in this volume). In the absence of technology, we could say that the artistic medium comprises the materials employed in making the work as well as the ways in which the materials are employed.   5 In what follows, I am not assuming any particular account of art appreciation. But I am assuming that art appreciation generally involves both understanding and evaluation.   6 Gaut’s chapter in this volume develops a sophisticated argument for the dependency of artistic appreciation on technological understanding. This involves arguing that artistic appreciation is in part an appreciation of achievement; that achievement is to be understood partly in terms of the overcoming of difficulties; and that the overcoming of such difficulties is in part dependent on technology.   7 Dom Lopes describes the display as “a structure that results from the artist’s creativity and that we apprehend in order to grasp a work’s meaning and aesthetic qualities” (Lopes 2010: 4). David Davies argues that “an artist aims to articulate a particular artistic content by means of what we can think of as an artistic vehicle” (2013: 226).   8 Although most movies involve sound as well as moving imagery, the fact that movies need not have sound in order to be movies tends to lead theorists to identify the cinematic medium with the medium of moving images.   9 See, for example, Bolter and Grusin (2000), Doane (2007), Manovich (2002), and Rodowick (2007). 10 Although I am sure there must be some, I have not yet found actual historical examples of this kind of manual, algorithmic artwork-translation. Wassily Kandinsky is famous for his attempts to paint music, but he did not follow an algorithm in doing so. Rather, he relied on his synesthesia, painting sounds, as he experienced them, in colors. Then there is Vera Molnar, who, in 1950s Budapest, before she had access to a computer, was writing code for

Movie Appreciation and the Digital Medium  •  53 the creation of abstract images that, she hoped, could be somehow purer than anything she could make by hand, “more completely detached from established patterns, knowledge, and cultural heritage” (Lieser 2010: 47). Here she describes her 1959 project, Machine Imaginaire: “I imagined I had a computer. I developed a program and then, step by step, I realized simple delimited series, which were self-contained, that is, they did not omit any combination of forms. As soon as possible, I replaced the imaginary computer with the real deal” (Molnar 1990: 16). Of course, some translation processes would take too long to do by hand; the speed and accuracy of computer processing are essential—for example, Andreas MüllerPohle, Digital Scores III (after Nicéphore Niépce), 1998: The artist translated the oldest preserved photograph, Niépce’s view from his study, into digital form. He digitized the image and converted the seven million bytes into alphanumeric code. The result, transcribed onto eight square panels, is indecipherable, yet contains an accurate binary description of the original (Paul 2008: 48). 11 Of course lithography, as a method of printing, can also be used to inscribe instances of a musical score or a literary work. In other words, lithography is not solely a medium for image-based artworks. When lithography is used to create works of visual art, however, the medium of the artworks in question is identified with reference to the works being images created in a particular way—that is, by lithography. By contrast, the medium of bitmapping nests inside a medium that is not limited to images; digital technology is such that information in different formats (text, images, sounds) is conveyed in the same way (by a common code). 12 Of course, with handmade images, as opposed to mechanically recorded images, it has always been the case that anything can be pictured. The contrast in which I am interested, however, is relative to the history of live-action moviemaking. Within this tradition, it is striking the degree and extent to which digital technology changes the role of recording in creating images. The images in Avatar rely on the recording of patterns of bodily movement, which are then assigned to the computer-synthesized figures of the Na’vi. The result is images that look like photographs. So now we can have images that look like photographs but that, unlike true photographs, can picture anything, including nonexistent things. 13 Many thanks to Berys Gaut for his feedback on an earlier draft of this chapter.

References Bolter, J. D. and R. Grusin (2000) Remediation: Understanding New Media, Cambridge, MA: MIT Press. Davies, D. (2003) “Medium,” in J. Levinson (ed.), The Oxford Handbook of Aesthetics, Oxford: Oxford University Press, 181–191. ——— (2013) “Categories of Art,” in B. Gaut and D. Lopes (eds.), Routledge Companion to Aesthetics, 3rd ed., London and New York: Routledge, 224–234. Doane, M. A. (2007) “The Indexical and the Concept of Medium Specificity,” Differences: A Journal of Feminist Cultural Studies 18:1, 128–152. Galili, D. (2011) “The Post Medium Condition, circa 1895,” Presented at the Second International Colloquium of the Permanent Seminar on the History of Film Theory: The Impact of Technological Innovations on the Theory and Historiography of Cinema, Montreal. Gaut, B. (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. Kittler, F. (1999) Gramophone, Film, Typewriter, Stanford: Stanford University Press. Lieser, W. (2010) The World of Digital Art, Potsdam, Germany: h.f. ullman. Lopes McIver, D. (2010) A Philosophy of Computer Art, London and New York: Routledge. ——— (2014) Beyond Art, New York: Oxford University Press. Manovich, L. (2002) The Language of New Media, Cambridge, MA: MIT Press. Margolis, J. (1980) Art and Philosophy, Atlantic Highlands, NJ: Humanities Press.

54  •  Katherine Thomson-Jones Molnar, V. (1990) “Lignes, Formes, Couleurs,” Exhibition Catalog, Vasarely Museum, Budapest. Paul, C. (2008) Digital Art, 2nd ed., London and New York: Thames & Hudson. Rodowick, D. N. (2007) The Virtual Life of Film, Cambridge, MA: Harvard University Press.

Suggestions for Further Reading Binkley, T. (1977) “Piece: Contra Aesthetics,” Journal of Aesthetics and Art Criticism 35, 265–277. (An argument for the role of the medium in appreciation, where the medium comprises the conventions establishing the relative salience or aesthetic significance of a work’s perceptible features.) Carroll, N. (2003) “Forget the Medium!” in Engaging the Moving Image, New Haven, CT: Yale University Press, 1–9. (An argument against the role of the agential sense of the medium in appreciating movies.) Goodman, N. (1976) “The Theory of Notation,” in Languages of Art: An Approach to a Theory of Symbols, Indianapolis: Hackett, 127–173. (As a subset of the conditions for a symbol system being a notation, Goodman gives the conditions for a symbol system being digital. This has become the classical account of the analog-digital distinction.) McKernan, B. (2005) Digital Cinema: The Revolution in Cinematography, Postproduction, and Distribution, New York: McGraw-Hill. (A detailed but accessible guide to the technology of digital cinema.)

PART 

II

In What Ways Is Film a Realistic Medium?

This page intentionally left blank

CHAPTER 

3

Imagined Seeing and Some Varieties of Cinematic Realism GEORGE M. WILSON

Chapter Overview This chapter argues at length that when movie viewers “see” the elements of a movie’s fictional world, the “seeing” that is in question must be understood as “imagined seeing”—seeing through the imagination, prompted and structured by the evolving character of the image track. The “imagined seeing” thesis that is defended here is highly qualified and nuanced, and it is elaborated in considerable detail. On the basis of this overall position, I go on to explain the sense (or senses) in which our perceptual access to a movie’s fictional world can properly be understood as being “realistic” in a significant way.

Imagined Seeing and Some Varieties of Cinematic Realism In a rich, intriguing paper, Robert Hopkins raises the question, “What does a movie viewer see in a movie?” (Hopkins 2008: 149–159), and his discussion brings out some of the surprising complexity implicit in the question. At the end of this chapter, I will discuss an aspect of a key positive proposal that Hopkins offers on the subject, but I want to start by noting a central distinction he sketches. I will go on to point out several complications that are connected with the basic terms of the distinction and investigate at least briefly some ways in which these complications impinge on the messy notion of “cinematic realism.” For purposes of the present discussion, I will assume that traditional movies are in question here, movies that have been made by using motion picture photography to visually record the actors and sets that portray the fictional action. 57

58  •  George M. Wilson

It is, at first blush at least, both natural and appealing to say concerning movies that their viewers characteristically “see” items and events that belong to either of two categories. That is, (i) they “see” the largely fictional items and events that belong to the story told, and (ii) they “see” the actual items or events that belong to the “theatrical” portrayal of the story. In fact, this is a distinction that Hopkins makes, and he states it pretty much in these terms. For instance, in a Tarzan-movie, viewers may “see” Tarzan throwing spears at an enemy or swinging on a convenient jungle vine. The tier I items, belonging to the story told—for example, Tarzan and his vine swinging—are characteristically fictional. Of course, events and situations portrayed as belonging to the story are fictional as well. Viewers also “see,” as a part of tier I, a host of fictional facts. That is, they “see” that Tarzan is trying to ward off an enemy that threatens him or that he is swinging through the jungle foliage. Notice that the logical relations between these constructions can be a little complicated. A viewer might see (in the movie) Tarzan throwing a spear and thereby trying to ward off the enemy without seeing that (in the movie) it is Tarzan who is doing the throwing or even that he is trying to ward off the enemy. The viewer may see, for example, that someone they have yet to recognize as Tarzan is throwing his spear, and they see this much without seeing the cause (the threat from the enemy) that fictionally elicits his action. At the same time, it is likely that viewers will think of themselves as having “seen” things that belong to tier II—the level of material used to portray the fictional action with the aid of motion picture photography, editing, and so on. Thus, they will have also seen the actor, Johnny Weissmuller, going through various movements that enact Tarzan’s spear throwing or behavior that enacts Tarzan’s swinging on a jungle vine. In other words, they will have seen the actual items and events of tier II that have been used, in making the movie, to represent corresponding items and events of tier I. As natural as it may be to say that we “see” the actors and various facets of their acting and parts or features of the theatrical sets, there can be doubts about whether such perceptual claims are strictly true. Certainly, one might object that it is only the projected movie images that are before the viewers’ eyes, available to be seen, and that the viewers certainly don’t see the photographed subjects of the movie face-to-face. Those subjects are too distant from the putative viewers in time and space. Famous actors, for instance, may be long dead long before the movie is ever shown. Still, a number of theorists have responded that viewers nevertheless do literally see the items that belong to the tier of “theatrical” portrayal but see them only indirectly. They see these photographed subjects through the mediation of the projected motion picture photography on-screen. In doing so, they see these subjects quite literally just as a person might literally see her family behind her through the mediation of a conveniently located mirror.

Imagined Seeing and Cinematic Realism  •  59

This claim defines the familiar doctrine that photography (still and motion picture) is “transparent.” The general claim is this: TRANSPARENCY) If a viewer sees a photographic image of an actual object or event N, then, in seeing that image, the viewer thereby also sees N (indirectly). “Transparency” constitutes one of the several bases upon which film shots and larger movie segments have been affirmed to be “realistic” (Walton 1984: 246–277). Thus, it is thought to draw an essential contrast between photographs, on the one hand, and handcrafted images (e.g., paintings and drawings) derived from a model, on the other. The thesis has played a central role in many of the debates about cinematic realism. In discussing these matters, it will be convenient to speak of an image as being “transparently derived from” an item or situation O just in case O is a causal source of the visual appearance of the image, that visual appearance is counterfactually dependent in a dense way on the visible properties of O, and the counterfactual dependence has not been extensively set up through the intervention of human agency. Thus, a mirror image of a set of reflected circumstances is, in this sense, “transparently derived from” those circumstances, and arguably, a still snapshot or a motion picture shot of a photographed scene is “transparently derived from” the scene depicted. It does seem right to say that, in seeing the mirror image, one thereby sees the reflected situation indirectly, and also that, in seeing the photograph, one thereby sees the photographed scene albeit indirectly. However, I want to start out by describing some complications that are involved in the related notion that viewers literally “see” constituents of tier I—depicted elements in the story told. In particular, I will begin by focusing on two apparently simple questions about the largely fictional objects and events that are the denizens of tier I. I will ask, “Do movie viewers see the fictions in the movies they watch? And, if they do see these fictions, what makes it possible for them to do so?” Probably, most people believe that both of these questions have an obvious affirmative answer. After all, viewers, if asked what they “see” in the movie, will unreflectively report that they “see,” for example, Tarzan and that they “see” him swinging on a vine. Moreover, they are likely to report that what makes it possible for them to “see” these things is, in the first instance at least, the fact that it is Tarzan and his actions that have been “shown” to them on-screen. And yet, in thinking that the answers are trivial, these people are simply wrong. One sign of a level of hidden complexity here is illustrated by the following reflections. First, it is at least initially puzzling that moviegoers can simply and

60  •  George M. Wilson

straightforwardly “see” items that are, after all, merely artifacts of the filmmakers’ collaborative imaginations. Second, it will be widely agreed that viewers “see” Tarzan and that they also “see” the actor, Johnny Weissmuller. But do they simultaneously “see” both Tarzan and Weissmuller in the same shot and in the same sense of the word? On the one hand, the word “see” does not seem to have different linguistic meanings in the two occurrences. On the other hand, if viewers see a shot of Tarzan standing alone and are asked, “How many people do you see in the shot?” they usually will not answer “Two!” on the grounds that they are seeing both the fictional Ape Man and the actor who plays him. I will address these worries in a moment. Getting the answers to these questions right has important consequences for the theory of cinematic representation. It is often asserted, for example, that images and shots in the cinema are standardly “realistic” in an important, distinctive way. Similarly it is often claimed that movie narration or cinematic exposition in general has a natural affinity to some significant brand of psychological “realism.” However, issues of these sorts about the purported “realism” of movies are also notoriously obscure. The possible senses of “realism” that have application to the cinema are multiple.1 Many of these senses are connected to the idea that larger movie segments and movies as a whole make it possible to “see” their depicted subject matter in some characteristically “veridical” or otherwise privileged way. For example, if photographically derived shots in nonfiction films characteristically satisfy the condition of TRANSPARENCY) earlier, then the fact of transparency directly yields a sense in which the relevant nonfiction films might be thought to be “perceptually realistic.” But one has the impression that in familiar discussions of “realism in film” many (segments of most) fiction films—most movies—are also supposed to be instances of a similar kind of “perceptual realism.” Since, as I will explain, movies are not straightforwardly transparent to their tier I components, it is puzzling to understand how they can be “perceptually realistic” (construed in relation to these fictive components) in some analogous sense. It is crucial to make a beginning here to grasp what, if anything, “seeing fictions” might plausibly amount to. Moreover, given the historical centrality of essays by André Bazin to the discussion of “cinematic realism,” it proves useful, I believe, to distinguish what I will come to call “epistemic realism” from the sort (or sorts) of “perceptual realism” adumbrated earlier. In some contexts the problem of “realism” in film is raised as a question about whether cinematic narration or exposition—especially the narration of fiction films—has claims, normally or always, to be “realist” in some crucial way. It seems to me that this is related to but different from questions of “perceptual realism” that apply meaningfully to more local movie segments—for example, to individual shots. For instance, is it true that cinematic narration (or perhaps: narration only in cinematically suitable styles) prompts viewers to “see” objects and events in

Imagined Seeing and Cinematic Realism  •  61

a movie’s fictional world in a manner that is critically similar to the way in which these same viewers characteristically perceive the objects and events around them in the real world? This is closer to the topic that Bazin discusses in “The Evolution of the Language of Film” (Bazin 1967: 17–40). He famously maintains there that movies, in the range of “anti-montage” styles that he delineates, force movie viewers to register the natural significance of experienced continuities of space and time and to seek out actively the potential ambiguities, indeterminacies, and vagueness that may be implicit in their grasp of narrative (or expositional) sequencing in the cinema.2 The conditions that define for him an “anti-montage” style are, very roughly, the following: (a) the cinematic exposition involves the relatively extensive use of long takes; (b) there is an extensive use of shots that are composed in deep focus; and (c) there is a practice in the construction of the film of characteristically refraining from the interpretative use of editing—that is, the use of editing to assign the filmmakers’ special interpretation upon the depicted action and to resolve ambiguities of significance that would otherwise be present. In these respects, the “epistemic” position of movie viewers following a movie sequence may strike one as fundamentally like that of normal human perceivers who, situated within the concrete world as they are, find themselves regularly confronted with ambiguity, uncertainty, and vagueness in their environments. If some version of this idea is correct, then it yields an important way in which one would be justified in speaking of an “epistemic realism” that movies sometimes or even commonly exhibit. That is, sometimes or often, the ways in which movie viewers (fictionally) come to know about the world of the story are, so to speak, realistically presented and articulated. Later I will expand this proposal a bit, and I will accept a version of it.3 Let’s now go back to the narrower investigation of the crucial idea of “seeing” fictions in a movie. As noted before, this is not simply the thought that viewers “see” the characters in the story, but also the thought that they see them in various fictional circumstances, undergoing fictional events, and performing fictional actions elicited by the series of fictional situations that arise. Nevertheless, can we continue to accept that movie viewers “see” all of these elements of the story told when adequate ontological weight is given to the fact that the characters, their actions, and their circumstances in the diegetic world have, after all, only the status of fictions? Since these things are essentially fictive in nature it threatens to follow that the objects and characters don’t genuinely exist at all, the events and actions don’t really occur, and the situations and circumstances are never actually realized. They all belong to the realm of “make-believe” or “fanciful construction.” Even if it were allowed that fictive things may have some kind of existence— instantiate a particular mode of being—wouldn’t it have to be conceded further

62  •  George M. Wilson

that the mode of being is that of abstract items and abstract states of affairs? For example, when we speak about fictional characters, then we are speaking more accurately about certain dramatic roles, roles that different actors can embody. So, mustn’t it be concluded that, however we are naively inclined to think and talk about the matter, we do not—cannot—strictly speaking “see” such tier I constituents after all? This will be so, it seems, either because the fictions don’t exist at all or because they are not concrete and therefore cannot have a causal impact on the human senses. Perhaps, we don’t unequivocally and without further qualification “see” fictions in films. I am inclined to grant that point. Rather, I want to say instead that, speaking strictly, moviegoers imagine themselves seeing or, in some fashion, make-believe seeing the constructs in fiction film. This is a thesis—the Imagined Seeing Thesis—that a number of aestheticians—notably Kendall Walton, Jerry Levinson, and myself—have endorsed.4 This view faces some serious objections, several of which I will discuss in the later parts of this chapter. Nevertheless, with adequate refinements and defense, I continue to affirm the basic position. If we do accept it, we can go on to allow that when spectators at the movies claim to “see” Tarzan on the screen, they are thereby speaking a little loosely. They mean to be colloquially conveying, for instance, that it is make-believe for them that they are seeing Tarzan. Of course, if one adopts such a line of thought, one will incur the obligation of saying something helpful about how the newly favored concept of “imagined seeing” or “make-believe seeing” is properly construed. After all, there are several possible meanings of the phrase that are not in play when viewers report their experience of a movie story. For instance, there is a sense in which I might imagine (that I am) seeing my brother Mark in a street scene in Blade Runner. But that phenomenon, if it happened, would simply be the imaginative upshot of a certain kind of perceptual error. Surely, this sense of the phrase has little or nothing to do with the sense in which, as we commonly suppose, viewers imagine looking into the fictional world of a fiction film and seeing thereby what it is that is transpiring there. Finally, one can wonder, if a viewer’s “seeing” in the movies is only “imagined seeing,” how can it be that movie shots are “perceptually realistic” in relation to the fictions they depict? The phrase “imagined seeing,” as I am using it, does not apply to instances of purely imaginary seeing (like forming a private image of oneself seeing something in a certain visualized situation). It involves an actual seeing (of the movie images), a seeing about which the viewer imagines certain things. That is, it involves the actual seeing of a movie that the viewer imagines as a seeing of the elements in the fictional world of the story the movie tells. It is not at all uncommon that a person performs an activity or undergoes a process while imagining something false but engaging about the activity or process in question. Consider this odd but perfectly intelligible example. In drinking his

Imagined Seeing and Cinematic Realism  •  63

grape juice last night, surrounded by his collection of stuffed animals, Alfonso imagined about that episode of juice drinking that it was the imbibing of an exquisite but poisoned amontillado, consumed in the presence of his sympathetic family. This is a clear case of a person’s doing one thing (the juice drinking) and experiencing it (in his imagination) as something more and something different from what it actually is. Notice that in this instance and in many others the mode of imagining the pertinent activity is merely a manner of apprehending it, as it were, imaginatively. The imagining is not something else that is done in addition to and in synchronization with the viewing that is the subject of the imagining. The same is true, in my opinion, when the movie viewer imagines seeing fictions in the movies. That is, the viewer apprehends his own movie watching in a certain characteristic imaginative fashion. In fact, I believe that the viewing of movies is thoroughly imbued with a structuring narrative imagination. “Imagined seeing,” in the sense I have in mind, need not involve some further act or activity of imagining that is directed by the spectator at his actual visual experience. It is, roughly speaking, a matter of its seeming to the viewer “as if ” she were really seeing the segment of the fictional world on the screen. (For more on this, see my 2011: 68–74). This proposal offers the outlines of a sense in which movie images and shots may be thought to be “realist.” That is, they have a standard proclivity to prompt the sort of imaginative “as if ” impression of real seeing just described. The approach yields one reasonable interpretation of what a modest but plausible “perceptual realism” might be. The type of psychological/perceptual phenomenon in question here is not unusual, and it is not at all arcane. Think of the case in which viewers see the heroine (Serena) of a hypothetical movie fleeing through the rooms on several floors of a decaying mansion (the Mumford mansion) in which the story has mostly taken place. Serena is pursued by the villain, Old Man Mumford. In watching this movie, a viewer will be “seeing” the heroine running from Mumford through the complicated spaces of the mansion. The film’s expositional structuring will have carefully established for its audience just what the spatial layout of the Mumford mansion is supposed to be, and most competent viewers will have built up some kind of mental map or diagram that they imaginatively deploy in keeping track of the trajectory of Serena’s flight through the house. However, the visual articulation of space in the Mumford mansion is probably only a complicated and unreal construct that has been assembled over the course of the film by showing shots of varied spaces that probably occurred in very different photographed locales—in actual places that together have little of the geographical structure that the mental map exhibits.5 Viewers will “see,” for instance, a sequence of spaces in the lower floor of the mansion without knowing whether anything in the physical reality actually confronted by the movie cameras corresponds to the layout that

64  •  George M. Wilson

the movie apparently reveals. I want to stress: Movie viewers do “see” these narrative spaces in the movie—that is, they imagine seeing them much as they imagine seeing the fictional character Serena herself. And the viewers’ mental map of the elaborate terrain thoroughly infuses their “seeing” of the imbedded dramatic action in the fictional space. In my opinion, a viewer’s imagined seeing of a character (e.g., Tarzan) and his dramatic actions and circumstances is not very different from the way that the viewer’s imaginative experience of fictional spaces is activated and filled in. So, here is the basic idea. I propose to handle the puzzle about reports of movie viewing content by means of the following strategy. Instances of the various forms of words A i) The viewer sees Tarzan; A ii) The viewer sees Tarzan swinging on the vine; A iii) The viewer sees that Tarzan is swinging on the vine; often constitute true reports of what the agent is “seeing” in the film—true reports despite the fact that the words uttered, taken on their own while understood strictly and literally, say something false. How can this be? The puzzle is resolved, I am arguing, by the fact that the truth of the report depends on the fact that the explicit form of words in question is understood (and is intended to be understood) as contextually amplified in a certain way. That is, the explicit report is construed as if the speaker or thinker of a relevant instance of A) had said A i’) The viewer imagines seeing Tarzan; A ii’) The viewer imagines seeing Tarzan swinging on the vine; A iii’) The viewer imagines seeing that Tarzan is swinging on the vine. Moreover, all of these sentences, properly construed, can be used to say something strictly and literally true. So, in the first place, I accept a complex version of the Imagined Seeing Thesis. We can ask what viewers characteristically imagine about their visual experiences in watching a movie, and standardly it is a central aspect of their overall, imaginatively imbued impressions that they imagine them as instances of looking into the fictional world and seeing their fictional subjects as actual constituents of that world. Hence, normal viewers look into the world of Tarzan of the Apes and imagine seeing Tarzan swinging on the vine. When the viewers make the blunt claim formulated as, say, A ii), then they mean to express by this utterance a modified proposition like the one in A ii’). (See Soames 2009: 278–297.) A familiar example may make this proposal less artificial. Compare the instances of

Imagined Seeing and Cinematic Realism  •  65

A) with possible uses of B) The perceiver sees a unicorn wandering on his front lawn. On the plausible assumption that there are no unicorns on the person’s lawn for him to actually see, the statement made by a strict and literal use of B) will be strictly speaking false. But suppose that the perceiver in question is hallucinating a unicorn on his front lawn, and B) is implicitly used to report the perceiver (non-veridical) unicorn-inflected experience. In such a setting, the contextually amplified statement will be perfectly correct. This, it seems to me, is a very common use of perceptual reports like B). In such a use, B) says something like B’) It is visually for the perceiver as if he were seeing a unicorn wandering on his front lawn. It is my proposal that familiar reports of a subject’s movie viewing experience in construction like A i) through A iii) work in essentially the same manner. In both of these examples, it seems to me implausible that the word “see” is lexically ambiguous (like the word “bank”) in a way that gives rise to the true and false readings of B) or, by the same token, to the corresponding true and false readings of A i) though A iii). The explanation in terms of implicit contextual amplification is simpler and more credible, it seems to me, and it relies on a strategy for explaining the intuitive ambiguities here that applies widely to many intuitively ambiguous constructions in natural language. Thus, I am going to assume, subject to answering some key objections to the assumption, that viewers do “imagine seeing” the fictional objects and events in a movie. It is in this sense that the fictional objects and events are “seen.” So, for instance, in the story of Serena, let us stipulate additionally that viewers imagine seeing flying bats that are trying to attack her in the Mumford mansion. The bats are controlled by the villain, Old Man Mumford himself. Now, there is a good deal more to say about the viewers’ (fictive) perceptual relationship to the heroine, to Mumford and the flying bats, and to their various relations and interactions within the spaces of the Mumford mansion. If I imagine seeing the flying bats in a static shot, then I imagine seeing them from a fixed and qualitatively determinate perspective, a perspective that arises from within the fictional space in which they occur. I imagine seeing these creatures and their frenzied behavior at a certain rough distance, and I imagine seeing them from a more or less determinate angle in relation to the surface of the screen. Correlatively, if I “see” them in a tracking shot, then I imagine seeing them from a series of continuously changing perspectives, where each perspective, at every instant of the shot, is qualitatively specific in its visual details. I imagine

66  •  George M. Wilson

seeing the action from a perspective that moves, as it were, through the segment of space in the Mumford mansion where fictionally Mumford and the bats are attacking Serena. However, the provenance of these visual perspectives is itself perplexing when we try to think the matter through. If I imagine seeing the scene in the Mumford mansion, what is it that I imagine about how I come to see these fictions in these ways—in a perspectival manner whose “source” is alternatively fixed and mobile? Certainly I don’t imagine thinking that I am airborne, surveying the situation as I drift around among the flying bats. In one way, there seems to be no problem whatsoever. Or, at any rate, there is nothing mysterious about what we are seeing. We know very well that we are seeing the action from the perspective of a camera—the motion picture camera (or cameras) that shot the scenes in question. We know well enough how these motion picture shots came to be formed. And yet, that is not the question. The question is: What is it that movie viewers imagine about the mode or manner of the seeing that they imaginatively experience themselves as undergoing? What is it that they imagine about what makes it broadly possible for them to see these fictional items in this distinctive perspectival way? It is important to keep in mind that it is the question about what we imagine that is critical in this context, and this point will continue to be critical at several key junctures of the argument. An explanation of our apparent access to the fictions in terms of what we know about seeing movies is not in question. What is in question is an explanation about what we imagine about our access to the fictional worlds of movies. In a similar way, there is a contrast about what we are likely to know about how it is that the bats are able to fly around the rooms of the Mumford mansion. We know that the bat models that represent them have been, for instance, carried along on wires strung through the sets used to depict the mansion. But what we imagine about this question is very different. What we are likely to imagine is, say, that the bats have special other-directed flying powers and purposes bestowed on them by the villain. We may imagine further that they are flying after Serena because Old Man Mumford has contrived to have them harass the unfortunate heroine. Or take a different but very familiar type of example. In cases when backprojection has been employed, many viewers are likely to recognize that a special effect has been used even if they are not particularly clear about how it has been achieved. What they imagine seeing and what they are meant to imagine seeing is, for instance (in Henry Hathaway’s Circus World from 1964), the presence of certain of the characters surrounded on every side by huge visible flames. We know, of course, that the actors were not really threatened by fire in that way at all. We know that, when the dramatic action was filmed, the actors were actually performing in front of a screen upon which footage of a huge fire was back-projected. But the shot will have been constructed to look

Imagined Seeing and Cinematic Realism  •  67

as if it were a transparent image of the depicted characters contending with a surrounding fire. As a rule, we imagine ourselves looking into the fictional world of the movie and seeing the objects and events displayed in a succession of film segments, and we do so unless some intrusion of technical ineptness or artistic self-­consciousness disrupts our strong proclivity to imagine seeing in this way. Shots that have been built around “special effects” like back-projection will generally be constructed to be “photorealistic”—that is, to give the impression as much as possible that they are simply transparent images of the narrative space.6 They are thereby intended to prompt the viewer’s impression of its being as if she is (indirectly) seeing the depicted characters and the fire together within a continuous spatial region. Standardly, even stretches of CGI imagery, in the setting of straightforward narrative fiction films, will be designed to fulfill the objective of photorealism as well. In fact, the origin of the concept of “photorealism” arises in connection with this more specific type of case, but it can be applied to almost any shot or sequence in which special effects have been employed.7 In fact, one can state a pertinent analog to the principle of transparency for photographic images. Of course, movie shots are not really transparent to the items in the narrative world, but, at the same time, viewers characteristically imagine otherwise. Thus, a principle closely related to TRANSPARENCY seems to hold. That is, QUASI-TRANSPARENCY) if viewers “see” a motion picture image depicting a fictional item F and the image is photorealistic in appearance, then they are warranted in imagining themselves as thereby seeing the fiction F. This principle of “quasi-transparency” for movie shots plausibly grounds an extended sense of “perceptual realism” for movie images depicting cinematic fictions. It deserves to be thought of as a type of “realism” about fictional things precisely because of its intimate relation to the “realism” embodied in the more familiar idea of photographic transparency for nonfictions. Although movie images are not genuinely transparent to the movie’s fictional realm, the overall character and construction of the relevant film standardly help prescribe that its audience is to have the imaginative impression that they are. Paisley Livingston has highlighted an example that to my mind illustrates how complex and delicate these matters can be.8 In Buster Keaton’s (1925) film Go West, the main character, Friendless (played by Keaton himself), is chased by a bull, and the bull chases Friendless because he is wearing a red devil’s suit. The viewer imagines and is meant to imagine that the costume Friendless is wearing is red, although, given that the movie is in black and white, the viewer does not imagine seeing the redness of the outfit. Various contextual

68  •  George M. Wilson

clues may well prescribe an imagining that the costume is red, although the viewer may not be fully aware of how this prescription arises. For instance, it is patently supposed to be a “devil” costume in the story, and that is a central clue. In much the same way, if someone sees an actual person in a bright red suit, but sees that person and her suit only while looking through a window that is tinted blue, then the perceiver may be able to make out that the person observed is wearing a red suit, but the perceiver will not, on such an occasion, be able to see the redness of the suit. In the Go West example, when the viewer “sees” that Friendless is wearing red, this will be a matter of what the viewer “imagines seeing.” In point of fact, it could have been that the outfit Keaton actually wore was a certain shade of brown, and some viewers—people who had been members of the crew, for instance—might know that it was brown. Nevertheless, when they watch the movie, since the narrative context clearly conveys that fictionally the suit is red, they rightly imagine that they are seeing Friendless wearing a red suit. Moreover, prompted by other contextual clues, viewers imagine seeing Friendless being chased by the bull because he (Friendless) is wearing a red suit. But now, what do viewers imagine about why they are not seeing the redness of the outfit despite the fact that they do imagine seeing the outfit as having a definite color and seeing, moreover, that that color is red? More broadly, what do viewers imagine about why they do not see the distribution of colors in the scene despite the fact that they imagine themselves seeing a scene that has a natural array of colors? The answer here, I believe, is simple. Viewers imagine themselves seeing a naturally colored scene (or: seeing Friendless’s red outfit) through the mediation of moving images in black and white, images that have been transparently derived from the fictional subjects in the shot or scene. At the same time, such viewers do not normally imagine anything about a means or method within the fiction of the movie that has constructed or otherwise arranged the blackand-white mediation. No particular means or method for transparently deriving the screen images from the narrative reality (e.g., the wearing of the suit by Friendless) is imagined as the causal source of the segment. It is natural to worry that viewers do not imagine anything on this score, because they do not need to. A competent spectator knows that the actual filmmakers have shot the film in black and white and have contrived to convey that Friendless is wearing a red suit without being able to show the redness of the costume. The example of Friendless’s costume actually highlights once again the fact that there is a critical question about how the viewer imagines these matters, and that this question is not answered simply by appealing to what viewers know about what the filmmakers have and have not done. Thus, viewers are warranted in experiencing what I have dubbed an imaginative “as if ” impression of seeing Friendless in his red suit pursued by the threatening

Imagined Seeing and Cinematic Realism  •  69

bull, but they have no visual impression at all of seeing the redness of his suit. These delicate issues do not arise simply for instances of depicted color in movies. Similar questions and replies arise in connection with many of the visible properties that are presented on-screen in film. After all, in most movies the on-screen manifold of visual information will differ in several ways from what the face-to-face perceptual experience of normal human observers would have supplied, and the on-screen visual information may not be sufficient to determine which specific property of the entity in question gave rise to the information that the movie audiences consumes. Now it is this question about what makes our viewing of the movie fictions possible in the first place that can seem so daunting. Broadly, there appear to be two possible approaches to the question, and neither is feasible. FACE-TO-FACE) We imagine that we see directly the behavior of Serena and the pursuing bats and see them both face-to-face from positions we occupy within the fictional mansion. MEDIATED) We imagine that we see the behavior of Serena and the bats only indirectly and in an image-mediated way. That is, we imagine seeing the narrative occurrences by seeing them in “motion picture–like” shots that have been causally derived in some fashion from the actual presence of Serena, the bats, and the other elements in the fictional scene. Neither of these alternatives is even minimally plausible. I have already noted various reasons for rejecting the FACE-TO-FACE version. The apparent incoherence of the MEDIATED version is likely to strike one as even more obvious. In the Serena example, the audience shares visual perspectives on the actions that arise from within the rooms, corridors, and stairs of the Mumford house. And yet, they tacitly but unfailingly accept that there is nothing like a camera or a piece of recording apparatus located in the vicinity of the Mumford mansion. It would be absurd for viewers to worry whether fictionally Serena might bump into a piece of roving apparatus or hope that some of the bats might be fractured by flying into it. In the portion of the narrative in question, there are endangered heroines, creepy old houses, vicious bats, and the dangerous Old Man Mumford, but there definitely are no motion picture cameras on the scene. Hence, we don’t imagine ourselves seeing the narrative action indirectly in this way. The MEDIATED option also seems to be absurd. Hence, both potential refinements of the “Imagined Seeing Thesis” will probably seem to be untenable. How can it be that viewers “imagine seeing” the narrative action, even though it cannot be that they imagine their seeing of the fictions as being neither direct nor indirect—as neither face-to-face nor mediated? It looks like the overall position has run into a dead end. Maybe the very idea that we “imagine seeing” the narrative action has been confused from the beginning after all. A number of authors (e.g., Currie 1995: 38–42) have believed this to be the case.

70  •  George M. Wilson

Here is the solution I favor. On the one hand, I agree that if we imagine seeing fictions in the movies, we do not imagine seeing them face-to-face. And here is one other quick reason for rejecting the face-to-face version. Suppose I see a fiction film, and it is shot in black and white. As I have indicated, it is my position that movies are normally experienced as if they were transparently derived from the world of the story. Yes, we imagine and are meant to imagine seeing the fictional objects, events, and situations, but in the case of blackand-white movies we imagine ourselves seeing these fictional items in terms of black and white. Our imagined visual experience of things in the world of the story presents itself to us in a monochromatic visual manifold. Of course, we believe—and in fact we know—that we are seeing the actual photographed subjects, the actual actors and sets, for instance, through the mediation of black-and-white photographic images. But we do not believe and certainly don’t “know”—because the proposition in question is false—that we are seeing the constituents of the story through the mediation of black-and-white images that have been causally derived from the occupants of the fictional world. As I said before, we only imagine ourselves seeing into the fictional world through the mediation of the black-and-white images that have somehow— but in an unspecified manner—come to exist on-screen.9 And yet this just raises again the puzzle that I mentioned earlier. If we imagine seeing the fictional world through the mediation of qualitatively inflected images of some kind, what is it that we imagine about the means or method that might have supplied the images we supposedly experience? Surely we imagine something on this score! But, on the contrary, I want to say that normally we don’t imagine anything about this matter, and there is no rational expectation that we should construct an explanation, even schematically, in our imaginations. In the normal case, it is simply fictionally indeterminate for us how these transparent mediating images came to be. Many theorists have been absolutely correct to point out that we surely don’t imagine that a motion picture camera or some other recording apparatus has been present in the fictional world, serving to register the fictional occurrences of story. Moreover, this is precisely the point at which it is natural to insist that there must be something—some means or method—that served to register the visual profiles of the fictional items and events. To my mind, this temptation to insist—just because the question frames itself so insistently—that there “must” be something intelligible to say on this question is simply a philosophical illusion of the type that Wittgenstein repeatedly warned against.10 At some point, we have to say that this just is the imaginative framework we accept in watching our movie fictions. The imaginative framework lies at rock bottom, and no further explanation or justification is forthcoming. It is true that this position skirts self-contradiction pretty closely. But it is instructive to compare the conceptual situation in the movie case with the

Imagined Seeing and Cinematic Realism  •  71

situation in the case of literary fiction. Think particularly of a traditional novel that is narrated by a third-person narrator who is both effaced and, as we say, omniscient. That is, in the fiction, the narrator ranges freely over unlimited intervals of space and time and apparently has reliable access to outer and elaborately inner views of the characters. It would be possible, I suppose, to worry about how it comes to be the case that the narrator could have this implausibly extensive knowledge of the world that he or she describes. One could worry about this, but the worry would be silly. It is just a bedrock convention of certain novels, with a relatively unrestricted perspective, that if the narrator asserts that so and so, then, as a rule, it is true in the fiction that so and so. Moreover, the narrator fictionally knows this to be the case. If one insists on trying to reconstruct a method by means of which the narrator has reliably arrived at this more or less unconstrained knowledge, then one has embarked on answering a misconceived question. The fundamental conventions that get adopted in film and literature for the representation of fictional worlds can be both entirely reasonable and foundationally ungrounded at the same time. This is why, in my opinion, we are entitled to imagine our experience of fictional events in movies to have a certain qualitative character that partly arises from the fact that their appearance is mediated but that it is not mediated by any particular means or method we can name. We imagine seeing the fictions in transparent mediated images, but we imagine nothing or next to nothing about how the mediation came to be. Perhaps this could be dubbed a “radically modest” version of a mediated version of the Imagined Seeing Thesis. In any case, it constitutes, as I explained earlier, the basis of a novel brand of “perceptual realism” concerning movie images and shots.11 Thus, I accept this heavily qualified version of the Imagined Seeing ­Thesis— the thesis that, in watching fiction films, viewers do (normally) imagine seeing the characters, seeing the events, and seeing the circumstances depicted in the film.12 (These same viewers concurrently imagine “hearing” the sounds and noises that are produced from within the movie’s diegesis.) Indeed, I maintain that it is the standard function of the image track to mandate what the movie viewers are to imagine seeing in individual shots and collections of them. Similarly, the sound track mandates what the viewer is to imagine hearing from within the world of the story. I believe that it is because the sound track and the image track standardly have these complementary functions in a fiction film that we are justified thinking of them providing an audiovisual “narration” of the cinematic narrative. Finally, let me return then to the further question of explicating a kind of “epistemic realism” that I introduced earlier. If the images and shots of a given film preponderantly and systematically satisfy the condition of photorealism, and if the narration of the film conforms in addition to certain

72  •  George M. Wilson

Bazinian prescriptions concerning form and style—the prescriptions (a)–(c) that are stated briefly earlier—then one can say further that the movie as a whole or at least its overall narration is an instance of “epistemic realism.” On this approach, “perceptual realism,” as I use the term, is a necessary condition for this brand of Bazinian epistemic realism. To arrive at sufficient conditions requires that the additional, structural constraints on the pertinent cinematic exposition must be satisfied as well. But “perceptual” realism and “epistemic” realism, as I define those phrases, are conceptually connected. I mentioned at the start of this discussion that Robert Hopkins agrees that it is at least often the case that movie viewers “see” both (fictional) items at the level of story told (tier I) and “see” also the photographed items that are used to depict the story (tier II). For Hopkins, if I understand him, our “seeing” at both tiers is a matter of “seeing-in” as Richard Wollheim (1998: 217–238) has deployed that term of art. For me, of course, the viewer’s “seeing” of items, at least in the fictional domain, is a matter of a certain type of “imagined seeing.” I will not try to work out the significance of this apparent difference between us. But Hopkins takes a position on a question here that I have thus far not addressed. Presumably, viewers “see” at a given time both the relevant constituents from tiers I and II. But do viewers see elements from both tiers, at a given time, with each tier having an equal degree of salience and vividness? If theater employees are walking in front of the screen, the movie viewer is likely to see both the workers and the projected on-screen figures in pretty much the same manner. The viewer’s attention is likely to fall almost equally on each, and each will have about the same perceptual force. Is this even remotely like the situation of the viewers in relation to their normal simultaneous “seeing” of things from tiers I and II? Hopkins rightly thinks that this is not the case. He maintains that seeing into tier II standardly collapses into seeing into tier I. But what is the nature of this perceptual “collapse”? As I interpret him, Hopkins holds that the viewers’ “seeing” of tier I (their “seeing” of the world of the story told) tends to dominate and occlude their concurrent seeing of the theatrical objects and events at tier II. In any case, this is what he says: Two-tier film is often . . . made by creating a photographic record of the events filmed. Photographed pictures have a strong tendency to accuracy (the camera does not lie). But what is recorded in two-tier film is (often) a theatrical representation that is illusionistic. So we should expect to see in two-tier film only the story it tells. Our experience of two-tier film does indeed collapse. . . . When seeing-in collapses we are blind to a feature of the film’s making, i.e., that it was made by photographing a theatrical representation of the story told. But our experience does not misrepresent the facts.

Imagined Seeing and Cinematic Realism  •  73

We don’t see the picture as made in some way other than it was. . . . Thus, while illusion in the film’s making leads seeing-in to collapse, collapsed seeing-in is not in itself illusory. (2009: 71) Allowing for other differences between our frameworks, I think that I can make sense of these remarks. Certainly when viewers see Tarzan swinging on the vine, they also see Johnny Weissmuller on-screen and see him going through certain staged behavior. But normally their seeing of Johnny and his “theatrical” behavior is not something to which they notably attend as such. Usually, this aspect of their “seeing” is at best subliminal. What characteristically dominates the consciousness of movie viewers will be, for example, their imagined experience of looking into the (fictional) world of Tarzan and seeing the various things that (fictionally) occur around him in his jungle. If things go well in standard movie viewing circumstances, viewers will “identify” their conscious visual experience with their imagined visual experience of the fictional domain. It is probably no accident that things should work out this way. If Alfonso (see the example earlier) is trying to imagine his current juice tasting as some more vivid incident of drinking poisoned wine, then, if the sensed, non-­ imagined features of the tasting begin to dominate his awareness, the imagining is likely to falter and recede. Similarly, experiences of things that viewers imagine seeing on the screen will characteristically overwhelm the transparent perceptions that viewers simultaneously undergo. In any event, this is my reading of Hopkins’s idea that viewers’ seeing of things at tier I “collapses” into their seeing of things at tier II. It is for this reason that oversimplified versions of perceptual realism can wrongly falsify the complex phenomenology of spectatorship in the movies.

Notes   1 Of course, “realistic” is most often used in connection with questions about whether central aspects of the subject matter of a movie are treated with an adequate degree of accuracy and plausibility—for example, are the characters “realistic”? The questions about cinematic realism discussed in this chapter are not of this kind. The issues here are concerned with the relation of a film’s visual presentation to key aspects of ordinary human perception of the world. I hope that this will become clearer as the discussion proceeds.   2 How Bazin’s remarks in Evolution are to be construed has been hugely controversial, and the fairly standard formulations in (a) through (c) are chiefly meant to be invoke facets of his familiar but difficult discussion of the issues. In my opinion, it would do just as well for my purposes in this essay to modify (a) through (c) in significant ways and/or to add further conditions conceived in the same general spirit. Details are not crucial for present purposes.   3 The phrases “perceptual realism” and “epistemic realism” are both fairly widely used in the literature on realism in the cinema. In particular, both phrases are used in two of the most helpful surveys of the relevant literature I know—that is, Gaut (2010: 60–97) and Kania

74  •  George M. Wilson

  4   5   6

  7   8   9 10 11 12

(2010: 237–248). Despite the common occurrence of these phrases, however, there does not seem to be any stable, well-established usage for either term. As a consequence, I have felt free simply to stipulate the uses I adopt in this chapter. Both Gaut and Kania employ these phrases, but my use does not purport to agree with theirs. For a broader and very helpful overview of the topic, see Thomson-Jones (2008: 16–39). For Walton on this point, see his 1984; for Jerrold Levinson, see especially his 1996; for Wilson, see his 2011. It is well known that Hitchcock’s The Birds (1963) is like this. The real town of Bodega Bay is not laid out as it is presented in the movie. For the brief discussion of “photorealism” see Gaut (2010: 61–62). I believe that a concept more general than the standard one that Gaut cites can be profitably introduced and applied to a much wider range of cases. The more general concept is hinted at in my remarks in the main text. But this is too large a topic for a footnote. I am assuming that ordinary movie shots that do not employ special effects can trivially count as “photorealistic” themselves. Livingston (2011: 141–143). In the present chapter, I cover a number of points I first made in my 2011 reply to Livingston and to the other critics in this symposium. Of course, I’m speaking of the standard case. There are plenty of examples in which we are supposed to imagine that there is a recording device present in the story action and that the images projected on-screen are produced by the fictional device. Warnings against the urge to overgeneralize in various theoretical directions and to require explanations of inappropriate kinds are a theme that runs through a lot of the later Wittgenstein. See, in particular, Wittgenstein (2001). My use of the phrase “perceptual realism” is, to emphasize, pretty idiosyncratic. Gaut (2010: 71–78) introduces the same term, but uses it to denote something that is probably closer to what I am calling “epistemic realism.” This formulation applies only to objective shots at best, but subjective shots (of various kinds) can be captured if one were to adopt a more elaborate formulation.

References Bazin, André (1967) “The Evolution of the Language of Film,” in Hugh Gray (ed.), What Is Cinema? Vol. 1, Berkeley and Los Angeles: University of California Press, 17–40. Currie, Gregory (1995) Image and Mind: Film, Philosophy, and Cognitive Science, Cambridge, UK: Cambridge University Press, 38–42. Gaut, Berys (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. Hopkins, Robert (2008) “What Do We See in Film?” The Journal of Aesthetics and Art Criticism 66, 149–159. ——— (2009) Entry on “Depiction,” in Paisley Livingston and Carl Plantinga (eds.), The Routledge Companion to Philosophy and Film, London and New York: Routledge, 64–67. Kania, Andrew (2010) Entry on “Realism,” in Paisley Livingston and Carl Plantinga (eds.), The Routledge Companion to Philosophy and Film, London and New York: Routledge, 237–248. Levinson, Jerrold (1996) “Film Music and Narrative Agency,” in D. Bordwell and N. Carroll (eds.), Post-Theory: Reconstructing Film Theory, Madison: University of Wisconsin Press, 248–282. Livingston, Paisley (2013) “The Imagined Seeing Thesis,” in “Book Symposium on: Seeing Fictions in Film: The Epistemology of Movies, by George M. Wilson (Oxford: Oxford University Press, 2011),” Projections: The Journal for Movies and Mind 7:1, 141–143. Soames, Scott (2009) “The Gap between Meaning and Assertion: Why What We Literally Say Often Differs from What Our Words Literally Mean,” in Philosophical Essays: Vol. I, Princeton: Princeton University Press, 278–297. Thomson-Jones, Katherine (2008) Aesthetics and Film, London: Bloomsbury Academics, 16–39.

Imagined Seeing and Cinematic Realism  •  75 Walton, Kendall (1984) “Transparent Pictures: On the Nature of Photographic Realism,” Critical Inquiry 11, 246–277. Wilson, George (2011) Seeing Fictions in Film: The Epistemology of Movies, Oxford: Oxford University Press. ——— (2013) “Reply: Seeing through the Imagination,” in “Book Symposium on: Seeing Fictions in Film: The Epistemology of Movies, by George M. Wilson (Oxford: Oxford University Press, 2011),” Projections: The Journal for Movies and Mind 7:1, 155–171. Wittgenstein, Ludwig (2001) Philosophical investigations, 3rd ed., G.E.M. Anscombe (tr.), Oxford: Routledge. Wollheim, Richard (1980) Art and Its Objects, 2nd ed., Cambridge: Cambridge University Press. ——— (1998) “On Pictorial Representation,” Journal of Aesthetics and Art Criticism 56, 217–233.

Suggestions for Further Reading Branigan, Edward (2006) Projecting the Camera: Language Games in Film Theory, New York: Routledge. (A complex, intriguing discussion of cinematic narration and the epistemic relation of viewers to movie images and larger film segments.) Walton, Kendall L. (1990) Mimesis as Make-Believe: On the Foundations of the Representational Arts, Cambridge, MA: Harvard University Press. (A classic of contemporary aesthetics that articulates a general framework for understanding an audience’s imaginative access to fictional worlds in works of fiction.) Wilson, George M. (1986) Narration in Light: Studies in Cinematic Point a View, Baltimore: Johns Hopkins University Press. (A work that combines a study of general questions about cinematic narration and cinematic point of view with extended analyses of individual Hollywood films.)

CHAPTER 

4

Realism in Film (and Other Representations) ROBERT HOPKINS

Chapter Overview Is film distinctively realistic, is it realistic to a specially high degree, or in some unique way? The thought that it is has appealed to many, and realism has been found in various aspects of cinema. However, the phenomena offered are bewilderingly diverse, so much so that they push us toward the deeper question of what in general realism might be, if these are all to count. This is a question not only about film but about representations of any kind, be they moving pictures, still photographs, paintings and drawings, theatrical performances, sculptures, novels, poems, or other. It is relatively abstract, and answering it may seem to take us far from the concerns of critics and many film theorists. Rather than discussing particular cinematic techniques, such as the long take or deep focus, I will be talking about more abstract features, such as how much information is conveyed, or how far the movie sustains certain kinds of illusory experience. And when I attempt to find the underlying idea that renders those already abstract phenomena realist, things will get more abstract still. Nonetheless, my goal is to make out the idea of realism in such a way that it helps makes sense of talk of realism at every level, right down to the most concrete. There won’t be space here to connect what I say to the way critics and film theorists have used the term, but it is part of my ambition that such connections can be made. I begin this chapter by saying a little about what is meant by “film” here, and exploring its structure. I then present five kinds of realism that have

76

Realism in Film  •  77

been found in film, so understood: accuracy and precision, recessive form, illusion, transparency, and “collapse.” These are not in direct competition, since film might be realistic in all five ways. But what is realism, such that it can take these various forms? I answer by outlining a basic conception of realism, which I then use in the last section to test the claims of the five candidates.

The Structure of Photographic Fiction Film In what follows, I concentrate only on a particular, rather common, kind of movie: photographic fiction film. Such film is made by coupling two kinds of representation. The finished product involves pictures brought up on a screen. I’ll call these the Moving Images. Those images are the photographic record of various arrangements and events on set (the actors and their gestures, the props, scenery, etc.). I’ll call those the Events Filmed. And the Events Filmed are themselves representations of whatever objects, people, scenes, and events the movie ultimately seeks to put before its audience. I’ll call these last the Story Told.1 Thus cinema (of the kind that interests us) involves two tiers of representation:

Moving images

Events filmed

Story told

Figure 4.1  Tiers of cinematic representation

The first tier, representation of the Events Filmed by the Moving Images, is photographic. What of the second? Without begging too many questions, I think we can see it as at least very like representation in a play. Movie actors represent somewhat as actors on the stage do, movie sets represent somewhat as sets in the theater do, and so forth. This second tier is, then, more or less theatrical representation. In concentrating on film that has this structure, I set aside various other forms of cinema: animation, computer-generated imagery, and documentary. The justification for focusing so narrowly is that it is photographic fiction film that has been the focus of discussions of realism in cinema, and for which the widest range of claims for realism has been made. Since my hope is to use the diversity of realisms to motivate and guide the search for the nature of realism per se, this is the form of film that holds most promise.

78  •  Robert Hopkins

Given this structure, it’s tempting to wonder whether film’s realism derives from that of its constituents. This is certainly true of some cases. Some film is unrealistic because it involves unrealistic theatrical representations. For example, the German expressionist classic The Cabinet of Doctor Caligari shows the actors walking through streets represented by sharply angled and angular sets. The film doesn’t represent a world all akimbo. The movie is an unrealistic rendering of normal streets, not a realistic rendering of strange ones. Its unrealism stems from the unrealism of its theatrical representation: the Events Filmed do not share the appearance of the Story Told. Other movies achieve unrealism by involving unrealistic Moving Images. Richard Linklater’s Waking Life, for instance, records realistic theatrical representations of characters and their actions, but in Moving Images that have been doctored to look like less than fully realistic drawings. However, even in those cases in which we can attribute the (un)realism of the whole to that of its parts, doing so tells us where realism lies, but not what it is. When film is (un)realistic because the photographic and theatrical representations that compose it are, what does their (un)realism consist in? To say, we must examine some of the forms of realism that movies (as wholes) have been thought to exhibit. To that, our main business, I now turn.

Five Kinds of Realism I’m going to take you on a whirlwind tour of five kinds of realism that, it has been claimed, are found in film. I begin with forms of realism that are found in representations of any kind. Later, I describe realisms that apply to ever narrower ranges of representations. Only the last kind of realism has any prospect for being unique to cinema. Film might be specially connected to the others too, but only by displaying them to a specially high degree. Ask a philosopher what it is for a representation to be realistic, and you’ll probably be offered this: Accuracy & Precision: A representation is realistic to the extent that it represents its objects accurately, and to the extent that it represents them precisely. Accuracy and precision are distinct. I speak precisely but inaccurately if I say this section discusses seven kinds of realism, and accurately but imprecisely if I say it discusses a handful. Very roughly, precision concerns the amount of information conveyed about the represented item and accuracy concerns whether that information is correct. (For a more precise definition of precision, see Schier 1986: 176–177.) However, although the two are distinct, it’s convenient here to treat them as strands in a single kind of realism.2 A painted

Realism in Film  •  79

portrait of Oliver Cromwell might fail to show him “warts and all” in either of two ways. It might depict his skin as unblemished, and so be inaccurate; or it might depict his skin imprecisely (perhaps the image is rather blurry), so that the portrait sits on the fence about whether there are warts. Either way, the idea goes, the picture is less realistic (other things equal) than one showing the warts. Similar points hold for representations in language. While an academic history book might aim to be both accurate and precise about Cromwell’s character and actions, a children’s history might leave out the gory details (imprecision), and a satire might exaggerate them (inaccuracy). The first will be realistic as the other two are not. As with books, so with movies. It is all very well taking as an example representations of something real, such as Cromwell, but what should we say about pure fictions? Representations of made-up people, things, and events can be more or less realistic. A drawing of Eddard Stark may, for instance, be less realistic than a still image of him taken from the Game of Thrones TV series. Precision presents no difficulty here: the television image may simply tell us more about the fictional lord than does the drawing. Accuracy, however, is more problematic. Since there is no Lord Stark, if we want to compare the TV picture and the drawing for accuracy, to what should we take them to be true? (We could, of course, measure them against the descriptions given in the original Thrones novels, but that only postpones the evil hour: against what is the accuracy of the novel itself to be measured?) Now, perhaps the answer is simply that such comparisons cannot be made. Perhaps accuracy is part of realism only for representations of the real, and realism in fictional representations reduces to precision (along, perhaps, with other forms of realism described below). Even if that’s not right, there may be ways to make sense of accuracy when nothing real is represented.3 The issue is worth flagging, however, since it (or something like it) returns below. The first kind of realism is all about content, how things are represented to be. Precision concerns the content the representation has, and accuracy concerns the relation between that content and the world (Kulvicki 2014: 128–129). However, there’s more to representations than content. Any representation must also have other features, features we might for convenience bundle up as “form.” Our second kind of realism focuses on those: Recessive Form: A representation is realistic to the extent that its formal properties are recessive. What are the formal features of a representation? For present purposes, they are any features meeting two conditions: (i) they can be detected, in whatever form of engagement the representation is designed to support (e.g., watching

80  •  Robert Hopkins

a film, reading a novel, listening to a radio play); and (ii) they are not simply a matter of content, of what the representation represents. While form must be detectable, it might be more or less prominent to the representation’s “audience.” The less prominent it is, the more realistic the representation.4 The formal properties of painting include the brushstrokes, the thickness of the paint, and the facts about what color paint lies where. This last is a formal property that determines content: if the distribution of colors on the canvas had been otherwise, the painting would have depicted a different scene. Nonetheless, the distribution is not itself depicted by the picture, which represents a scene in 3-D, not its own flat surface. Such features can be more or less obvious to the viewer, depending on the picture. In some paintings, the brushwork is very prominent; in others it is far harder to see. Other things equal, the former will be less realistic than the latter, or so Recessive Form claims.5 Similarly in literature. Words have formal properties: their sound, syntax, and arrangement. (Again, some of these may determine what the words describe, but they are not themselves a matter of meaning.) A linguistic representation can highlight these properties, as poetry does through rhyme, alliteration, and distinctive word order; or it can let them drop into the background. These choices affect its realism. For a cinematic example, consider the point of view from which the action is shown. There are various ways to make this prominent. The movie might shift in rapid staccato between different angles on the same scene (the gunfight at the close of The Good, the Bad and the Ugly). It might hold a single viewpoint for unusually long (Rope). It might split the screen, showing the scene simultaneously from several angles (the polo match in the original Thomas Crown Affair). Or it might move smoothly and very quickly around the action in surprising ways, as in the “fly-arounds” developed for the Matrix movies, and used widely since. (This last effect is achieved in part via CGI, and so strictly speaking takes us beyond our remit of purely photographic fiction film. Even so, it illustrates how form in cinema can be more or less prominent.) The suggestion is that the more prominent are this and other formal features, the less realistic the film.6 What, though, if every formal feature were completely recessive, so that the viewer no longer saw them at all? (This need not contradict the definition of form—what is completely recessive at one time might, with a shift in viewpoint or attention, be readily detected at another.) Then she would be aware only of the representation’s content. And at least if the representation in question is visual (e.g., a picture, sculpture, or film) that, we might think, would leave her having the impression that what she is looking at is not the representation but the represented scene itself. She would be under the illusion of seeing that scene. This points to a third kind of realism: Illusion: A (visual) representation is realistic to the extent that it sustains the illusion of seeing what it represents.

Realism in Film  •  81

No doubt illusion is rarely, if ever, complete. Even the most realistic trompe l’oeil painting does not completely fool the eye. We may see the dome of the church as far deeper than it really is, but we nonetheless see clearly enough that the cherubs floating through it are painted, not real. Nor, at least in general, do we succumb completely to illusion in the movie theater. We see the border of the screen, the size of the projected images, and that what is before us is light shone onto a flat screen (though for disagreement see Allen 1995: ch. 3). Even so, the thought goes, our experience can mislead us in certain respects. When, for instance, 3-D cinema is most effective, we really do seem to see things before us arranged in depth. And this is true even if those things don’t appear to be the characters and events in the story, but only projected images of them. Even partial illusion such as this makes for greater realism, according to Illusion. While the first two kinds of realism apply to any representation, the third holds only for some: those that are in some way “visual.”7 Our fourth kind of realism holds for a narrower range still. Kendall Walton claims that photographs are transparent—they allow us to see, quite literally, the objects they depict (Walton 1984). Of course, we see only the photograph directly. Walton’s view is that seeing the marked surface directly allows us to see what it represents by indirect means, just as in a mirror we see someone indirectly by directly seeing his reflection. Indeed, one way to motivate this view is to construe photographs as fixed reflections: an image is formed inside the camera, just as it might be in a mirror or a camera obscura; and photography’s achievement is to have found a way to cement that image, so that it can still be seen long after the things whose image it is have gone. Walton claims that transparency is a kind of realism: Transparency: A (visual, photographic) representation is realistic when (directly) seeing the representation is (indirectly) seeing what it represents. According to Walton, photographs are transparent because they depend in a special way on their objects. They are counterfactually dependent on them (had the object been different in various ways, the photograph would have been). Moreover, that dependence is “natural,” i.e., the way the photo would vary with the object does not depend on what anyone believes about the latter. (Contrast pictures drawn from life. They too may vary counterfactually with the nature of the objects being drawn. If they do so, however, it is because changes in the objects would be noticed by the artist and thereby incorporated into the picture.) But natural counterfactual dependence is not sufficient for transparency. If we connect a TV camera to visual recognition software, and get a computer to generate descriptions of the scenes before the lens, the resulting verbal representations exhibit natural counterfactual dependence on their objects, but reading them is not a way to see those scenes. What is missing is

82  •  Robert Hopkins

what we might think of as a visuality condition: reading the descriptions is not enough like looking at the described objects in the flesh. To fill the gap, Walton also requires that our interactions with transparent representations model in particular ways our interactions with objects seen directly. For instance, there must be parallels in terms of the order in which we are liable to notice things in representation and scene, and in the sorts of mistakes we might easily make (Walton 1984: 270–273). It is important to be clear about how transparency relates to illusion. In illusion we seem to see directly the represented object. In transparency we really see, indirectly, that thing. Walton is not claiming that photographs are hard to see as photographs, or that they give us the illusion that what is before us is the represented object. On the contrary, he insists that photographs are seen for what they are, marks on flat surfaces. So transparency doesn’t involve illusion. And nor does illusion involve transparency. Seeming to see a heavenly vault before me is not a way really to see one, not even when the effect is achieved by really seeing a clever trompe l’oeil, for a trompe l’oeil can be (and usually is) painted, but only photographic pictures are transparent. Thus the third and fourth kinds of realism are distinct. Since film, at least the film we are concentrating on, involves (moving) photographic images, Walton’s view will apply as readily to cinema as to still photographs. Film will be realistic in his sense. However, it will be so only with respect to one of the elements identified earlier. When we visit the cinema, what is before us is the moving photographic record only of the Events Filmed: the Story Told is represented by non-photographic means. Since transparency depends on photographic representation, film will be transparent to the Events Filmed, but not to the Story Told. However, while this kind of realism reaches only so far, what it reaches at all it reaches completely. Unlike the first three kinds of realism, transparency is an all or nothing affair. Seeing the representation is either a way to see the object, or it is not. Transparency admits of no degree.8 Our final kind of realism finds something to agree with in each of the two preceding views. Like Walton, it rejects the idea at the heart of Illusion, that in the cinema we ever cease to be aware of the moving images before us. Like Illusion, and unlike Transparency, however, the realism it describes reaches to the Story Told. The idea is that, although film has the complex structure described in the first section of this chapter, we do not always see matters that way. One level of representation drops out of our visual experience:

Moving images

Figure 4.2  When Events Filmed drop out of the viewer’s experience.

Story told

Realism in Film  •  83

We see the moving images, but in them we no longer see the actors, sets, scenery, and so forth. All we see in the pictures before us is the Story Told. What does it mean to say that we “see” these events “in” the pictures? The idea is not that those events seem to be before us—rather, we are visually aware of what is before us as a mere pattern on a surface. (There is no return here to Illusion.) But nor is the idea that seeing that pattern is a way of really, albeit indirectly, seeing the events composing the Story Told. Only what exists can really be seen, even indirectly, but these events may well be fictional. (Thus the proposal here really does differ from Transparency.) But if not these options, what? Consider an ambiguous picture, such as the famous duck-rabbit. When you shift between seeing the marks one way and seeing them the other, you always see the marks: at no point does a duck or a rabbit appear to be before you. Nor is looking at this line drawing really a way of seeing—not even indirectly—a duck or rabbit. Nonetheless, switching between interpretations does alter your experience in radical ways. Seen one way, the marks are organized rabbit-wise, and seen another they are organized duck-ly. In each case, you are visually presented with two things at once (marks and duck, marks and rabbit), even if only one of those things (the marks) looks really to be there. And while ambiguous pictures are special in that we can see either of two things in them, in other respects they are quite ordinary. Every representational picture allows for something similar. Every such picture—be it moving or still, photographic or not—allows us to see the pattern that composes it as organized in such a way that we are thereby visually presented with something else. To say someone is seeing X in Y is just to say that that her visual experience has this complex structure (see Wollheim 1987: ch. 2; Hopkins 1998; Kulvicki 2014: ch. 1). Sometimes seeing-in is complex. If you show me a drawing of a statue of a man, then in the drawing I’ll see the statue, and in the statue I’ll see the man. Often our experience before traditional film has a similar complexity. In the Moving Images before us we see the Events Filmed (actors, sets, etc.), and in those in turn we see the Story Told (the events that are acted out). The idea behind the current proposal, however, is that things need not be this way. Sometimes we cease to see the middle term, and simply see in the Moving Images before us the Story Told. Since one of the two tiers of representation here has disappeared from our experience, we might think of the result as a kind of collapsing of levels. The level that remains is made up of moving photographic images. We now see those as images of the events in the story being told. We may well continue to see them as photographic (Hopkins 2008: 156–158). If we do, we see what is in fact a photographic record of the Events Filmed as a photographic record of the Story Told. And this, even though the events that make up the latter may

84  •  Robert Hopkins

well be entirely fictional. So we experience the film, in this respect at least, as if it were a documentary of those fictional events (cf. Perkins 1972: 121; Allen 1995: 90–97). (Of course, there are many differences in the conventions governing documentary and fiction film. I’m not claiming we are blind to those.) Since documentary is one of our touchstones for contact with the real, the result might plausibly be thought a kind of realism: Collapse: A film is realistic when we see it as the photographic record not of the Events Filmed but of the events that make up the Story Told. It’s instructive to compare this view with some of the claims of my fellow contributor, George Wilson. In a series of rich and important writings, Wilson has explored a position similar in spirit to Collapse. The point of closest contact in his present contribution lies here: Quasi-Transparency: If viewers “see” a motion picture image depicting a fictional item F and the image is photorealistic in appearance, then they are warranted in imagining themselves as thereby seeing the fiction F. (This volume, p. 67) That is, when viewers see Moving Images that look photographic, they imagine their seeing them to be their indirectly seeing the Story Told. They imagine the images to be transparent to the events in that story. Despite the similarity in spirit, there are several important differences between Wilson’s position and Collapse. First, while transparency lies at the heart of Quasi-Transparency, it plays no part in Collapse, for, though Collapse says we sometimes seem to be watching a moving photographic record of the Story Told, it does not say anything about seeing photographs being (or being imagined to be) a way of really seeing things. Collapse simply uses the notion of photography. Quasi-Transparency, in contrast, involves a theory of what photography involves, a theory it borrows from Transparency. Second, the two views rest on different central notions: for Collapse, seeing-in; for Quasi-Transparency, imagining seeing. To some extent, these offer two ways to conceive of the same phenomenon—that is, our experience of pictures when we understand them. We might describe the switch between the two readings of the ambiguous picture in terms of our seeing in it first, duck, then rabbit; or in terms of our alternating between imagining seeing those things. Both notions are intended to capture a familiar phenomenon, but both do so in terms that are themselves theoretical. (Consider in this respect Wilson’s attempts (this volume, pp. 62–3) to specify exactly what imagining seeing is supposed to be.) We might try to use one set of terms to analyze the other, and in particular to analyze seeing-in in terms of imagining seeing

Realism in Film  •  85

(Walton 1990: ch. 8). However, Wilson doesn’t do that, and there are reasons to be skeptical about the attempt (Kulvicki 2014: 84–86). So it seems best to treat the two claims as framed in independent terms. Provided seeing-in is not analyzed as imagining seeing, a third difference opens up. Many of the questions that Quasi-Transparency generates simply do not arise for Collapse. If we imagine that the Moving Images are transparent to the events in the Story Told, then what do we imagine about how those events came to be recorded photographically? This is a question that Wilson’s theoretical framework forces him to confront, and on which he expends considerable energy. In contrast, Collapse faces no such issue. Sometimes we see Moving Images before us, but in them we see only the Story Told, not also the Events Filmed. What more needs to be said? There are questions we could ask about how the images were made, but none that press simply in virtue of our seeing the film this way, and none that threatens to require us to start fleshing out the fictional world of the Story Told with some account of how we have access to it. Fourth, and finally, there are differences over the source of any realism hereabouts. While Quasi-Transparency doesn’t say this, Wilson does take it to describe a kind of realism in film. In explaining why, he says it counts as realism because of its intimate connections to transparency, which, he accepts, is itself a kind of realism (this volume, p. 67). Since Collapse makes no appeal to transparency, if the phenomenon it describes counts as realistic, it must do so on other grounds. But is collapse a kind of realism?9 For that matter, is transparency? (And if not, where does that leave its derivative, quasi-transparency?) To say, we must raise the deeper question of what realism in general might be.

What Is Realism? We have reviewed five claims about realism in film: Accuracy  & Precision, Recessive Form, Illusion, Transparency, and Collapse. The five are not in direct competition. It is true that some are mutually incompatible. For instance, if a film generates the illusion that the events in the Story Told are before us, it cannot simultaneously support collapse. (The latter requires, while the former precludes, our being visually aware of the Moving Images.) But the five do not offer competing accounts of a single phenomenon called “realism.” Rather, each aims to describe a particular form that realism takes. Realism in cinema might involve them all. Still, it is striking how diverse the five are. Some are a matter of degree, others not. Some can be found in representations of all kinds, others are restricted to representations of a certain type. Some are about content, others about form; some concern our experience, others do not; some concern how things seem in that experience, others whether it really relates us to objects in

86  •  Robert Hopkins

a particular way. Given this diversity, it’s natural to ask what realism must be, if these all count as varieties of it. What do they have in common? What is the genus of which these are the species? This question appears to have received little attention. There are certainly other reviews of the varieties of realism exhibited by film (Gaut 2010: ch. 2; Kania 2009; Thomson-Jones 2008: ch. 2) or by pictures in general (Kulvicki 2014: ch. 6). (These include helpful discussions of forms of realism I have not the space to consider and more nuanced treatments of some of those I have explored.) Some of those reviews attempt to explain how different forms of realism relate to one another. For instance, John Kulvicki and Berys Gaut distinguish between realisms of content (roughly, what is represented) and realisms of manner (roughly, how it is represented). But exploring how various kinds of realism stand to one another is quite different from asking what makes them count as forms of realism at all. Now, it may be that others have not asked the question because they doubt it can be answered. There are certainly issues about how ambitious any answer can reasonably hope to be. However, abandoning the attempt before even starting is surely defeatist. So let us see if we can make some progress. How are we to find the core of realism, to give an account of realism per se? In giving accounts of each of the five kinds of realism, we described features a representation must exhibit to count as realistic in the relevant way. Now we are asking a question at a higher level: What makes exhibiting those features (a kind of) realism? Thus while in effect we earlier sought to fill out the following schema: R exhibits realism of kind K iff . . . ; the task now is NOT to complete a similar schema for realism in general: R exhibits realism per se iff . . . To look for that would be to treat the genus as just another, particularly widespread, species. That would be confused. Representations exhibit realism in general in virtue of exhibiting some species of it, not in virtue of meeting some more general condition. Rather, the question now is what makes the species count as species of realism. I don’t think it reasonable to expect the answer to this question to take the form of a theory or definition, at least not one with even the relatively modest rigor of those theories and definitions offered at the lower level. Rather, we should look for an answer in some conception or image, some Ur-picture, that captures our deepest sense of what realism amounts to. The foregoing phenomena will be vindicated as realisms provided they can be related in some suitably perspicuous way to that conception or image. I propose that what lies at the heart of our thinking is a conception of the ideal case, the maximally realistic representation. Wherever there is representation, of any type, there must be content (what is represented) and there must be something like form (other features of the thing that bears that content). Moreover, representations do not spring from nothing, but must somehow

Realism in Film  •  87

be made. We can think of particular representations as products of the interaction between ways of making representations and elements in the world. When some such element—a thing, scene, fact, state of affairs, an existing representation, or even an experience or a thought—is fed into some such source of representations, a representation results. The products of this interaction are realistic to the extent that their nature aligns with that of the element in question. What does “aligned” mean here? Well, certainly not that the representation itself shares the nature of the represented element. We’ve said nothing about the type of representation involved, and while some types (e.g., pictures) do generally share some of the properties of their objects, others (e.g., words) do not. Rather, alignment is a matter partly of content and partly of form: content, in that no aspect of the element is misrepresented, and no aspect is missed (remember, we are describing an ideal); and form, in that the representation’s other features reduce to those necessary for it to have the content just described. The ideally realistic representation is, as it were, completely consumed in capturing whatever it represents. It helps to distinguish two strands in this conception. The first is the idea of perfect alignment between a representation R and what it represents O. This first strand is, broadly speaking, semantic. It is a matter of the content of R, how that content relates to the nature of O, and the other features of R necessary for it to have that content. But this strand cannot exhaust our conception of realistic representation, for to say that R and O are aligned in this way is not yet to assign priority to either. For all we’ve said, they might align because O has no nature independently of its being represented by R. It might be a mere shadow cast by R, an “object” only in the sense in which every representation has an object, be it real or otherwise. The semantic strand in our conception of realism is thus supplemented by a second, explanatory strand: the alignment occurs because R conforms to how O is, and not vice versa. (Indeed, the nature of ideally realistic representations admits of explanation of a special kind. The more realistic R is, the more we are able to explain its features by a particular kind of appeal to O’s features. Every aspect of its content can be explained as a result of O’s being the way it is represented as being.10 And every aspect of its form can be explained as necessary to that content, and thus also due to the way O is. The nature of O thus dictates, in a particularly straightforward way, the nature of R.) Since our conception of realism contains two strands, we have a decision to make. Is it enough for some phenomenon to count as a kind of realism that we can intelligibly relate it to just one strand, or must we relate it to both? I propose we take the liberal line: an intelligible connection to either strand will do. Adopting this line will make it easier to give sympathetic treatments of the claims of various phenomena to count as realisms. As we will see, the game will remain sufficiently difficult to be worth the candle.

88  •  Robert Hopkins

On this conception of realism, the primitive case is one in which there is something independent of the representation that provides its object. This might seem to reintroduce the Game of Thrones problem. Representations of the purely fictional can certainly be realistic. How can this be, if realism per se involves capturing how something is, the nature of which causes the representation to be a particular way? However, remember that our basic conception of realism is not supposed to define just another species of it, one that is particularly widespread, and that in stating that conception we are not laying down conditions that must somehow be met by anything that is to exhibit realism of any form. Rather, the conception makes intelligible how various phenomena might count as kinds of realism, by their relating to the elements in that conception in some perspicuous way. If a given phenomenon—say, precision—can be intelligibly related to the conception, then precision’s claims to be a form of realism will have been vindicated. Precision will be a form of realism wherever it occurs: in representations of the real, and in representations the “objects” of which are purely internal to them.

Five Kinds of Realism? Does this conception of realism allow us to legitimate the realist pretensions of the five phenomena described earlier? The easiest cases are Accuracy & Precision, and Recessive Form, so let’s begin with those. To assess their claims, we should appeal to the first, semantic, strand in the conception. The First Two Candidates Accuracy and precision are really very close to the relations in content the semantic strand involves. Perfect alignment requires R not to misrepresent any features of O, and not to miss any features either. The notions of accuracy and precision just are these ideas, or perhaps these ideas made slightly more precise. If the conception captures our basic idea of realism, accuracy and precision will thus surely count as species of it. Now, we might worry that the connection is too close: that we are not really vindicating their claims to be realisms by appeal to something distinct, but have merely written them into the conception from which their legitimacy is to flow. However, this worry is misplaced. Accuracy and precision are forms of realism, if anything is. The right account of realism per se must vindicate their claims to count. If it does so by, in effect, building them into the basic conception or image that guides our thinking in this area, so what? Why think this amounts to a failure of theory, rather than showing that we built our theory on the right foundations? Making the connection to Recessive Form is a little more complicated. According to the conception, the ideally realistic representation has only those

Realism in Film  •  89

formal features necessary for it to have its perfectly accurate and precise content. This is not quite the idea of recessive form. Earlier we divided formal features into those that determine content and those that do not. Recessive Form requires that neither be prominent to the person engaging with the representation R. The conception precludes the ideally realistic representation from having any formal features other than those that determine content. Since what doesn’t exist can’t be prominent, Recessive Form and the conception agree that the fewer content-neutral formal features we are aware of, the more realistic we will find the representation to be. But what of content-determining features? The conception allows R to have them, but if they are prominent R’s form will not be recessive. Does Recessive Form thus demand more than the conception requires? Here we might make either of two moves. We might take the moral to be that we should refine Recessive Form, replacing it with something more closely tailored to the semantic strand in our basic conception of realism.11 Or we might note that, if form is limited to what content requires, even that form is likely to be recessive. Attention is liable to be drawn to the content supported, form itself being treated as mere means to that end, and thus overlooked. So much for the easy cases. What of Illusion, Transparency, and Collapse? I will tackle these in two phases, first laying out some first thoughts on each, and then revisiting them to consider residual options or problems. The Last Three Candidates: Initial Thoughts Let’s begin with Collapse. There is a superficial obstacle to its fitting the conception. Collapse concerns our experience of R and its relation to O, whereas perfect alignment concerns how R and O are in fact related. However, it is surely not too great a stretch to think that a phenomenon counts as realism when it involves either things really being a certain way, or their seeming to be that way. Collapse will count provided we experience a relation between R and O of the kind the conception describes. And surely that condition is met. The effect of collapse is to remove a level of representation from the world as experienced. In general, where R’s representation of O is mediated by M, R must capture features of both O and M. It cannot capture only those of M’s features that O also shares, for then what makes it a representation of both O and M, rather than O alone? But if it must capture features of M that are not also O’s, to that extent it must align with O less perfectly. Some of R’s features will be there only to reflect M, and so will dilute its capturing of O. Removing the mediating level of representation thus removes obstacles to R’s aligning with O as well as it might. This is true of mediated representation in general. It will also be true of the particular case in which Events Filmed mediate the relation between Moving Images and the Story Told. In the world as experienced, the onset of collapse leaves the Moving Images aligned more closely with the Events Filmed. Collapse promotes alignment, and so counts as realism.

90  •  Robert Hopkins

Next, consider Illusion. We might hope to vindicate this too by appeal to the semantic strand in our conception of realism. Where representation is illusionistic, the nature of R, at least as experienced, is so closely aligned with that of O that we no longer see the difference between them: R looks to be O itself. What closer alignment could we desire? Unfortunately, things are not so straightforward. Indeed, the conception and illusion are in tension. Our conception treats realism as a modification of representation: the alignment at its heart is between the form and content of a representation and the represented item. And surely in this respect the conception gets things right. Representations of things can be realistic, the things themselves cannot. A picture or description of my cat can be realistic, but what would it mean for my cat to be? Illusion, in contrast, is antipathetic to representation. To the extent that experience is illusory, representation disappears from the world as we experience it. The alignment it offers is instead that of identity. If, for instance, there were fully illusory cinema, it would involve our no longer seeing the Moving Images or the Events Filmed. We would not be aware of anything before us as a representation: all we would seem to see is the events the film narrates (the Story Told). It takes two to align representationally, but here only one remains. Of course, illusion need not be complete. But even where it is partial, the point stands: the more illusory our experience of R, the less we are aware of what is before us as alignment between a representation and its object. Every extra bit of illusion removes a bit of representation. Illusion seems fundamentally at odds with a key element in our conception of realism. Are we missing a trick here? Illusion banishes representation from experience, not from the world. Even a fully illusory movie would be just as much a representation as any other film. Perhaps, then, we can solve the problem by appeal to a combination of how things are and how they seem. Realism requires representational alignment; in illusory cases, representation is really present, and alignment is apparently so. However, this won’t do. What realism requires is not just representation plus some alignment, but the sort of alignment in which a representation can stand to its object. For all that has been said, that is present neither in reality nor in experience. In reality there is only representation; in experience, to the extent that illusion occurs, there is only identity. Finally, let us consider Transparency. It fails to connect with the semantic strand in our conception of realism. Photographs can be transparent regardless of whether they are accurate or precise, and whether their obvious features are limited to their content (Walton 1984: 258–259). A photograph might be transparent to its object even if the image is very faint and blurry, and even if the picture has many visible properties that have nothing to do with what it depicts—creases, tears, stains, gloss or matt patches, and so on. Even if we concentrate on the picture’s content, the point holds. A photo taken through

Realism in Film  •  91

a powerfully distorting lens is still transparent to its object, even if the setup makes many such objects look more or less alike. Given this, it seems a picture could be transparent, and yet fail to align with its object at all closely. Transparency and representational alignment are independent of one another. What of the causal strand in our conception of realism? Does it offer Transparency a firmer hold? At the heart of transparency lies natural counterfactual dependence, and what is that, if not a special kind of causal relation between the photograph and the object on which it depends? When a representation is dependent in this way on its object, the former arises from the latter without essentially depending on anyone’s mental states.12 In the terms offered by our conception of realism, the object confronts a means of making representations in such a way that a representation emerges without anyone’s mental states directly determining what its content or formal features will be. This purely mechanical way of generating a representation may not fit the result in our conception of realism—as we just saw, it doesn’t secure alignment between representation and object. Nonetheless doesn’t it somehow constitute a particularly pure instance of the process that conception describes? Here is a promising place to look for a vindication of Transparency. However, it’s another question whether that promise is fulfilled. First, note that it is far from obvious quite how the specialness of natural counterfactual dependence counts toward realism. Of course, philosophers are used to associating the idea of independence from the mental with some notion of realism. However, the realism in question is metaphysical: roughly, the idea that some thing, property, fact, or state of affairs is as it is independently of our mental states. Our interest throughout has been in the rather different notion of representational realism: the realism exhibited by external representations, such as pictures, novels, poems, movies, sculptures, or plays. It is an interesting question why we use the term “realism” for these two very broad ideas.13 Whatever the answer, however, there can be no doubt that they are distinct. How, then, does the fact that transparency involves independence of the mental bolster its claim to count as a species of representational realism? Second, and more importantly, there is more to transparency than natural counterfactual dependence. It also involves the visuality condition. Walton is quite clear that either condition can be met without the other being, and that only when both hold does transparency occur. Any connection between our conception of realism and natural counterfactual dependence promises (at most) to show that the latter counts as realism. That realism would be present wherever natural counterfactual dependence is, regardless of whether the visuality condition is also met, and so regardless of whether transparency occurs. Thus, even if the connection between the conception and natural counterfactual dependence can be made, transparency’s claim to be realism remains unproven.

92  •  Robert Hopkins

The Last Three Candidates Reconsidered My treatment of illusion and transparency may seem too quick. In particular, some might wonder whether it relies on trying to do the impossible. The conception of realism offered earlier is quite general—it applies to representations of any type (pictures, words, theatrical presentations, etc.). Illusion and transparency, in contrast, are features only of representations of certain types—those that, like pictures, movies and sculptures, are in some key way visual (or perhaps auditory, or at least in some sensory mode or other). Can we reasonably hope to make sense of what is realistic about transparency or illusion, while limiting ourselves to ideas applicable to representation quite generally? As put, the complaint can’t be quite fair. Collapse too is a phenomenon limited to visual representations (indeed, it is probably limited only to film), yet we had no trouble using the conception to vindicate its claim to count as realism. So it is not impossible to use an entirely general conception of realism to make sense of realisms found only in representations of a certain type. Even so, the complaint raises an important question: Does our sense of what is possible here shift if we allow ourselves resources specific to particular types of representation? Let’s consider just one attempt to use type-specific resources to help Illusion and Transparency. Proponents of either view may, and indeed are likely to, accept something like the following: when we grasp what a picture (any picture) represents, we are in some way visually presented with its object. They may also accept that it is the job of all pictures to present us with things in this way. Now, as we saw in discussing seeing-in and imagining seeing in the second section of this chapter, in general the visual presentation of objects by pictures is very different from that involved in seeing things in the flesh. Still, what is not true in general might nonetheless be true in special cases. Illusionistic pictures visually present us with O in a way indistinguishable from seeing O face-to-face, and transparent pictures visually present us with O in a way that counts as really seeing it, albeit indirectly. The former matches seeing O in the flesh in terms of phenomenology; the latter matches it in terms of the relation in which we are placed to O (perceptual contact). But seeing O face-to-face is not just any old way of being visually presented with O, but the central, paradigm case. If pictures aim to visually present us with their objects, and if illusion and transparency share features with the paradigm case of such presentation, each fulfills in a special way the mission that defines pictorial representation generally. Perhaps this is what makes them realistic. They count as kinds of realism in pictures because they count, in different ways, as ideal fulfillments of picturing’s mission. (Cf. John Kulvicki’s discussion (2014: 124–126) of “kind realisms.”)

Realism in Film  •  93

This is an interesting line, but it faces a question. If illusion and transparency count as realisms because they are in some sense limiting cases of pictoriality, where does this leave realism in general? There are at least some forms of realism that can be found in representations of any type: accuracy and precision, and recessive form. There is at least one other kind of realism found only in a specific type of representation (film): collapse. We made sense of how collapse counts as realism without appeal to the ideas just sketched. And it is hard to see how appeal to them could begin to vindicate the claims to realism of accuracy and precision, or recessive form. So vindicating Illusion and Transparency by these means threatens to leave us unable to make sense of the realm of realism as a whole. If some phenomena count as realist for one kind of reason, and others do so on entirely different grounds, nothing unites that realm. It is left partitioned into permanently isolated and mutually incomprehensible pockets, a theoretical Balkans. The question for the present proposal is whether it can avoid this outcome and, if not, how far this should motivate us to explore other options—including abandoning the idea that illusion or transparency counts as realism after all. I close by returning to Collapse. Just as some might worry that my treatment of Illusion and Transparency was insufficiently sympathetic, others might wonder whether I’ve given Collapse too easy a ride. There are really two ideas in Collapse: that we cease to see the Events Filmed, and that we continue to see the Moving Images as photographic.14 It was the former I appealed to earlier: collapse involves loss of mediation and so closer alignment, and that’s why it counts as realism. But it is surely the latter that made it plausible to treat collapse as a form of realism in the first place. The thought was that we experience fiction films as documentaries—that is, as the moving photographic record, of events that are in fact merely fictional. Can we also vindicate the claims of the photographic element to count toward collapse’s realism? Since photography is a special way of making pictures, we might expect its connection to realism to go via the conception’s causal strand. However, that was a path we explored in discussing Transparency, without much result. I propose we take a different line. In offering transparency as what is special about photographs, Walton chose a feature that comes apart from accuracy and precision. But on other accounts of photography’s distinctive nature, its connection to those semantic features is much closer. Jonathan Cohen and Aaron Meskin (2004), for instance, argue that we are at least inclined to treat photographs as more accurate and precise than other pictures, even if we are not right to do so. And I have suggested that, while many photographs in fact fail to be accurate and precise, photographic systems are governed by the ideal of promoting those features (Hopkins 2012). Either of these views ties being photographic to the first kind of realism in our list of five. So one way to secure the link between realism and both aspects of collapse would be to appeal to

94  •  Robert Hopkins

views such as these. Collapse is doubly realistic: it not only removes an obstacle to alignment by banishing from experience a mediating representation, but also positively promotes it by importing into experience a way of making pictures that is closely tied to accuracy and precision.

Conclusion The moral of all this is that we should perhaps exercise more caution in speaking of realism in film. All five phenomena discussed earlier are interesting. At least most are found in film, and many are important to its aesthetic and emotional impact. Whether all count as forms of realism is, however, not so clear. I’ve suggested a way to try to make progress with that question, in terms of the conception of realism in the third section of this chapter. That does not clearly vindicate the realist pretensions of all five. However, as things stand, it is quite unclear what could do that. Film may indeed be realistic in special ways, or to a specially high degree. Whether it is realistic in every way that has been proposed is quite another matter. And this not because it fails to have the features claimed (about that, I have said almost nothing) but because some of the features it really has may not count toward its realism, after all.

Notes   1 The term “the Story Told” names the events the story relates, not the telling of them. That last is as much a matter of the Events Filmed and (given editing, etc.) the Moving Images.   2 There are, of course, questions about how to trade one off against the other in assessing how realistic (in this sense) any representation is. I’m going to ignore those. It is also not compulsory to see accuracy and precision as on a par: John Kulvicki (2014: 119ff.) explores views on which the two play distinct roles in a definition of (a single kind of) realism. My purposes license ignoring these (and other) important issues, since I am interested in the pair only so as to contrast them with the quite different realisms I go on to discuss.   3 For instance, we might try to make sense of accuracy by reference to kinds of things, rather than particular objects. There are no particulars that pure fictions represent, but there are kinds of things, with real instances, that those nonexistent particulars are represented as belonging to. Perhaps it is these real kinds to which the representation is accurate (Gaut 2010: 61; Hyman 2005: 40–41).   4 A broadly similar view is defended in Lopes (2006: ch. 1). For a rather different way to use form to define a kind of realism, see Wollheim (1987: 72–75).   5 Throughout I use capitalized names (e.g., “Recessive Form”) for claims about realism, and uncapitalized versions (e.g., “recessive form”) for the phenomenon each concerns.)   6 Of course, what is prominent to an audience will in key part depend on their expectations. Realism as recessive form thus has two features that realism as accuracy and precision lacks. It is relative to an audience—what is prominent for one group of viewers might not be so for another. And it is dependent on context, including conventions governing the relevant representations. To take a cinematic example, what is prominent for someone used to long-take work may be very different from what is prominent to someone brought up in a different filmic tradition.

Realism in Film  •  95   7 There may be analogs in other sense modalities—perhaps the sound track in cinema induces auditory illusions; but some representations, such as written descriptions, seem ill-suited to illusion of any kind.   8 A representation might be transparent with respect to some objects, but not others. (It might be a doctored photograph, with some items added by hand.) That gives a sense in which transparency comes in degrees. The point is that for a given represented object, transparency is all or nothing, while the preceding forms of realism are not.   9 In my former work on collapse, I make no mention of realism. 10 Note this covers precision as well as accuracy. If R fails to be precise, it represents O as merely F: that is, as bearing some determinable F, without there being some determinate f that R also ascribes to O, in virtue of which it is F. Unless O is in fact merely F (i.e., it is F, but not in virtue of exhibiting any determinate of F), then the way R represents O as being (i.e., merely F) cannot be explained by appeal to O’s being that way. So imprecision involves loss of realism—except in the cases (if there are any) where O is in fact merely F. 11 Alon Chasid (2007) has advocated an account of pictorial realism that is, in effect, just this. 12 Walton talks only of beliefs. But it helps Transparency’s case to broaden the notion of natural dependence to include other mental states too. 13 A side benefit of the conception of representational realism I am exploring is that it offers some kind of answer. These days it is a familiar idea that our mental states are themselves representations. Metaphysical realism about a subject matter can thus be reframed (still crudely) as the idea that the existence and nature of the relevant items are not determined by our representations of them, be those representations mental or external. Representational realism, on the current conception, requires that an external representation be determined, to the highest degree possible, by things the existence and nature of which are independent of that representation itself. (Note that representational realism does not entail metaphysical. A realistic representation needs something independent of itself to capture, but not something independent of all representations. Indeed, what is captured might be another representation.) In metaphysical realism, things are not determined by representations; in representational realism, representations are determined by things. The forms of determination in question will differ (constitution, in the metaphysical case, causation in the representational); the relations run in different directions; and one involves absence of determination, the other its presence; but, these differences aside, the two involve broadly the same relations and relata, and form mirror images, or complements, of one another. 14 In earlier work (Hopkins 2008; 2010) I used the term “collapse” so as to refer only to the first idea. Here I’ve used it to cover both.

References Allen, R. (1995) Projecting Illusion: Film Spectatorship and the Illusion of Reality, Cambridge: Cambridge University Press. Chasid, A. (2007) “Content-Free Pictorial Realism,” Philosophical Studies 135, 375–405. Cohen, J. and Meskin, A. (2004) “On the Epistemic Value of Photographs,” Journal of Aesthetics and Art Criticism 62, 197–210. Gaut, B. (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. Hopkins, R. (1998) Picture, Image and Experience, Cambridge: Cambridge University Press. ——— (2008) “What Do We See in Film?” Journal of Aesthetics and Art Criticism 66:2, 149–159. ——— (2010) “Moving Because Pictures?” Midwest Studies in Philosophy 34:1, Film and Emotion, 200–218. ——— (2012) “Factive Pictorial Experience: What Is Really Special about Photographs?” Nous 46:4, 709–731.

96  •  Robert Hopkins Hyman, J. (2005) “Realism and Relativism in the Theory of Art,” Proceedings of the Aristotelian Society 105, 25–53. Kania, A. (2009) “Realism,” in P. Livingston and C. Plantinga (eds.), Routledge Companion to Philosophy and Film, London: Routledge, 237–248. Kulvicki, J. V. (2014) Images, Abingdon, UK: Routledge. Lopes, D. M. (2005) Sight & Sensibility, Oxford: Oxford University Press. Perkins, V. F. (1972) Film as Film, New York: Penguin. Schier, F. (1986) Deeper into Pictures, Cambridge, UK: Cambridge University Press. Thomson-Jones, K. (2008) Aesthetics and Film, London: Bloomsbury. Walton, K. (1984) “Transparent Pictures: On the Nature of Photographic Realism,” Critical Inquiry 11:2, 246–277. ——— (1990) Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Wollheim, R. (1987) Painting as an Art, London: Thames and Hudson.

Suggestions for Further Reading Gaut, B. (2010) A Philosophy of Cinematic Art, Cambridge, UK: Cambridge University Press. (Ch. 2 offers extensive discussion of various forms of realism in film.) Kulvicki, J. V. (2014) Images, Abingdon, UK: Routledge. (A helpful discussion of many issues concerning pictures that lie in the background to the present question.) Williams, C. (1980) Realism and the Cinema: A Reader, London and Henley: Routledge & Kegan Paul. (A useful, if eccentrically presented, selection of writings on realism from film criticism and film theory.) Wilson, G. (2011) Seeing Fictions in Film: The Epistemology of Movies, Oxford: Oxford University Press. (A rich elaboration and defense of the idea of quasi-transparency.)

PART

III

How Do Films Work as Narrative Fictions?

This page intentionally left blank

CHAPTER 

5

Fictional Indeterminacy, Imagined Seeing, and Cinematic Narration ANGELA CURRAN

Chapter Overview This chapter focuses on the debate over two central claims regarding cinematic narration: the claim that there are implicit cinematic narrators and the thesis that when we watch movies, we imagine seeing the events and characters in the film fiction. I examine what a consideration of the indeterminate nature of fictional narration—that is, what is specified by the fiction about how we come to imagine the story events—can contribute to the debate on these issues. It is argued that consideration of fictional indeterminacy can be used to show that positing an implicit cinematic narrator is not only unnecessary but also incompatible with appreciating the film fiction. Meanwhile, the opposite result is reached regarding the claim about imagined seeing: Considerations of indeterminacy suggest that we can suppose, without absurdities, that audiences at the movies sometimes engage in imagined seeing.

Introduction Fiction films tell a story, but how more exactly do they do this? One widely accepted answer is that movies invite viewers to imagine that certain things have occurred, as they are reported and shown in the image and sound track of the movie (Carroll 1990; 1998; 2004; this volume; Currie 1995; Lamarque and Olsen 1994; Walton 1990). But what, more exactly, is the nature of viewers’ imaginative engagement with film? Here there is a great division of opinion. Some notable philosophers have argued that it is standard for a fiction film to 99

100  •  Angela Curran

prompt or invite viewers to imagine that they are being shown the events in the story by means of an agency that is part of the world of the film fiction. Some philosophers refer to such a fictional agent as the “implicit cinematic narrator.” In the first part of what follows, I focus on the indeterminacies with respect to fictional narration in movies, and what a consideration of this aspect of fictional indeterminacy can contribute to the debate over the implicit cinematic narrator. It has been noted that fiction, especially literature, features examples where nothing in the fiction specifies how it is that we come to imagine the story events (Walton 1990: 174–182; Wilson 1997: 309). I argue that the indeterminacy of fictional narration governs our imaginings in relation to our appreciation of movies as well. I will argue that unless it is explicitly given in the fiction, viewers are prompted to make-believe nothing about what in the fiction makes their imaginings about the story events possible. I will use this point to argue against several prominent lines of argument for the claim that every fiction film has an implicit cinematic narrator. In the second part of the chapter I turn to a discussion of the Imagined Seeing Thesis (IST), which is the claim that viewers of mainstream fiction film imagine that they are seeing sections of the story world from a series of definite visual perspectives. Whether audiences at the movies do engage in imagined seeing and whether an acceptable version of IST can be found are the subject of much debate. I try to show that that imagined seeing at the movies is neither an extraordinary occurrence nor the default mode of experiencing film fictions, but is one way in which viewers come to know about story events. I consider a central objection to the Imagined Seeing Thesis, that it should be rejected because it entails that the audience must engage in absurd imaginings (Currie 1995: 170). I suggest that the claim that we imaginatively fill in the implications of what is explicitly fictional does not apply to “meta-fictional” imaginings, including our imagined seeing of characters. So there is no absurdity in supposing that we can imagine seeing characters and events in fiction.

Kinds of Narrators in Movies There is no disagreement that sometimes it is true that we are to imagine that we come to know about the story events by means of a narrator who belongs to the fictional world that the film prompts the audience to imagine. The uncontroversial case involves what are known as explicit verbal narrators. One such verbal narrator is character narrators who tell the story through the use of voice-over narration. Explicit character narrators who are voice-over narrators are found, for example, in Sunset Boulevard (Billy Wilder, 1959) and in American Beauty (Sam Mendes, 1999), which feature character narrators who are dead and tell the story in voice-over narration from beyond the grave.

Fictional Indeterminacy  •  101

It is also agreed that films also have voice-over narrators who are not characters in the film, and offer an omniscient, third-person reporting of the events, as in The Assassination of Jesse James by the Coward Robert Ford (Andrew Dominik, 2007), Vicky Cristina Barcelona (Woody Allen, 2006), and Magnolia (Paul Thomas Anderson, 1999). The controversy over cinematic narrators arises with respect to the claim that every fiction film has an implicit narrator who is responsible for conveying the story to the audience. We can call this the “Ubiquity Thesis” (see Kania 2005: 47). The most direct way to pose the possibility of such a narrator is to consider that movies make use of both a moving-image and a sound track, and so have a “two-track” system of conveying a story to the audience (see Thomson-Jones 2009: 77–78). This involves a visual showing and verbal telling, giving rise, according to one line of argument, to the implicit cinematic narrator, an agent that is responsible for using editing, cinematography, and sound to convey the story to the audience (Chatman 1990: 211; see Levinson 1996: 251–256 for the argument that the cinematic narrator is responsible for the non-diegetic music). The cinematic narrator, so understood, is an agent who belongs to the story and conveys the events as if they were true. This means the cinematic narrator is someone distinct from the filmmaker, the actual individual who created the film. The cinematic narrator is an agent that is internal to the story world, while the filmmaker is an actual, not fictional, being who is responsible for creating the film.

“Access” Arguments for the Cinematic Narrator One basic form of this line of argument for the implicit cinematic narrator starts with what is known as either the “A Priori Argument” or the “Analytic Argument” (Gaut 2004: 235–236; Kania 2005: 47–48). This is the argument, which originated with Seymour Chatman, that narration entails a narrator. 1. Narration is the activity of telling or showing a story. 2. There can be no narration without a narrating agent. 3. The author of a narration is its narrator. 4. Therefore, there is a narrator for every narration, including film narration (Chatman, 1990: 115–116; Levinson 1996: 252). The essential problem with the A Priori or Analytic Argument is that it fails to establish that narration requires a fictional narrator, for even if we restrict the argument to fictional narratives, and we suppose that the claim: (a) there is a telling or narration of a fictional story, entails the claim: (b) there is someone who tells the story, it does not follow without some further argument that: (c) there is

102  •  Angela Curran

fictional narrator or teller of the story. For it could be the author who is the one who tells the story (Kania 2005: 48; Köppe and Stürhring 2011: 64; see also Gaut 2004: 235–237; Wilson 1997: 299–300). To illustrate, if a father makes up a story about a heroic cat that saves a small child to tell as a bedside story to his children, then the father is the person who tells the story and is the narrator. And in the case of a historical narrative, such as Thucydides’ History of the Peloponnesian War, the author, Thucydides, is the one who is telling the story of the war that was fought between the Peloponnesian League and the Delian League. Who is “the author” of a film? Movies are the creative product of a coordination of a number of agents, including the director, the screenwriter, the actors, the cinematographer, the set designers, and so on. How can we address this problem? It may be possible, when one individual, such as the director, exercises the greatest creative control over the film, to speak in this case of the filmmaker as the analog of the author (Livingston 1998). Or when it makes sense to think of multiple individuals exercising creative control over the final product we can talk about the film as a collaborative project of multiple filmmakers (Gaut 1998a; 2010: 98–152). To establish the need for a fictional narrator, some other argument is required, one that focuses on fictional narratives, in particular. The argument that has been brought in to secure this conclusion is known as the ontological gap argument (for a defense see Levinson 1996: 254–256). Ontological gap argument: 1. Reason demands an answer to the question of what makes possible our knowledge of the story events. 2. Only fictional beings can have access to events in the world of the fiction. 3. Therefore only a fictional narrator can convey to us the knowledge of the events in the world of the fiction. 4. We do have knowledge of the story events in film fiction. 5. Therefore, there is a fictional being, an implicit, cinematic narrator, who is responsible for conveying the knowledge of the events in the story. A central problem with the ontological gap argument is that it is undermined by its own assumptions (Kania 2005: 48–49). Premise 2) says that only fictional beings can have access to the world of the fiction, and so it follows that the audience cannot have access to the fictional narration carried out by the cinematic narrator. If the premises of the ontological gap argument are correct, then a fictional cinematic narrator cannot be the means by which the audience comes to know about the events in the story (for additional criticisms of the ontological gap argument, see Carroll, this volume, and Köppe and Stürhring 2011: 64–65; for a response to this line of criticism see Wilson 2011: 116–117; and Zipfel 2015: 52–61).

Fictional Indeterminacy  •  103

In addition to questioning the coherence of the ontological gap argument, we can probe the ground-floor assumption (premise 1) that gets the argument off the ground: It is reasonable to expect answers to questions about what it is that makes our access to the story world possible. Jerrold Levinson puts this point as follows: “Reason . . . demands an answer as to how it is that a world is being made visible to us, and that demand, it appears, is only satisfied by the assumption of an agency responsible for that” (Levinson 1996: 256). In response Andrew Kania poses a slippery slope objection. If we think it is legitimate to ask who is the fictional narrator who is showing us the story events, then there are other questions to ask, such as what sort of agent this implicit narrator is, how it happens to know so much about the story events, and how it is able to impart this information (Kania 2005: 49). Seymour Chatman, a defender of the fictional narrator in film, thinks that while we are to imagine being shown the events by a cinematic narrator, the question as to how the narrator came to have the information about the story events is a “non-question” (Chatman 1990: 130). In turn Kania wonders why there is a sudden “loss of interest” in questions about access (Kania 2005: 49).

Fictional Indeterminacy and Cinematic Narration In what follows, I want to press what I think is an implication of Kania’s argument: that there is no principled way for defenders of the fictional narrator to wonder how we come to know the story events, but then defer asking other embarrassing questions about fictional contact, such as how the narrator gets her knowledge, which have no answers within the terms of the story. Reflections on the basic mechanism of fictional narration suggest that absent an explicit narrator, the viewer or reader is mandated to imagine or make-believe nothing about how she comes to imagine the story events. So, according to this line of thinking, the audience watching The Wizard of Oz (1939, Victor Fleming) is prescribed to imagine that there is a girl, Dorothy Gale (Judy Garland), who, along with her dog, Toto, is swept away to the magical land of Oz and embarks on a quest to find the Wizard in the hopes that he can return her to her home in Kansas. But there is no explicit narrator in The Wizard of Oz, and so the audience is mandated to imagine nothing about how it is that they come to imagine the goings-on in the story. Instead, in the case of impersonal fictional narratives, the question of how the audience comes to imagine the story events does not arise. One example of how this works is philosophical thought experiments. In Judith Jarvis Thomson’s classic thought experiment regarding abortion, you are to imagine that you wake hooked up to a famous violinist, who needs the use of your healthy kidneys for nine months: If you unplug yourself, the violinist dies (Thomson 1971). The philosophical reader is asked to make-believe or

104  •  Angela Curran

imagine that certain things are true, but there is no need to posit an implicit fictional narration as the means by which we come to access the story events (Davies 2010: 392). Now the proponent of the ontological gap argument may reply that for our imaginings about what goes on in the story events to be warranted, we need to suppose that we come to know of them via a narrator who takes the events to be “real and reportable” (Köppe and Stürhring 2011: 76, n. 21; Levinson 1996: 256). The idea here may be that we are warranted in imagining that certain things are true in the world of the fiction only if we also imagine that there is someone on the inside of the fiction who is reporting the story events to us as if they are true. The filmmaker does not think the story events actually took place: Only someone who belongs to the fictional world can report the events as if they are really so. Hence, there is a need to posit an implicit fictional narrator. However, this line of reply is incompatible with how it is that appreciators engage with fiction, for unless it specified how it is that the story events get conveyed, the audience is prompted to make-believe nothing about what it is that makes possible their imaginings. We can call this the principle of the indeterminacy of fictional imaginings (it derives from Walton 1990: 174–182). This idea is most often illustrated with respect to literary narration. It is fictional in some novels that certain terrible events occur, but no one lives to tell about them. Yet we, the readers, are learning about it. How is this possible (Walton 1990: 238)? Other forms of fiction also share this same indeterminacy at the level of narration, or the way the story events are conveyed to us: For example, the audience at the opera does not ask why it is that the characters sing rather than state what they want to say (Walton 1990: 177). Fiction films are also governed by the principle of indeterminacy of fictional imaginings. The omniscient narrator who appears in voice-over in Vicky Cristina Barcelona knows about the goings-on of the characters, as well as their attitudes and personal traits. But viewers are not meant to imagine how it is that the narrator got such knowledge or how it is able to present the information to viewers. Sunset Boulevard (Billy Wilder, 1959) and American Beauty (Sam Mendes, 1999) feature explicit voice-over narrators who tell their stories from beyond the grave. The audience would get entangled up in imagining absurdities if they asked how a character who is dead can tell a story, so considering what makes this possible cannot be part of what the film prompts the audience to imagine. Movie musicals also feature aspects of narration the audience is not supposed to question as part of their imaginative appreciation of the story. No competent viewer of Singin’ in the Rain (Stanley Donen and Gene Kelly, 1952) would imagine why it is that Don Lockwood (Gene Kelly) expresses how he feels about his budding romance with Kathy Selden (Debbie Reynolds) by

Fictional Indeterminacy  •  105

breaking out in song and dance in the famous scene in which Kelly sings the title song! And viewers are not mandated to imagine why it is that the movie musical Oklahoma (1955) opens up with Curly McLain (Gordon MacRae) singing “O What a Beautiful Mornin’ ”: Rather than finding this way of beginning the story stilted or even absurd, the audience understands that they are not to imagine anything about why it is that Curly is singing rather than talking. If this question occurs to them at all, the answer is found outside their imaginings about the story world, in the conventions of motion picture musicals that mandate that characters sing as part of the way the story is told. These examples regarding fictional indeterminacy support a line of reply to the ontological gap argument. Unless it is explicit in the fiction, questions about how the audience comes to imagine the story events are not part of what the audience is mandated to imagine. Indeed, to pursue such imaginings would be to get taken up in questions that detract from the audience’s appreciation of the film fiction. Friends of the implicit cinematic narrator have a response. They argue that movies do not provide the sense of getting direct access to a story, as is the case with philosophical thought experiments. Rather, the use of camera work, editing, particular lenses, and other aspects of cinematography gives the impression that there is some intervening visual medium at work in conveying the events to the audience. This creates the sense of being presented with an entry point into the story events that is mediated through some narrating agency. These claims give rise to the following argument. The Argument from Filmic Mediation: 1. Film fiction narratives display some sort of mediation, through the use of camera work, editing, and so on. 2. There is no mediation without the intervention of a mediating agent, a narrator who is responsible for conveying the story events from a certain perspective to the audience. 3. Therefore, every film fiction has a fictional narrator. This argument suffers from the same problem that the ontological gap argument faced. Even if we accept that movie narration displays mediation, and there is no mediation without a mediating agent, it does not follow that the sense of mediation should be accounted for within the scope of the audience’s imagining of what makes their access to the story events possible. The experience of mediation can instead be accounted for by considering the filmmakers’ decisions to shoot a scene from a particular angle, with a certain lens, and so on. The choices of the filmmakers can also be used to account for those situations where it is said that the film contains another sort of “mediation”: an

106  •  Angela Curran

evaluative attitude (empathy, sympathy, and so on) toward the characters. So, for example, the final scene of Bresson’s moving film Au Hasard Balthazar (1966) is said to offer an “empathetic commentary” on its long-suffering titular character, the donkey Balthazar (Wilson 2011: 132–133). It might be said that an implicit cinematic narrator is the source of these attitudes toward the donkey. But positing the cinematic narrator in this scene is unnecessary. We can instead suppose that the filmmakers stage the scene in the way they do so as to invite us, the audience, to imagine that it is unjust that Balthazar lies down to die, having lived a life suffering inhumane treatment.

Conclusion to Part I The essential problem with the Filmic Mediation Argument is one it shares with the other arguments for the implicit cinematic narrator that we have considered. There is no need to posit an agency that is inside the fiction, reporting the events as true, to explain how it is that we come to imagine the story events as happening. Such a narrator does nothing to enhance our understanding of the story, and considerations of indeterminacy of access suggest that questions of how we come to imagine the story events are not, unless it is explicitly made fictional, part of what the audience is prompted to imagine when they watch fiction films.

The Imagined Seeing Thesis As noted, the claim that audiences “imagine seeing” the story events in the film fiction, the Imagined Seeing Thesis (IST), is a subject of great debate among philosophers of film. There are a number of questions that philosophers have raised concerning the IST. One basic question is: Just what is imagined seeing? Is it just a manner of speaking? (For this suggestion, see Carroll, this volume). Should the phenomenon we call “imagined seeing” be analyzed in terms of or reduced to other sorts of imagining—for example, perceptual imagining, where one imagines oneself perceiving a certain event or state of affairs; or “seeing-in,” where one imagines of one thing (a photographic image of Gary Cooper) that it is another (an image of Marshall Will Kane)? (Stecker 2013) Imagined seeing is often contrasted with impersonal imagining, or imagining that certain things are true in the fiction—for example, that Harry Potter is a student at Hogwarts, the school for wizards. In contrast, imagined seeing is a form of personal imagining, for I place myself into the content of what I imagine—for example, I imagine that I see Harry Potter at Hogwarts taking his wizardry final exam (for personal versus impersonal imaginings see Currie 1990: 181–185). Imagined seeing is thought to be a kind of experiential imagining, because when a visual representation induces imagined seeing in

Fictional Indeterminacy  •  107

the viewer, it is said to induce an experience one thinks of as seeing the object that is represented. There is no reason, I think, to suppose that imagined seeing is the default mode of how we imagine the story events. Some scenes in movies seem to be designed simply to prompt us to imagine what is true in the world of the story world—for example, a shot of the front of a house that establishes that Sherlock Holmes lives at 221B Baker Street, London. So I agree with Noël Carroll, among others, who argues that the default position for movie watching is that viewers engage in impersonal imaginings (see Carroll, this volume). However, a number of films present some good examples where it seems that certain scenes are designed to prompt imagined seeing in the audience. So I favor a “mixed view” that movies standardly prompt viewers to engage in impersonal imagining, but that on occasion, some movies, in certain scenes, invite viewers to engage in imagined seeing as part of their appreciation of the film (see also Gaut 1998b: 333–334; 2010: 217). For example, Alfred Hitchcock’s Vertigo (1958) contains multiple scenes that are constructed to induce imagined seeing in the audience. The story is about a police detective, John “Scottie” Ferguson, who has to quit his job because of intense vertigo or fear of heights. His disability leads (he thinks) to the death of the woman he loves, Madeline Elster (Kim Novak) because he is not able to ascend the stairs in time to prevent her apparent suicide. Throughout the film, there are repeated point-of-view shots that show us what it is that Scottie sees when he experiences vertigo. The first occurs in the opening sequence, in which Scottie is in pursuit of a criminal, and slips down the side of the roof. We are given a point-of-view shot of the sidewalk far below, and here the aim seems to be not simply to have us imagine that Scottie is experiencing vertigo: The effect is to have us imagine seeing what he sees, to get an understanding of the experience that he is having. Even better examples are found later in the film, where Hitchcock uses a technique that is known as the “dolly-zoom” effect: The camera is used in a “push-pull” action that involves zooming in and tracking backwards. The effect is to create disorientation in the viewer so that the viewer imagines experiencing what Scottie feels when he looks down the stairwell. This technique is used in multiple scenes throughout the film, especially when Scottie is attempting to ascend a staircase at the San Juan Bautista Mission to stop Madeline from jumping from the rooftop. Here the point seems to be to induce imagining the experience of vertigo that Scottie experiences. There are also other point-of-view scenes that prompt the audience to imagine seeing what Scottie sees. After Madeline’s apparent suicide, Scottie is distraught and keeps thinking that he sees Madeline, but every time he gets closer it turns out not to be her. He first thinks he sees Madeline coming out of his apartment, and we also are prompted to imagine that we are seeing Madeline,

108  •  Angela Curran

because of the resemblance between the woman Scottie sees and Madeline. But it turns out not to be her. This is repeated at the museum, and again Scottie (and the audience) is wrong. Then when he is at a flower shop, he sees a woman who looks like Madeline, only her name is Judy Barton (also played by Kim Novak). At this point he does not realize that Madeline is Judy and that he was tricked into believing Madeline committed suicide so that the husband of the real Madeline could kill her. Later in the film, when Scottie is kissing Judy, there is a 360-degree pan from Scottie’s point of view that moves from what he actually sees to his own fantasy. All these point-of-view scenes seem to be designed not just to engage the audience in impersonal imaginings: They invite the audience to imagine that they are experiencing what Scottie sees and experiences so as to prompt us to get in the imaginative grip of his situation. If the examples from Vertigo are not isolated or highly unusual ones (and the repeated use of dolly-zoom and related effects in other films suggests that they are not), then imagined seeing in films is not an extraordinary occurrence, nor is it the default or primary way in which the audience imagines what happens in the story. But it is one way in which they do. If imagined seeing does occur in watching films, the problem is to understand just what this entails and whether it is coherent. This is the project in George Wilson’s recent book, Seeing Fiction in Films: The Epistemology of Movies (2011), which advances the arguments in Wilson’s earlier work (see Wilson 1997; 2007; 2013). The problem with imagined seeing arises from the following claims: If the audience imagines that they are seeing the story events, then they imagine seeing them from a series of definite visual perspectives. If they imagine seeing from a visual perspective, then they also imagine that they see from a vantage point that is within the story world. This would be the account of the IST that George Wilson calls “face-to-face imagined seeing.”

Face-to-Face Imagined Seeing Thesis When the audience watches a fiction film, they are prompted to imagine that they are seeing the story events by standing face-to-face with them (Wilson 2011: 36). But it is implausible to think that Vertigo prompts the audience to imagine that they are there in the fictional world, on the stairwell of the bell tower of Mission San Juan Bautista, experiencing vertigo. Such an imagining would provide an unwelcome distraction, rather than enhance the audience’s appreciation of Scottie Ferguson’s situation (but see Thomson-Jones 2012a, who argues that some films prompt the audience to imagine that they are moving within the world of the film). How do we block the inference from imagining seeing the bottom of the stairwell to imagining being located in the fictional world, at the vantage point implied by the perspective of the shots?

Fictional Indeterminacy  •  109

One solution to the problem would be to employ the principle of the indeterminacy of fictional imaginings. Since, I have argued, fiction films are governed by the same indeterminacy that other forms of narrative fiction are, we could then appeal to indeterminacy of the fictional means of access to block the inference from imagined seeing from a perspective to imagined seeing from a vantage point within the story. This would suggest the following account of the Imagined Seeing Thesis.

Modest Imagined Seeing (MIS) The audience imagines seeing the events in the fiction, but it is left open how it is that comes about (Wilson 2011: 77–78). Wilson rejects face-to-face imagined seeing, because he thinks that when the audience imagines seeing the depicted events, they do not imagine they are located within the depicted space of the story (Wilson 2011: 38). But modest imagined seeing is silent on the means of access to the fictional world, and this means, Wilson argues, that it cannot account for what is imagined when the camera moves or there are multiple shots that incorporate different perspectives on the action. So Wilson instead favors the following version of imagined seeing.

Mediated Imagined Seeing (MeIS) The audience imagines that what they are seeing is an accurate recording of the events in the story world that has been photographically derived in some undetermined way (Wilson 2011: 89). According to mediated imagined seeing, the audience imagines that they are seeing the story events indirectly, through seeing a true recording of them (Wilson 2011: 89–91). This recording conveys a “fictional showing” of the events to the audience that is carried out by a “minimal narrating agency,” a fictional being that has no discernible character traits and so is distinct from the “robust” narrators that are said by some to be found in literary fiction (for the argument for literary narrators, see Chatman 1990: 109–123). Its only capacity is its ability to narrate and thus make perceptible access to the story events possible (Wilson 2011: 112). Wilson has a battery of arguments in favor of mediated imagined seeing (see Wilson 2011: 88–105). One of them is that it, like modest imagined seeing, blocks the inference from saying that the audience imagines seeing the story events from a definite visual perspective to the conclusion that they imagine seeing from a vantage point within the fiction, for the audience imagines seeing the incidents in the story indirectly, via a recording of them, and not by being present in the fictional world.

110  •  Angela Curran

There is another central argument in favor of mediated imagined seeing: It is able to explain the way that aspects of cinematic construction, such as color, grain, focus, camera angle, and editing, mediate the audience’s imagined seeing of the characters and action. Because modest imagined seeing is silent on the means of access, Wilson argues, it cannot account for these aspects of the audience’s experience, which are evident to any good appreciator of cinema. So, Wilson concludes that mediated imagined seeing is the preferable account of imagined seeing, and with it, we also have an argument to the conclusion that there is a work-internal “fictional showing” of the story events carried out by a “minimal narrating agency” (Wilson 2011: 112). However, I fail to see why modest imagined seeing is not compatible with an account that seeks to explain important characteristics of the movie-watching experience, such as the impression of editing, color, and so on. These aspects would not be accounted for with reference to what the audience imagines to be true in the fiction: Instead, they would be accounted for in terms of the properties of the movie, as an audiovisual presentation. Certainly someone, for example, who watches the scenes shot from the stairwell in Vertigo could appreciate the creative cinematography that goes into making these motion picture shots. But according to modest imagined seeing it would not be part of the audience’s warranted imaginings to suppose that the character or scene looks the way it does because some being in the fiction recorded it via a camera held at a certain angle, with a certain lens, and so on (see Livingston 2013: 143). To suppose otherwise is to go down the “slippery slope” that we encountered in discussing Kania’s objections to the ontological gap argument. If we suppose the audience is prompted to imagine that there is a work-internal “fictional showing” of the events via a recording of them, then there will be embarrassing questions that will need to be answered. These would include questions such as how the recording took place, how the fictional recorder/ narrator was able to record the incidents and go unnoticed, and a host of other concerns (see Carroll, this volume; 2004; Gaut 2004: 242–246; 2010: 207–209). Wilson is quite aware of these objections and says that we may imagine that we are watching segments of the story world via “naturally iconic shots,” shots that are causally dependent on the scene but need not be supposed to be produced by a camera. So imagining the presence of a camera or other recording device need not be part of what it is that the audience imagines (Wilson 1997: 113; 2011: 48). There is another, preferable way to avoid getting entangled in these questions: That is to think that something like the Modest Imagined Seeing Thesis is the better way to begin to understand imagined seeing at the movies. Here we suppose that unless it is explicitly fictional, the audience is to imagine nothing about how it is that they come to imagine seeing the characters and incidents in the fiction.

Fictional Indeterminacy  •  111

Concluding Thoughts A recurring line of argument in this chapter has been to make use of the principle of indeterminacy to show there is no need to posit an implicit cinematic narrator. This principle has also been appealed to in trying to make out how we can suppose, without encountering absurdities, that the audience, at times, imagines seeing the incidents in the story world. The main idea that is a viewer can imagine seeing a character or scene in the film fiction, without it following that the viewer imagines that she is seeing the character or scene from a vantage point within the story world. But the principle of the indeterminacy of fictional narration has also been appealed to by George Wilson to argue in favor of the mediated version of imagined seeing. In response to the objections against mediated imagined seeing raised by Berys Gaut and Noël Carroll, Wilson maintains that the viewer can imagine that he is seeing a true recording of the events in the story world, but need not imagine anything about how the recording was derived (Wilson 2011). Thus, if Wilson is right, we can imagine that there is a fictional showing of the events by a work-internal “minimal narrating agency,” but we need not imagine anything further about how the recording was derived. This is because, Wilson argues, there are a number of questions about fictional narration that are “silly questions” to expect an answer to, and a question about how the fictional showing can be carried out by the minimal narrating agency is one of these silly questions. I disagree with Wilson’s appeal to the indeterminate nature of fictional narration to defend the Mediated Imagined Seeing Thesis: So it is incumbent upon me to explain why I think his appeal to fictional indeterminacy is not warranted, but the reliance on indeterminacies to defend modest imagined seeing is legitimate. Berys Gaut and Noël Carroll argue that what has come to be known as the “realistic heuristic” governs our imaginings about fiction. This is the idea that when we engage with a work of fiction we “fill in” and draw implications from what is explicitly true in the fiction based on how things are in the real world, unless it is explicitly stipulated to be otherwise (Carroll, this volume; 2004; Gaut 2010: 212–214). It is explicit in the Harry Potter novels and movies, for example, that the fictional world has magic, so we imagine this is so. But in other respects we imagine that things operate in the story world of Harry Potter as they do in real life. The realistic heuristic has an ancient pedigree in Aristotle’s Poetics, though he discussed the issue in terms of what the poet should prompt the audience to imagine happens in the story, and not what the audience is supposed to “fill in” as they engage with the story (see Poetics 9 and 24). The basic idea, however, in both approaches is the same: For the audience to engage with a story, things

112  •  Angela Curran

should happen in the fiction as they do in real life, or it undermines the sense that what the audience is watching is believable and true to life. I take the realistic heuristic to apply to what is explicitly fictional—that is, to what is supposed to be true in the fictional world. This means that we fill in the implications of what is explicitly fictional as part of our appreciating the work (see Gaut 2010: 213). Opponents of the mediated version of imagined seeing use the realistic heuristic to argue against this account of imagined seeing; if we are to imagine that there is such a minimal narrating agency at work in the fiction, we then will need to imagine the implications of this—for example, that the narrator is there recording the events, hanging off the side of the Empire State Building as King Kong does, and so on. While I think the realistic heuristic is, in general, a sound one, it seems that it should apply only to what is true in the story, and not to the audience’s “metafictional” imaginings. When the audience imagines seeing a character or incident from the story, it need not be true in the fiction that the audience sees these things: Their imagined seeing would belong to the category of extra-fictional imaginings that the audience engages in as part of their appreciation of the work. Movies, more generally, invite their audiences to engage in a number of meta-fictional imaginings, and some of these seem to be quite unrealistic. For example, characters often speak directly to the audience, as in Ferris Bueller’s Day Off (1986) or Woody Allen’s Annie Hall (1977). It is not true in the fiction that we, the audience, are listening to Ferris Bueller (Matthew Broderick) or Alvy Singer (Woody Allen), but it seems that to appreciate these two films the audience must imagine that the characters are talking with them, and if we are to imagine the implications of how this could be, we would get tangled up in absurdities. These considerations show that the realistic heuristic should not be taken to govern the audience’s “meta-fictional” imaginings, including the audience’s imagined seeing of story events (for a related point see Levinson 1993: 71). A further question is whether the realistic heuristic governs aspects of the narration, or the way in which the story is told and the specific way in which it is told. If readers of Huckleberry Finn are to imagine that they are reading the words of a rather illiterate young man, Huck Finn, are they then mandated to imagine all the implications of this—for example, that such a person could compose such a finely crafted narrative? It seems unlikely, for if they do so imagine this will clearly interfere with their appreciation of the work. It seems, then, that the realistic heuristic, the filling in of all the implications of what is explicitly fictional in terms of how things work in the real world, does not necessarily apply in all cases to questions of the means and manner of how we come to imagine the story events (Thomson-Jones 2012a: 37) (but one aspect of narration that the realistic heuristic does apply to is the plot, which is the temporal ordering of the story events into a coherent causal structure: This

Fictional Indeterminacy  •  113

must be believable in order to prompt the right response in the audience). What are the implications for the principle’s use in the argument against the mediated version of imagined seeing? I think this is a topic that invites further thought. On the one hand, it could be said that because the minimal narrating agency is a feature of the fictional narration, or how we come to access the story events, we should not suppose that questions about it (e.g., how it gains knowledge of the story events) have determinate answers. However, once it is introduced as an explicit part of the fiction, then it seems reasonable for us to fill in the implications of its presence in the fictional world. And when we do, we go down the “slippery slope” of embarrassing questions that we have rehearsed in the chapter, calling into question why we should want to introduce such a narrator in the first place. So I am inclined to think there is no conflict between the use of the realistic heuristic to argue against mediated imagined seeing and the argument from fictional indeterminacy that I have relied on in this chapter. We have looked at how considerations regarding the indeterminacy of fictional access can illuminate two central claims in the debate about cinematic narration: the claim that there are implicit cinematic narrators and the thesis that when we watch movies, we imagine seeing the events and characters in the film fiction. Ultimately, the principle of indeterminacy of access is based on considerations of what it is that appreciators of fiction must imagine when they engage with fiction. The argument that there are implicit cinematic narrators in every fiction film fails because it asks us to imagine things that are unnecessary and that interfere with our appreciation and interpretation of a work. On the other hand, the claim that we imagine seeing at the movies holds because there is reason to think audiences at times engage in imagined seeing as part of their appreciation of the film fiction.1

Note   1 I thank Carol Donelan for very helpful discussion on the examples of films used in this chapter, and I thank Kate Thomson-Jones very much for inviting me to contribute to this volume and for her very helpful comments on drafts of this chapter.

References Carroll, N. (1990) Philosophy of Horror, or Paradoxes of the Heart, London: Routledge. ——— (1998) “Fiction, Nonfiction and the Film of Presumptive Assertion,” in R. Allen and M. Smith (eds.), Film Theory and Philosophy, Oxford: Oxford University Press, 173–202. ——— (2004) “Introduction” to “Part IV: Film Narrative/Narration,” in N. Carroll and J. Choi (eds.), Philosophy of Film and Motion Pictures: An Anthology, Malden, MA, and Oxford: Blackwell, 175–184. Carroll N. and J. Choi (2004) Philosophy of Film and Motion Pictures: An Anthology, Malden, MA, and Oxford: Blackwell.

114  •  Angela Curran Chatman, S. (1990) Coming to Terms: The Rhetoric of Narrative in Fiction and Film, Ithaca, NY: Cornell University Press. Currie, G. (1995) Image and Mind: Film, Philosophy, and Cognitive Science, Cambridge: Cambridge University Press. Davies, D. (2010) “Eluding Wilson’s ‘Elusive Narrators,’ ” Philosophical Studies 147, 387–394. Gaut, B. (1998a) “Film Authorship and Collaboration,” in R. Allen and M. Smith (eds.), Film Theory and Philosophy, Oxford: Oxford University Press, 149–172. ——— (1998b) “Imagination, Interpretation and Film,” Philosophical Studies 89, 331–341. ——— (2004) “The Philosophy of the Movies: Cinematic Narration,” in P. Kivy (ed.), The Blackwell Guide to Aesthetics, 1st ed., Hoboken, NJ: Wiley-Blackwell, 230–253. ——— (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. Kania, A. (2005) “Against the Ubiquity of Fictional Narrators,” Journal of Aesthetics and Art Criticism, 63, 47–54. Köppe, T. and J. Stürhring (2011) “Against Pan-Narrator Theories,” Journal of Literary Aesthetics 40, 59–80. Lamarque, P. and S. H. Olsen (1994) Truth, Fiction, and Literature: A Philosophical Perspective, Oxford: Oxford University Press. Levinson, J. (1993) “Seeing, Imaginarily, at the Movies,” The Philosophical Quarterly 43, 70–78. ——— (1996) “Film Music and Narrative Agency,” in D. Bordwell and N. Carroll (eds.), PostTheory: Reconstructing Film Studies, Madison: University of Wisconsin Press, 248–282. Livingston, P. (1998) “Cinematic Authorship,” in R. Allen and M. Smith (eds.), Film Theory and Philosophy, Oxford: Oxford University Press, 132–148. ——— (2013) “The Imagined Seeing Thesis,” Projections 7, 139–146. Stecker, R. (2013) “Film Narration, Imaginative Seeing and Seeing-In,” Projections 7, 147–154. Thomson, J. J. (1971) “A Defense of Abortion,” Philosophy and Public Affairs 1, 47–66. Thomson-Jones, K. (2009) “Cinematic Narrators,” Philosophy Compass 4:2, 296–311. ——— (2012a). “Narration in Motion,” British Journal of Aesthetics 52, 33–43. ——— (2012b) “Review of George M. Wilson, Seeing Fictions in Film: The Epistemology of Movies,” Journal of Aesthetics and Art Criticism 70, 393–395. Walton, K. (1990) Mimesis and Make-Believe: On the Foundations of the Representational Arts, Cambridge, MA: Harvard University Press. Wilson, G. (1997) “Le Grand Imagier Steps Out: The Primitive Basis of Film Narration,” Philosophical Topics 25, 295–318. ——— (2007) “Elusive Narrators in Literature and Film,” Philosophical Studies 135, 73–88. ——— (2011) Seeing Fictions in Film: The Epistemology of Movies, Oxford: Oxford University Press. ——— (2013) “Seeing through the Imagination in the Cinema,” Projections 7, 155–171. Zipfel, F. (2015) “Narratorless Narration? Some Reflections on the Arguments for and against the Ubiquity of Narrators in Fictional Narration,” in D. Birkee and T. Köppe (eds.), Author and Narrator: Transdisciplinary Contributions to a Narratological Debate, Berlin: De Gruyter, 45–80.

Suggestions for Further Reading Gendler, T. (2013) “Imagination,” The Stanford Encyclopedia of Philosophy (Fall 2013 Edition), E. N. Zalta (ed.), http://plato.stanford.edu/archives/fall2013/entries/imagination/. This helpful essay discusses philosophical controversies about imagination, including the topic of visualizing the unseen (section 5.6) that is relevant for imagined seeing at the movies. Stock, K. (2013) “Imagining and Fiction: Some Issues,” Philosophy Compass 8:10, 887–896. This article provides an excellent discussion of some key topics relating to imagination and fiction (see especially section 3).

CHAPTER

6

Motion Picture Narration

NOËL CARROLL

Chapter Overview There is a very influential view in the philosophy of motion pictures that conjectures that all fictional movies have implicit fictional narrators. George Wilson has proposed the most sophisticated version of this position, based on the notion of mediated fictional narration. Since the view encompasses all fictional motion pictures, it can be said to claim that implicit fictional narrators are ubiquitous across the relevant cinema. Wilson maintains that this follows from the idea that viewers imagine seeing mediated images of fictional characters. Other proponents of the ubiquity of implicit fictional narrators have invoked what Andrew Kania has called “the ontological gap” argument. In this chapter I attempt to deny the ubiquity of implicit, mediated fictional narrators in fictional cinema by challenging both the claim that imagining seeing is ubiquitous and the ontological gap argument. I deny that we typically imagine seeing fictional characters in favor of the hypothesis that we literally see images of actors on-screen that we use to imagine the fiction. I deny the ontological gap argument on the grounds that it is insufficiently motivated. Typically, motion picture fictions tell a story. That is, to be more precise, on the one hand, there is a story and, on the other hand, there is the telling of the story. Both the story and the telling occur in time, but they need not follow the same temporal ordering. With regard to the story, event A follows upon event B follows upon event C; the sequence of events has a linear structure; time’s arrow flies forward. But the temporal order of the telling need not mirror the linear temporal ordering of the events of the story—of the fictional 115

116  •  Noël Carroll

story world. In the telling, we may begin in medias res with event B, then flash back to event C and, lastly, flash forward to event A. This distinction between the story and its telling is often marked in terms of the distinction between the narrative and the narration. That is, there is the chain of fictional events in the order in which they putatively occurred (the narrative). And there is the recounting of those fictional events, which may or may not reconstruct the temporal order of the narrative (the narration). One of the major debates in the philosophy of cinema concerns the nature of motion picture narration with respect to fictions. Presumably, where there is narration, there is a narrator; a telling, that is, assumes a teller. Indeed, some theorists, like Seymour Chatman, appear to regard this inference as a strict entailment (Chatman 1978, 1990). But when it comes to motion picture narration, who exactly is the narrator? The straightforward, unforced answer would appear to be the motion picture maker. Charlie Chaplin is the narrator of The Gold Rush. However, some philosophers, such as Jerrold Levinson, have doubted the feasibility of this proposal (Levinson 1996). One consideration against regarding the motion picture maker as the narrator of the relevant film has been called “the ontological gap argument” (Kania 2005). The other argument we can label the seeing/showing argument. It has been defended most recently by George Wilson (Wilson 2011).

Two Arguments in Favor of the Ubiquity of Implicit Fiction Narrators1 The ontological gap argument According to the ontological gap argument, a fiction is a mandate to imagine. More specifically, an author mandates or proscribes an audience to imagine that thus and so. In consequence, the audience imagines that it is true in the fictional world that thus and so—for example, that Frankenstein created a living being from dead flesh and bones. In other words, it is true, when asserted in the fiction, that Frankenstein created a living being from dead flesh and bones. But who could possibly be making this assertion? Well, it might be made by one of the villagers who is being terrorized by Frankenstein’s monster. Or it might be made by an omniscient fictional narrator embedded in the story world, like Mary Shelley (as played by Elisa Lanchester) in Bride of Frankenstein. Or it might even be uttered by an omniscient, disembodied fictional narrator on the voice-over commentary or by means of inter-titles. All these are fictional denizens of the story world. They are so situated ontologically that they are able to make assertions about what is happening in the fictional world, unlike the actual author, since the actual motion picture maker is outside of the fiction. Indeed, he/she created the fiction so he/she cannot

Motion Picture Narration  •  117

assert that Frankenstein created a living being from dead flesh and bones, since assertion requires belief and the motion picture maker does not believe it for the simple reason that he/she knows he/she made the whole thing up. But recall: Narration entails a narrator, putatively someone who can assert that it is true in the fictional world that thus and so. The actual motion picture maker cannot do this for the reason just given. Allegedly, only someone or something positioned inside the fictional world is ontologically placed to do this. But if there are no explicit fictional agencies of the sort just canvassed earlier available to play the role of narrator, who is telling the story of Frankenstein as true? Who is asserting that Frankenstein created a living being from dead flesh and bones? The ontological gap argument posits that it is an implicit fictional narrator. That is, it supposedly must be inferred that there is an implicit teller of the tale who is asserting thus and so from, so to speak, inside the story world. Furthermore, with respect to motion pictures, we are primarily “told” what is thus and so by being shown it. We are shown by way of a visual presentation what is thus and so in the environs of Frankenstein’s laboratory. The motion picture narrator, thus, is a presenter. But the actual motion picture maker cannot be such a presenter, since he/she can show us only, for example, the sets on the lot of Universal Pictures; he/she cannot present us with views of Castle Frankenstein. There is no such place. So, we must presuppose that there is an implicit fictional presenter operating from inside the fictional world and offering us veridical views of what is thus and so in that world. Indeed, it should be our default assumption that all movies have such implicit, albeit disembodied, fictional presenters in order to guarantee that each relevant view of the fictional world is being presented as veridical so that we, the audience, can imagine it as thus. Consequently, according to the ontological gap argument, implicit fictional presenters are ubiquitous across the gamut of fictional movies. They are the agencies that show us how things truly stand in the world of the fiction. Supposedly, just as narratives presuppose narrators, views of events presented as veridical require implicit presenters who, in the case of fictions, must be fictional, since they must present from within the fiction operator (the “it is fictional that . . .” operator), a place where actual filmmakers cannot stand. With respect to the first version of King Kong, the moviemakers showed us models of the eponymous Kong and various dinosaurs, often in combat. But they could not present us with the battle between Kong and the Tyrannosaurus rex because there were no such things in the actual world. So who presented the battle as it unfolded? An implicit fictional presenter. The actual moviemakers are responsible for constructing the motion picture, an object in the actual world. They make it fictionally the case that “Kong loves Anne Darrow” by mandating that audiences imagine that this state of

118  •  Noël Carroll

affairs obtains in the story world. But they do not believe that “Kong loves Anne Darrow” is true simpliciter, since they confected the two of them. Yet, it is argued that if viewers suppose that this is true in the fiction, then someone in the fiction must be asserting that and/or presenting it as veridical, which proponents of this argument appear to presume entails that someone or something in the fiction must believe it. It cannot be the actual moviemakers. Apparently, there is an ontological gap between them and the fictional world that they cannot cross. So it must be some implicit fictional narrator/asserter/ presenter. The seeing/showing argument A second line of argument on behalf of the ubiquity of fictional narrators rests on the premise that seeing imaginarily or the Imagined Seeing Thesis entails the Fictional Showing Hypothesis, where the Fictional Showing Hypothesis amounts to the claim that there is an implicit fictional agency in all motion pictures, albeit a very minimal one. It is in operation in the presentation of every motion picture fiction. Though minimal, this narrating agency is ubiquitous. We can call this argument “the seeing/showing argument.” This argument may be linked to the ontological gap argument, or it may be advanced independently. In order to follow the seeing/showing argument, we must first grasp the notion of seeing imaginarily or imagining seeing—aka the Imagined Seeing Thesis (see Curran, this volume). The idea of seeing imaginarily begins with the commonplace observation that viewers often speak as if they see fictional characters. For example, it appears natural to viewers to say that they see Atticus Finch cross-examine a witness or that they saw Captain Ahab harpoon Moby Dick. In neither case do they feel compelled to say they saw Gregory Peck doing so. The proponent of the seeing/showing argument urges us to take this talk seriously. Thus, his/her task becomes to attempt to make sense of it, for, on the face of it, it does not seem plausible. It seems as unlikely that one could see a fiction as does seeing a dream walking. So, at this point, the advocate of the seeing/showing argument introduces the notion of imagining seeing or seeing imaginarily. When one says that “I saw Atticus Finch defend Tom Robinson,” since one cannot literally be seeing that event (since it never happened), we should interpret it to mean “I imagine that I am seeing Atticus Finch defending Tom Robinson” or “I am seeing Atticus Finch defending Tom Robinson imaginarily.” To unpack this in the language made popular by Kendall Walton, a fiction, including a movie fiction, is a prop in a game of make-believe (Walton 1990). When children play a game of pirates, they make-believe that their sticks are

Motion Picture Narration  •  119

cutlasses. They make-believe that they are dueling to the death, using their sticks as props. Similarly, when watching a fictional motion picture, we makebelieve that we are seeing such things as Atticus Finch defending Tom Robinson. Here “making believe that I am seeing,” “imagining seeing,” and “seeing imaginarily” all amount to the same thing. Perhaps needless to say, the suggestion that we are seeing imaginarily events unfold before us as they might in the ordinary course of terrestrial events leads to certain anomalies. What are we to make of the various cinematic devices— such as cuts, fades, dissolves, filters, lens distortions, black-and-white cinematography, and so on—that are typically in evidence in motion pictures? When the images dissolve in the shots of Xanadu in Citizen Kane are we to imagine seeing the world deliquescing? Or do we imagine of the black-and-white films The Artist and The Man Who Never Was that we are seeing black-and-white worlds? In order to avoid anomalies such as these, the friend of seeing imaginarily maintains that we are not imagining seeing the world of the movie directly as we might look upon a monkey in a cage at the Bronx zoo. That is, we are not imagining gazing upon the world of the motion picture unmediated. Rather we are imagining seeing that world through certain processes of mediation. Speaking heuristically, we are imagining seeing something like a documentary of, say, the plight of King Kong, marked by such features as black-and-white cinematography, dissolves, superimpositions, flicker effects, and so forth. Here it is important to stress that this “something” is only “something like” a documentary. The analogy is meant to avoid the various metaphysical oddities cited earlier. But this “something” is not literally a documentary. If the final scenes of Greed were shot by a documentary crew, why, we would be forced to ask ourselves, did they stand by so callously as Marcus and McTeague perished? So instead of watching a standard-issue documentary, we imagine seeing something else. But what exactly? First of all, we are imagining a shower or presenter, for if we imagine seeing something, we putatively must imagine that we are being shown it by something. Moreover, this showing is mediated; we do not imagine seeing it directly; we see it through a medium. Furthermore, this showing is not being executed by a documentary film crew. That would cause not only moral conundrums, as in the case of Greed, but also logical ones. For example, some motion pictures involve scenes and events of which it is said, within the world of the fiction, that no one witnessed them. But if we are imagining seeing a documentary film, then didn’t the documentary film crew see whatever we are imagining seeing? So in order to avoid such logical contradictions, the defender of the seeing/showing argument maintains that viewers are imagining seeing

120  •  Noël Carroll

something shown to them by some absolutely minimal agency like a natural surveillance camera or a mirror, albeit one that moves somehow, getting about from here to there in the world of the fiction. Presumably this must be posited because there must be a presenting agency to account for our seeing imaginarily, for instance, Atticus Finch defending Tom Robinson. That is, seeing imaginarily supposedly requires a fictional presenter or shower. Thus, the Imagined Seeing Thesis is said to entail the Fictional Showing Hypothesis, though the fictional presenters here are quite minimal. How minimal? At this point the defender of the ubiquity of the minimalfictional-presenter thesis argues that this is a silly question on a par with the notorious “How many children does Lady MacBeth have?” Likewise, it would be silly to ask after the physics of the crystals that power the Starship Enterprise. So similarly it is silly to ask about the nature of the implicit fictional narrators in cinema. In order to motivate this claim, proponents of the minimal version of the ubiquitous implicit narrator/presenter remind us of the video device in the Flash Gordon serials of yore. This mechanism enabled Flash, Dale Arden, and Dr. Zarkov to see anything anywhere in the universe at the flick of a wrist. But how was that possible? Where were all the recording devices that would be required to carry off such a feat? Wouldn’t they have to be virtually everywhere? Or maybe the video worked by some other means. But if no recording machines were necessary, how was it accomplished? How exactly? Of course, these questions are unanswerable. But that is beside the point, for it is given in the fiction that the video works. Enough said. To request more information about the nature of the device is as silly as asking how King Kong managed to have a son with no Queen Kong in evidence. Fictioneers get to introduce whatever improbabilities/impossibilities they need to make the story go. Ours is not to reason why or wherefore. Proponents of the minimal implicit fictional narrator do not claim that all motion pictures are equipped with Flash Gordon’s video scanner. That is merely an analogy meant to serve as an intuition pump, securing the plausibility of an implicit, naturally occurring, nonhuman, fictional agency that presents us with veridical views from the story world, which we then go on to see imaginarily. Moreover, since the implicit narrating device is not human, it eludes the counterexample of someone, like a documentary crew member, seeing something that is unseen according to what is given in the fiction. Perhaps needless to say, few viewers, save maybe a handful of metaphysicians, are aware of the supposed presence of these minimal, implicit fictional presenters. Most viewers would undoubtedly deny that there are such things. But the friends of the idea urge that we ignore the protestations of plain viewers on this matter in order to make sense of their claim that they see Atticus

Motion Picture Narration  •  121

Finch defending Tom Robinson. To make sense of that putative conviction nothing less than the postulation of a minimal, implicit, fictional narrator/ presenter is allegedly necessary.

Resisting Ubiquitous Implicit Fictional Narrators/Presenters Against the ontological gap argument According to Andrew Kania, a critic of the ontological gap argument, proponents of this view hold that the narrators/presenters of moving picture fictions must be on the same ontological level as the fictions they present (Kania 2005). This is an obscure idea, but if it implies that there can be no traffic between ontological “levels,” then it would be self-defeating because it would render problematic the implicit narrator’s contact with us, the living, the audience. Also it does not seem true that actual filmmakers cannot narrate/present their fictions. Sometimes, in a reflexive gesture, motion picture makers have included images of themselves shooting the very moving picture that we are seeing. For example, toward the end of Bergman’s Persona, we see the film crew in action. How are we intended to understand this? That it is fictional that Bergman made Persona? Surely not. Should we treat this as a documentary? If so, it is one of the shortest on record. Rather he was acknowledging that he and his crew were responsible for the images that we have been watching, where this sort of reflexive self-acknowledgment is virtually conventional. If the ontological gap argument presupposes that this was impossible, then so much the worse for the argument, since what Bergman did was actual. Perhaps it will be urged that the image of Bergman itself must be attributed to an implicit fictional presenter, since the film crew was not shown filming itself. So let us vary the example somewhat. Imagine that we are watching a movie by a director like Hitchcock, who enjoys appearing in his own films. So he shoots scenes from time to time with mirrors in the background so that we see them being shot at the same time we see their results. Here there is no need to posit an implicit presenter, since the filmmaker is filming himself in the process of presenting the scene to us. Must we say that the presentation of the images we are seeing is fictional? Suppose the filmmaker or one of his crew has a heart attack and dies while filming and we see that? Is it fictional that he is dead? Tell that to the coroner. Moreover, once again it makes little sense to interpret the film as a documentary, not only because it has not been presented as such but also because, as just noted, such self-acknowledgment is conventional in contemporary cinema. In any event, it seems at least controversial to maintain that actual people cannot function as the presenters of the sights and sounds of fictional worlds.

122  •  Noël Carroll

Didn’t Bergman in effect claim to do precisely that at the end of Persona? Nor did viewers find this in any way absurd or contradictory. Proponents of the ontological gap argument may charge that is absurd; that it must be the case that the best interpretation of the shot is that it is fictional that Bergman is the narrator/presenter of Persona. But that is the takeaway of few, if any, viewers. Perhaps the deepest motivation for the ontological gap argument concerns its recommendation for how viewers are to understand that thus and so is true in the fiction—for example, that it is true in the fictional movie The Thing that a large blood-thirsty carrot is on the rampage. The proponent of the ontological gap argument maintains that the only way to understand this is in terms of there being someone or something within the scope of the fiction operator asserting it. However, this is highly debatable. Other accounts of “what is true in the fiction” are available. For example, it may be argued that what is true in a fiction is whatever the maker or makers of the fiction mandate the audience to imagine. So if James Whale, among others, mandates the audience to imagine that Frankenstein created a living being from dead flesh and bones, then it is true in the fiction that Frankenstein created a living being from dead flesh and bones. On this account, there is no reason to posit a fictional asserter and therefore no pressure to invoke an implicit one. But the defender of the ontological gap argument is not likely to give in to this account easily. He is apt to argue that it is not a necessary account of truths in fiction because there are facts in fiction that the creators themselves intend us to believe (Matravers 2014). David Lean mandated that we imagine that Dr. Zhivago loved Lara, but he pretty explicitly intended us to believe that there was a civil war between the Bolesheviks and the White Russians. How will the imagination account handle the fact that there are truths in fiction that we are supposed to believe? Moreover, there are not only explicit truths in fictional motion pictures that moviemakers mandate that viewers believe but also an indeterminately large number of implicit truths that moviemakers intend us to believe—implicit truths that must be believed, it might be argued, if the film is to remain intelligible. In order to see this, recall that no fiction can proceed without presupposing that the viewers will fill in many of the details that remain unstated or un-exhibited in the motion picture. For example, it is never said or exhibited that Marlowe has a heart, that if a bullet pierces it, he will die, and that knowing this, Marlowe will be cautious when a gun is pointed his way. Viewers just suppose this in order to, among other things, make Marlowe’s behavior intelligible. Viewers are constantly filling in or completing the fictional world of the film. Our doing this in the right way is crucial to rendering the story intelligible. When the detective halts and raises his hands when the villain points

Motion Picture Narration  •  123

a revolver at him at point-blank range, we infer that the detective does this because he realizes that if he doesn’t do this, he will be dead. If we did not infer something along these lines, the situation would not make much sense. But how does the viewer fill in the fiction in the right way? She has a number of heuristics at her disposal. The first is the realistic heuristic; this tells her to fill in the story world in terms of the kinds of details that she believes obtain in the actual world. Of course, the realistic heuristic is not always pertinent; sometimes fictional worlds depart from the actual world; vampires rise from the dead; spaceships travel faster than the speed of light. In cases like these, the viewer falls back upon the conventions and presuppositions of the presiding genre. That is, they rely on what we may call “the genre heuristic.” And when this fails to render the story intelligible, we have other heuristics. For example, we may avail ourselves of the beliefs about the actual world shared by viewers and moviemakers at the time the fiction was composed. Call this “the historicist heuristic.” Which of the heuristics that are available to viewers they opt for depends upon which one or which combination of heuristics seems, upon reflection, to make the fiction maximally intelligible. But what does this have to do with the supposed inferiority of the imagination account of truth in fiction to the approach recommended by the friend of the ontological gap argument? Simply this: In light of the realistic heuristic, there are truths with virtually any conceivable movie fiction that its makers mandate us to believe, not imagine. So the imagination account is supposedly not a comprehensive account of what constitutes truth in fiction, whereas this is allegedly not a problem for the view that there is a fictional narrator/ presenter who is asserting, albeit often implicitly, that thus and so obtains in the fiction. That is, with any fictional movie, it is most likely that we are intended not only to imagine a, b, c . . . but also to believe x, y, c . . . Therefore, the notion that the creator’s mandate to viewers to imagine thus and such is not a full account of what is true in a fiction. Can the imagination approach survive this objection? There seem to be two strategies open to the friend of the imagination approach for responding to this objection. The first is to argue that whether we believe the explicit and/or implicit truths in the fiction is irrelevant. What we have been mandated to do is to imagine what we are shown. We are mandated to imagine that Moscow is in Russia when watching Gorky Park. That we also believe that Moscow is in Russia is beside the point. But this gambit comes with some obvious liabilities. Many social protest movies will miss their mark if we only imagine that a certain minority group is being treated unjustly.

124  •  Noël Carroll

Another response to what we might call the problem of belief in fiction might be to maintain that imagining is the default attitude when it comes to fiction. That is, with fictions we are mandated to imagine, unless there are reasons to think that belief is what is intended. That is, the overall, defining attitude with fiction is imagining, although episodes of believing may be nested in the overarching, complex, authorial intention for us to imagine. Here, of course, the question will arise about when viewers have grounds to think that believing is mandated. The answer to that question must be decided on a case-by-case basis in terms of what suppositions—what imaginings— make sense of the movies in question—that is, what seems the most plausible way of rendering the movie maximally intelligible. So the imagination approach can deal with the fact that sometimes we are intended to believe what is presented in the fiction at the same time that its basic account of truth in fiction forgoes the postulation of an implicit, fictional narrator/presenter by offering the actual filmmakers’ default mandate to imagine such and such as what makes things true in the fiction (e.g., that Scarlett O’Hara lives in Tara), while also allowing that sometimes that mandate will be suspended in favor of a mandate to believe z (e.g., that Atlanta is in Georgia) just in case upon reflection belief makes more sense in context. Does the implicit narrator approach or the imagination approach offer a better account of the problem of truth in fiction? Perhaps at least one thing in favor of the imagination account, to be pursued in greater detail in the next section, is that plain viewers will be prone to identify the actual moviemakers as the source of the views they are presented with, on the one hand, and they will deny any knowledge of the implicit, fictional narrator/presenter, on the other hand. Against the seeing/showing argument According to the seeing/showing argument, the Fictional Showing Hypothesis is said to follow from the Imagined Seeing Thesis. Seeing imaginarily implies a fictional presenter—if not an explicit one, then an implicit one, although an extremely minimal one. Strictly speaking, this is not right. Seeing does not ordinarily imply a presenter nor is it obvious that seeing imaginarily does; using a leaf floating in my swimming pool as a prop in my game of make-believe, I may imagine seeing a warship headed my way. But I need not posit that this is the work of some presenter. Nor is it evident that we need to hypothesize an implicit, fictional mediator/presenter with respect to theatrical performances. Rather the Fictional Showing Hypothesis with regard to cinema emerges only after one, in the face of the battery of anomalies and counterexamples cited earlier, remains committed to the notion of seeing imaginarily events like Atticus Finch defending Tom Robinson.

Motion Picture Narration  •  125

The implicit fictional presenter is a strange entity—a putatively naturallike process that somehow inexplicably, automatically presents views from the fictional world. In order to persuade us of the acceptability of such a posit, we are reminded of Flash Gordon’s all-seeing video device. We have no problem accepting that. So we should have no problem with the implicit fictional presenter. Just as it would be silly to raise questions about the nature and workings of Flash’s video, so likewise with the implicit fictional presenter. However, the analogy here is flawed. Flash Gordon’s video machine is explicitly given as part of the furniture of Flash Gordon’s fictional universe, as is Ming the Merciless and the planet Mongo. It is silly to ask for an explanation of the way in which it works, because the moviemakers have made it true in the fiction that it works in Flash’s universe, just as Bram Stoker made it true in the fiction Dracula that the count came from Transylvania. Both Flash’s video and Stoker’s vampires would not be items that audiences would supply in light of their default recruitment of the realistic heuristic. In order to accommodate them, the realistic heuristic must be suspended. And the reason that it is so suspended is that the relevant fictioneers have explicitly bracketed it in favor of their inventions. The audience has been overtly alerted that their default realistic assumptions are to be put in abeyance. This tells us something important about retiring the realistic heuristic. In order to do so, the audience needs to be clearly made aware that they are no longer confronting a world like their own. There are different entities and/or rules operating in places like Flash Gordon’s universe and Dracula’s castle. It is silly to ask about their nature, since they have been made up with scant respect for reality. But there is a big difference between Flash Gordon’s video mechanism and the implicit fictional presenter. Because audiences have been saliently apprised of the existence of Flash’s video, they suspend their default realism and sit back and enjoy the show. But they have not been clued into the alleged existence of the implicit fictional presenter. After all, it’s implicit. This has several consequences. First, since it has not been explicitly posited as a feature of the world of the fiction, questions about it cannot be rejected as silly on the grounds that “because the filmmakers say so.” Because the moviemakers haven’t said so. Instead, the audience would need some reason to posit the implicit fictional presenter, which would require the theorist attracted to the idea to be somewhat forthcoming about the nature of the beast. Indeed, told that they must be harboring some notion of the implicit fictional presenter—if only to make sense of their protestations that they see Atticus Finch defending Tom Robinson—most plain viewers would probably deny it. But in response, defenders of the seeing/showing argument are likely to suggest that we ignore the testimony of the plain viewers. Yet, at the very least, this seems arbitrary. Why are the friends of the seeing/showing argument

126  •  Noël Carroll

so ready to lend credence to plain viewers when they say they see Atticus Finch defending Tom Robinson, but so fast to dismiss them when they deny positing fictional implicit presenters? Why not be suspicious of the plain viewers’ talk about seeing Atticus Finch? Why accept it at face value? Isn’t it likely that when plain viewers speak this way, they are not speaking literally? Isn’t it highly probable that they are speaking elliptically? Instead of constantly repeating, “In the fiction film To Kill a Mockingbird, I see Atticus Finch doing this and that,” it is far more economical to abbreviate that down to “I see Atticus Finch doing this and that.” Moreover, the advocates of the implicit fictional presenter should not bridle at this interpretation, since they, too, will have to concede that plain viewers are speaking elliptically when they claim to see Atticus Finch. Expanded, what they really mean, according to the seeing/showing argument, is “I am seeing imaginarily (or, I am imagining seeing) Atticus Finch defending Tom Robinson.” In contrast, then, to the seeing/showing approach, the imagination approach is ready to accept both the plain viewer’s assertion (as elliptical) that he sees Atticus Finch and the plain viewer’s denial that he posits an implicit fictional presenter, whereas the seeing/showing approach arbitrarily accepts the first testimonial but not the second. Rather than characterize the audience’s engagement with what is on the screen as a matter of seeing imaginarily, another alternative is to say is that when we look at the screen we literally see an image of x, say, an image of the actor Gregory Peck, and then use that image to imagine something, say, Atticus Finch defending Tom Robinson. We don’t imagine seeing. We literally see actors and sets and then use that to imagine the story. Friends of the seeing/ showing argument may object to this on the grounds of the argument from belief rehearsed earlier—namely, that sometimes we are intended to believe what we see represented on the screen. But in response, it may be argued that although the default attitude mandated is the imagination, episodes of proscribed belief may be nested within it, as we suggested previously. Friends of the seeing/showing argument may find this a strained and unnatural way of proceeding, but it is not clear that it is self-contradictory, contrary to spectators’ experiences, or otherwise obviously false, whereas the notion of ubiquitous, implicit fictional presenters seems neither grounded in the plain viewer’s experience nor necessitated by logic. Although this may appear to add up to a standoff, at least the imagination approach is consistent with the phenomenology of the plain viewer. Additional problems One factor that intuitively draws some philosophers to the idea of an implicit fictional mediator/presenter seems to be the constantly changing point of view of most movies without explicit narrators. Someone or something has

Motion Picture Narration  •  127

to account for that.2 Theater, on the other hand, seems a different matter. Even if we describe our engagement in terms of seeing imaginarily, we are seeing imaginarily directly. But, then, what are we to make of movies that basically record theatrical performance from a stationary camera position after the fashion of the early twentieth-century films d’art? Would they be without implicit fictional mediators/presenters like the stage performances they slavishly, mechanically documented? But then the assertion of the ubiquity of implicit fictional presenters even of minimal agency in all motion pictures would be false. In a related vein, there may be movies in which it is given that there is no presenter, even of the minimal variety. In the recent motion picture Unfriended, it is given that we are viewing five or six interconnected (haunted) computers. Everything we see, save for the very last shot, fictionally is there because of the activity of one of the characters (including the ghost). They skype each other, they send each other e-mails and text messages, google news items and forward them to each other, and so on. There is no implicit presenter nor are the characters explicit narrators; they are not telling their story to anyone. They are (fictionally) enacting it. Admittedly, this is an imperfect example, due to the last shot, which occurs outside of computer-land. But it is easy to see how one could have made the movie without that last shot. However, if it is possible to make a motion picture that seems to eschew the idea of a narrator of even the minimal sort, then once more the claim that implicit fictional narrators are ubiquitous is dubious. Nor would it help matters to claim that they are more ubiquitous than not, since “ubiquity” is not a degree concept.

Notes   1 This chapter is solely concerned with the claims for the ubiquity of implicit fictional narrators in motion pictures. It is not my intention to deny that there could be an implicit fictional narrator in a given motion picture.   2 This has been denied (Thomson-Jones 2012).

References Carroll, N. (2013) “Cinematic Narration,” in N. Carroll (ed.), Minerva’s Night Out, Oxford: Wiley-Blackwell, 122–132. Chatman, S. (1978) Story and Discourse, Ithaca, NY: Cornell University Press. ——— (1990) Coming to Terms, Ithaca, NY: Cornell University Press. Kania, A. (2005) “Against the Ubiquity of Fictional Narrators,” The Journal of Aesthetics and Art Criticism 63:1, 47–54. ——— (forthcoming) “Our Wildest Imaginings: Two Theories of Our Engagement with (Narrative) Fiction.” Levinson, J. (1996) “Film Music and Narrative Agency,” in D. Bordwell and N. Carroll (eds.), PostTheory, Madison: University of Wisconsin Press, 248–282. Matravers, D. (2014) Fiction and Narrative, Oxford: Oxford University Press.

128  •  Noël Carroll Thomson-Jones, K. (2012) “Narration in Motion,” British Journal of Aesthetics, 52, 33–43. Walton, K. (1990) Mimesis as Make-Believe, Cambridge, MA: Harvard University Press. Wilson, G. (2011) Seeing Fictions in Film, Oxford: Oxford University Press.

Suggestions for Further Reading Gaut, B. (2010) A Philosophy of Cinematic Art, Cambridge: Cambridge University Press. Along with the articles by Andrew Kania, chapter 5 contains the most trenchant challenge to the notion of the ubiquitous implicit fictional narrator. Wilson, G. (2006) “Le Grand Imagier Steps Out: The Primitive Basis of Film Narration,” in N. Carroll and J. Choi (eds.), Philosophy of Film and Motion Pictures, Oxford: Blackwell, 185–199. The key article defending the idea of ubiquitous, minimal, mediated, implicit fictional narrators in fictional cinema; this is the article that launched a thousand rejoinders, including my own.

PART

IV

How Do Films Engage Our Emotions?

This page intentionally left blank

CHAPTER

7

Putting Cognition in Its Place Affect and the Experience of Narrative Film CARL PLANTINGA

Chapter Overview While increased attention to the body and to affect in film viewing is a welcome development, it threatens to eclipse cognition in film theory and philosophy. This chapter argues that a lack of attention to cognition is a serious error, and that we must put cognition in its place. This chapter argues for a broad conception of cognition that incorporates bodily and affective processes, and pays particular attention of the place of affect within cognition. The chapter demonstrates this in relation to film spectatorship, with discussions of “twofoldedness,” cognitive frames, the various sorts of emotions spectators have while viewing, moods in film and their relation to cognition, and the cognitive nature of empathy.

Introduction In the last thirty years scholarly understanding of the viewer’s experience of narrative film and television (what will be called here “film”) has changed significantly. The so-called cognitive revolution of the 1980s, which modeled the mind as a computer-like information processor, has been challenged by an intense interest in the body, the senses, affect, features of brains such as mirror neurons, and research projects rooted in neuronal mapping. The focus on affects and emotions that blossomed in the 1990s had already begun to move scholars away from the computer metaphor. While both computers and minds engage in inference-like activity, computers apparently don’t have affects or 131

132  •  Carl Plantinga

emotions, suggesting that as a model of human cognition, computation may be at best partial, and at worst misleading. Neither do computers have bodies. Theories of embodied cognition supplement and in some cases critique traditional cognitive theories (Wilson and Foglia 2011). In film and media studies (hereafter referred to as “film studies”) the interest in affect and embodiment is sometimes seen as leading away from cognition itself. Influential strains of phenomenological film theory emphasize the perceptual and embodied aspects of the experience of films (Barker 2009; Marks 2000; Sobchack 1992). I did an informal count of certain words as they appeared in titles of the talks and panels at the 2015 Society for Cinema and Media Studies conference in Montreal. “Body” and “embodiment” appeared in titles forty-three times and “affect” nineteen times. The word “emotion” appears three times, while “cognition” and “mind” each appear twice (SCMS Conference Program 2015). One could make the case that discussions of the body and affect have displaced examination of the higher cognitive functions in much of film studies. I have heard film theories criticized as “too cognitive,” which I take to mean overly focused on conscious, rational thought at the expense of other viewer experiences, activities, and responses. This begs the question of what cognition is in the context of film viewing, and of how cognition relates to desire, emotion, the senses, and the body. The study of affect, emotion, desire, embodiment, and the senses promises to expand and deepen our understanding of the viewer’s experience. But as I shall argue, this is so only if we relate these to cognition. My argument here has two parts. First, I will claim that we must broaden our conception of cognition to include bodily and affective processes, such as perception, emotion, and affect, in addition to those mental activities such as inference making and schematic categorization that in this context I will call “computational” activities. I will also argue that cognition must not be necessarily associated with rationality, nor is it primarily an activity of the conscious mind. Second, I claim that perceptual, affective, and computational components of cognition mutually influence each other, such that when examining the film viewing experience, considering them in strict isolation will often lead us astray. The broader point is that cognition, broadly and accurately conceived, is central to the experience of narrative films. Recent disciplinary trends in both media theory and aesthetics have justifiably recognized the importance of affect and the body, but threaten to lead us astray by ignoring or denying the centrality of cognition. We must put cognition in its place.

What Is Cognition? Steven Shaviro, in a blistering attack on cognitive film theory, writes that cognitive theorists “exclude anything that might be described as unconscious” and

Affect and Narrative Film  •  133

make it “impossible” to “ask any questions about desire, fantasy, passion, and emotion” (Shaviro 2008: 50–51). Were this true, cognitive film theory would have a problem. The claim is patently false, however. The questions purportedly impossible to ask had become mainstream research programs in cognitive film theory several years before Shaviro’s (2008) critique (see, e.g., Carroll 1998; Grodal 1997; Plantinga and Smith 1999; G. Smith 2003; M. Smith 1995a). Such bromides about cognitive film theory, however inaccurate, may have some currency in film and media studies and beyond, and put an interest in the cognitive activities of the spectator at a disadvantage, threatening to make such studies seem like a rear-guard action that ignores emotion, desire, the unconscious, and that most recent focus of interest, the body. The spectator’s cognitive activities, however, are central to all film viewing. Cognition is the mental activities of gaining knowledge and understanding through thought, experience, and the senses, plus the results of such ­activities— comprehension, intuition, insight, perception, and so on. Cognition is neither singular nor linear; the modularity of the mind suggests multiple cognitive processes occurring simultaneously and connected in complex ways. Daniel Dennett, for example, conceives of the mind as a “locus of multiple, quasiindependent processing streams” (quoted in Clark 2000: 178). Neither is cognition necessarily either conscious or rationalistic. Psychologists speak of the “cognitive unconscious” to differentiate it from Freudian conceptions of the unconscious (Plantinga 2009: 49–50). Moreover, cognition can be either rational or irrational depending on the individual and circumstance. Just as emotion in the abstract is neutral with regard to practical rationality, so is cognition. Their rationality must be gauged in context. The so-called cognitive film theorists themselves, though saddled with a name that seems so 1980s, are interested in much more than cognition narrowly conceived of as conscious thought. Cognitive scholars have turned to studies of film spectatorship that emphasize the senses, embodiment and body metaphors, preconscious affects, moods, and neurological activity as measured by fMRIs, and so forth. Cognitive film theorists have been fascinated with mirror neurons, one topic of Dan Shaw’s chapter in this volume, and have attempted to understand how their workings can be integrated into cognition (Plantinga 2009: 124; 2013: 100–104; M. Smith 2016; Vaage 2006). Computers, most would agree, have no senses, bodies, neurons, affects, moods, or brains, and thus are at best an incomplete model for spectators in all their fleshy corporeality. Cognitive media theorists who write centrally about emotion, for example, put emotion in the context of the cognitive activities of the spectator. Emotions have a “searchlight” function; they direct the attention of the viewer and elicit various inferences and anticipations. When viewing a narrative film, the emotion of disgust, for example, can lead to various sorts of moral intuitions or evaluations (Plantinga 2009: 198–220). These evaluations in turn may

134  •  Carl Plantinga

influence desires for future narrative outcomes. Moods, similarly, have characteristic cognitive styles, and can draw together memory and associations (Plantinga 2012: 467–470). Thus desire, emotion, and mood are aspects of cognition rather than separate from it. We feel our way through the world as much as think it. In discussions of film and media viewing, then, cognition cannot be ignored for long. Cognition is not separate from affect and bodily processes; rather, affect and bodily processes are aspects of cognition.

Viewer Affects and Cognition That is not to say that all aspects of the viewer’s experience involve the “higher” cognitive processes in equal measure. Consider the distinction between affect and emotion. Affect in its broadest sense covers all bodily disturbances, including pains, indigestion, sexual desire, hunger, moods, physiological sensations, and emotions. Many affects have their genesis in automatic and non-cognitive processes (though their experience does affect subsequent cognition). Affects in this sense are the bodily sensations, feelings, and moods that have no objects, as emotions do, and that do not seem to arise from appraisals or perceptions (in a sense that I will define ahead). For instance, I wake up in a foul mood for mysterious reasons, despite life circumstances that seem otherwise fine. Or certain primitive mirror responses cause me to synchronize my facial expressions with the teary-eyed character on the screen, and perhaps feel a sad twinge similar in valence (if not intensity) to what I suppose the character feels. What Dan Shaw calls “basic empathy” in his chapter in this volume is one sort of primitive affect—what I would call “mirror responses”—that is usefully distinguished from emotions. Emotions, on the other hand, are the result of appraisals that often come in the form of “perceptions.” I follow Robert C. Roberts in calling emotions “concern-based construals.” Emotions, in other words, are perceptual states by which the perceiver construes a situation in relation to her concerns. As such, emotions have cognitive content. Roberts writes that emotions “are perceptual states that can in principle be true or false perceptions of the values that inhere in situations” (Roberts 2013: 67; see also Roberts 2003). I may construe that the garter snake in my garden threatens my life and thus I experience fear; the fear in this case would be a false perception of the threat posed by the harmless snake. Emotions have the quality of “aboutness,” since they are implicit evaluations of an object in relation to one’s concerns. The kinds of emotions people have at the movies are varied; I offer an extended typology in Moving Viewers: American Film and the Spectator’s Experience (2009). For now I would mention three important varieties: artifact, direct, and sympathetic/antipathetic emotions. Artifact emotions take as their object the film as an artifact, as the product of makers who present the artifact

Affect and Narrative Film  •  135

for our consumption. Artifact emotions include disdain or admiration, gratitude or impatience, all directed at the film as an artifact made by someone. The direct emotions take as their object narrative content and the unfolding narrative, and would include curiosity, suspense, anticipation, and surprise or startle. Sympathetic/antipathetic emotions have characters as their object, and include compassion, pity, admiration, anger, and socio-moral disgust. As I demonstrate ahead, the elicitation of these emotions in the context of viewing screen stories presumes various antecedent cognitive processes. Just as important, since both film narrative and cognition are time-bound, having emotions also affects and directs subsequent cognition. Emotions often come quickly and in a nearly automatic fashion. Yet they are nonetheless an aspect of cognition because this quality of “aboutness” firmly implicates them in the “sense-making” activities of cognition. Yet even if one denies all cognitive precedents for emotions, one cannot easily deny cognitive antecedents. Not only emotions but also affects in general have important cognitive antecedents.

Twofoldedness and Cognitive Frames Of the sympathetic/antipathetic emotions, we might ask, as Colin Radford did in 1975, whether it is rational to be moved by the fate of fictional characters, since readers (and viewers) understand, even while reading or viewing, that such characters do not in fact exist. Observing actual spectator emotions is enough to decisively demonstrate that viewers assume differences between the immersive experience of a fictional world, on the one hand, and their actual lives, on the other. Rampaging monsters, dangerous gunfights, and devastating social humiliations would all be too much to take were the viewer not aware of the fictional, institutional, and nonthreatening nature of the viewing situation. If, during the sinking of the Titanic in Titanic (1997), the viewer were for a single moment to mistake the fiction for reality, the terror would be gutwrenchingly powerful. We abide the narrative because we understand that it is a representation and not the actual thing. Though we might say we experience terror while viewing the film, this emotion is different in both kind and degree from the terror we presume the characters to experience. It is different in kind because the viewer presumes not only that she is viewing a representation and not the real thing, and not only that she is an observer and not a participant, but also that she is viewing a fiction. If we know it is a fiction, however, then why do we respond with strong emotions? Is an emotional response thereby irrational? These questions have been extensively debated, and there is no space here to rehearse the various proposed answers (Plantinga 2009: 64–65). Noël Carroll’s thought theory is close to my position (Carroll 1990: 79–87), with a caveat. Carroll claims that

136  •  Carl Plantinga

people respond emotionally to unasserted thoughts—that is, to imaginings of entities and situations that are not thought to exist in the actual world. In the case of fiction films, however, neither the word “imagining” nor “thought” seems to adequately describe the object that elicits the response. As the Titanic sinks in Titanic, the viewer is assaulted with vivid images and sounds. The experience of a narrative film is sensual. The film viewers respond affectively to visceral stimuli that they implicitly understand are representations. This understanding depends on cognitive schemas, as I will elaborate further ahead. The usual viewer responses to fictional characters—sympathy for the good character or disdain for the fictional villain—are rational. Humans are not conscious of the bulk of the cognitive processing of external stimuli that occurs throughout waking life. Automaticity refers to the quick and non-conscious types and patterns of response that allow us to interact with our environment quickly and efficiently. Many of these patterns of response are beyond our conscious control and unmediated by conscious deliberation, but subject to the workings of the body and the “cognitive unconscious” (Plantinga 2009: 49–53). Tears may form in our eyes at the end of Stella Dallas, when Stella views her daughter’s wedding ceremony through a window, on a rainy night. We are responding to the “suction” of storytelling, as Jonathan Gottshcall puts it (Gottshcall 2012: 3), to the emotional contagion of facial expressions as viewed in close-up, to the emotionality of the music, to the perception of the imagined human experience of joy mixed with regret and sorrow. Outside of the experience of screen stories as well, we consistently respond to the world around us with quick and sudden affect. The sudden loud noise startles us. We see a beautiful man or woman enter the room, and our heart rate increases and palms become sweaty. We become annoyed with the antics of a friend, but don’t understand why. What is “rational” in relation to such events differs according to context, but consists in the “appropriate” harnessing of and meta-response to primitive affects, such as startle, mood, physiological feelings, and so forth, or to the flashes of approval and disapproval that often constitute the moral emotions. By “harnessing” affect I mean the regulation of behavior and thought to generate an “appropriate” response. By metaresponse I mean the subsequent response to the initial response. For example, I may revel in the fear and disgust initially caused by the sight of the repulsive Borg in an episode of Star Trek: The Next Generation. I do not stand up and run away, however, because my reverie depends on the cognitive schema that tempers my response; I presume that I am viewing a fictional representation. This cognitive frame also frees me to admire the fine work of the creators in thinking up and then bringing to fruition such magnificent examples of monsterhood. This harnessing and meta-response are aspects of multiple cognitive streams functioning simultaneously. So-called primitive and automatic responses are managed by higher cognitive processing. It is the nature of what

Affect and Narrative Film  •  137

we do with our immediate, automatic, non-cognitive affects that determines whether the behavior is rational (by whatever standard). Just as our tendency to respond physiologically to musical rhythm or to flinch at loud sound is neither rational nor irrational in itself, so is our tendency to respond emotionally to fictional stories and characters. One might object that my response to the paradox of emotion may account for low-level affective responses to films better than it does emotional responses rooted in high-level cognition. Yet even in the case of complex sympathetic responses, for example, we cannot separate the deliberations of thinking from the workings of the body. My sympathy for a character may be in part automatic, when, for example, through facial mimicry I have mirror responses to a close-up of a character’s face. In other regards, however, my sympathy results from a complex understanding of the character’s situation. Even the “higher” emotional responses involve automatic and bodily processes that to some extent escape conscious control. In relation to the viewer’s interaction with a painting, Richard Wollheim writes of “twofoldedness,” our simultaneous awareness of the surface of the painting and the objects it depicts (Wollheim 1987: 46–47, 72–75). When we view a painting of an apple, we may recognize the apple and the brush marks simultaneously. It is hardly irrational that the depicted apple looks appetizing to us if it appears to be shiny and fresh, or disgusting if it appears to be rotten. The depiction triggers certain baseline responses that are a feature of our psychologies. Yet we don’t reach out for the apple because we understand that it is not a real apple, but merely a pattern of brushstrokes on a canvas. Murray Smith has extended this thought to viewer responses to fictional characters and to narrative films generally (M. Smith 2011; see also Plantinga 2013: 96–98). We respond to characters simultaneously as people, as actors playing roles, and in some cases as familiar movie stars; we respond with emotion to narrative films while simultaneously recognizing their artifactuality and their having been produced within certain institutions of fiction (M. Smith 1995b). One of the foundations of cognitive science is the examination of mental categories or frames, called schemas. To comprehend a narrative film, viewers employ dozens of schemas to contextualize what they see, enable the processing of information, and direct anticipations and hypothesis making. In film viewing, these schemas range from real-world schemas (e.g., procedural schemas about how crooks behave during a bank robbery) to extrinsic filmic schemas (e.g., conventions of the crime drama and other genre schemas) to intrinsic schemas (a character schema that develops as we get to know a particular character in the film). Schemas are more than simple background information; they can consist of templates or categorizations that develop over time. One example is the type or stereotype, by which we respond to characters as exemplars of a “sort of person.” The employment of schemas makes our

138  •  Carl Plantinga

response to the world around us quick and efficient. (Obviously, schemas can have problematic implications, as they do in the case of harmful stereotypes.) Like cognition itself, schemas are neither rational nor irrational in themselves. Among film scholars, David Bordwell has written the most extensively about schemas in relation to narrative comprehension (Bordwell 1985; 2008). As I mentioned earlier, cognition is generally thought of as the mental activities of gaining knowledge and understanding through thought, experience, and the senses, plus the results of such activities—comprehension, intuition, insight, perception, and so on. The nature of the emotions we have at the movies depends fundamentally on cognition, because cognition extends far beyond conscious inference making to the largely non-conscious schemas that guide our thinking and response. Earlier I identified three sorts of spectator emotions—direct, sympathetic/antipathetic, and artifact emotions. Artifact emotions depend on schematic understandings of excellence in filmmaking— for example, on knowledge of the director or actors or other talent, on genre, and on much else. Direct emotions such as anticipation and suspense rely on complex understandings of the narrative situation as it unfolds in time (as I will further explain ahead). Sympathetic and antipathetic emotions depend in part on various types of moral evaluations, themselves dependent on moral schemas of right and wrong. It seems that many of the emotions of the sort we have at the movies are fundamentally cognitive in nature. Cognition, as we have seen, should not be associated solely with conscious mental activity. Neither should it be assumed to be a priori rational. Cognition itself is a psychological process of perceiving, framing, contextual understanding, inference making, hypothesis making, anticipation, questioning, reasoning, and so on. In itself the word cognitive is neutral with regard to rationality. Cognitive activity can be disordered. It can be motivated by mysterious passions and desires. Our inference making is subject to all kinds of biases, affecting our social judgments, memory, and a wide range of beliefs, as I will further describe in the section on sympathy and empathy (Rational Wiki 2014; J. Smith 2014). Thus my insistence that cognition plays a central role in film affect should not be taken to imply that all elicited film affect is rational in nature. When I say that film affect depends heavily on cognition, what I mean is not that response is intrinsically rational, but merely that it presupposes and depends on the sorts of mental activities I list earlier.

Narrative Progression and the Mind Emotions result from cognition, broadly conceived, and having an emotion directs subsequent cognition. Various forms of affect strongly contribute to cognition as well—for example, mood. Orson Welles’s Touch of Evil (1958), for example, is a particularly moody black-and-white film, bathing its many

Affect and Narrative Film  •  139

night scenes in the low-key lighting characteristic of films noir and setting the action against rusted oil drilling machinery and the garish neon lights of a seedy border town. The film’s pessimistic tone is assisted by the tough and cynical banter of many of the characters, together with suggestions of violence and corruption. Each element feels a certain way due to its phenomenal qualities and cultural associations. This amalgam of elements sets the tone or mood for the film—its affective character (Plantinga 2012). Humans have moods too, but obviously they aren’t the same thing as moods in cinema. A human mood is a diffuse mental and bodily state characterized by a pervasive feeling, usually with a negative or positive valence. We sometimes say that we are in a good or a giddy mood or in a bad or sad mood. Films cannot have mental states, of course, so when we attribute moods to a film we must mean something else. A film mood is its tonal or affective character, considered as an intentional construct designed to provide both affect and perspective. The opening scene of Alfred Hitchcock’s Rebecca (1940), to consider another example, features a hushed and nostalgic voice-over narration describing the narrator’s return to the past, low-key lighting on a moonlit night, the camera slowly moving down an overgrown path through wafts of mist and fog, toward the ruined mansion, Manderly, standing in silhouette against the night sky. Attempting to describe a mood quickly leads to the heresy of paraphrase; ultimately one must experience it to understand it. A mood in a narrative film doesn’t necessarily lead to a corresponding mood in the viewer, but it may. The causality of human moods is somewhat mysterious. Yet a mood in film can lead to human moods by eliciting emotions associated with that mood, foregrounding aesthetic qualities of form and content, and encouraging the sorts of cognition or cognitive styles associated with particular moods (Plantinga 2012: 464–470). The casual direction between affect and cognition runs both ways, such that the elicitation of a particular mood while viewing can lead to particular ways of thinking. Moods have marked influence on memory, social judgment, thinking, and perceiving. Psychologists have been researching these phenomena for years. Thus Joseph Forgas writes, Affect and cognition are integrally linked within an associative network of cognitive representations . . . Thus, material that is associatively linked to the current mood is more likely to be activated, preferentially recalled, and used in various constructive cognitive tasks, leading to a potential mood congruency in attention, learning, memory, associations, evaluations, and judgments. In other words, affect is not an incidental, but an inseparable, part of how we see and represent the world around us; how we select, store, and retrieve information; and how we use stored knowledge structures in the performance of cognitive tasks. (Forgas 2000: 11)

140  •  Carl Plantinga

More specifically, certain moods are associated with cognitive styles. Interestingly, people in sad moods, Forgas and Eric Eich write, are “less prone to judgmental errors . . ., more resistant to eyewitness distortions. ., and are better at producing high-quality and effective persuasive messages” (Forgas and Eich 2012: 78). The authors suggest that perhaps over the span of evolutionary history, affective states became triggers for cognitive processing patterns appropriate to specific situations. Mood in film is thus an important facilitator of various sorts of cognition in the spectator.

Sympathy and Empathy In narrative film, among the strongest forces in generating spectator affect are various sorts of attitudes toward characters. Screen stories play on our fascination with and emotional responses to characters to provide the kind of immersive experience we desire. The viewer’s fascination with a particular character does not necessarily entail a strong allegiance or identification with the character. Viewers sometimes take a distanced but intense interest in protagonists they are not sure whether to like—for example, Elizabeth Vogler (Liv Ullmann) in Bergman’s Persona (1966) or Guido (Marcello Mastroianni) in Fellini’s Eight and a Half (1963). Audiences take an interest in bizarre antagonists, as with Hannibal (the Cannibal) Lecter (Anthony Hopkins) in The Silence of the Lambs (1991) or the Joker (Heath Ledger) in The Dark Knight (2008). The kind of viewer/character relationship most theorists and philosophers have been interested in, however, is what we call “identification” or “allegiance,” in which (minimally) viewers develop a strong liking for a single protagonist. The nature of such identification is a matter of dispute. Noël Carroll has claimed that the very idea of identification is too broad, noting that the term could mean any of the following: . . . that we like the protagonist; that we recognize the circumstances of the protagonist to be significantly like those we have found or find ourselves in; that we sympathize with the protagonist; that we are one in interest, or feeling, or principle, or all of these with the protagonist; that we see the action unfolding in the fiction from the protagonist’s point of view; that we share the protagonist’s values; that, for the duration of our intercourse with the fiction, we are entranced (or otherwise manipulated and/or deceived) so that we fall under the illusion that each of us somehow regards herself to be the protagonist. (Carroll 1990: 89) Some views of identification presume that the viewer shares an identity with the identified-with character, or has exactly the same emotions as the

Affect and Narrative Film  •  141

character identified with. Neither of these assumptions is tenable, Carroll claims, because viewers always take an “external” perspective on characters. Viewers cannot help but see characters from the outside, from their own perspectives. The viewer’s construal of a narrative situation will differ from that of a character because the viewer almost invariably has more information than the character. For example, we fear for the life of the young woman splashing in the ocean because we know that a shark is approaching; she, on the other hand, is unaware of the shark and is delighted to be having a pleasant swim. Or the protagonist in a romantic comedy may be hopelessly attracted to a certain man, but the audience, unlike her, has been made privy to the man’s selfish motives. Carroll suggests that we drop the term “identification” altogether, since on his view it is either hopelessly ambiguous or conceptually mistaken. Murray Smith does just this in his work on “character engagement,” abandoning the term “identification” and offering instead a useful three-part distinction between what he calls recognition, alignment, and allegiance, all-encompassing what Smith calls the viewer’s “engagement” with characters. Recognition describes the process by which viewers construct a fictional character from various textual cues. Alignment is developed through two processes: spatiotemporal attachment and subjective access. Spatiotemporal attachment occurs when we are literally provided with the character’s point of view or with the sounds she/he hears; subjective access refers more generally to the access viewers are provided to the character’s emotions, thoughts, and desires—her/his mental life. Allegiance, finally, is a strong pro feeling or attitude that develops toward the protagonist. Smith claims that this is rooted in the viewer’s moral judgment of the protagonist (M. Smith 1995a). These pro and con attitudes that viewers take toward characters are fundamental to the work of a narrative, and to the elicitation of affect. Pro attitudes can range from mere liking, to sympathy, to allegiance, to what I call “projection” (the desire to emulate a character). When a filmmaker can elicit strong sympathies or allegiances for particular characters, this makes for the synchronization of viewer affective response. Pro attitudes and feelings work to align viewers with character goals and to generate desires about particular outcomes. In either case, the synchronization of concerns, together with the particular construal of those concerns developed through the strategies and techniques of narrative filmmaking, increases the likelihood that audiences will respond in intended ways. Much of what I have argued also applies to “con” attitudes toward characters, ranging from dislike to antipathy to opposition (Plantinga 2010). I know of no one who denies the role of cognition in character engagement; it is rather the nature of that cognition that may be in dispute. In the foregoing discussions, I noted that cognition is not necessarily either conscious or rational. Smith, in my opinion, shades too much toward the rational nature of

142  •  Carl Plantinga

cognition when he claims that audiences form allegiances with characters based on moral judgment. Smith would certainly agree that the concept of allegiance is too general to account for all of the pro attitudes taken toward characters (M. Smith 1999: 220–221). For example, we may form initial good impressions of a character because she is amusing or quirky, without knowing enough about the character to evaluate her morally. Or we may sympathize with a victim, despite the victim not having acted (or having had the opportunity to act) in a way that could be evaluated morally. The determinants of human sympathy for others can be wide-ranging indeed, including such amoral characteristics as perceived similarity, proximity, vividness, and a match between one’s own feeling state and that of the target individual (Plantinga 2010: 39). I would argue that these sorts of character traits can be mixed up with those we might think of as evaluable solely on moral terms, such that good looks or charm may influence our judgment of virtue and moral behavior. This is the case in Legends of the Fall (1994), I argue, where audiences are encouraged to develop allegiance to Tristan (Brad Pitt) despite behaviors conventionally considered to be immoral (revenge killing, bootlegging, drug use, consorting with prostitutes, etc.). If allegiance were solely the result of reasoned appeals to moral precepts, it would be very difficult for audiences to have allegiance to Tristan. It is various amoral qualities and associations that lead some viewers to have such allegiances to Tristan, such as his good looks, his passionate personality, and his association with the wilderness and with a certain conception of masculinity as unfettered, untamed, and self-governing. In this case, cognition in the form of moral judgment is certainly involved in the formation of allegiance to Tristan. Character engagement is always a cognitive process; it isn’t always a rational process. One should note that this is not necessarily a case where affect leads cognition astray, although that can surely happen. The viewer’s erotic attraction to a character, for example, could cause a viewer to evaluate the character less harshly, or to allow fascination to submerge or attenuate moral evaluation. This is a case in which affect biases cognitive processing. But cognition does not require affect to be led astray. Hasty generalizations, for example, or the confirmation bias (by which one searches for information that confirms one’s prior beliefs), for another, are not necessarily affect-driven. They seem to be mistakes in thinking, not feeling. The bottom line, however, is that this discussion highlights the artificiality of separating affect from cognition in everyday mental life. Affect is part and parcel of cognitive processes. Within character engagement, the notions of sympathy and empathy are centrally important. The two terms—sympathy and empathy—are shapeshifters, and are either given various distinct and stipulative definitions or used interchangeably, depending on the writer. When they are argued to be distinct, sympathy is held to be feeling for another person, while empathy is

Affect and Narrative Film  •  143

feeling with. Such conceptions of empathy as feeling with require more than caring for another person; they require some sort of similarity of feeling, or even identical feeling. On this view, I can sympathize with you without sharing your feelings. You are angry; I am sad that you are angry and for the injustice done to you, and thus I sympathize. Empathy, on the other hand, would mean that I share your anger. Amy Coplan defines empathy as a cognitive and affective process with three necessary features: affective matching, other-oriented perspective taking, and self-other differentiation (Coplan 2011). Affective matching occurs when the observer’s “affective states are qualitatively identical” to the target’s affective states, though they may differ in degree (ibid.: 6). Other-oriented perspective taking occurs when I imagine the target (rather than myself) undergoing the target’s experiences. This requires a suppression of the ego, a kind of “effort and regulation” suggesting that empathy is a “motivated and controlled process” (ibid.: 14). Self-other differentiation is the maintenance, while empathizing, of a separation of an awareness of one’s own experiences from mental representations of the other and the other’s experiences. This conception of empathy considers it to be a higher-order cognitive and affective state, more cognitively complex than a mirror response or a response rooted in mimicry. It is important to note that Coplan’s definition of empathy is stipulative and quite narrow in scope. It is stipulative because Coplan is not describing how the term is in fact used by most English speakers, but rather proposing her definition as one that ought to be used. The definition is narrow in scope because in requiring the three features she proposes as constitutive of empathy, she provides an honorary definition that would make empathy an inherent virtue. As Coplan writes, only “empathy that combines affective matching, other-oriented perspective-taking, and self-other differentiation provides experiential understanding” (2011: 17), providing knowledge of another person’s thoughts, feelings, and behavior. While scientific understanding may provide knowledge of causes and effects, she writes, it is only empathy that helps us to understand the experiences of others. Coplan cannot be accused of eliminating the cognitive aspects of empathy. Yet in insisting on the qualitative identity of the empathizer’s affect and the affect of the one empathized with, she eliminates the possibility of empathy for fictional characters. This is true because the affective states elicited by films are typically always different than those we presume any given character has in the world of the fiction. Coplan’s assertion that empathy requires an identity of affect (in kind if not degree) between observer and target either (1) misconstrues the place of affect in what she calls experiential understanding, or (2) renders screen stories wholly incapable of eliciting empathy, and thus experiential understanding. The latter would be anathema to Coplan; elsewhere she has written about the nature of empathy in the film viewing experience,

144  •  Carl Plantinga

and clearly believes it to be an important component of the viewer’s experience of screen stories (Coplan 2006). To see why Coplan’s conception of empathy is challenged by the film viewing situation, consider the nature of emotion. As I argued earlier, an emotion is a concern-based construal. We are jealous, for example, when we see that our romantic partner is flirting with another person. Our concern is to preserve the exclusive romantic attachment of our partner. Our construal is that our partner is attracted to another, and that our attachment may be threatened. Given this conception of emotion, the viewer of a screen story and a character in the fictional world of the screen story cannot have full-blown “affective states [that] are qualitatively identical” (Coplan 2011: 6). This is especially true if those affective states include emotions, which consist of concern-based construals that are often cognitively complex. First consider a problem that may in fact be specious, but nonetheless needs addressing. Fictional characters do not have actual affective states, but are in fact creative constructs that viewers respond to in some ways as though they were real people. If the character has no affective state, and if empathy requires affective matching, then empathy in film viewing is thrown into question. My sadness and the character Louella’s sadness do not match, because Louella is fictional and not real, and because as such the construct Louella is not actually sad. The claim that one’s response can be identical in kind to that of a fictional character needs unpacking. At the very least, Coplan would have to extend and/or revise her definition of empathy to account for the fictional situation. What would it mean to have an affective state identical in kind to that of a fictional character? It might be possible to work around this problem by positing the identity as one between an affective state of the observer and the representationally prescribed affective state of a character. That is, various conventions of representation (and of human behavior) would prescribe that a character shown crying in close-up is sad, and thus the spectator who recognizes this, and is also sad (and fulfills the other two criteria), could be said to empathize with that character. Suppose that we resolve the issue in some way similar to this. This still leaves problems for the theory. For one, there is the issue of twofoldedness. I suggested earlier that the film spectator reacts to fictional characters as though they were simultaneously both persons and artificial entities presented through the institutions of fiction. Thus the viewer’s concern-based construal would take the fictional situation into account, while the fictional character would presumably believe that the events were really happening (unless the characters found themselves in a reflexive Pirandellian drama). To put it bluntly, Louella in the fiction is sad because her pet dog has a broken leg. The viewer is sitting in a movie theater watching a film in which a character is sad because her pet dog has a broken leg. These cannot help but lead to very different construals.

Affect and Narrative Film  •  145

To my mind, these radically different positions in the world affect both the kind and degree of affective experience. To be sure, I may share Louella’s sadness, but I never forget that I am a separate person from Louella, that Louella is a fictional character, and that I am viewing a fiction. Cognitive frames and schemas different from those of Louella govern my responses. My sadness will share a valence with Louella’s sadness, but it may also be accompanied by some of the following emotions, none of which will be shared by Louella: (1) anticipation or curiosity about how Louella will react (direct emotions), (2) disdain or embarrassment for having responded in the ways prescribed by the narrative (meta- or self-conscious emotions), and (3) admiration or disdain for the way in which the filmmakers constructed the scene (artifact emotions). If these emotions are mixed with my sadness over Louella’s dog’s broken leg, then my affective state does not match that of Louella. My argument would be that while viewing fictions, these sorts of emotions are typically mixed with the emotions directed to the plight of characters. These sorts of considerations also pose a problem for Dan Shaw’s theory of mirror neurons and simulation in the context of empathy for fictional characters (Shaw, this volume). Shaw implies that cognition plays little role in the spectator’s “basic empathy” for characters, but rather to understand empathy, we should focus on the effects of mirror neurons in the spectator’s simulation of the character’s situation. Although our understanding of the functions of mirror neurons is far from settled, it would be hasty to deny the importance of mirror neurons in empathy. It would be equally hasty, however, to deny the role of cognition in empathy. The spectator’s simulation of a fictional character, as I understand it, is not opposed to cognition but is itself a cognitive activity (using the expansive sense of cognition I have been arguing for). Were empathy merely a mirror effect, it would be difficult to account for twofoldedness or to explain why spectators tend not to empathize with villains, even those whose faces are shown for extended periods on-screen. Shaw claims that spectators do empathize with the villainous Hannibal Lecter (Anthony Hopkins) in The Silence of the Lambs (1991) because they are drawn in by the intensity and expressiveness of his face. We also see the faces of serial killer Buffalo Bill (Jame Gumb) and the hideous Dr. Chilton (Anthony Heald) in close-up, however, and few would argue that viewers empathize with them. If audiences do empathize with Hannibal Lecter, it may well have something to do with his impeccable speech, mysterious intelligence, and creativity; his sympathetic affiliation with the protagonist Clarice Starling (Jodie Foster); his victimization (at the hands of the self-serving Dr. Chilton); and thus his comparative attractiveness in comparison with Buffalo Bill and Dr. Chilton. These are all possible contributors to empathy that extend beyond mirror effects and must be processed cognitively.

146  •  Carl Plantinga

Conclusion These thoughts on empathy highlight the centrality of cognition in film viewing. The film spectator is constantly monitoring, comprehending, making inferences, estimating, anticipating, and evaluating. These activities are not carried out in a purely rational fashion, and neither are they necessarily conscious. Cognition works in tandem with the body and with the passions. Cognition, properly conceived, includes affective components as well as “computational” activities, and cognition should not be held to be the opposite of or opposed to affect. Cognition, properly defined, is at the heart of human experience, and no less in the viewing of narrative films. Film theory and the philosophy of film should not minimize the centrality of cognition, but may need to consistently monitor evolving ideas about just what cognition is, and how it emerges from and relates to the body.

References Barker, J. (2009) The Tactile Eye: Touch and the Cinematic Experience, Berkeley: University of California Press. Bordwell, David (1985) Narration in the Fiction Film, Madison: University of Wisconsin Press. ——— (2008) Poetics of Cinema, New York: Routledge. Carroll, N. (1990) The Philosophy of Horror, or, Paradoxes of the Heart, New York: Routledge. ——— (1998) The Philosophy of Mass Art, Oxford: Clarendon Press. Clark, A. (2000) Mindware: An Introduction to Cognitive Science, Oxford: Oxford University Press. Coplan, A. (2006) “Catching Character’s Emotions: Emotional Contagion Responses to Narrative Fiction,” Film Studies 8, 26–38. ——— (2011) “Understanding Empathy: Its Features and Effects,” in A. Coplan and P. Goldie (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Clarendon Press, 3–18. Forgas, J. P. (2000) “Introduction: The Role of Affect in Social Cognition,” in J. P. Forgas (ed.), Feeling and Thinking: The Role of Affect in Social Cognition, Cambridge: Cambridge University Press, 1–28. Forgas, J. P. and E. Eich (2012) “Affective Influences on Cognition,” in A. F. Healy and A. W. Proctor (eds.), Handbook of Psychology, vol. 4, Hoboken, NJ: Wiley, 61–82. Gottshcall, J. (2012) The Storytelling Animal: How Stories Make Us Human, Boston and New York: Houghton Mifflin Harcourt. Grodal, T. (1997) Moving Pictures: A New Theory of Film Genres, Feelings, and Cognition, Oxford: Clarendon Press. Marks, L. (2000) The Skin of the Film: Intercultural Cinema, Embodiment, and the Senses, Raleigh, NC: Duke University Press. Plantinga, C. (2009) Moving Viewers: American Film and the Spectator’s Experience, Berkeley: University of California Press. ——— (2010) “ ‘I Followed the Rules, and They All Loved You More’: Moral Judgment and Attitudes toward Fictional Characters in Film,” Midwest Studies in Philosophy 34, 34–51. ——— (2012) “Art Moods and Human Moods in Narrative Cinema,” Projections: The Journal for Movies and Mind 43, 455–475. ——— (2013) “The Affective Power of Movies,” in A. Shimamura (ed.), Psychocinematics: Exploring Cognition at the Movies, Oxford: Oxford University Press, 94–111.

Affect and Narrative Film  •  147 Plantinga, C. and G. M. Smith (eds.) (1999) Passionate Views: Film, Cognition, and Emotion, Baltimore: Johns Hopkins University Press. Radford, C. and M. Westin (1975) “How Can We Be Moved by the Fate of Anna Karenina?” Aristotelian Society Supplementary Volume 49, 67–93. Rational Wiki (n.d.) “List of Cognitive Biases,” accessed November 19, 2014, http://rationalwiki. org/wiki/List_of_cognitive_biases. Roberts, R. C. (2003) Emotions: An Aid in Moral Psychology, Cambridge: Cambridge University Press. ——— (2013) Emotions in the Moral Life, Cambridge: Cambridge University Press. Shaviro, S. (2008) “The Cinematic Body Redux,” Parallax 14, 50–51. Smith, G. M. (2003) Film Structure and the Emotion System, Cambridge: Cambridge University Press. Smith, J. (2014) “Filmmakers as Folk Psychologists: How Filmmakers Exploit Cognitive Biases as an Aspect of Film Narration, Characterization, and Spectatorship,” in L. Zunshine (ed.), The Oxford Handbook of Cognitive Literary Studies, Oxford: Oxford University Press, 483–501. Smith, M. (1995a) Engaging Characters: Fiction, Emotion and the Cinema, Oxford: Clarendon Press. ——— (1995b) “Film Spectatorship and the Institutions of Fiction,” Journal of Aesthetics and Art Criticism 53, 113–127. ——— (1999) “Gangsters, Cannibals, Aesthetes, or Apparently Perverse Allegiances,” in C. Plantinga and G. M. Smith (eds.), Passionate Views: Film, Cognition, and Emotion, Baltimore: Johns Hopkins University Press, 217–238. ——— (2011) “On the Twofoldedness of Character,” New Literary History 42, 277–294. ——— (2016) Film, Art, and the Third Culture, Oxford: Oxford University Press. Sobchack, V. (1992) The Address of the Eye: A Phenomenology of Film Experience, Princeton, NJ: Princeton University Press. Vaage, M. B. (2006) “The Empathetic Film Spectator in Analytic Philosophy and Naturalized Phenomenology,” Film and Philosophy 10, 21–38. Wilson, R. A. and L. Foglia (2011) “Embodied Cognition,” in Edward N. Zalta (ed.), The Stanford Encyclopedia of Philosophy (Fall Edition). http://plato.stanford.edu/archives/fall2011/ entries/embodied-cognition/. Wollheim, R. (1987) Painting as Art, London: Thames and Hudson.

Suggestions for Further Reading Carroll, N. (2008) The Philosophy of Motion Pictures, Malden, MA, and Oxford: Blackwell. (Work by the leading philosopher of film, including a substantial chapter on affect and the moving image.) Hogan, P. C. (2011) Affective Narratology, Omaha: University of Nebraska Press. (A theory of the relationship between stories and emotions.) Nanicelli, T. and P. Taberham (eds.) (2014) Cognitive Media Theory, New York: Routledge. (Important collection of essays on cognitive media theory and criticism.) Plantinga, C. (2009) Moving Viewers: American Film and the Spectator’s Experience, Berkeley: University of California Press. (On the means by which mainstream Hollywood films elicit emotion and the implications of this.) Shapiro, L. (ed.) (2014) The Routledge Handbook of Embodied Cognition, New York: Routledge. (Overview of developments in embodied cognition.) Shimamura, A. (ed.) (2013) Psychocinematics: Exploring Cognition at the Movies, Oxford: Oxford University Press. (Foundational essays on cognition in movies, including discussions of perceptual and affective processes.)

CHAPTER

8

Mirror Neurons and Simulation Theory A Neurophysiological Foundation for Cinematic Empathy DAN SHAW Chapter Overview One of the most influential theories of our emotional responses to fictional characters in film is simulation theory, which is urged by several aestheticians and philosophers of film. Its chief competitor is a cognitivist theory proposed by Noël Carroll and defended by Carl Plantinga, who have both argued that we sympathize with such characters because of cognitive appraisals that are primarily moral in nature. The thesis of this chapter is that recent developments in neurophysiology provide convincing evidence for some version of simulation theory. The discovery of the existence and emotive function of mirror neurons confirms that we simulate other people’s emotions in a variety of ways, even in cinematic contexts. This is made possible by these components of the human brain firing identically when we observe others engaged in an activity (or exhibiting a facial expression or body language) and when we do so ourselves. The mirror neurons form shared circuits with the emotive centers of the brain, triggering emotional reactions immediately and without any cognitive appraisals being involved. One of the most exciting discoveries in recent neuroscience is the identification of mirror neurons in monkeys, unique brain cells that fire identically whether one is engaging in an activity or watching someone else do so. Many neuroscientists now believe that mirror neurons lead human beings to anticipate the actions associated with an object (e.g., the tendency to grasp and/or 148

Cinematic Empathy  •  149

chew when one sees someone grab something and eat it), and also allow us to feel what other persons are feeling. In the last two decades, mirror neurons have been linked to a wide variety of emotional phenomena. This chapter will argue that evidence for the existence of mirror neurons provides a compelling reason for preferring simulation theories of our emotional response to characters in narrative films over cognitive judgment theories. While cognitivist accounts generally privilege conscious processes, and situational preconditions for feeling such emotions as sympathy, simulation theories highlight our precognitive tendencies to mirror the feelings of others, mimic their expressive behavior, and empathize with them thereby. Accordingly, the chapter will be divided into three sections. Section 1 will give a brief overview of the scientific research into the existence and nature of mirror neurons in the last twenty years. Section 2 will survey several rival accounts of our emotional response to cinematic fictions. Finally, in section 3, I will argue that the existence and function of mirror neurons in human beings provide a neurological foundation for simulation theories of our emotional experience of motion pictures. The controversies this chapter will address are far from trivial. Do we truly empathize with fictional characters in films? If so, what does that mean, and how are such shared emotions engendered? Why on earth would we ever want to feel with such characters, and what makes the cinema so effective in causing us to do so? The question throughout is whether recent developments in neuroscience have implications for aesthetic theorizing about films, and for philosophical theories of the emotions. I will be using the sense of the term “empathy” that Kendall Walton recently proposed: Let’s count as instances of empathy cases in which a person experiences a target as feeling like this . . . without requiring that she say to herself anything like “he feels like this”. This will help to accommodate what has sometimes been called “automatic empathy”, empathy based on emotional contagion, for instance. (Walton 2015: 40) What Walton calls “automatic empathy” is what I will refer to as “basic” empathy in what follows. Marco Iacoboni, one of the premiere researchers in this area, ably states my thesis here: “We have empathy for fictional characters—we know how they are feeling—because we literally experience the same feelings ourselves. And when we watch the movies stars kiss on-screen? Some of the cells firing in our brain are the same ones that fire when we kiss our lovers” (Iacoboni 2009: 33). According to Iacoboni, empirical tests from various sources “. . . strongly support the hypothesis that we understand the mental states of others by

150  •  Dan Shaw

simulating them in our brain, and we achieve this end by way of mirror neurons” (Iacoboni 2009: 34). So, what are mirror neurons, and how are they crucial to the simulation processes that allow us to empathize with others?

Mirror Neurons: Nature and Function In the early 1990s, neuroscientists in Parma, Italy, were monitoring neural firings in the brains of macaque monkeys with electrode implants. The team, headed by Dr. Giacomo Rizzolati, was seeking information on how the brain exerts motor control over the hand, in hopes of finding ways that humans with brain damage could recover at least some degree of manual function. In the process, they made an astonishing discovery: Certain brain cells of the macaque fired identically if the monkey was watching another grasp an object as when it was grasping the object itself. The idea that such cells would fire at the mere perception of another animal’s action exactly as they would when the animal engaged in the action itself was totally unexpected, and had profound implications. To their surprise, the researchers found that 20 percent of the cells in the motor cortex of the brain of the macaque functioned this way, as mirror neurons. The discovery of the existence of such neurons shattered a paradigm of brain function that had been widely accepted for decades. On that view, perception and action were governed by separate brain areas, and cognition was somehow the “go-between” that allowed us to select and engage in appropriate motor behavior. The startling implication of the existence of mirror neurons was that the brain responds to the world in a far more holistic fashion than had hitherto been appreciated. Initially the focus of this research was on motor mirror neurons, and how they explain that our basic understanding of physical objects already involves the various actions we can take in relation to them, offering an example of a cognitive activity that is at least partially unconscious.1 But Iacoboni extended the study, and discovered that mirror neurons interact with the emotive centers of the brain (primarily the amygdala), and that the insula acts as a gobetween, facilitating such interactions. He contends that this allows mirror neuron systems to lead us to simulate the emotions of others, primarily in response to their facial expressions, body language, and tones of voice. Christian Keysers, in his book The Empathic Brain, calls these interactive systems “shared circuits.” Drawing on the work of early neuroscientist Donald Hebb, Keysers argues that neurons that fire together wire together, and that eventually they “. . . connect visual, auditory, somatosensory and premotor areas together because the brain has to plan actions based on what it sees, hears and feels. Empathy is then the inevitable consequence of Hebbian plasticity in these connections” (Keysers 2011: Kindle location 2799).

Cinematic Empathy  •  151

It is a cliché that monkeys are good imitators (“monkey see, monkey do”), and we humans surpass them in the degree and complexity of our imitative skills. These powers of imitation are crucial to understanding our fellow creatures and empathizing with them. While earlier accounts of this process assumed that it must be the higher cognitive powers of monkeys and humans that enabled us to do so, they were mistaken: Wouldn’t we expect that in order to imitate an observed facial emotional expression, we first should be able to recognize it? Only if we assume that deliberate, explicit recognition must necessarily precede mimicry. Our only proof of this is an age-old theory, and the existence of mirror neurons provides an alternative explanation: the idea that the mimicry actually precedes and helps the recognition. (Iacoboni 2009: 111) Motor mirror neurons cause us to pre-reflectively mimic the actions of others. The “shared circuits” discussed earlier do the same thing for facial expressions, and the feelings they evince. In a development that seems to confirm the contentions of William James, many neuroscientists now believe that we take on the expression and then feel the emotion, rather than the other way around. Following Charles Darwin’s lead, James argued that emotions are caused by the physiological changes with which they are associated, rather than the emotion causing those changes. Iacoboni describes the process as follows: . . . mirror neurons provide an unreflective, automatic simulation (or “inner imitation,” as I have sometimes phrased it here) of the facial expressions of other people, and this process of simulation does not require explicit, deliberate recognition of the expression mimicked. . . . mirror neurons send signals to the emotional centers located in the limbic system of the brain. The neural activity in the limbic system triggered by these signals from mirror neurons allows us to feel the emotions associated with the observed facial expressions—the happiness associated with a smile, the sadness associated with a frown. Only after we feel these emotions internally are we able to explicitly recognize them. (Iacoboni 2009: 111–112) A few important points to note here: (1) Iacoboni speaks of “inner imitation” here, so explicit external mimicry of the facial expression is not necessary for emotional contagion; (2) mirror neurons connect with the emotive centers of the brain (primarily in the limbic region), and trigger the associated emotions (presumably by getting the neurons in that region to fire in the appropriate

152  •  Dan Shaw

fashion); and (3) this is what allows us to feel the emotions associated with the facial expressions in question. He has confirmed these findings many times over the course of his empirical research. For Iacoboni, the implications of this hypothesis are nothing short of revolutionary: These results clearly supported the idea that mirror neuron areas help us understand the emotions of other people by some form of inner imitation. According to this mirror neuron hypothesis of empathy, our mirror neurons fire when we see others expressing their emotions, as if we were making those facial expressions ourselves. By means of this firing, the neurons also send signals to emotional brain centers in the limbic system to make us feel what other people feel. (Iacoboni 2009: 119) Empathy is hence the most natural thing in the world, and crucial to the success of social creatures like monkeys and human beings. “This simulation process is not an effortful, deliberate pretense of being in somebody else’s shoes. It is an effortless, automatic, and unconscious inner mirroring” (Iacoboni 2009: 120). Our understanding of what others feel is grounded in our feeling the same way. Now, some caveats need to be registered at this point. Given that not all emotions can be read in a person’s face, and given that emotions are not always easy to distinguish by their feeling alone, Iacoboni’s account of the way mirror neurons support empathy may well be limited to a particular range of (basic?) emotions. I will draw the distinction between such “basic” emotions (which can be felt directly and immediately) and more sophisticated ones a little bit later on. But first an example of the most basic form of empathy that I am describing. In the summer of 2012, my son Patrick and I were watching a weightlifter at the Olympics in London. As the athlete jerked an immense weight above his head, his shoulder dislocated, and he dropped the bar behind him and began writhing in agony. Patrick, who lifts weights himself, immediately grabbed his own shoulder, and groaned empathetically. I could clearly see that he was feeling the weightlifter’s pain. Iacoboni describes the process convincingly: at such times . . . our brain produces a full simulation—even the motor component— of the observed painful experiences of other people. Although we commonly think of pain as a fundamentally private experience, our brain actually treats it as an experience shared with others. This neural mechanism is essential for building social ties. It is also very likely that these

Cinematic Empathy  •  153

forms of resonance with the painful experiences of others are relatively early mechanisms of empathy, from an evolutionary and developmental point of view. (Iacoboni 2009: 119) Such a capacity would clearly be of value in the struggle for survival and ascendency in social groups. It is socially appropriate to reflect the sad look of someone relating her sufferings to us, and we naturally do so because of the mimicry triggered by our mirror neurons. We form social bonds with our fellow humans through such moments of empathic response, and these bonds are evolutionarily adaptive. The conclusions that Iacoboni reached have two other possible types of empirical confirmation. If mirror neurons function as he claims, then (1) people who are particularly empathetic should have a higher concentration of mirror neurons, and highly facilitated pathways between their mirror neurons and the emotional centers of the brain; and (2) people who are deficient in the ability to feel empathy (like victims of autism) should also be deficient in mirror neurons and/or such facilitated pathways. Research into (1) has been conducted for some time, trying to establish what Keysers called “shared circuits.” Unfortunately, evidence cannot be collected by surgically implanting electrodes in people’s brains, and must be gathered by fMRI imaging, a far less exact measure. For this reason, studies that seek to correlate mirror neuron activity with social competence and empathy are far from definitive. What fMRI imaging has shown is (as Keysers predicted) that areas of the human brain in which mirror neurons are thought to exist fire simultaneously with segments of the insula and of the amygdala when empathic subjects are exposed to photos of the faces of people who are clearly expressing certain emotions (Iacoboni 2009: 117). Evidence for reduced mirror neuron activity in persons with autism is clearer, in studies like the one conducted by Oberman and Ramachandran at the University of San Diego in the mid-2000s (Oberman and Ramachandran 2007: 310–327). Now, it may be useful at this point to distinguish between a more “basic” form of empathy, which results unconsciously from immediate mimicking tendencies that can be traced back to the activity of mirror neurons, and a more reflective form of empathy, which, at its most complex, involves the conscious activity of the imagination running a detailed simulation involving complex cognitive components. But these are both “simulations” in the sense of the Merriam Webster online dictionary’s second definition of the term: Both offer an imitative representation of the functioning of one system or process by means of the functioning of another. Both give rise to legitimate emotional states that need to be recognized as such.

154  •  Dan Shaw

The distinction can roughly be drawn as follows: Basic empathy results from an immediate process triggered by our mirror neurons, primarily in response to human faces. These neurons fire identically when we are in the emotional states that are being expressed as when we observe such expressions on the faces of others. This allows us to “feel with” the other person, by arousing similar emotions within us. Developing a more complex and reflective sense of empathy surely involves what the cognitivists call “appraisals,” when, for example, we recognize more and more similarities to us in the fictional characters with whom we most strongly identify, or when they gain our emotional allegiance by doing good deeds (since the most significant appraisal of such characters is moral in nature). But what I have called “basic empathy” is a genuine “feeling with” fictional cinematic characters, by virtue of our immediate neuronal response to their facial and bodily expressions. In such cases of “basic empathy,” I wish to emphasize the automatic nature of our responses based on mirror neuron firings, and the way those responses occur even (as in the case of the still photographs) when we lack any information about the situation of the person whose emotional expression we are registering. We feel the fear of the people fleeing from Indominus rex in Jurassic World (Colin Treverrow, 2015) because the looks on their faces and their behavior immediately give rise to a fearful reaction in us. Cognitivists like Carl Plantinga might inject at this juncture that there is no real character to “feel with” in the fictional context. But my point is that if actors are convincing enough to trigger our mirror neurons, it is because they successfully simulate the outward signs of emotion (unless they are successful method actors, in which case they actually feel the emotions from within themselves and then evince them automatically). Having explained (all too briefly) the nature and function of mirror neurons as I understand them, let me turn to the controversies in philosophical aesthetics to which they are relevant.

Identification, Criterial Prefocusing, and Simulation Theory Uncritical filmgoers speak all the time about being able to identify with many of the fictional characters that people our cinematic experiences, noting how they often feel for or feel with such figures. One of the major “good-making characteristics” that viewers cite when explaining why they like certain films is the presence of characters with whom they were able to identify. But, as Noël Carroll highlights in his case against audience identification, the emotions the audience feels are not often identical to what the character is feeling at the time (more on this soon).

Cinematic Empathy  •  155

Kendall Walton ignited the contemporary philosophical dialog on the issue of our emotional response to films with his groundbreaking essay “Fearing Fictions,” which appeared in the Journal of Philosophy almost four decades ago (Walton 1978). In his view, fictions (both filmic and otherwise) provide us with a set of props that we utilize in a game of make-believe. We don’t experience real emotions in response to fictional characters, but quasi-emotions caused by us making believe that we are in those characters’ situations. There is no paradox involved in our feeling real emotions for fictional characters, because we simply don’t do so. Walton’s “make-believe account” has faced substantial criticism from those who contend that we do feel genuine emotions for fictional cinematic characters. The dialog really took off in the mid- to late 1990s, where two schools of thought developed over how to account for this apparent fact, a bifurcation that still dominates the conversation to this day. In “Film, Emotion and Genre” (Carroll 1995), Carroll argued that a good film can evoke real feelings of sympathy for its characters, by virtue of a process that he calls “criterial prefocusing.” This is achieved through the film’s skillful use of cinematic and narrative techniques to direct the audience’s attention, shape its attitudes, and thereby prime the audience for having particular emotions. Carroll argued that emotions necessarily involve cognitive components like beliefs. My fear of I. rex is, in this sense, genuine (despite the fact that I disbelieve in its existence) because it involves both feelings and thoughts about how dangerous it is that are prompted by the movie. I believe, on the contrary, that emotional reactions like fear can be accounted for without involving any kind of cognition—the characteristic affect, aroused at the neuronal level by my observing the reactions of the on-screen fictional characters to its threat, is enough to trigger genuine fear of the I. rex monster without having to be fully appraised of the situation. As noted earlier, Carroll’s primary grounds for rejecting traditional notions of audience identification had to do with the asymmetry between the emotions characters feel and those the audience feels simultaneously. One of his favorite examples is of a presidential candidate who doesn’t know he is about to be assassinated, who does not feel the suspense that the audience does while watching the attempt. In Carroll’s opinion, we feel sympathy for these characters despite all the noted differences, generally because we think of them as “good guys.” Films can hence cause us to sympathize with, and root for, the “good guys” by their skillful adherence to tried and true cinematic conventions. In the aforementioned essay, Carroll proposed a cognitive judgment theory of emotion, based on the claim that there are certain objective states of affairs that predictably cause certain feelings. When we, for example, judge that our children are

156  •  Dan Shaw

being threatened by the neighbor’s Doberman pinscher, we become angry at the dog. He argues that a necessary condition for a particular feeling state to qualify as a genuine emotion is that it be caused by at least one of a range of appropriate cognitive states (Carroll 1995). Certain types of situations give rise to certain kinds of beliefs, and those beliefs help generate the appropriate emotional responses. Someone claiming to be in an emotional state that does not fulfill the appropriate cognitive preconditions is, in his view, simply deluded. Carroll distinguishes between fleeting feelings and reactions, and what he calls “emotions proper” (or “garden variety emotions”), which must involve cognitive components that cause us to feel them (Carroll 1995). The notion that we first bodily react to things and then recognize the emotions we feel thereby gets the causal order backwards, in Carroll’s opinion. I will return to this crucial claim. If the emotions, as Carroll argues, are governed by conditions of appropriateness, then filmmakers must draw upon these conditions to arouse our emotional response. Unlike our experience of everyday life, films are organized self-consciously in ways designed to direct our perceptual attention: Thus, a film text can move us emotionally by being criterially ­prefocused—that is, by being so structured that the descriptions and depictions of the object of our attention in the text will activate our subsumption of the relevant categories and events under the categories that are criterially apposite to the emotional state in question. (Carroll 1995: 30) Films arouse our passions by leading us to adopt pro or con attitudes toward characters and the possible scenarios that they face. We are led to appraise the characters the way the film would have us appraise them. We sympathize with characters that are shown in a positive light (like George Bailey in It’s a Wonderful Life; Frank Capra, 1946), and root for the success of the projects in which they are engaged. For Carroll, then, the task of the good critic is “. . . to determine which features of the film are designed to engender pro attitudes in viewers, along with determining what those pro attitudes are. By following this procedure, one can pith the emotional structure of the film” (Carroll 1995: 33). If successful, that structure will be seen to satisfy the “necessary criteria” for being in the emotional state the filmmakers want to elicit in the audience. Film genres (like melodrama, horror, or suspense) can be distinguished by the emotions they typically seek to arouse. In all such cases “. . . the audience’s faculties of cognition and judgment are brought into play in the process of eliciting an emotional response to film” (Carroll 1995: 47).

Cinematic Empathy  •  157

Predictably, Carroll discounts the purported implications of the discovery of mirror neurons in one of his more recent books, Art in Three Dimensions. In his view, theorists such as Alessandro Giovannelli and Marco Iacoboni “. . . write as though mirror reflexes and related patterns of mirror neuron firings are tantamount to empathizing. But if empathy is an emotion, that seems unlikely, since these reflexes don’t involve appraisals, whereas emotions do necessarily” (Carroll 2010: n. 242). As we have seen, these “appraisals” are primarily moral in nature, although they can also involve the recognition of similarities that ground our imaginative identification with particular cinematic characters (e.g., that he was also an only child, unpopular in high school, somewhat overweight, etc.). Criticisms of Carroll’s account have been leveled from several quarters. The most strident critics defend a robust sense of empathy for fictional cinematic characters, some of them (like Berys Gaut 2010) on the basis of simulation theory. In her excellent overview of empathy theory in The Routledge Companion to Philosophy and Film, Amy Coplan sketches out the philosophic landscape here perfectly: . . . cognitive film theorists and philosophers of film have focused primarily on modes of character engagement that involve high-level processing. This is in part a reaction to what they regard as an overemphasis in psychoanalytic film theory on unconscious processes and in part a consequence of their theoretical frameworks. The pattern is beginning to shift, as a small number of cognitive film theorists and philosophers have begun to take seriously less sophisticated ways of responding to characters, such as motor mimicry, affective mimicry, low-level simulation, mirror reflexes and emotional contagion. (Coplan 2012: 104) She usefully groups the various forms of response cited at the end of the foregoing quote under the broad category of mirroring processes. For Coplan, mirror neurons play a crucial role in the affective matching that is a necessary, but not a sufficient, condition of empathy. Since emotions like empathy, in her view, result from complex processes that involve both cognitive and affective elements, “empathy proper” involves three distinct necessary conditions: affective matching, other-oriented perspective taking, and self-other differentiation (Coplan 2012: 6–9). Like Carroll, Coplan would rule out what I have called “basic empathy” (or affective matching) from qualifying as a form of “empathy proper,” because it doesn’t fulfill the second and third necessary conditions for being such that she proposes. (See Plantinga’s chapter in this volume for a more detailed discussion and critique of Coplan’s account of empathy.)

158  •  Dan Shaw

One of the fundamental forms of such mirroring processes is emotional contagion, defined by several prominent psychologists as “the tendency to automatically mimic and synchronize expressions, vocalizations, postures and movements with those of another person, and consequently to converge emotionally” (Coplan 2012: 105). This explains why watching a film is far more contagious than reading a work of literature, because a substantial portion of our emotional response to the movies is triggered by direct sensory stimulation. Literature can be just as criterially prefocused as film (providing us with all the appropriate narrative and literary prompts for how to respond emotionally), but it cannot be nearly as contagious, in part because there are no facial expressions for us to mimic and be infected by. As noted earlier, Carroll contends that the primary variable that influences our pro or con attitudes toward characters is moral. In his view, when we do sympathize with apparently evil characters, it is because of the compensatory virtues they possess (e.g., Hannibal Lecter’s intelligence, aesthetic taste, sense of humor, concern for Clarice Starling, etc.). This is a cognitive process of moral evaluation, through which a character either wins our sympathy and allegiance or fails to do so. I believe it is fair to say that most recent attempts to account for our empathy for cinematic characters do so by claiming that we simulate the emotions of those characters, adopting their perspective, and experiencing their world as we would if we were in their position. At the heart of such simulations are not cognitive evaluations but mirroring processes that lead our emotional states to converge with those of the characters. This trend concerns Plantinga; as he notes, “One could make the case that discussions of the body and affect have displaced examination of the higher cognitive functions in much of film studies” (Plantinga, this volume). Simulation theory gets around many of the problems that bother Carroll about the notion of identification. We do not totally lose ourselves in such simulations, nor do we have to be literally identical to the subjects of our empathy in order to identify with them. We do feel some emotions that are different from what the on-screen characters feel, especially when we have knowledge (e.g., of a pending assassination attempt) that they do not share. But our mirror neurons fire identically when we watch fictional characters show pain, grab objects, or face situations of peril as when we do so ourselves, causing us to feel the same emotions. To cite a common example, when we respond to the look on a fictional character’s face who is about to have a serious auto accident on-screen, we feel his terror (though that feeling may be mitigated by our appraisal of whether he deserves to die). In such cases, we may, or may not, physically mimic their actions and facial expressions; in other, more complex situations, we may run a detailed psychological simulation that involves appraisals of various sorts. Both processes lead

Cinematic Empathy  •  159

us to feel emotions that are similar enough to the ones the characters are feeling to say that we empathize with them. Empathy, in my view, is also the basis of what can legitimately be called our emotional identification with fictional cinematic characters. Neither camp in this controversy denies the relevance of the element that the other side highlights in its account of our feeling for films. Carroll has come to admit that contagious mimicry affects the emotional intensity of our sympathy, and simulation theorists (like myself) acknowledge that cognitive processes have a significant role to play, especially in detailed psychological simulations. What the two camps primarily differ over is what qualifies as “having an emotion,” as well as which set of elements is primary in determining our most basic emotional responses to cinematic characters. The upshot of my analysis thus far is that we naturally tend to empathize with our fellow humans (and their cinematic counterparts), without this matching of human feelings having to be caused by any conditions of appropriateness other than the possession of a human face. The growing consensus about the existence and function of mirror neurons disconfirms the categorical claim of most cognitivists, and suggests that at least some genuine emotional reactions to the cinema do not involve a cognitive component.

Cognitivism Refuted Jesse Prinz has captured the distinction sketched earlier in the most succinct of terms: “. . . a cognitive theory of the emotions is a theory that maintains that all true emotions involve cognitions essentially. Noncognitive theories maintain that emotions do not necessarily involve cognitions” (Prinz 2003: 69). Notice that only one of these claims is categorical. While cognitivists contend that all “true” emotions have an essential cognitive component, non-cognitivists are willing to admit that some of our emotional responses to cinematic characters do involve cognitive appraisals, while claiming that at least some do not. I have been arguing that some of our most basic empathic responses to cinematic characters are immediate and intuitive. They are non-cognitive affects caused by mirror neurons, which give rise to emotions independently of the appraisal processes cognitivists privilege as necessary conditions to the feeling of an emotion. The implications for philosophy of film are striking. As Plantinga himself observed over a decade ago, the human face is the primal scene of cinematic empathy (Plantinga 1999: 156). As he put it then, “Viewing the human face can move beyond communication to elicit an emotional response in the viewer” (ibid.: 242). He also noted that “. . . we have a tendency to ‘catch’ others’ moods and emotions,” and that we “. . . sometimes mimic the facial expressions of people we see on film and video” (ibid.: 253). Most significantly, for my purposes,

160  •  Dan Shaw

Plantinga acknowledged that “Today most theorists agree that our subjective emotions are influenced by facial feedback, such that one who mimics a facial expression actually catches the emotion of the one mimicked” (ibid.). Yet, in the accompanying article, Plantinga defends the cognitivist claim that all “emotions proper” involve appraisals, because they are fundamentally what he calls “concern-based construals” (this volume). Jealousy is a particularly apt example for Plantinga to cite, since it involves construing one’s beloved as being sexually interested in another, or another in them. This construal requires us to mobilize a variety of cognitive beliefs. But empathy can be much more direct. In Hannibal (Ridley Scott, 2001), for example, we witness the glee that Hannibal takes in serving up the brains of Starling’s nemesis, and the hungry look on her face as she asks (like a Dickensian waif), “Please, may I have some more?” They infect us with their anticipation, as those of us who love horror find ourselves enjoying the prospect as well. I’ll never forget taking my extremely nonviolent girlfriend to see A Clockwork Orange when it first came out in 1971. Expecting she would be appalled, I found her to be enraptured with little Alex and his strikingly aesthetic brand of ultraviolence. Close-ups of our humble narrator are prominently featured from the beginning, contributing to our surprising identification with a savage criminal. This, in turn, allowed us to share his perverse exhilaration at his appalling acts. The hypothesis that the shared circuits that give rise to empathetic reactions are especially triggered by facial expressions explains why so much of the compelling emotional power of classical Hollywood cinema derives from its characteristic use of facial close-ups. One of my favorite examples is Casablanca (Michael Curtiz, 1942), with its long, lingering shots of a tearful Ingrid Bergman and a tortured Humphrey Bogart. Blown up onto the big screen, their facial expressions cause us to have a strong tendency to identify with them, at least in the sense of sharing their visceral torments. As alluded to earlier, this theory also explains how we can empathize with such evil film characters as Hannibal Lecter. Most audiences were introduced to Lecter in The Silence of the Lambs (1991), where director Jonathan Demme used an extensive series of shot–reverse shot close-ups to depict the dialogs between him and Clarice, which constitute the dramatic and emotional heart of the film. These shots help secure our empathy for both characters, apart from our moral attitudes toward them. Despite their obvious moral differences, we end up rooting for both. Before making the cognitive appraisals of Hannibal’s intelligence, aesthetic tastes, or protective attitude toward Starling (in terms of which Carroll accounts for our sympathy for him), we are drawn in by the intensity and expressiveness of his face. Finally, the importance of acting quality is intriguingly accounted for by this approach as well. Good actors can convincingly simulate facial expressions (and body language, and tones of voice) that convey genuine emotions; bad

Cinematic Empathy  •  161

actors simply don’t get our mirror neurons to fire in the appropriate manner, and are hence unable to trigger facial mimicry and/or emotional contagion. Take, for example, an actor trying to fake a smile. As Christian Keysers has noted, doing so is notoriously difficult. He explains why: The answer lies in the fact that the brain areas that control the voluntary movement of your face are different from those that cause the emotional generation of facial expression. The premotor cortex and primary motor cortex . . . are part of the voluntary motor system. If you fake a smile without generating the emotion, you use these two cortical areas. I term this system the “cold” facial expression system, because it does not require the heat of emotions. (Keysers 2011: location 1925) Keysers contrasts this system with the one that produces involuntary emotional behavior, which he designates as “. . . the ‘hot’ motor system, because it transforms the heat of emotional affect into observable behavior of the face and body” (ibid.). Faking a smile requires us to self-consciously mimic facial movements that usually occur automatically. No wonder it is hard to do so convincingly. We smile with our eyes as well as our mouths, and an effective actor must learn to do both. To conclude, I have argued that we feel genuine emotions in response to cinematic characters, and that evidence for the nature and function of mirror neuron systems tends to confirm the account of such feelings offered by simulation theory. This should lead us to question the cognitivist contention that all “true” emotions must involve a substantial cognitive component. If viewing the human face can, of itself, elicit an emotional response in the viewer, and if facial mimicry can directly cause us to “catch” the emotions of people on film, then cognition, while relevant to some of our emotional responses to film, does not play a crucial role in all such responses. Mirror neurons offer a plausible neurological foundation for the precognitive simulation account of our basic empathy for cinematic characters that I have offered here. The shared circuits that they form with the emotive areas of the brain also explain why close-ups are so engaging to audiences, and why good actors can be so emotionally compelling. It turns out that Norma Desmond (Gloria Swanson) in Sunset Boulevard (Billy Wilder, 1950) showed profound insight when she defiantly claimed about the great silent film stars that “We didn’t need dialog. We had faces!”2

Notes   1 I agree with what Plantinga contends in the accompanying piece, that the cognitive/noncognitive distinction is not coextensive with the conscious/unconscious realms. As my editor put it in a note on an earlier draft of this chapter, perhaps the distinction is not so much between conscious and unconscious processes as between ones that are more or less

162  •  Dan Shaw automatic, more or less open to reflection, and more or less dependent on beliefs and other mental states.   2 I wish to thank my editor, Katherine Thomson-Jones, for all her helpful suggestions, especially for pointing out where I had failed to prove what I claimed to have proven.

References Carroll, Noël (1995) “Film, Emotion and Genre,” reprinted in Carl Plantinga and Greg Smith (eds.), Passionate Views: Film, Cognition, and Emotion, Baltimore: Johns Hopkins University Press, 1999. ——— (2010) Art in Three Dimensions, Oxford: Oxford University Press. Coplan, Amy (2011) “Understanding Empathy: Its Features and Effects,” in Amy Coplan and Peter Goldie (eds.), Empathy: Philosophical and Psychological Perspectives, Oxford: Oxford University Press, 3–18. Coplan, Amy (2012) “Empathy and Character Engagement” in Paisley Livingston and Carl Plantinga (eds.) The Routledge Companion to Philosophy and Film, London: Routledge. Gaut, Berys (2010) “Identification and Emotion in Narrative Film,” Midwest Studies in Philosophy 34, 136–157. Iacoboni, Marco (2009) Mirroring People, New York: Macmillan. Keysers, Christian (2011) The Empathic Brain, Sydney: Social Brain Press, Kindle Edition. Oberman, Lindsay and Vilayanur Ramachandran (2007) “The Simulating Social Mind: The Role of the Mirror Neuron System and Simulation in the Social and Communicative Deficits of Autism Spectrum Disorders,” Psychological Bulletin of the APA, 133:2, 310–327. Plantinga, Carl (1999) “The Scene of Empathy and the Human Face on Film,” in Carl Plantinga and Greg Smith (eds.), Passionate Views: Film, Cognition, and Emotion, Baltimore: Johns Hopkins University Press, 239–256. Prinz, Jesse (2003) “Emotion, Psychosemantics and Embodied Appraisals,” in Anthony Hatzimoysis (ed.), Philosophy and the Emotions, Cambridge: Cambridge University Press, 59–86. Walton, Kendall (1978) “Fearing Fictions,” Journal of Philosophy, 75:1, 5–27. ——— (2015) In Other Shoes: Music, Metaphor, Empathy, Existence, Oxford: Oxford University Press.

Suggestions for Further Reading Carroll, Noël (2007) The Philosophy of Motion Pictures, Hoboken, NJ: Wiley-Blackwell. (A comprehensive overview and update of Carroll’s philosophy of film, including his account of sympathy.) Coplan, Amy and Davies, David (2015) Blade Runner (Philosophers on Film), London: Routledge. (Coplan on mood and meaning and Gaut on empathy in Ridley Scott’s film.) Currie, Gregory (2003) Recreative Minds: Imagination in Philosophy and Psychology, Oxford: Clarendon Press. (An influential account of the branch of simulation theory that foregrounds imaginative activity.) French, Peter and Wettstein, Howard (eds.) (2010) Film and the Emotions, Hoboken, NJ: WileyBlackwell. (An excellent anthology representing a diverse set of approaches to our emotional response to cinema.) Gaut, Berys (2009) Art, Emotion and Ethics, Oxford: Oxford University Press. (Disentangles the issues surrounding emotional responses and their relation to ethical appraisals.) Plantinga, Carl (2009) Moving Viewers: American Film and the Spectator’s Experience, Berkeley: University of California Press. (Informed by the latest work by philosophers and psychologists about the emotions, Plantinga offers his clearest account of spectator experience here.)

PART

V

Can Films Philosophize?

This page intentionally left blank

CHAPTER

9

Film as Philosophy The Pro Position THOMAS E. WARTENBERG

Chapter Overview This chapter discusses the debate among philosophers and film theories as to whether films have made, and continue to make, significant contributions to the discipline of philosophy. In presenting a defense of the notion of cinematic philosophy, the chapter canvases a number of objections that have been raised against this notion. In the course of this defense, a number of films that actually do philosophy are discussed. Prior to investigating the objections, the notion of film as a medium for doing philosophy is proposed and supported. The central objections addressed are: a. The generality objection, that asserts that films can never be general in the way that philosophical claims and theses are; b. The explicitness objection, based on alleged differences between film as an artistic medium and philosophy as an intellectual endeavor; c. The imposition objection that attributes the philosophy done in an interpretation of a film to the interpreter rather than the film; d. The “no author” objection, according to which the plurality of filmmakers undermines the attribution of the doing of philosophy to the film. Finally, the idea of a moderate pro-cinematic philosophy position is clarified and defended.

165

166  •  Thomas E. Wartenberg

Under the rubric of film as philosophy, film-philosophy, filmosophy, and cinematic philosophy, film theorists and philosophers have been arguing that films have made, and continue to make, significant contributions to the discipline of philosophy. Let me here just cite Stephen Mulhall’s On Film (2008), in which he argues, among other things, that the films in the Alien Tetralogy contribute to our understanding of what makes one film a sequel of another film, and my own Thinking on Screen: Film as Philosophy (2007), in which I outlined a number of different ways in which films have made contributions to the discipline of philosophy. Although Mulhall and I both support the idea of films doing philosophy, there is a significant difference in our views. Mulhall claims that films do philosophy “in just the ways that philosophers do” and specifies only that philosophizing involves “serious and systematic thinking” about important philosophical issues (Mulhall 2008: 4). I characterize Mulhall’s view as endorsing the extreme pro-cinematic position in regard to films doing philosophy, a position he shares with Giles Deleuze (1986; 1989), Stanley Cavell (2004), and Daniel Frampton (2006), among others. My own view—the moderate pro-cinematic philosophy position—focuses upon specific techniques that filmmakers can employ to do philosophy on film, most centrally the thought experiment. So while I do see philosophers and filmmakers as employing a common technique to address philosophical issues, I am less sanguine about their doing so in just the same way, hence my characterization of my own view as moderate. Claims like Mulhall’s and mine have given rise to a good deal of controversy. There are many who find the claim that films can actually do philosophy preposterous, while other, less global critics of the notion of cinematic philosophy argue that there are strict limits to what films can achieve philosophically. Murray Smith’s chapter in this volume exhibits both of these traits: Although he argues that films should be recognized as works of art and not philosophy, he does endorse a very limited sense in which certain films might do philosophy. In this chapter, I shall investigate this dispute, canvasing the views of both the advocates and the opponents of the idea that philosophy can be done on film in what I will call the film-as-philosophy debate. In the course of my defense of films doing philosophy—which will focus mainly upon the claims made by critics of the idea—I will discuss a number of films that I take to be doing philosophy. So my argument will not be entirely defensive: At the same time that I undermine the claims of the critics of cinematic philosophy, I will also examine some ways in which films do philosophy.

Film as a Philosophical Medium As a first step in my argument, it is important to clarify exactly the thesis that animates the cinematic philosophy debate. There has been a great deal of

Film as Philosophy: The Pro Position  •  167

confusion caused by the manner in which some of the advocates of cinematic philosophy have presented our case. What we have claimed is that the films we are interested in are actually doing philosophy. This has led critics of cinematic philosophy to assert that films themselves are incapable of doing anything and that only people can do philosophy. In order to clarify the issue here, let’s think about the traditional media for doing philosophy, acknowledging that people are indeed the only entities we know of that can actually do it. (I realize that some philosophers believe that texts are the entities that, in some sense, do philosophy, for the notion of an author of the text is no longer valid. Aside from indicating my disagreement, this is an issue I cannot go into here. But see Barthes [1977] and Foucault [1977].) There are two such media: oral discussion and written texts. To see this, consider the early Dialogues of Plato (2002), in which we encounter a portrait of Socrates doing philosophy by discussing issues such as the nature of virtue with a group of young men, his interlocutors. (Socrates is so portrayed in Raphael’s famous painting School of Athens.) Although we may have some problems with the manner in which Socrates conducts his ­discussions—he often appears an intellectual bully—they do provide an illustration of one way in which philosophy can be done: in a face-to-face oral discussion among individuals who are seeking understanding. So, oral discussion is clearly one traditional medium in which philosophy has been done, and many of us therefore use Plato’s Dialogues to teach students exactly what the practice of philosophy involves. Of course, no one would nowadays assert that oral discussion is the only medium for doing philosophy. The Western philosophical tradition has resorted to written texts as a means for doing philosophy since before the time of Plato. Each of the two traditional mediums for doing philosophy has its advantages. For example, the immediacy of an oral discussion has to be contrasted with the availability of a written text for repeated readings. And the advent of the Internet, with its potential to easily disseminate recorded oral discussions, may have altered our usual assessment of these two mediums. Still, the fact remains that there are two traditional mediums in which philosophy has been done. With this background, we now can clearly state the claim made by the advocates of films as doing philosophy: Film or the cinema, we say, is a medium for doing philosophy that should be placed alongside oral discussion and written texts. Our critics, those skeptical of the existence of philosophy in films, deny that film should be granted this status. But this raises another question: What exactly is the medium of film as discussed in the film-as-philosophy debate? Since my concern is with film as an artistic medium, I will not limit the film medium to a specific set of physical materials and tools, such as celluloid film stock that is edited by the physical splicing together of pieces of that stock. Rather, I conceive of the film medium in an inclusive manner, so that all uses of displayed moving images and sounds,

168  •  Thomas E. Wartenberg

however these are physically realized, are included. This means that video, television, and digital media are all different physical realizations of film in this inclusive sense. So the question of whether film can do philosophy is about the art form and not any of its particular physical exemplars.

Rejecting the “Easy” Solution Before turning to the actual film-as-philosophy debate, I want to focus my attention on what appears to be an “easy” solution to the question of whether the notion of film as philosophy is coherent. The claim is that it is obvious that there can be philosophy done on film. All you need to do is find a philosopher who is giving a lecture in which she is making an original claim and film the lecture. Clearly, it is maintained, when we watch the film of the lecture, we see someone philosophizing, so the resulting film is an example of cinematic philosophy and the whole debate is settled in the affirmative. Unfortunately, the debate cannot be settled in favor of film as philosophy so easily. Sure, the hypothetical film I have just described would show someone doing philosophy, but the general consensus among all the participants in the debate about cinematic philosophy is that this film would not count as an instance of philosophy done on film. The reason for this is that the philosophizing that is presented in the film actually took place extra-cinematically, at a real place and time. While the film’s recording of those events disseminates them to a wider audience than actually witnessed them, the film itself has not contributed to the philosophy that was actually done. If this is correct, then we need to ask, “What must a film do in order to be at least a candidate for being an instance of cinematic philosophy?” This is a question that is best answered not by a general discussion of the nature of the film medium and how it can contribute to philosophy, but through a detailed examination of individual films and what their philosophical contributions are. The question of how films do philosophy cannot be settled a priori, but only through a consideration of what films have done philosophically. In order to get a better sense of how the film medium does philosophy, it will be helpful to look at some actual films that have philosophical content but that don’t qualify as instances of cinematic philosophy. For example, in Eric Rohmer’s A Tale of Springtime (1990), Jeanne (Anne Teyssèdre), a philosophy teacher at a lycée, gives a pretty accurate summary of some of the basic claims of The Critique of Pure Reason (1999) as part of a general discussion of her teaching practices. But I don’t know that anyone would want to claim that just having someone summarize the Critique qualifies a film as a work of cinematic philosophy. A better candidate for cinematic philosophy is a scene from Waking Life (Richard Linklater, 2001). In it, we see the late philosopher Robert Solomon

Film as Philosophy: The Pro Position  •  169

discussing the contention that existentialism is a philosophy of despair. He argues against this claim on two grounds: first, that Sartre denied it; second, that the realization that one controls one’s own destiny should result in exhilaration and not despair. Certainly, these are interesting claims and the sort of thing that one might find in a philosophy text, but I still don’t think that we should count this as an instance of cinematic philosophy. The reason is that, despite its use of rotoscoping (a technique that transforms a photography-based film into an animated one), Waking Life here, and throughout the film, is doing no more than recording philosophy being done orally. Solomon is shown giving a lecture to a class that the film records. So the medium in which philosophy is being done is Solomon’s lecture—an example of oral philosophy—and not the film itself, which, because of its technological capacities, only records and transmits that lecture to people who were not present at it. But that is insufficient for characterizing the film as an instance of cinematic philosophy. I can imagine someone responding that, for all that, Waking Life does get its audience to think about a philosophical issue, say, the meaning of life, by means of its narrative about someone seeking answers to that question. Doesn’t this qualify the film as cinematic philosophy? Once again, we find a general consensus among both advocates for and critics of the notion of film as a medium for philosophy that films can raise philosophical questions for their audiences. The question at issue in the debate about cinematic philosophy is what more films can do philosophically. Some critics of film’s doing philosophy, like Bruce Russell (2000), argue that there are strict limits to what the philosophical contributions of film can be. In addition to raising philosophical questions, the only thing films can do philosophically, according to Russell, is to present narratives that can be used by philosophers as counterexamples to philosophical theses. To understand the rationale behind the claims of critics of cinematic philosophy like Russell, it is useful to investigate some objections to the notion of cinematic philosophy, for considering them will help us think about the issue in a more general and systematic light.

Objections to the Idea of Cinematic Philosophy In this section of the chapter, I consider four general objections that have been made to the possibility of cinematic philosophy. In each case, I will argue that the objections are misguided. The Generality Objection The first objection focuses on an important difference between film and philosophy. Specifically, the generality objection emphasizes the narrative

170  •  Thomas E. Wartenberg

character of fiction films, in contrast to the universalistic aspirations of philosophy. While a film like Woody Allen’s Crimes and Misdemeanors (1989), the objection maintains, tells us the story of a person who attains happiness despite acting in a reprehensible manner, philosophy is concerned not with such individual cases but general truths, such as Plato’s claim that virtue is necessary for happiness. The objection concludes that there is an unbridgeable gulf between narrative fiction film and philosophy. There are a number of ways in which one might respond to the generality objection. One could deny that philosophy involves universal claims. In defending the idea that fictional literature is a medium for doing philosophy, Cora Diamond (1991) has emphasized the transformative effect of philosophy, its ability to get people to see the world in a different, unfamiliar way. When Socrates presents his view of the state in the Crito (Plato 2002), Diamond asserts, he is trying to get his interlocutors to reconceive of the nature of the state, to fundamentally alter their understanding of their relationship to it. This is something that cannot be achieved by a deductive argument from agreed-upon premises to a conclusion. If we accept Diamond’s contention that this type of transformation is the true mission of philosophy, the teeth are removed from the generality objection, for its contention that philosophy involves the attempt to defend universal truths has been rejected. Alternatively, one could argue, as Catherine Constable (2009) does, that we should emphasize the common figurative nature of film and philosophy. There is no doubt that the texts of the great philosophers of the Western tradition are studded with literary tropes that play a role in elaborating their views. When Kant conceives of reason as a small island “surrounded by a wide and stormy ocean, the native home of illusion, where many a fog bank and many a swiftly melting iceberg give the deceptive appearance of farther shores, deluding the adventurous seafarer ever anew with empty hopes” (Kant 1999: 258), he is clearly using a figure of speech—a metaphor—to convince the reader of the need for a critique of pure reason. If one takes such figures of speech to play an essential role in philosophical argumentation—without which the explicit argumentation of a text like the first Critique would be incomplete—then the objection to cinematic philosophy based on an understanding of philosophy as inherently universal simply fails. Such contested understandings of philosophy fail to acknowledge the usefulness of the generality objection in pushing us to think about the ways in which film can be doing something like philosophy on the usual understanding of it. Here, I agree with Smith’s contention in his chapter in this volume that we need a less contentious conception of philosophy as the basis of the debate about cinematic philosophy. So the generality objection presents a challenge to the advocates of cinematic philosophy, the challenge of explaining how films can replicate philosophy,

Film as Philosophy: The Pro Position  •  171

traditionally conceived, a challenge that Smith contends cannot be met. My contention is that this challenge can be met by the advocates of cinematic philosophy by showing that philosophy, with its interest in establishing universal truths, employs a similar method to narrative fiction films: namely, the thought experiment. Philosophical literature from Plato onwards is replete with imaginary scenarios that play a role in the attempt to ground general claims. Think, for example, of Plato’s “Allegory of the Cave” (Plato 1992) or Judith Jarvis Thomson’s case of the famous violinist (Thomson 1971). These imaginary scenarios have a cinematic quality to them. So it’s not surprising that films can employ this particular philosophical technique. We even find Plato’s Allegory given a cinematic rendering in Bernardo Bertolucci’s The Conformist (1970), where it is used to ground a philosophical critique of Italian fascism. The presence of thought experiments in philosophical texts, then, disarms the generality objection. Philosophy, like narrative fiction films, includes narratives and stories. We will see the importance of this shared feature as we turn to the next objection to film as a medium for philosophy. The Explicitness Objection The explicitness objection shares with the generality objection a concern for the differences between film as an artistic medium and philosophy as an intellectual endeavor. Philosophy, it is maintained, proceeds by making explicit claims, such as the principle of utility’s assertion that an action or practice is justified if and only if it contributes more to the general welfare than any available alternative. Even if we think that there is some unclarity to this formulation of the principle, philosophers would respond by simply trying to remove the unclarity or ambiguity. So, philosophers have, for example, distinguished between act and rule utilitarianism in an attempt to disambiguate the principle of utility (Rawls 1955). Films, however, because they are works of art, are different, the objection maintains. When we find ambiguity in Shakespeare’s description of himself as “bare ruined choirs” (Sonnet 73), we are not inclined to remove it, but rather, as William Empson (1930) asserted some eighty years ago, to value that ambiguity as contributing to our understanding and appreciation of the poem. So what is anathema to philosophy—for example, ambiguity—is highly valued in art. And film, being a type of art, shares that characteristic. One way in which this claim has been developed is in regard to thought experiments, the feature I isolated as common to films and philosophy a moment ago. Murray Smith (2006) argues that cinematic thought experiments and philosophical thought experiments are used for fundamentally different purposes, so the presence of a thought experiment in a film does not entail that the film is using the thought experiment to make a philosophical point. Smith claims that philosophical thought experiments are very brief and

172  •  Thomas E. Wartenberg

sketchy in their elaboration of details, while cinematic thought experiments are full of rich details and are highly elaborated. What Smith takes this difference to show is that filmmakers and philosophers employ thought experiments for quite different purposes, and this vitiates the appeal to them as grounds for accepting the existence of cinematic philosophy. Although Smith is correct to characterize many philosophical thought experiments as sketchy and brief, particularly those in contemporary analytic philosophy of mind, not all philosophical thought experiments share those properties. Plato develops a very elaborate thought experiment in The Republic (1992) to show that a person who is immoral but has a reputation for being moral would nonetheless be unhappy. Brevity is thus not a necessary condition for a philosophical thought experiment. And even if it were, it’s not clear why a difference in length and detail should undermine the possibility of cinematic thought experiments doing philosophy. The one example of a brief philosophical thought experiment that Smith discusses is brief for a specific reason other than it being used philosophically. This thought experiment, proposed by Bernard Williams, asks us to consider the case of the emperor and the peasant, based on a similar case discussed by John Locke (1979). Here is the imaginary scenario of Williams’s thought experiment: Suppose a magician is hired to perform the old trick of making the emperor and the peasant become each other. He gets the emperor and the peasant in one room, with the emperor on his throne and the peasant in the corner, and then casts the spell. What will count as success? . . . (Williams 1973: 11) We are to imagine, roughly speaking, that the emperor’s mind is placed in the peasant’s body and vice versa. What’s unusual about this thought experiment is Williams’s claim that we can’t really carry it out. It is a failed thought experiment, one whose failure is intended to show that personal identity depends upon bodily identity. Here is what Williams says by way of explanation of the failure: The requirement is presumably that the emperor’s body, with the peasant’s personality, should be on the throne, and the peasant’s body with the emperor’s personality in the corner. What does this mean? In particular, what has happened to the voices? The voice ought to count as a bodily function; yet how would the peasant’s gruff blasphemies be uttered in the emperor’s cultivated tones, or the emperor’s witticisms in the peasant’s growl? A similar point holds for the [facial] features. (Williams 1973: 11–12)

Film as Philosophy: The Pro Position  •  173

Williams’s claim is that we can’t fill out the details of the case as described in the thought experiment except in an arbitrary way. The reason for this is his belief that the case violates an important criterion for personal identity— namely, bodily continuity—and, hence, cannot actually be carried out. The case cannot be specified in any more detail without undermining the claim Williams wants to make by means of it. Although Williams’s argument may convince some people that the thought experiment he proposes cannot be done, it is pretty easy to imagine a science fiction scenario in which the transfer of consciousness takes place. All that would be needed is for the creator of such a fiction to vividly present the imaginary case. While the choice of the case’s details—what sorts of accents the two transformed beings will have, for example—may be somewhat arbitrary, a filmmaker or novelist could provide a rendering of the thought experiment that would make it seem possible, contra Williams. As aestheticians have been at pains to argue, we can imagine many things without being able to specify the precise grounds for our imaginings (see, e.g., Walton 1990). What, then, of Smith’s claim that films and philosophy employ thought experiments for such fundamentally different purposes that it does not make sense to treat thought experiments in film as making philosophical claims or arguments? Smith makes his case by referring to what he claims is the same thought experiment, in a philosophical text—Bernard Williams’s mind transfer thought experiment I have just discussed—and in a film—Rob Reiner’s All of Me (1984), starring Steve Martin and Lily Tomlin.1 In the film, a rich woman with a terminal illness, Edwina Cutwater (Lily Tomlin), is planning to have her consciousness transferred into the body of a young and beautiful woman who has agreed to let her consciousness be dispersed over the universe as part of the transfer. (The details of how this is to happen are not important for our purposes, though the film does attempt to make it at least minimally plausible.) Through a glitch, Cutwater finds her consciousness co-inhabiting the body of the Steve Martin character, Roger Cobb, with his consciousness. Many comic antics follow in which we see Martin’s body being ruled by the consciousness of Cutwater or by her and Cobb simultaneously. Smith claims that, while Williams uses the consciousness transfer thought experiment to substantiate his claim that bodily continuity is necessary for personal identity, the film uses the identical thought experiment purely for comic purposes. The idea is that the film gets us to laugh at various situations that ensue once a woman’s consciousness cohabitates with that of a man in the man’s body. We are to conclude that the film employs the thought experiment purely for entertainment while the philosophy text uses it to establish a philosophical truth about personal identity. A first thing to note is that All of Me does not actually replicate the scenario that Williams presents because it involves the cohabitation of two different

174  •  Thomas E. Wartenberg

minds in one body, rather than the consciousness-switching scenario presented by Williams. So Smith is simply wrong when he says that it is the identical thought experiment in both cases. At the same time, the film does show that it is possible to imagine the scenario that Williams says we can’t by simply using one actor’s voice when the other actor’s mouth is moving to register the presence of the mind of the first actor in the body of the second. But what about the film’s presentation of the Cobb-Cutwater melded being? Is it simply an occasion for comic riffs, as Smith maintains, or might the film being doing something more with its admittedly comic antics? There is an important similarity between Williams’s claim that his envisioned thought experiment fails and the film’s presentation of the unsuitability of Cobb’s body for Cutwater’s consciousness. That is, the humorous antics of Martin’s parodic physical comedy demonstrate that a woman’s consciousness cannot find a suitable home in a man’s body. In so doing, the film makes a point related to the one that Williams makes, albeit about gender: All of Me presents gender identity as having a bodily component. By presenting Cobb’s attempts to act like a woman and Cutwater’s to act like a man, satirically using gender stereotypes, the film demonstrates that gender is registered in the body and is not just an idea or biological fact, for a woman’s consciousness cannot be transferred into a man’s body seamlessly. (For a fuller discussion of the film and the issue of personal identity, see Falzon 2015: ch. 2.) The theoretical point I wish to make by means of my admittedly brief discussion of All of Me is that the use of a cinematic thought experiment to entertain is not incompatible with it also having a philosophical function. Conversely, there are many authors of philosophical texts who stud their writing with puns and other word play. Do such attempts at humor vitiate the philosophical content of these works? Of course not. So why assume that a film’s goal of entertaining an audience entails that it cannot also establish a significant cognitive claim? The Imposition Objection The imposition objection is to my mind the most difficult one for the cinematic philosophy advocate to handle. The claim made by this objection is that, interesting as interpretations of films put forward by philosophers may be, it is a mistake to see these interpretations as supporting the notion that the philosophical insights so uncovered are actually present in the film itself. Instead, it is maintained, the insights are due to the philosophical interpreter, who is able to use the film as a vehicle for her own philosophical insights. Now it certainly is true that films can stimulate philosophers to make interesting claims. For example, Michelle Mason (2003) has developed an account of contempt as a moral attitude that takes as its point of departure a specific understanding of the origin of Camille’s (Bridget Bardot) contempt for her

Film as Philosophy: The Pro Position  •  175

husband Paul (Michel Picolli) in Jean-Luc Godard’s Contempt (1963). One might characterize Mason’s project as that of developing an account of contempt that validates the film’s presentation of it, at least on one understanding of the film. On this interpretation, the film presents Camille’s contempt as a justified response to Paul’s ignoble actions. But philosophers have not viewed contempt as an admirable moral attitude, so Mason develops an account of that attitude that shows when and why it is a justified moral attitude to have. But, despite its origin in the film, the theory that Mason develops is certainly nothing that one would want to claim was present in the film itself. Many philosophical interpretations of films, however, have a fundamentally different character in that their authors maintain that they are exploring the philosophy actually done by the film. The imposition objection denies that there is any difference. All philosophical interpretations of films involve philosophers taking off from films that can only inspire, but not actually do, the philosophizing. In my earlier work (2007), I responded to this objection by distinguishing between two different types of interpretation: those that are creatororiented and those that are audience-oriented interpretations. The former type of interpretation attempts to reconstruct the meaning that the author of a work intended; the latter type of interpretation is not restricted in this way. So if someone wants to investigate, say, gender relationships in High Noon (1952), to choose a film pretty much at random, that is certainly a legitimate interpretive project. But it would be an example of an audience-oriented interpretation of the film. The interpretations that lie at the heart of the cinematic philosophy debate are creator-oriented ones. Paisley Livingston has criticized my account of creator-oriented interpretation. In particular, he criticizes me for loosening the constraints on what the creators of a film might have had knowledge of in order for an interpretation to count as showing that the philosophy is being done by means of a film. He quotes me as saying, All that is necessary for a creator-oriented interpretation to be acceptable in this regard is that the creator might have been acquainted with the philosophical ideas, etc., because of, for example, their general circulation within a culture. (Livingston 2008: 5) To which he responds, The audio-visual evidence underdetermines the initial choice of philosophical positions, so the interpreter’s selection of philosophical background sources is insufficiently motivated in the absence of additional

176  •  Thomas E. Wartenberg

evidence pertaining to the film-maker’s actual sources and attitudes, and, of course, the features of the audio-visual display (such as the rhetorical patterns constitutive of its discursive coherence). (Livingston 2008: 6) I think that Livingston misunderstands my claim. I was arguing that a filmmaker need not be aware of the source for the philosophical question he intends his film to illustrate. At the time, what I was thinking was that an issue could be so widely disseminated within the culture that a filmmaker could present it without being aware that it was an issue that had been discussed by a particular philosopher. For this reason, I said that the filmmaker need not have been aware of the specific source of the philosophical claims or theories she addressed in her film. But I did not deny that the interpreter was not obligated to seek evidence for the attribution of such a philosophical concern to the filmmaker. Perhaps where Livingston and I disagree is on the significance of evidence internal to the film. Consider, for example, the claim that I made in Thinking on Screen, that The Third Man (Carol Reed, 1948) both illustrates and supplements some of the claims that Aristotle makes about loyalty to a friend in The Nichomachean Ethics. Holly Martins (Joseph Cotton) is a pulp novelist who comes to Vienna at the behest of his friend, Harry Lime (Orson Welles), only to discover that Lime has died as a result of being struck by a car. After attending Lime’s funeral, he accepts from Major Calloway (Trevor Howard), a British officer, both a ride back into town and a drink. As Martins gets drunk, he tells Calloway how much Lime meant to him: “Best friend I ever had,” he slurs, establishing his view of Lime as a true friend. Lime had, it seems, filled an important void in his life: “Never so lonesome in my life till he showed up.” Calloway’s response—“that sounds like a cheap novelette”—is rude and unsympathetic. As viewers, we are taken aback by what appears to be Calloway’s repeated baiting of Martins. When Martins opines that Lime’s death (apparently the result of being run over by a car) was “a shame,” Calloway retorts, “Best thing that ever happened to him,” and goes on to explain why: “He was the worst racketeer who ever made a dirty living in this city” (Greene 1988: 26). When Martins responds by conceding that Lime could have been involved in minor black marketeering—an admission we were prepared for by the film’s initial newsreel-like sequence in which Vienna was characterized as a city in which everyone dealt on the black market—Calloway counters that Lime’s illicit dealings involved murder and not just petty racketeering. When he then dismisses Martins as a “scribbler with too much drink,” Martins responds by accusing Calloway of not taking Lime’s death seriously, but only using it to solve crimes he is too lazy to investigate. This time, Calloway’s contemptuous response mocks Martin’s profession as an author of popular fiction: “Going to find me the real criminal? It sounds like one of your stories” (Greene 1988: 27).

Film as Philosophy: The Pro Position  •  177

It seems uncontestable that the film is here showing us that loyalty is one of the benefits of friendship, a benefit that becomes apparent in the face of a perceived slander to a friend. Martins takes Calloway to be slandering his dead friend, Lime, and he responds by demonstrating his loyalty to Lime in both words and actions. Now, it so happens that Aristotle makes this very claim about friendship in The Nichomachean Ethics. Furthermore, it is only the friendship of good people that is immune to slander. For it is hard to trust anyone speaking against someone whom we ourselves found reliable for a long time; and among good people there is trust, the belief that he would never do injustice, and all the other things expected in a true friendship. (Aristotle 1999: 1157a21ff.) So the film, it seems to me, illustrates a thesis of Aristotle’s. However, in order for this to be true, it does not have to be the case that the filmmakers—here, most centrally its screenwriter, Graham Greene—intended to illustrate Aristotle’s theory about loyalty to a friend. Rather, in portraying Martins’s response to what Martins takes to be Calloway’s slander of his friend, the film shows us the very phenomenon that Aristotle discusses in the passage I quoted. So, there is an ambiguity in my claim that the film illustrates Aristotle’s claim about loyalty to a friend. I could be taken—mistakenly—to be asserting that the filmmakers intend to illustrate a claim that they realize Aristotle articulated in The Nichomachean Ethics. That is not the claim that I intended to make. What I did want to assert is that the film illustrates a claim about the benefits of friendship that is one part of Aristotle’s account. The film presents us with an instance of loyalty that Aristotle also describes without the filmmakers having to be aware that they are treading on territory first explored by Aristotle. Moreover, the film goes beyond Aristotle by showing us that loyalty, although a benefit of a true friendship, can have devastating consequences when the person to whom one is loyal is a vicious criminal. Livingston himself provides a way of explicating my view more clearly than I did. He talks about a “philosophical problèmatique,” meaning by that term a genuine philosophical puzzle or issue (Livingston 2010: 25). Using this terminology, my claim can be formulated this way: The Third Man investigates a philosophical problèmatique about the benefits of friendship that we can find clearly articulated in The Nichomachean Ethics, and in so doing illustrates Aristotle’s theory, but also supplements it by considering the dangers of misplaced loyalty. So although I do refer to the intentions of the filmmakers, I do so to establish their interest in the philosophical problèmatique. Whether a film illustrates a philosophical theory is a factual question that need not depend on the filmmakers’ intentions.

178  •  Thomas E. Wartenberg

The “No Author” Objection The final objection to treating a film as itself a philosophical work is what I call the “no author” objection. Although scholars often identify the director of a film with its “author” or creator, anyone who has sat through the closing titles of a film realizes that films are a cooperative enterprise. Even if we just admit that the screenwriter, cinematographer, production designer, and the actors all play a role in the creation of the final product, it becomes clear that one cannot immediately identify the director as the one individual who is solely responsible for the final product. But then, the objection counters, how can you claim that the film exhibits the sort of intentionality required of works in order for them to be philosophical? It certainly is true that we need to pay attention to the joint nature of cinematic creation. The final product may exhibit evidence of conflict and disagreement between the different filmmakers even as it may exhibit their agreement. Indeed, in discussing Jungle Fever in Unlikely Couples: Movie Romance as Social Criticism (1999), I argued that it was essential to our understanding of the film that we be aware of the conflict between the film’s director, Spike Lee, and one of the actors, Annabella Sciorra. But many films, such as The Third Man, can be treated as the product of a group that shared an intention. Although many of its central ideas stem from its screenwriter, Graham Greene, the other central players in the film’s creation, from its director, Carol Reed, to its British producer, Alexander Korda, and the actors, including Orson Welles and Joseph Cotton, helped create a film that realized Greene’s intention of presenting us with an arresting case of misplaced loyalty. Although some of the other players in the creation of the film, such as the American producer David O. Selznick, had other ideas, the central filmmakers, especially Korda and Reed, kept him at bay, allowing them to realize the film that they wanted (Drazin 1999). In this respect, film is no different than at least one of the traditional mediums for doing philosophy, oral discussion. If we reflect on how Socrates philosophized—that is, through engaging in oral discussion with partners who were not always in accord with him on the aims or means for conducting their discussion—we can see both that this form of philosophizing requires at least two participants and that the participants need not share a view about the aim of their interaction. Film thus shares significant characteristics with oral philosophy that put to rest the worries raised by the “no author” objection.

Why a Moderate Defense There is a curious asymmetry in the film-as-philosophy debate. While critics of the notion of cinematic philosophy present general arguments aiming to

Film as Philosophy: The Pro Position  •  179

demonstrate the impossibility of cinematic philosophy or, at least, strict limits to what films are able to accomplish philosophically, advocates for cinematic philosophy are placed in a defensive position, having to argue, as I have done in this chapter, that the critics have not established their case. So it’s important to realize that there are a number of positive claims that have emerged from my examination of four objections to the possibility of cinematic philosophy. I have maintained that, if we regard films as thought experiments, we can see how narrative fiction films can do philosophy. In addition, I have argued that the fact that a film, as a work of art, seeks to entertain its audience does not entail that it cannot also make a significant philosophical claim. I have also rejected a strongly intentionalist position, so that philosophical interpretations of films do not have to demonstrate that the filmmakers were acquainted with a specific work of philosophy, but only the philosophical problèmatique raised by a work. Finally, I have argued that the contributions of many different people to the making of a film do not undermine the possibility of the film doing philosophy. In the course of establishing these claims, I have also looked at a number of films that I take to be actually doing philosophy. It is only by canvasing the different philosophical interpretations of films presented by philosophers that one can really see the extent to which films are capable of doing philosophy. In making my argument, I have emphasized my disagreements with the critics of cinematic philosophy. But I also do not accept the claims of some advocates of cinematic philosophy. Film is not able to do everything that has been done philosophically in its two traditional media, face-to-face conversation and written texts. There are some areas of philosophy, formal logic being just one clear one, that seem unlikely to find a cinematic realization. Gödel’s incompleteness theorem, to choose one famous example, seems not to be the sort of thing that could be rendered on film, particularly a fiction film, no matter how arcane and artsy. For this reason, I think that the defender of the possibility of cinematic philosophy must be moderate in his or her claims. Making overly strong claims for cinematic philosophy only renders the case for its existence problematic. On the other hand, the claims of critics of cinematic philosophy—I have discussed Smith, Russell, and Livingston here, but there are others—often are also too extreme and one-dimensional. In order to counter their assertions, in addition to criticizing their arguments against cinematic philosophy, we need to explore the philosophical insights developed by specific films. As I hope this chapter has made clear by its example, only the detailed discussion of works of cinematic philosophy, such as The Third Man, can actually establish the validity of the claim that films can actually make a contribution to the discipline of philosophy. Let me end with an acknowledgment that I have not been able to show you here all the different ways in which I think films can make a contribution to

180  •  Thomas E. Wartenberg

philosophy. I have focused on film’s ability to present a philosophical thought experiment, to illustrate a philosophical thesis, and to offer a novel philosophical insight as some exemplary accomplishments of cinematic philosophy. But there are others, some of which I discuss in Thinking on Screen. Only the actual interpretations put forward by philosophically inclined interpreters of the cinema can establish all the philosophically relevant things that can be done on film.

Note   1 Throughout my discussion of All of Me, I ignore the problematic presentation of Eastern religion.

References Aristotle (1999) Nichomachean Ethics, Terence Irwin (tr.), Indianapolis, IN: Hackett. Barthes, Roland (1977) “The Death of the Author,” in Stephen Heath (ed. and tr.), Image— Music—Text, New York: Hill and Wang, 142–148. Cavell, Stanley (2004) Cities of Words: Pedagogical Letters on a Register of the Moral Life, Cambridge, MA: Harvard University Press. Constable, Catherine (2009) Jean Baudrillard and “The Matrix Trilogy,” Manchester: Manchester University Press. Deleuze, Giles (1986/1989) Cinema 1 and 2, Hugh Tomlinson and Barbara Habberjam (tr.), Minneapolis: University of Minnesota Press. Diamond, Cora (1991) The Realistic Spirit: Wittgenstein, Philosophy, and the Mind, Cambridge, MA: MIT Press. Drazin, Charles (1999) In Search of “The Third Man,” London: Methuen. Empson, William (1930) Seven Types of Ambiguity, London: Chatto and Windus. Falzon, Christopher (2015) Philosophy Goes to the Movies: An Introduction to Philosophy, London: Routledge. Foucault, Michel (1977) “What Is an Author?” in Donald F. Bouchard and Sherry Simon (tr.) and Donald F. Bouchard (ed.), Language, Counter-Memory, and Practice, Ithaca, NY: Cornell University Press, 113–138. Frampton, Daniel (2006) Filmosophy, London: Wallflower Press. Greene, Graham (1988) The Third Man, London: Faber and Faber. Kant, Immanuel (1999) Critique of Pure Reason, Paul Guyer and Allen Wood (tr.), Cambridge: Cambridge University Press. Livingston, Paisley (2008) “Recent Work on Cinema as Philosophy,” Philosophy Compass 3, 1–14. ——— (2010) Cinema, Philosophy, Bergman: On Film as Philosophy, Oxford: Oxford University Press. Locke, John (1979) An Essay Concerning Human Understanding, Oxford: Oxford University Press. Mason, Michelle (2003) “Contempt as a Moral Attitude,” Ethics 113, 234–272. Mulhall, Stephen (2008) On Film, 2nd ed., London: Routledge. Plato (1992) The Republic, G.M.E. Grube (tr.), Indianapolis, IN: Hackett. ——— (2002) Five Dialogues, G.M.E. Grube (tr.), Indianapolis, IN: Hackett. Rawls, John (1955) “Two Concepts of Rules,” The Philosophical Review 64, 3–13. Russell, Bruce (2000) “The Philosophical Limits of Film,” in Sander Lee (ed.), Film and Philosophy: Special Edition on Woody Allen, n.p.: Society for the Philosophic Study of the Contemporary Visual Arts, 163–167.

Film as Philosophy: The Pro Position  •  181 Smith, Murray (2006) “Film Art, Argument, and Ambiguity,” in M. Smith and T. E. Wartenberg (eds.), Thinking through Cinema: Film as Philosophy, Malden, MA: Blackwell, 33–42. Thomson, Judith Jarvis (1971) “A Defense of Abortion,” Philosophy and Public Affairs 1, 47–66. Walton, Kendall (1990) Mimesis as Make-Believe: On the Foundations of the Representational Arts, Cambridge, MA: Harvard University Press. Wartenberg, Thomas (1999) Unlikely Couples: Movie Romance as Social Criticism, Boulder, CO: Westview. ——— (2007) Thinking on Screen: Film as Philosophy, London: Routledge. Williams, Bernard (1973) “Personal Identity and Individuation,” in Bernard Williams (ed.), Problems of the Self: Philosophical Papers 1956–72, Cambridge: Cambridge University Press, 1–18.

Suggestions for Further Reading Carel, Havi and Greg Tuck (2011) New Takes of Film-Philosophy, New York: Palgrave MacMillan. (An anthology that explores the possibility of films doing philosophy.) Carroll, Noël (2002) “The Wheel of Virtue: Art, Literature, and Moral Knowledge,” The Journal of Aesthetics and Art Criticism 60, 3–26. (Argues for seeing literary works as thought experiments, using the example of the screenplay for The Third Man.) ——— (2008) The Philosophy of Motion Pictures, Malden, MA: Blackwell. (A comprehensive anthology about the philosophy of film as developed within analytic philosophy.) Cavell, Stanley (1981) Pursuits of Happiness: The Hollywood Comedy of Remarriage, Cambridge, MA: Harvard University Press. (A study of 1930s film comedies that shows them to be addressing the philosophical issue of skepticism about our knowledge of other minds.) ——— (1996) Contesting Tears: The Melodrama of the Unknown Woman, Chicago: University of Chicago Press. (A companion to the previous book that looks at melodramas from the 1930s and 1940s and interprets them as also confronting issues of skepticism.) Cox, Damien and Michael P. Levine (2012) Thinking through Film: Doing Philosophy Watching Movies, Hoboken, NJ: Wiley-Blackwell. (An introduction to philosophy that uses films as a vehicle for introducing a range of different philosophical issues.) Gilmore, Richard A. (2005) Doing Philosophy at the Movies, Albany: State University of New York Press. (Shows how films supply insights that mirror those of important philosophers.) Litch, Mary M. (2002) Philosophy through Film, New York: Routledge. (A useful introduction to philosophy that uses films to introduce philosophical issues.) Phillips, James. (2008) Cinematic Thinking: Philosophical Approaches to the New Cinema, Stanford: Stanford University Press. (An anthology about how films do philosophy from the point of view of Continental philosophy.) Reed, Rupert and Jerry Goodenough (2005) Film as Philosophy: Essays on Cinema after Wittgenstein and Cavell, New York: Palgrave MacMillan. (A collection of essays that show how specific films do philosophy.) Russell, Bruce (2008) “Film’s Limits: The Sequel,” Film and Philosophy 12, 1–16. (An article by one of the critics of the idea that films can do philosophy.) Smith, Murray (1995) Engaging Characters: Fiction, Emotion, and the Cinema, Oxford: Oxford University Press. (A study of how viewers emotionally engage with fictional characters in a film that demonstrates the cognitive dimensions of the interaction.) Stoehr, Kevin L. (2002) Film and Knowledge, Jefferson, NC: McFarland. (An anthology that explores how specific films can contribute to philosophical knowledge.) Stolnitz, Jerome (1992) “On the Cognitive Triviality of Art,” British Journal of Aesthetics 32:3, 191–200. (The classic statement of the limits of what art can teach us.)

CHAPTER

10

Film, Philosophy, and the Varieties of Artistic Value MURRAY SMITH

Like persons, works of art are a great deal richer than philosophy can or should want to capture. —Arthur Danto, Encounters and Reflections (Danto 1991: 9)

Chapter Overview The philosophy of film is now a well-established subdomain of the philosophy of art. So film can be an object of philosophy; can it also be a vehicle of philosophy? This is the question posed in the “film-as-philosophy” debate. Many authors have answered affirmatively, claiming philosophical significance for a variety of films. But how plausible are these claims? According to what conception of “philosophy” can these films be said to be “doing” philosophy? In this chapter I tease out the different senses in which a film might be said to be “philosophical.” Although it seems entirely possible for the techniques of film to be used for strictly philosophical purposes, the actual cases drawn upon by the friends of film as philosophy do not meet the norms of philosophy. This should not surprise us, given that the films in question are artworks rather than works of philosophy, and as such are characterized by the norms of the particular artistic traditions from which they emerge. The task is then to provide an analysis of the nature and place of the cognitive (“philosophical”) value that artworks may possess. In tackling this task, I draw upon Arthur Danto’s arguments concerning the philosophical dimensions of modern art. 182

Varieties of Artistic Value  •  183

Introduction Our intellectual landscape is not an indivisible whole, but one crosscut by various boundaries marking out specific domains and subdomains of activity: We distinguish science, philosophy, the arts, and theology, for example, and we make dozens of finer-grained discriminations within each of these fields, even as we recognize kinships and interconnections straddling the boundaries. To what extent are these inherited categories mere accidents of history? Or to what extent are they justifiable distinctions, ones worth preserving now that we have them, even if the path toward them has been crooked and progress along it erratic? It is probably not sensible to answer these questions in such abstract terms; why assume that there is a single answer fit for all the categories concerned? Consider the distinction between science and philosophy. Aristotle would not have recognized the distinction; empirical curiosity about the stuff of the world is as fundamental to his endeavor as is the kind of conceptual work that we would recognize as properly philosophical. Things hadn’t changed so much for David Hume, who also regarded empirical enquiry as an integral part of his own project; it is no surprise that Hume is classified in textbooks as a “British empiricist.” And as recently as 1876, the new journal Mind presented itself as “A Quarterly Review of Psychology and Philosophy” (Robertson 1876). Since that moment, however, the idea that science and philosophy are distinct practices has grown into orthodoxy, driven by the growth and increasing authority of modern science as a whole, and the proliferation of new scientific fields, each defined by their focus on some specific aspect of the material world. (Tellingly, Mind became “A Quarterly Review of Philosophy” in 1972, a change symbolizing the very shift at stake here.) This orthodoxy now faces a challenge, at least within philosophy, where practitioners of X-Phi—experimental philosophy—and others within the lineage of Aristotle, Hume, and W.V.O. Quine press the view that the concerns of philosophy cannot be made wholly independent of the concerns of science; that conceptual and empirical questions are so intimately and intricately entwined that one cannot practice philosophy with a sublime disregard for the insights of science. And neither, one might add—though this point has received much less attention in recent debate—can any scientific discipline get very far without an awareness of its conceptual underpinnings: the concepts it posits, the questions it asks, and the assumptions it makes in framing these questions with these concepts. As Edouard Machery puts it, philosophy is the pursuit of empirical knowledge by (typically, though not exclusively) conceptual means: philosophy is in the business of examining, criticizing, reforming the findings, theories, methods developed

184  •  Murray Smith

by scientists and of grasping the implications of the sciences for our understanding of the world and our place in it. (Machery 2009: 6) Regardless of whether one accepts Machery’s particular conception of philosophy, with its emphatic orientation toward the sciences, the key point here is that there are real grounds to question whether the intellectual landscape we inherit, and its associated division of labor, is the best—the most justified, the most useful—of all possible worlds. It is not obvious that the relationship between science and philosophy as it has been conceived for the past century is a fully justified one, and one that should not be revised. What of the relationship between philosophy and the arts? Can similar questions be raised about the neatness of the distinction between philosophy and art as it operates in much of our thinking? Should there be any doubt that we do assume a clear distinction between these enterprises most of the time, consider the following list offered by Paisley Livingston (2008: 1): Edmund Husserl, Logical Investigations Ludwig Wittgenstein, Philosophical Investigations W.V.O. Quine, Word and Object John Rawls, A Theory of Justice Thomas Kuhn, The Structure of Scientific Revolutions Saul Kripke, Naming and Necessity Frank Capra, It Happened One Night Laurence and Andrew Wachowski, The Matrix As Livingston notes, most of us will feel that something striking happens when we reach the names of Capra and Wachowski, for we do not customarily treat filmmakers—let alone Hollywood filmmakers—as philosophers. But this is precisely the starting point and intuition underlying the “film-as-philosophy” (FAP) thesis—the contention, that is, that the medium of film, or at least some individual films, may not only serve as objects of philosophical enquiry but also act in some sense as vehicles of philosophical activity (Mulhall 2008; Wartenberg 2007). And this argument is echoed by debates in other traditions of film theory, and by parallel debates on the philosophical role of literature and the plastic arts.1 My own take on these arguments is skeptical: The distinction between “art” and “philosophy” is not simply an accident of history, but a real and important one. It serves neither art nor philosophy to construe either practice in terms of the other, given that they serve different functions. As the “New Critic” Cleanth Brooks put the point in his famous essay “The Heresy of Paraphrase,” in a concise formulation that is hard to improve upon, we must be wary of

Varieties of Artistic Value  •  185

bringing—perhaps “forcing” is the right word—“the statement [of a poem— by extension, an artwork] into an unreal competition with science or philosophy or theology” (Brooks 1968: 164). Developing Brooks’s thought and putting it in even stronger terms, we might say that to assimilate art and philosophy is to make a kind of category mistake, confusing the distinct imperatives that each practice observes. While philosophy seeks to clarify our understanding of the world, the vocation of art is to deliver an adventure in perception, cognition, and emotion. Artworks embody an array of values—aesthetic, moral, epistemic—that interpenetrate one another. Art may deliver knowledge of the world, but philosophy is much more monotheistically wedded to epistemic goals. That is, works of philosophy are more narrowly oriented toward epistemic values; we judge works of philosophy primarily by the quality of the arguments they advance and the insights they afford, not in terms of their beauty (or other aesthetic attributes). We may note that a work of philosophy is well written or entertaining, but if the arguments are poor these achievements will not redeem it as a work of philosophy. But this is just to state, in the baldest terms, my conclusion and the intuition on which it is based. How do we get from the intuition to the conclusion, and what qualifications and concessions are forced upon us on that journey? As is ever the case with philosophy, the value of the exercise derives as much from the ride as from the destination. Many exercises in “film as philosophy”—and its Deleuzean cousin, “filmphilosophy”—skirt the underlying philosophical question at stake here, viz, “what would count as an instance of a film acting as a vehicle of philosophy?”, opting instead to plunge directly into a philosophically informed interpretation of one or more films. As a method for addressing the issue at stake, such an approach is of very limited value, as it simply begs the question, proceeding with the act of philosophical paraphrase (often on a grand scale)—explicating verbally the philosophical content that the work is claimed to represent ­filmically—with none of the caution urged by Brooks. So although I will engage in some critical engagement with particular works, to begin with we need to draw back from such specificity, in order to ask some more general (and properly philosophical) questions. One thing that unites this essay with the companion piece by Thomas Wartenberg is a shared concern with clarifying the very idea of “film as philosophy,” and of specifying in what sense we believe film to exhibit philosophical (in)competence. The most basic of these questions concerns the definitions or characterizations of our two points of comparison—philosophy and the art of film. Any serious engagement with the question “can films act as vehicles for philosophy?” will quickly find itself thrown back on particular conceptions of “film-as-art” and of “philosophy,” each term complex and laden with historical baggage. Indeed, much of what remains of this essay will take the form of an

186  •  Murray Smith

argument in favor of particular conceptions of film and of philosophy, conceptions that make salient the differences rather than the similarities between them. Given the fact that the natures of philosophy and film are both disputed, I can scarcely provide a full defense of my theory of each practice in the space available here. All I aim to offer in this essay is an approximation sufficient to reveal some problems besetting the FAP thesis. Before we enter these labyrinths, however, a few further preliminary observations are worth making. First, there is one sense in which it is obviously true that the medium of film possesses the capacity to act as a vehicle of philosophy. If what we are engaged in here counts as philosophy—I don’t say good philosophy—then we can easily imagine a filmed version of this essay. But such examples miss the true target of the FAP debate. Here there is common ground between Wartenberg and myself: He considers the example of a filmed philosophy lecture, but notes that in such a case “the philosophizing that is presented in the film actually took place extra-cinematically . . . the film itself has not contributed to the philosophy that was actually done” (this volume: 168).2 Film here plays the role not of an artistic medium but of a tool of reproduction, like a photocopier. Such films amount to a kind of “canned philosophy,” where the medium of film functions to disseminate, but not to articulate, the philosophical ideas or arguments in the lecture. What interests us here, rather, is whether there are traditions of film art— from popular narrative cinema through to art cinema and the avant-garde— that are able to sustain philosophical thinking. (Again, Wartenberg and I are in agreement on this point: See his comments on film as an artistic medium, this volume.) Do any of our existing traditions of film art, or ways in which we can imagine them being extended and reinvented, function as effective vehicles for philosophy? Certainly the case for FAP has been made with respect to the whole range of types of film art, from Stanley Cavell and Stephen Mulhall on classical and contemporary Hollywood filmmaking, to András Bálint Kovács on Antonioni and Michael Witt on Godard, to Noël Carroll on Ernie Gehr and Jinhee Choi on Kurt Kren. Wartenberg’s Thinking on Screen: Film as Philosophy (2007) encompasses this whole sweep of possibilities, from Modern Times (1936) to The Matrix (1999), The Third Man (1949) to Eternal Sunshine of the Spotless Mind (2004), through to Andy Warhol’s Empire (1964) and Tony Conrad’s The Flicker (1965) (Carroll 2006; Cavell 1981; Choi 2006; Kovács 2006; Mulhall 2008; Wartenberg 2007; Witt 2014). While we are still in the business of clearing the ground, we should also note that there are some humdrum senses of the word “philosophical” that are not the target of the debate here. A film might be said to be “philosophical” just in the sense that it adopts a serious tone, or in the sense that it expresses a Stoic attitude (“I am trying to look upon my humiliation in Groningen philosophically”—i.e., with detachment and a degree of resignation). These

Varieties of Artistic Value  •  187

senses of “philosophical” are certainly connected with philosophy proper— philosophy as a system of thought, as a practice of reflection—but they are hardly sufficient to ground what the advocates of FAP have in mind. Similarly it is inadequate simply to claim that a film represents subject matter that raises philosophical questions, or can be approached in a philosophical spirit—for that is true of anything and everything. You can look into a fireplace and begin to doubt the reliability of your knowledge of the external world, for example (Descartes 1996). What we are seeking, rather, is i) a suitably ambitious and complete conception of philosophy as a practice (rather than an attitude or stance), ii) the execution of which practice is solicited by the design of a film. And here we may find that advocates of FAP are caught between two unappealing stools: the trivially true claim that films-as-artworks can be philosophical in a weak, ill-defined, and/or “adoptive” sense (i.e., we can appropriate them for philosophical discussion, as we can appropriate anything), and the patently false claim that films can be philosophical in a more robust sense. But again, I am getting ahead of myself.

The Norms of Philosophy What might that more robust notion of “philosophy” or “the philosophical” amount to? Here it is customary to survey the kinds of techniques used by philosophers: argument, of course, but other methods as well, such as thought experiments and counterexamples (my interlocutor Tom Wartenberg has considered both in great detail; see his contribution to this volume, and his Thinking on Screen). Here I want to pursue a different approach. I want to propose that philosophical thinking and knowledge are characterized in particular by three epistemological norms or “regulative ideals”: rational warrant, empirical adequacy, and reflective maturity.3 Philosophical techniques like the thought experiment are best conceived as means to these normative ends. I’ll explore each of these norms in turn, before considering how they might or might not map onto works of film art.4 But let me add that, in describing these criteria for philosophical thinking, I have a secondary goal in mind, that of laying out what I take to be the key desiderata of a robust philosophy of film. In this sense there is a reflexive dimension to this exercise: Whatever we might learn about the question already on the table—can films act as vehicles of philosophy?—I hope in addition to identify some of the requirements, routines, and ingredients for good theorizing or philosophizing about film in general. I will use the debate concerning the cinematic “illusion of reality”—that is, arguments concerning the nature of our experience as cinematic spectators—as an example.

188  •  Murray Smith

By “rational warrant,” I mean to pick out perhaps the most obvious feature of the philosophical as it is understood in most traditions: its attention to the rational—or non-rational or irrational, as the case may be—underpinnings of our beliefs and practices. The tools of logic, mapping out the abstract architecture of thought, are the core means of such rational assessment. For example, we speak of the illusion of cinema, motivated by the powerful emotional responses that films commonly prompt. From the fact that we are moved by emotion, however, does it follow—by strict logical entailment—that we are subject to an illusion of reality? Several of us in the cognitive tradition have insisted that it does not. The entailment holds only if we sign up to a hidden premise—namely, that to be moved emotionally requires that we believe in the existence of the events that move us.5 The rational assessment of the belief that we are subject to an “illusion of reality” also involves close examination of the concepts on which our beliefs depend: Thus we need to ask, just what is an “illusion”? Does the term mask significant ambiguity? Similarly, what do we mean when we speak of “film”: a specific technology, or any technology that can deliver a moving image? Or something else again? Can we tease out of these terms more precisely defined concepts that will allow us to think more clearly about the phenomena at hand—namely, those aspects of behavior that appear to be continuous between engagement-with-reality and engagement-with-filmic-depictions? In all these ways, we seek both to ascertain what degree of rational warrant we have for our existing beliefs, and where possible to revise or replace them with beliefs commanding greater rational warrant. In sketching “rational warrant,” I have already touched upon my second normative principle—empirical support—for in practice these principles will act in concert with one another. Logic needs material to act upon, and that material comes in general in the form of empirical observation, casual and more formal. Thus, we observe that we are moved by watching films: There goes my heart racing again; another teardrop falls into the popcorn. But when we act as theory-building philosophers, we should be in the business of putting pressure on these empirical assumptions: Is everyone moved by watching films? Am I, or is my tribe, idiosyncratic? And how good is my evidence for the belief that we are moved by films? Perhaps our states of mind and body are better thought of as “quasi-emotions,” to use Kendall Walton’s phrase; perhaps we need to look at all the ways in which our behavior at the movies does not conform with ordinary emotional responses (Walton 1978; 1990: ch. 7). In other words, we need to examine and reexamine the evidence at our disposal, and if necessary seek out further evidence to support or (if we are in a Popperian mood) refute our beliefs and conjectures. Just as my second normative principle, empirical soundness, began to emerge in the discussion of the principle of rational warrant, so my third

Varieties of Artistic Value  •  189

normative principle of philosophy, reflective maturity, has been adumbrated in my remarks on the first two principles. Thus, for example, when I suggest that in approaching a topic philosophically we should examine and reexamine the evidence available to us, I am gesturing toward this third principle. I have in mind a specific and widely recognized procedure, as well as a more general spirit of enquiry. The specific procedure is reflective equilibrium, christened and most famously developed by John Rawls in A Theory of Justice (Rawls 1971). Reflective equilibrium for Rawls involves the comparative movement back and forth between either (i) specific principles and particular cases to which they apply (narrow reflective equilibrium), or (ii) different principles or perspectives that we take ourselves to be committed to, or at least regard as principles or perspectives that deserve to be taken seriously—Kantianism and utilitarianism, for example, in the context of ethical debate (wide reflective equilibrium). In general, the idea is that we seek, through this process, coherence among our beliefs and commitments. The notion of “reflective maturity” I am advancing here is heavily indebted to Rawls’s concept, but I adopt a revised phrase in order to emphasize an additional (or at least, specific) feature of the process: the fact that whatever conclusions we reach when thinking philosophically will be the product of a sustained process of reflection, of testing our beliefs and assumptions by examining them from a multitude of angles. The reflective maturity that this process delivers is no guarantee of coherence, at least not of absolute consistency, nor of epistemic confidence in some belief or assumption; indeed reflective maturity sometimes delivers us into the arms of skeptical doubt. In this sense “maturity” does not always lead to “equilibrium,” with its overtones of stability and harmony. So I am concerned to bring out the hard-won quality of the views we arrive at through philosophical reflection, regardless of the degree of confidence we have in them and the degree to which we have achieved coherence among our beliefs. How might the principle of reflective maturity apply in the case of claims regarding the illusionistic power of cinema? We take instances of this power, from our own experience and the testimony of others, and try to make our understanding of them cohere with whatever more explicit beliefs or theories we hold about the nature of cinema and of perception more generally. That is, we examine what degree of fit there is between the views we currently hold, or the ones that warrant serious consideration, and what the putative examples of illusion seem to suggest. Reflecting on the theories may make us rethink our descriptions of the examples—thus what we initially described as an emotion we might relabel a “quasi-emotion”—but equally the examples may make us wonder about the adequacy of an established theory, or incline us toward accepting one theory over another. My point here is not to enter into a discussion of the particulars of this debate, but to play up the nature of the debate

190  •  Murray Smith

as a philosophical one. Regardless of whether this reflective movement back and forth, between examples and principles and between competing theories, leads us to an illusionistic or to an anti-illusionistic perspective or to some intermediate position, to skepticism about the question or our ability to arrive at an answer to it, the key point is that it will be characterized by a sustained process of reflection—that is, by reflective maturity. Philosophy, then, is a hard taskmaster—arguably the toughest taskmaster of all, in the multiple demands it makes upon us rationally, empirically, and reflectively: more demanding than science, with its emphasis on the empirical and on delimiting specific domains of enquiry, and more demanding than art—too demanding for art, as I shall argue. Or to be fairer and more precise: Philosophy is more demanding than the neighboring endeavors of science and art in terms of the epistemic norms it seeks to observe. In other respects, scientific and artistic practice may be much more demanding—in the case of science, for example, in terms of how one turns hypotheses into experiments, and how one analyses data; in the case of art, how one gives compelling, imaginative embodiment to a given perspective on the world. Even with that concession in place, one immediate objection to the picture of philosophy given here is that it is too parochial, too narrowly bound to an analytic conception of philosophy. I’m not sure this is a fair criticism. Plato established a foundational contrast between art and philosophy, and my three norms are inspired as much by the spirit of Socratic dialog as by modern analytic practice. I imagine philosophers of diverse stripes would think of themselves as observing these norms, or something very similar to them. Thus while my framework may not catch all models of philosophy, it does have pretty wide reach. To this point we can add a parallel one: The picture of philosophy as a distinctive practice governed by norms different from those appropriate to art is a central plank of Hegel’s philosophical system. For Hegel, art works through “picture-thinking” and emerges from the religious phase in the development of Geist (Hegel 1977). Whatever its value, it falls short of the purely conceptual realization of Absolute Spirit embodied by philosophy. Hegel thus holds that art should not stray from its own vocation, acting as a vehicle for politics, morality, or philosophy. (Later on we will see how Hegel’s views are revived and developed, in a contrary direction, by Arthur Danto.) Paisley Livingston discerns in Hegel what he calls the “rationality objection” to the idea that art can function as a vehicle of philosophy: Given the nature of art (and the philosophical limitations of “picture-thinking”), it is irrational to pursue philosophy through art—just as it would be irrational to secure a screw with a coin if a screwdriver is available.6 To put it in the terms of my argument, the practices that we recognize as artistic are less effective as a means to secure the epistemic ends of philosophy than are the recognized practices of philosophy. Thus, even if there is a strong analytic flavor to my characterization of the

Varieties of Artistic Value  •  191

nature of philosophy and the norms governing it, the idea of a kind of constitutive incompatibility between art and philosophy is by no means unique to the analytic tradition, as we see in its strong resonance on this issue with Hegel’s philosophy.

Philosophy and the Plurality of Artistic Value I claimed earlier that the epistemic demands of philosophy as I have characterized it are too high for film-as-art. We do not typically scrutinize a film, be it Casablanca, Cat’s Cradle, or Caché, for its rational warrant, empirical adequacy, and reflective maturity (in the senses discussed earlier). Rather when we regard a film as an artwork we assess it for its capacity to engage us perceptually and imaginatively, to afford us a compelling cognitive and emotional experience, its ability to offer us a vivid experience of the world. While art may not be “cognitively trivial” (Stolnitz 1992), neither is art governed by cognitive value in the same way or to the same extent as philosophy. Rather than defend this claim further, I will assume that it is clear enough that there is a significant and substantial difference between the norms governing art and those governing philosophy—notwithstanding the debates concerning the correct characterization of each practice—and that works of film art typically do not and should not be expected to meet demands appropriate to philosophy proper. Does our investigation run aground at this point? Have we succeeded in establishing that the hypothesis that films may act as vehicles of philosophy— defined in a robust way—is, as I put it earlier, patently false? I don’t think so. There is another, more plausible, fruitful, and illuminating way forward. But it involves lowering the stakes and abandoning the baldness (and boldness) of the FAP hypothesis as it is typically framed. Once again we need to pull back before we can move forward. On one orthodox view within the philosophy of art, artworks embody an array of different kinds of value—not only aesthetic value, most often regarded as the mainstay of artistic value, but cognitive, ethical, and perhaps also political and historical value. Arguments in this area focus on the interrelationship among these types of value, the manner of their embodiment within works of art, and their relative importance within different types of art. Set in the context of these debates, we can understand the FAP hypothesis as a very strong version of a claim in favor of the cognitive value of (some) films. The strength of the claim derives from the demanding epistemic norms that I have claimed philosophy seeks to observe. But we do not need to claim that a film is “doing philosophy” in order to recognize that it has cognitive value. And neither is it appropriate to construe the cognitive value of a film in terms of it “philosophizing,” as this is apt to distort the other kinds of value obtaining in the work, and

192  •  Murray Smith

the way in which the various kinds of value hang together. Note that the second claim—bearing on what we lose, or at least lose sight of, when we draw together film-as-art and philosophy—is critical.7 Films can engage us on every level of our being: perceptually, imaginatively, emotionally, and cognitively. We not only see, hear, and—as Michel Chion has rightly insisted—“audioview” films (Chion 2009) but also experience them, at least to some extent, in a more thoroughgoing multimodal fashion. That is, the integrated nature of our sensory systems ensures that the audiovisual stimuli that constitute films are able to trigger a wider array of sensory-perceptual experiences, including those of touch, taste, smell, and even proprioceptive orientation. And even more basic than these experiences are the physiological reflexes that cinema exploits, including such things as the startle response, flicker effects, and the perceptual experience of motion itself, the very bedrock of cinema. We do not, of course, experience films only at these “basic” and bodily levels: Reflexes, sensations, and perceptions lead to affective appraisals and emotions, and emotions proper are themselves constituted in part by cognitive judgments. As Torben Grodal has underlined, these various types of responses work in a recurrent “flow” or cycle, our higher-order emotions and cognitions being shaped by “bottom up” perceptual processes, while in turn conditioning from the “top down” the next wave of visual and sonic material put before us (Grodal 2009). It is also important to underline that film spectatorship is an imaginative enterprise: What we see and feel allows us to represent in our embodied minds situations, perspectives, and characters that are either historically past, spatially remote, or wholly fictional; so nonfictions as much as fictions engage us imaginatively.8 Cinema works on all these fronts, the exact balance and interplay among these levels and types of responses varying across both modes of filmmaking and individual works. The varied and layered experience that the cinema affords underpins the varied kinds of value that film—or individual films—can embody: once again, from the aesthetic to the cognitive, ethical, and political. Earlier I alluded briefly to the ideas of Arthur Danto; at this point, we need to bring Danto’s arguments properly into play. Danto might seem an odd person for me to invoke as an ally, given that he is perhaps best known for his arguments concerning the “philosophical turn” in the art world of the twentieth century—the claim, that is, that what distinguishes at least a major lineage of twentieth-century and contemporary art practice is an acutely selfconscious turn, whereby artists themselves come to investigate, through their art, the traditionally philosophical question: “what is art?” (Danto 1981). But when we examine Danto’s argument in detail, and especially when we remind ourselves of its Hegelian origins, the proximity of Danto’s position to my own will, I hope, become apparent.

Varieties of Artistic Value  •  193

Danto draws directly from Hegel the idea that art can be defined as “the sensuous embodiment of the idea” (Danto 1991). For Hegel, as I note earlier, this is a way of recognizing the ideational content of art, while insisting upon the difference between art and philosophy—the latter occupying a later and more advanced stage of development in the Hegelian narrative of Geist. Danto reframes Hegel’s account so that the formula for art—“the sensuous embodiment of the idea”—takes on a very different value. The grand Hegelian narrative of Geist is set aside in favor of the (relatively!) modest claim that the major project of Western visual art from the Renaissance onwards—the rendering of visual appearance—had drawn to a close in the nineteenth century, and that art had been thrown into a new period of self-consciousness about its own nature. It is this self-consciousness, embodied above all first by Duchamp and later by Warhol, that Danto characterizes as “philosophical.” In this new context, the Hegelian formula still serves to underline the ideational content of art, but now as a way of insisting on what is shared by art and philosophy, rather than what divides them. If Hegel sees art as “the sensuous embodiment of the idea,” then Danto shifts the stress to “the sensuous embodiment of the idea.” In the work of Duchamp, Warhol, in much of Pop and in Conceptual art in general, we see the withering away of “sensuous embodiment” in favor of minimally expressed ideas (Danto 1998a; 1998b). What bearing does all this have on the question of film, the types of value films can instantiate, and the FAP hypothesis? The answer to this question is not straightforward. There is a tradition of filmmaking—avant-garde filmmaking, and the contemporary monster known as “artists’ film and video”—which is directly connected with the art world that Danto focuses upon; Duchamp and Warhol, after all, both made films with strong “conceptual” dimensions. But outside of that specific tradition, the relevance of Danto’s ideas for film, I want to suggest, depends rather on restoring the balance between the two points of comparison in Hegel’s formula: on acknowledging both the ideas and their “sensuous embodiment” in works of film art. I want to explore, briefly, two examples in this light—examples spanning fiction and nonfiction, as well as mainstream and art house modes of filmmaking. It is important to keep in mind throughout my discussion of these cases that we have moved beyond the starkest formulation of the FAP question: Is it possible for the medium of film to act as a vehicle of philosophy? I have granted that it is possible, but argued that the more important question is: In what sense can actual works of cinematic art exemplifying established modes of film art practice be regarded as “doing” philosophy? Given the philosophical claims made on behalf of mainstream cinema by Cavell, Mulhall, Wartenberg, and others, let me begin with the much-discussed example of Blade Runner (1982). Ridley Scott’s film dramatizes a trio of “existential-ethical” concerns: the limits of self-knowledge (symbolized by Rachel’s initial ignorance of her

194  •  Murray Smith

status as a replicant, with merely implanted memories, and then by the suggestion that Deckard too may be a replicant); the subordination of a class of agents (the replicants) who share the same capacities as the subordinating class (the humans—or at least those in positions of power); and the condition of living in the shadow of imminent, premature death. The power and the value of Blade Runner, however, derive not from the fact that it articulates what we recognize as philosophical ideas, and still less from its establishing the truth or plausibility—in a way meeting the philosophical standards I’ve discussed—of these ideas. As a futuristic fictional narrative, the film offers very little in the way of systematic empirical support for the philosophical ideas it represents, nor does it subject these ideas to searching rational tests, and thus it cannot deliver reflective maturity. But again, these are absurd demands to make of a fiction film. Instead, following Danto’s lead, we need to look at the film’s sensuous embodiment of its ideas to locate the source of its value (where “sensuous” should be understood as shorthand for the various levels of response and appreciation I note earlier: from bodily reflexes to imaginative reflection). Consider, for example, the protracted pursuit of Deckard by Roy in the climactic scene of Blade Runner. Roy’s claim that he has shown Deckard “what it feels like to be a slave”—and by extension the implication that the film has enabled us to feel this in imagination too—hinges on our registering Deckard’s terrified state through such bodily mechanisms as affective mimicry, and not merely through a detached understanding of Deckard’s situation.9 Note also the significance here of sheer visual and sonic texture: Scott seems invested in affording us an intense, exhilarating experience at the sensoryperceptual level, not merely as a vehicle to articulating an idea or a philosophical claim. Consider, for example, Scott’s use of richly detailed mise en scène, and the motif of strong backlighting in Blade Runner (Figures 10.1 and 10.2). While these techniques certainly play a role in articulating the philosophical ideas in the film, their presence and value in the film are not exhausted by this role. This is one reason that Scott is regarded as an ostentatious visual

  Figures 10.1 and 10.2  Backlighting and intricate mise en scène in Blade Runner (Ridley Scott, 1982).

Varieties of Artistic Value  •  195

stylist: As we watch Blade Runner, it is clear we are being invited to savor the visual and auditory textures of the film, and the fictional world it depicts, as an end in itself. And the fact that the film devotes so much of its energy to the aesthetics of the image is not a cost-free choice with respect to its cognitive (“philosophical”) value: By making the sheer visual design of its images so salient, the role of these images in articulating an existentially charged narrative is compromised.10 Moreover, an analysis of the film that fails to acknowledge the interplay between cognitive and aesthetic value—whatever form that interplay takes—is apt both to overstate the significance of the film’s philosophical ideas and to fail to register the distinctive, sensuous embodiment of these ideas that it is the special vocation of art to render (Livingston 2008; Smith 2006). Now let me turn to a very different film: Joshua Oppenheimer’s The Act of Killing (2012), a documentary concerned with the massacres of (presumed) communists by Indonesian paramilitary groups in the 1960s (and by extension, the violent nature of the histories and regimes underpinning the capitalist development of many parts of the world, and the moral corruption that generally accompanies authoritarian power). The Act of Killing lays claim to a kind of cognitive value: The film provides detailed factual knowledge about Indonesia, and it invites us to reflect upon the moral and political state of affairs represented. To the extent that the film succeeds in these goals, it thus possesses specifically ethico-political value. But if The Act of Killing raises ethical questions—questions about how one ought or ought not to live; about what one owes to one’s immediate and more remote neighbors—why not then simply grant that the film is doing philosophy? Again the answer hinges on identifying properly what The Act of Killing is doing and what it is not doing. What it is not doing is probing the evidence presented from numerous angles, offering counterexamples and counterarguments, and seeking to establish reflective maturity in the philosophical sense described earlier. Rather The Act of Killing seeks to present the specific critical perspective that it advances in a vivid and aesthetically engaging manner: The film thereby seeks to make its ethical-political claims via an aesthetic experience.11 Thus The Act of Killing—described by Oppenheimer as a “documentary of the imagination”—takes us into the minds of its gangster-assassins via the lurid sets and performances for the film they are making about their own activities (Figures 10.3 and 10.4). A gobsmacking mix of the surreal, the kitsch, and the morally depraved, the sensuous embodiment of the ideas in this film—the particular figures, moments, gestures that comprise it—is critical to the experience the film affords and the value we thereby (might come to) attach to it. It is not so much that the provision of aesthetic experience by The Act of Killing

196  •  Murray Smith

  Figures 10.3 and 10.4  Surreal and kitsch imagery in The Act of Killing (Joshua Oppenheimer, 2012).

necessarily undermines the epistemic norms of philosophy but rather the fact that, as a work of art, the film observes a different set of norms from those we would bring into play were we engaging with philosophical discussion of these events. Let me now draw the threads of this second part of the chapter together. Artworks embody ideas, as Danto proposes, but in doing so they come to have various kinds of value, and the responses upon which their appreciation depends work on a multitude of levels, from reflexes to highly abstract reflections. Moreover, we need to resist the temptation to hierarchize our responses and the different kinds of value they underpin, from the sensuous to the perceptual, the imaginative, the emotional, to the cognitive, with the philosophical perched at the apex. This perspective accords with Viktor Shklovsky’s well-known arguments on the “defamiliarizing” capacity of art, in which simple perceptual experiences are put on a par with existential threats and complex, abstract ideas: Everything from the “stoniness” of stones to the fear of war can be a target of defamiliarization (Shklovsky 1990: 5–6). Our sensuous-perceptual responses are not a mere means to conveying philosophical content; in the context of art, they are (also) valued for their own sake, just as the standard formula tells us. If I can offer a gustatory metaphor: When we engage with a work of art, we savor the individual tastes and textures as well as the overall flavor. We recognize and take pleasure in the distinct kinds of value embodied by the work, as well as the way in which these come together in the work. With respect to the FAP proposal, my argument is thus deflationary rather than eliminativist in character: Many works of art are rightly valued in cognitive terms. They get us thinking and they give us knowledge; we learn about the world through them, usually in modest but sometimes in profound ways. But artworks do not pursue or offer us knowledge in the manner of philosophy, and we honor neither art nor philosophy by conflating or cross-dressing them.

Varieties of Artistic Value  •  197

Objections and Alternatives In closing, I can offer the FAP evangelist two final concessions. First, because many artworks have cognitive value—because they embody ideas—they can readily be appropriated for philosophical purposes. Earlier in the chapter I complained that simply using a film for philosophical purposes offered no proof of the philosophical character of a film, because anything and everything can prompt philosophical questioning and reflection. That point holds, but now we can add that the cognitive dimension of works of art predisposes them for philosophical appropriation in a way that does not hold for ordinary, “untransfigured” objects: Understood in their distinctive contexts, Warhol’s Brillo box provokes us cognitively in a way that the perceptually indiscriminable Brillo box in the local Walgreens does not. The second concession is one that has been hovering just offstage through much of this chapter. Why not accept that there is an area of overlap between the domains of art and philosophy—an area occupied by such genres as conceptual art, the novel of ideas, and (why not?) the film of ideas? Why not indeed. As I stated earlier, there seems to be no impediment in principle to the use of the techniques of cinema—the combination of moving depiction, editing, music, and language—for philosophical purposes, and even meeting the exacting demands of rational warrant, empirical adequacy, and reflective maturity. I’m not sure how heavily populated this zone of overlap is, but in some educational television, some documentary filmmaking, and some examples of the “essay” film, we see works that draw creatively on an array of cinematic techniques of cinema to craft effective expositions of philosophical topics and debates. Such films contrast with the case of the primitive single-camera recording of a philosophy lecture, where the medium of film is, so to speak, inert. Perhaps some of the works of Errol Morris and Adam Curtis would fit within this category; perhaps, as Wartenberg suggests, John Berger’s film version of Ways of Seeing is an example.12 Whatever the answer, there is a crucial rider to this concession: Accepting that there is an area of overlap between art and philosophy does not undercut the general validity and significance of the distinction. Similarly, there may be an indeterminate area between red and yellow, where judgments about the color vary; but red and yellow are still different colors. In the passage from which I have drawn the epigraph for this essay, Danto declares that “Like persons, works of art are a great deal richer than philosophy can or should want to capture” (Danto 1991: 9). Given the tenor of Danto’s theory, this is a surprising utterance. Perhaps this statement is a kind of corrective or caution against too literal a construal of the idea that art can be or has become philosophical. Similarly in his essays on literature, Danto seems at pains

198  •  Murray Smith

to emphasize that philosophy and literature should not be wholly or simply identified with one another, no matter how significantly they overlap (Danto 1986: 21, 161, 163, 186). At a minimum, the statement reveals that Danto recognized the complex and often messy layering of types of value inherent in most art forms, no matter how “philosophical” some artworks might have become. Here again I am inclined to agree and disagree with Danto. Works of art should not be reduced their ideational or philosophical contents, but why should we not aim to formulate a framework apt to capture and illuminate precisely the multifaceted nature of art that he points to? For good or ill, and with whatever degree of success, that has been one of the goals of the present essay.

Notes   1 For example, within film theory, debates around so-called film-philosophy inspired by the writings of Gilles Deleuze; beyond the study of film, debates around the philosophical capacities of literature, as in the work of Martha Nussbaum (e.g., Nussbaum 1992); the plastic arts, exemplified by the work of Arthur Danto (e.g., Danto 1986, where Danto also discusses philosophy and literature); and even music, as in the work of Philip Kitcher (Kitcher 2014). I touch upon some of these debates later in this essay.   2 See also Wartenberg’s remark on Waking Life, which he describes as “doing no more than recording philosophy being done orally” (this volume).   3 On the notion of regulative ideals, see McKeever and Ridge (2006).   4 I am not sure that there is a precise fit between any of the four objections to FAP that Wartenberg attempts to rebut in his essay and my line of thought here, but there is a kinship between my argument and his “generality” and “explicitness” objections. These objections map more straightforwardly onto my essay “Film and Philosophy” (Smith 2008), where I discuss three problems faced by advocates of FAP—the problems of narrative, fiction, and depiction. Similarly, as Wartenberg’s discussion of All of Me in relation to his “explicitness” objection indicates, my “Film Art, Argument, and Ambiguity” (Smith 2006) is also relevant here.   5 The idea that film viewing involves illusion is critically examined in Allen (1997), Carroll (1989), and Smith (1995b). See also the exchange between Allen and Smith in the Journal of Aesthetics and Art Criticism (Allen 1998; Smith 1998).   6 Livingston (2009: 43–44); see also my review of Livingston (Smith 2010).   7 So here again my argument intersects with, but moves off in a different direction to, Wartenberg’s argument. Like Wartenberg, I am happy to recognize that many works of film art (of diverse types, including popular narrative films) will have a “cognitive component” (Wartenberg, this volume) or possess cognitive value. I query, however, whether this type of value is aptly labeled “philosophical,” in part because of the relationship between cognitive and other kinds of value in art. While both art and philosophy may instantiate various types of value, the two practices differ in the placement and prominence of cognitive value within their respective schemes. See Smith (2006: 39).   8 Gregory Currie and Ian Ravenscroft discuss the importance of “motor imagery” in perceptual imagining (Currie and Ravenscroft 2002: 97). They focus in this passage on an embodied imaginative experience as prompted by a verbal description, but it is easy to see how the same process would be carried over to film viewing, and perhaps even enhanced by the role of visual and auditory perception in film viewing. The various forms of imagination at play within cinematic spectatorship are explored by Currie (1995: part II) and Smith (1995a: ch. 2).

Varieties of Artistic Value  •  199  9 For more on affective mimicry, and more generally the orchestration of our emotional responses to film, see Smith (1995a; forthcoming 2016). 10 Ridley Scott himself downplays the philosophical interest of the film. Although it is hard to take his outright denials at face value, it is probably fair to say that the philosophical ideas in the film derive from Philip K. Dick’s novel and the various versions of the screenplay. Scott’s achievement seems to lie in the vivid realization of the story; it is less clear that he can be credited with further shaping and fine-tuning of the specifically philosophical implications of the script. For an excellent recent discussion of the film in philosophical terms, see Shanahan (2014). 11 For more on the complex interweaving of aesthetic and political value in art, see my essay on Patrick Keiller’s London: “Aesthetics and the Rhetorical Power of Narrative” (2000). 12 See the discussion on nonfiction film in Wartenberg (2007: 79–80); Wartenberg cites John Berger’s television documentary Ways of Seeing (1972) as an example of a nonfiction work making philosophical arguments.

Acknowledgements Thanks to Julian Hanich and Judith Vega for the invitation to give an early version of this essay in the ‘Thinking Cinema: New Approaches in Film Philosophy’ series of talks at the University of Groningen in May 2014. I am grateful to Julian, Judith, and the audience on that occasion, as well as Kate ThomsonJones, for their challenging questions and helpful feedback.

References Allen, R. (1997) Projecting Illusion: Film Spectatorship and the Impression of Reality, Cambridge, MA: Cambridge University Press. ——— (1998) “Film Spectatorship: A Reply to Murray Smith,” Journal of Aesthetics and Art Criticism 56:1 (Winter), 61–63. Brooks, C. (1968) “The Heresy of Paraphrase,” in The Well Wrought Urn: Studies in the Structure of Poetry, London: Methuen, 157–75. Carroll, N. (1989) Mystifying Movies: Fads and Fallacies in Contemporary Film Theory, New York: Columbia University Press. ——— (2006) “Philosophizing through the Moving Image: The Case of Serene Velocity,” in M. Smith and T. E. Wartenberg (eds.), Thinking through Cinema: Film as Philosophy, Cambridge: Wiley-Blackwell, 173–185. Cavell, S. (1981) Pursuits of Happiness: The Hollywood Comedy of Remarriage, Cambridge, MA: Harvard University Press. Chion, M. (2009) Film, A Sound Art, Claudia Gorbman (tr.), New York: Columbia University Press. Choi, J. (2006) “Apperception on Display: Structural Films and Philosophy,” in M. Smith and T. E. Wartenberg (eds.), Thinking through Cinema: Film as Philosophy, Cambridge: WileyBlackwell, 165–172. Currie, G. (1995) Image and Mind: Film, Philosophy, and Cognitive Science, Cambridge, MA: Cambridge University Press. Currie, G. and I. Ravenscroft (2002) Recreative Minds: Imagination in Philosophy and Psychology, Oxford: Clarendon Press. Danto, A. C. (1981) The Transfiguration of the Commonplace: A Philosophy of Art, Cambridge, MA: Harvard University Press. ——— (1986) The Philosophical Disenfranchisement of Art, New York: Columbia University Press. ——— (1991) Encounters and Reflections: Art in the Historical Present, New York: Noonday Press.

200  •  Murray Smith ——— (1998a) Beyond the Brillo Box: The Visual Arts in Post-Historical Perspective, Berkeley: University of California Press. ——— (1998b) “Blam! The Explosion of Pop, Minimalism, and Performance, 1958–64,” in G. Horowitz and T. Huhn (eds.), The Wake of Art: Criticism, Philosophy, and the Ends of Taste, Amsterdam: G + B Arts International, 57–62. Descartes, R. (1996) “First Meditation,” in J. Cottingham (ed.), Meditations on First Philosophy, Cambridge, MA: Cambridge University Press, 12–15. Grodal, T. (2009) Embodied Visions: Evolution, Emotion, Culture, and Film, Oxford: Oxford University Press. Hegel, G.W.F. (1977) Phenomenology of Spirit, A. V. Miller (tr.), Oxford: Oxford University Press. Kitcher, P. (2014) “Philosophical Music? A View from Late Mahler,” Royal Music Association Music and Philosophy Conference keynote, London. Kovács, A. B. (2006) “Sartre, the Philosophy of Nothingness, and the Modern Melodrama,” in M. Smith and T. E. Wartenberg (eds.), Thinking through Cinema: Film as Philosophy, Cambridge: Wiley-Blackwell, 135–145. Livingston, P. (2008) “Recent Work on Cinema as Philosophy,” Philosophy Compass 3, 1–14. ——— (2009) Cinema, Philosophy, Bergman: On Film as Philosophy, Oxford: Oxford University Press. Machery, E. (2009) Doing without Concepts, Oxford: Oxford University Press. McKeever, S. and M. Ridge (2006) Principled Ethics: Generalism as a Regulative Ideal, Oxford: Oxford University Press. Mulhall, S. (2008) On Film, 2nd ed., London: Routledge. Nussbaum, M. (1992) Love’s Knowledge: Essays on Philosophy and Literature, New York: Oxford University Press. Rawls, J. (1971) A Theory of Justice, Cambridge, MA: Harvard University Press. Robertson, G. C. (ed.) (January 1876) Mind: A Quarterly Review of Psychology and Philosophy 1:1, 1–156. Shanahan, T. (2014) Philosophy and Blade Runner, Houndmills, UK: Palgrave Macmillan. Shklovsky, V. (1990) Theory of Prose, Benjamin Sher (tr.), Elmwood Park: Dalkey Archive Press. Smith, M. (1995a) Engaging Characters: Fiction, Emotion, and the Cinema, Oxford: Clarendon Press. ——— (1995b) “Film Spectatorship and the Institution of Fiction,” Journal of Aesthetics and Art Criticism 53:2 (Spring), 113–127. ——— (1998) “Regarding Film Spectatorship: A Reply to Richard Allen,” Journal of Aesthetics and Art Criticism 56:1 (Winter), 63–65. ——— (2000) “Aesthetics and the Rhetorical Power of Narrative,” in I. Bondebjerg (ed.), Moving Images, Culture and the Mind, Luton: University of Luton Press, 157–166. ——— (2006) “Film Art, Argument, and Ambiguity,” in M. Smith and T. E. Wartenberg (eds.), Thinking through Cinema: Film as Philosophy, Cambridge: Wiley-Blackwell, 33–42. ——— (2008) “Film and Philosophy,” in J. Donald and M. Renov (eds.), The SAGE Handbook of Film Studies, London: SAGE, 147–163. ——— (2010) “Review of Livingston,” Cinema, Bergman, Philosophy, accessed July 4, 2014, https://ndpr.nd.edu/news/24361-cinema-philosophy-bergman-on-film-as-philosophy/ ——— (forthcoming 2016) Film, Art, and the Third Culture, Oxford: Oxford University Press. Smith, M. and T. E. Wartenberg (eds.) (2006) Thinking through Cinema: Film as Philosophy, Oxford: Blackwell. Originally published as a special issue of the Journal of Aesthetics and Art Criticism 64:1 (Winter). Stolnitz, J. (1992) “On the Cognitive Triviality of Art,” British Journal of Aesthetics 32:3, 191–200. Walton, K. (1978) “Fearing Fictions,” Journal of Philosophy 75:1 (January), 5–27. ——— (1990) Mimesis as Make-Believe: On the Foundations of the Representational Arts, Cambridge, MA: Harvard University Press. Wartenberg, T. E. (2007) Thinking on Screen: Film as Philosophy, New York: Routledge. Witt, M. (2014) Jean-Luc Godard: Cinema Historian, Bloomington: Indiana University Press.

Varieties of Artistic Value  •  201 Suggestions for Further Reading Livingston, P. and C. Plantinga (2009) The Routledge Companion to Philosophy and Film, London: Routledge. (Comprehensive overview of debates in philosophy of film; part 4 is devoted to film as philosophy.) Nannicelli, T. (2012) “Review of H. Carel and G. Tuck, New Takes in Film-Philosophy,” Journal of Aesthetics and Art Criticism 70:3 (Summer), 326–328. (Skeptical review of anthology in the tradition of “film-philosophy,” highlighting the vagueness of the project.) Routledge “Philosophers on Film” series. (Includes volumes on Blade Runner, Eternal Sunshine of the Spotless Mind, Memento, Talk to Her, and The Thin Red Line, exploring the films from a variety of philosophical perspectives.) Sinnerbrink, R. (2011) New Philosophies of Film, London: Bloomsbury. (Wide-ranging overview of recent trends in the philosophy of film, identifying cognitive-analytic philosophy of film and “film-philosophy” as the two major movements.) Smuts, A. (2009) “Film as Philosophy: In Defense of a Bold Thesis,” Journal of Aesthetics and Art Criticism 67:4 (Fall), 409–420. (Defense of the FAP thesis that engages closely with its critics.)

This page intentionally left blank

Index

accuracy, in realism 76, 78 – 9, 88 – 9 achievement argument: achievement and difficulty 25 – 8; appreciation and achievement 24 – 5; appreciation and technology 29 – 30; difficulty and technology 28 – 9 The Act of Killing (Oppenheimer) 195 affect: and cognition 134, 138 – 9; study of 131 – 2; of the viewer 134 – 5 affective matching 143 affective mimicry 157, 199n8 Ahab, Captain (character) 118 Alex (character, A Clockwork Orange) 160 Alien Tetralogy (Scott) 166 All of Me (Reiner) 173, 174, 198n4 allegiance 140 – 2 “Allegory of the Cave” (Plato) 171 Allen, Woody 101, 112, 170 American Beauty (Mendes) 100, 104 Anderson, Paul Thomas 101 animation, photorealistic 29 Annie Hall (Allen) 112 Antonioni, Michelangelo 186 Argument from Filmic Mediation 105 – 6 Aristotle 111, 183 Arnheim, Rudolf 2, 31

art and art forms: as cluster concept 19; Conceptual 193; as different from media 21; and philosophy 11 – 12, 184 – 5, 197 – 8; philosophy of 182; Pop 193; relationship to technology 21; and technology of film 3 – 4; as a work of philosophy 185; see also artworks Art in Three Dimensions (Carroll) 157 artifacts 19 – 21, 24, 49, 60 artworks: and the agential sense of media 41 – 2, 48; cognitive value of 197; embodiment of ideas by 12; indifference to materials of 40; see also art and art forms The Assassination of Jesse James by the Coward Robert Ford (Dominik) 101 Atticus Finch (character) 118 – 21, 124 – 6 Au Hasard Balthazar (Bresson) 106 automaticity 136 backlighting 194 – 5 back-projection 66 – 7 Bailey, George (character) 156 Bardot, Brigitte 174 Barthes, Roland 167 Basalla, George 19

203

204  • Index Bazin, André 2, 4 – 5; on cinematic realism 60 – 1, 73n2 Berger, John 197, 199n14 Bergman, Ingmar 121 – 2, 140 Bergman, Ingrid 160 Bertolucci, Bernardo 171 The Birds (Hitchcock) 74n5 bitmaps and bitmapping 18, 20, 45 – 6, 53n11 Blade Runner (Scott) 62, 193 – 4 Blue (Jarman) 18 Boden, Margaret 24 Bogart, Humphrey 160 Bordwell, David 138 Breaking Bad 22 – 7, 29, 31, 32 Bresson, Robert 106 Broderick, Matthew 112 Brooks, Cleanth 184 – 5 Brown, Christy 27 Buffalo Bill (character) 145 Burj Khalifa 32 – 3 The Cabinet of Doctor Caligari (Wiene) 78 Caché (Haneke) 191 Calloway, Major (character; Contempt) 176 – 7 cameras: digital 37, 47, 48, 50; sound52n3; stationary 127; video 37, 50; see also Steadicam camera work: and the articulation of space 63, 66, 105, 107, 109 – 10; deep focus 36, 61, 76; “dolly-zoom” effect 107; in-camera 23 – 4, 26 – 7, 29, 37; naturally iconic shots 110; setting mood 22, 139; single-camera 197; stationary position 127 Camille (character) 174 Capra, Frank 156, 184 Carroll, Noël 3, 7, 30, 107, 111, 135, 140 – 1, 148, 154, 155 – 7, 186 Casablanca (Curtiz) 160, 191 Cat’s Cradle (Brakhage) 191 Cavell, Stanley 2, 11, 166, 186, 193

CGI imagery see computer imaging Chaplin, Charlie 116 character engagement 136, 140 – 2, 157 – 8 characters: fictional 144; viewer relationship with 140 – 1; see also character engagement; viewer experience Chatman, Seymour 101, 116 Chion, Michael 192 Choi, Jinhee 186 cinema: defined 18; as mass art 3; media constituting 18; medium-specificity claims in 30; and the nesting of media 42, 45; non-art form use of 22; see also cinematic art; digital cinema; film(s); filmmaking cinematic art: appreciative dependence on technology 22 – 30, 34; ontological dependence on technology 21 – 2, 34; relationship with technology 17 – 20 cinematic exposition 61 cinematic narrators: “access” arguments for 101; implicit 100; ontological gap argument for 102 – 3, 104 cinematic philosophy 166, 168 – 9; explicitness objection to 165, 171 – 4; generality objection to 165, 169 – 71; imposition objection to 165, 174 – 7; and the “no author” question 165, 178 Circus World (Hathaway) 66 Citizen Kane (Welles) 36 Clarice Starling (character) 145, 148, 160 Clash of the Titans (Leterrier) 51 A Clockwork Orange (Kubrick) 160 Cobb, Roger (character) 173, 174 cognition: affect and body processes as part of 134, 138 – 9; and character engagement 141 – 2; in film theory and philosophy 131; embodied 132; and emotions 138 – 9, 155 – 6, 159; and empathy 145; nature of 133 cognitive film theory 132 – 3, 157 cognitive revolution 131 cognitive unconscious 133, 136

Index  •  205 cognitivist theory 132 – 3, 148, 154, 159 Cohen, Jonathan 93 collapse, and realism 76, 84 – 5, 89, 93 – 4 Collingwood, Robin George 40 computer imaging 39 – 40, 67; real-time motion-capture systems 37 Conceptual art 193 The Conformist (Bertolucci) 171 Conrad, Tony 186 Constable, Catherine 170 Contempt (Godard) 175 Coplan, Amy 10, 143 – 4, 157 Cotton, Joseph 175, 178 Crimes and Misdemeanors (Allen) 170 Critique of Pure Reason (Kant) 168, 170 Crito (Plato) 170 Croce, Benedetto 40 Cruise, Tom 32 – 3 Curran, Angela 7 Currie, Gregory 199n8 Curtis, Adam 197 Curtiz, Michael 160 Cutwater, Edwina (character) 173, 174 Danto, Arthur 12, 182, 190, 192 – 3, 196, 198, 198n1 Dargis, Manohla 51 The Dark Knight (Nolan) 140 The Dark Knight Rises (Nolan) 27 Darwin, Charles 151 Davies, David 41 Day-Lewis, Daniel 27 – 8 deep focus 36, 61, 76 Deleuze, Gilles 11, 166, 185, 198n1 Delgado, Marcel 32 Demme, Jonathan 160 Dennett, Daniel 133 Desmond, Norma (character) 161 Dialogues (Plato) 167 Diamond, Cora 170 Dick, Philip K. 199n11 Dickson, W.K.L. 22 difficulty principle 26

digital cameras 37, 47, 48, 50 digital cinema 18, 29, 36 – 8; and the achievement argument 22 – 30; appreciation of 50 – 1; challenges of 46 – 7, 49; defending, 42 – 7; and digital dystopia 30 – 4, 38 – 9; recalcitrance of 39 – 40, 47 – 52; relationship to media 43; as technology 20; understanding 40 – 2; see also special effects digital dystopia 4, 17, 30 – 4, 38 – 9 documentaries 2, 22, 77, 84, 93, 119 – 21, 195, 197; see also film “dolly-zoom” effect 107 Dominik, Andrew 101 Donen, Stanley 104 Dr. Chilton (character; Silence of the Lambs) 145 Dr. Zhivago (character) 122 Duchamp, Marcel 193 editing, digital 47, 50 Edwina Cutwater (character) 173, 174 Eich, Eric 140 Eight and a Half (Fellini) 44, 140 Elizabeth Vogler (character) 140 embodiment 132, 133; sensuous 194 – 6 emotion and emotions: artifact 134 – 5, 138, 145; and cognition 138 – 9, 155 – 6, 159 – 60; direct 134 – 5, 138; and facial expressions 151; related to film genre 156; “searchlight” function of 133; study of 131 – 2; sympathetic/antipathetic 134 – 5, 138; three types of 134 – 5, 138; of viewers 135 – 8; see also empathy; sympathy emotional contagion 157 – 8 The Empathetic Brain (Keysers) 150 empathy 142 – 5; automatic/basic 149, 153 – 4, 157; as emotion 157 and emotional identification 159; in fiction film 10; and Hebbian plasticity 150 Empire (Warhol) 186 empirical adequacy 187, 188 – 91

206  • Index Empson, William 171 Eternal Sunshine of the Spotless Mind (Gondry) 186 ethico-political value 195 – 6 “The Evolution of the Language of Film” (Bazin) 61, 73n2 existentialism 169 experimental philosophy (X-Phi) 183 explicitness objection 165, 171 – 4, 198n4 exposition, cinematic 60 – 1, 63, 72, 197 Face-to-Face Imagined Seeing Thesis 108 – 9 facial expressions 10, 151, 160 facial feedback 160 FAP see film-as-philosophy (FAP) “Fearing Fictions” (Walton) 155 Fellini, Federico 140 Ferris Bueller’s Day Off (Hughes) 112 fiction: imagination approach 123 – 4; problem of belief in 123 – 4; what is true in 122 fiction film: empathy in 10; imaginative engagement with 99 – 100 fictional indeterminacy 7, 99, 100, 104, 111, 113 Fictional Showing Hypothesis 124 film(s): as art form 1, 41 – 2; audience response to 8 (see also viewer experience); creator-oriented vs. audience-oriented interpretations 175 – 6; engagement of emotion by 8 – 10; five types of realism in 76 – 85, 88 – 94; medium of 167; as narrative fiction 7 – 8; as philosophic medium 166 – 8; photographic fiction 77 – 8; as realistic medium 4 – 6; see also cinema; digital cinema; fiction film; filmmaking film art 186 film-as-art 185, 187, 192 film-as-philosophy (FAP) 2, 10 – 12, 165 – 6, 182, 184 – 7, 191, 193, 197, 198n4; “easy” solution to 168 – 9; objections to 165 – 78

“Film, Emotion and Genre” (Carroll) 155 Filmic Mediation, Argument from 105 – 6 filmmakers 60, 61, 68, 101, 102, 104 – 6, 117, 124, 125, 141, 145, 165; as artists 1, 3; intentions of 11, 176 – 9; as narrators 8, 121; as philosophers 184; problemsolving by 17, 22 – 3, 25 – 9, 31 – 3, 37 – 8; and the thought experiment 166, 172 – 3; and viewer emotions 8 – 9, 156 filmmaking, 3 – 5, 11, 29, 31 – 3, 38, 138, 141, 186, 192 – 3, 197; digital 4, 37 – 8 filmosophy 166 film-philosophy 166, 185, 198n1 film realism 4 – 5 films d’art 127 film studies 9, 132, 158 film theory 132 – 3, 198n1 Finch, Atticus (character) 118 – 21, 124 – 6 Flash Gordon (character) 125 Fleming, Victor 103 The Flicker (Conrad) 186 fMRI imaging 153 Forgas, Joseph 139 – 40 Foster, Jodie 145 Foucault, Michel 167 frames-per-second (fps) 3 Frampton, Daniel 166 functional magnetic resonance imaging (fMRI) 153 Galili, Doron 44 Game of Thrones (TV series) 79, 88 Gaut, Berys 4, 37, 38, 42, 51, 73 – 4n3, 86, 111; on nesting 45, 46 Gehr, Ernie 186 Geist (Hegel) 193 generality objection 165, 169 – 71 George Bailey (character) 156 Giovannelli, Alessandro 157 Go West (Keaton) 67 Godard, Jean-Luc 175, 186 The Gold Rush (Chaplin) 37, 116 The Good, the Bad, and the Ugly (Leone) 80

Index  •  207 Gordon, Flash (character) 125 Gorky Park (Apted) 123 Gramophone, Film, Typewriter (Kittler) 43 Gravity (Cuaron) 51 – 2 Greene, Graham 177, 178 Griffith, D.W. 1 Grodal, Torben 192 Guido (character; Eight and a Half) 140 Gumb, Jame 145 hand-drawn cinema 18 Hannibal (Slade) 160 Hannibal Lecter (character) 140, 145, 158, 160 Harry Lime (character) 176 – 7 Harry Potter novels and movies, 106, 111 Hathaway, Henry 66 Heald, Anthony 145 Hebb, Donald 150 Hegel, Georg Wilhelm Friedrich 190, 192 – 3 heuristics 111 – 13, 119, 123, 125 High Noon (Zinnemann) 175 historicist heuristic 123 The History of the Peloponnesian War (Thucydides) 101 Hitchcock, Alfred 107, 121, 139 Holly Martins (character) 176 – 7 Holmes, Sherlock (character) 107 Hopkins, Anthony 140, 145 Hopkins, Robert 6, 57 – 8 Huckleberry Finn (Twain) 112 Hume, David 183 Iacoboni, Marco 10, 149 – 53, 157 identification 140 – 1, 154, 155, 158 – 9 illusion 76, 80 – 1, 90, 92 – 3, 187 images and imagery: black-and-white 67 – 8, 70; digital 45, 49; handmade 53n12 “imagination account” of truth 8, 122 – 4 imagined seeing 99; and cinematic realism 57 – 8; details 57; Mediated (MeIS) 109 – 10, 111, 113; Modest (MIS) 109

Imagined Seeing Thesis (IST) 5, 7 – 8, 62 – 7, 69, 71, 100, 106 – 8, 124; Face-to-Face 108 – 9 imagining: fictional 104; “meta-fictional” 112; perceptual 199n8 IMAX format 33, 35n7 imposition objection 165, 174 – 7 imprecision 79 inaccuracy 79 indeterminacy, fictional 99, 100, 104, 111, 113 IST see Imagined Seeing Thesis It’s A Wonderful Life (Capra) 156 James, William 151 Jarman, Derek 18 Jeanne (character, A Tale of Springtime) 168 Joker (character) 140 Jungle Fever (Lee) 178 Jurassic World (Trevorrow) 154 Kandinsky, Wassily 52n10 Kania, Andrew 73 – 4n3, 103, 115, 121 Kant, Immanuel 170 Keaton, Buster 67 Kelly, Gene 104 Kermode, Mark 51 – 2 Keysers, Christian 150, 153, 161 King Kong (character) 112 King Kong, two versions compared 31 – 2 Kitcher, Philip 198n1 kitsch imagery 195 – 6 Kittler, Friedrich 43 – 4 Korda, Alexander 178 Kovács, András Bálint 186 Kren, Kurt 186 Kroes, Peter 19 Kulvicki, John 86 Lane, Anthony 51 language, as medium 20 Lean, David 122

208  • Index Lecter, Hannibal (character) 140, 145, 158, 160 Ledger, Heath 140 Lee, Spike 178 Legends of the Fall (Zwick) 142 Levinson, Jerrold (Jerry) 62, 116 Lime, Harry (character) 176 – 7 Linklater, Richard 78, 168 literature, as artistic medium 45 lithography 46, 53n11 Livingston, Paisley 67, 175 – 6, 177, 179, 184, 190 Locke, John 172 The Lord of the Rings: The Fellowship of the Ring (Jackson) 32 Lubezki, Emmanuel 51 Lumière brothers 22 Machery, Edouard 183 – 4 Magnolia (Anderson) 101 Martin, Steve 173 Martins, Holly (character) 176 – 7 Mason, Michelle 174 – 5 Mastroianni, Marcello 140 The Matrix (A. and L. Wachowski)186 McRae, Gordon 105 media: “agential” sense of 41 – 2, 48; nesting of 18 – 19; overlapping of 19 Mediated Imagined Seeing (MeIS) 109 – 10, 111, 113 Medium-Specificity Value (MSV) claims 30 MeIS see Mediated Imagined Seeing Méliès, Georges 1, 22 Mendes, Sam 100 Meskin, Aaron 93 metaphysical realism 91, 95n13 mimicry: affective 157, 199n8; contagious 159; see also mirror reflexes Mind (journal) 183 mirror images 59 mirror neurons 131, 133, 159, 161; nature and function of 150 – 4; and simulation theory 148 – 9, 161

mirror neuron theory 145 mirror reflexes 157 – 8 MIS see Modest Imagined Seeing mise en scène 194 – 5 mismatches 33 Mission Impossible –– Ghost Protocol (Bird) 32 – 3 Mitte, R.J. 27 – 8 Moby Dick 118 Modern Times (Chaplin) 186 Modest Imagined Seeing (MIS) 109 Molnar, Vera 52n10 mood(s) 134, 139 – 40 Morris, Errol 197 motor imagery 199n8 motor mimicry 157 movie editing software 47, 49 movie musicals 104 – 5 moviemaking see filmmaking Moving Viewers: American Film and the Spectator’s Experience (Plantinga) 134 moving-image media 2; see also film MSV (Medium-Specificity Value) claims 30 Mulhall, Stephen 166, 186, 193 Müller-Pohle, Andreas 53n10 music, as medium 20 My Left Foot (Sheridan) 27 narration: cinematic 60, 99 – 101; implicit 101, 104, 115 – 16; literary 104; by “minimal narrating agency” 110, 111; third-person 101; voice-over 100 – 1, 104, 139 narrative comprehension, schemas in 138 narrative film 2, 5; see also film narratives, impersonal fictional 103 narrators: additional problems regarding 126 – 7; against the ontological gap argument 121 – 4; against the seeing/ showing argument 124 – 6; kinds of 100 – 1; omniscient 104; ontological gap

Index  •  209 argument for 116 – 18; seeing/showing argument for 118 – 21; see also cinematic narrators; narration nesting 18 – 19, 42, 45 neuronal mapping 131 neurophysiology, and the simulation theory 148 – 9, 161 Nichomachean Ethics (Aristotle) 175, 177 “no author” objection 165, 178 Nolan, Christopher 27 Norma Desmond (character) 161 Novak, Kim 107, 108 Nussbaum, Martha 198n1 O’Brien, Willis 31 O’Hara, Scarlett (character) 124 Oklahoma! (Zinnemann) 105 On Film (Mulhall) 166 “ontological gap” argument 115, 116: for the cinematic narrator 102 – 3, 104, 110; slippery slope objection to 103, 110, 113 Oppenheimer, Joshua 195 oral discussion, and philosophy 167 Oz the Great and the Powerful (Raimi) 51 Paul (character; Contempt) 175 Peck, Gregory 126 perceptual imagining 199n8 perceptual realism 60, 63, 67, 73, 74n11; see also realism Persona (Bergman) 121, 122, 140 phenomenology 92, 132 philosophical problèmatique 177, 179 philosophy: experimental (X-Phi) 183; norms of 187 – 91; overlap with art 11 – 12, 197 – 8; and the plurality of artistic value 191 – 7; Western tradition of 167; see also cinematic philosophy; film-as-philosophy; philosophy of art; philosophy of film philosophy of art 182 philosophy of film 1, 5, 157; on cognition and emotion 159

photographs, transparency in 5 – 6, 93, 95n8 photorealism 67, 74nn6 – 7; in animation 29; digital 50 Picolli, Michel 175 Pitt, Brad 142 Plantinga, Carl 9, 148, 154, 158, 159 – 60, 161n1 Plato 167, 170, 171, 172 Poetics (Aristotle) 111 point of view 19, 37, 80, 107 – 8, 126, 140, 141, 153 Pop art 193 Potter, Harry (character) 106, 111 principle of the indeterminacy of fictional imaginings 104 Prinz, Jesse 159 projection: back- 66 – 7; digital 38; emotional 141; 3D 3 protagonists, viewer relationship with 140 – 1 quasi-emotions 189 quasi-transparency 67, 84 – 5 Quine, W.V.O. 183 Radford, Colin 135 rational warrant 187 – 8, 191 Ravenscroft, Ian 199n8 Rawls, John 189 realism: in accuracy and precision 76, 78 – 9; cinematic 57 – 8, 60; and collapse 76, 84 – 5, 89, 93 – 4; epistemic 60, 61, 71, 74n11; examination of 85 – 8; film 4, 5; illusion, 76, 80 – 1, 90, 92 – 3; metaphysical 91, 95n13; perceptual 60, 63, 67, 73, 74n11; precision in 76, 78 – 9, 88 – 9; as recessive form 76, 79 – 80, 88 – 9, 94n6; and representation vs. the represented 87; representational 91, 95n13; and transparency 76, 81 – 2, 90 – 3; types of 6, 76 – 85, 88 – 94 realistic heuristic 111 – 13, 123

210  • Index reality-distance principle 27 – 8 Rebecca (Hitchcock) 139 recalcitrance 39 – 40, 47 – 52 Reed, Carol 176, 178 reflective equilibrium 189 reflective maturity 187, 189, 190 – 1 reflexes 192 Reiner, Rob 173 The Republic (Plato) 172 Reynaud, Émile 18 Reynolds, Debbie 104 Rizzolati, Giacomo 150 Robinson, Tom (character) 118 – 21, 124 – 6 Roger Cobb (character) 173, 174 Rohmer, Eric 168 rotoscoping 169 Routledge Companion to Philosophy and Film (Plantinga and Livingston) 157 Russell, Bruce 169, 179 Sartre, Jean-Paul 169 Scarlett O’Hara (character) 124 schemas 137 – 8 School of Athens (Raphael) 167 Sciorra, Annabella 178 Scott, Ridley 160, 193, 199n11 Seeing Fiction in Films: The Epistemology of Movies (Wilson) 108 seeing/showing argument 116 self-acknowledgment 121 self-other differentiation 143 Selznick, David O. 178 shadow plays 18, 20 Shakespeare, William 171 Shaviro, Steven 132 – 3 Shaw, Dan 10, 133, 145 Sherlock Holmes stories and movies, 107 The Shining (Kubrick) 37 Shklovsky, Viktor 196 The Silence of the Lambs (Demme) 140, 145, 160 simulation theory 157 – 8; neurophysiological support for 148 – 9, 161

Singin’ in the Rain (Kelly and Donen) 104 – 5 smart phone, editing apps on 50 Smith, Murray 11 – 12, 141, 142, 170 – 3, 179 Society for Cinema and Media Studies 132 Socrates 167, 170, 177, 178, 190 software, movie editing 47, 49 Solomon, Robert 168 – 9 sound, addition of 38, 52n3 sound effects 194 – 5 special effects 3, 22 – 3, 67, 367; computergenerated 39 – 40, 67; postproduction 37; see also digital cinema spectatorship see viewer experience Star Trek: The Next Generation 136 Starling, Clarice (character) 145, 148 Steadicam 37 Stella Dallas (Vidor) 136 stereotypes 137 – 8 Stoker, Bram 125 structure, in photographic fiction film 77 – 8 Sunset Boulevard (Wilder) 100, 104, 161 surrealism 195 – 6 suspense 8, 135, 138, 155, 156 Swanson, Gloria 161 sympathy 134 – 5, 138, 142 – 5 A Tale of Springtime (Rohmer) 168 Tarzan movies 58, 59 – 60, 64 technology: and cinematic art 17; digital 3, 44 – 5, 49, 51; of filmmaking 3 – 4; as knowledge 19; relationship to art forms 21 television 18, 34n1, 131, 168, 197 television serials 2, 79 Teyssèdre, Anne 168 A Theory of Justice (Rawls) 189 Thinking on Screen: Film as Philosophy (Wartenberg) 166, 176, 180, 186 The Third Man (Reed) 176 – 9, 186 Thomas Crown Affair (Jewison) 80 Thomasson, Amie 20

Index  •  211 Thomson, Judith Jarvis 103, 171 thought experiment 11, 103, 105, 166, 171 – 4, 179 – 80, 187 thought theory 135; see also thought experiment Thucydides 101 Titanic (film) (Cameron) 135 To Kill A Mockingbird (Mulligan) 126; characters from 118 – 21, 124 – 6 Tom Robinson (character) 118 – 21, 124 – 6 Tomlin, Lily 173 Touch of Evil (Welles) 138 – 9 Transformers series 31, 38 transparency 81; and cinematic realism 59, 76, 81 – 2, 90 – 3; in photography 93, 95n8; see also quasi-transparency Tristan (character; Legends of the Fall) 142 trompe l’oeil painting, 81, 82 truth, “imagination account” of 8, 122 – 4 twofoldedness 137, 144 Ubiquity Thesis 101 Ullmann, Liv 140 unconscious, cognitive 133, 136 Unfriended (Gabriadze) 127 Unlikely Couples: Movie Romance as Social Criticism (Wartenberg) 178 utilitarianism 171 Vertigo (Hitchcock) 107 – 8, 110 Vicky Cristina Barcelona (Allen) 101, 104 video art 2 – 3; see also film video cameras 37, 50 video games 3, 18 viewer experience: and the act of seeing fictions 61 – 4; affective matching 143; allegiance 140 – 2; “audioviewing” 192; character engagement 140 – 2, 157 – 8; and cinematic transparency 59 – 60; and the cognitive revolution 131; and emotion 135 – 8; and Face-toFace Imagined Seeing Thesis 108 – 9;

face-to-face vs. mediated approach 69; identification 155; imaginative engagement with film 99 – 100; and the Imagined Seeing Thesis 62 – 7, 69, 71, 99 – 100, 106 – 8; and Mediated Imagined Seeing 109 – 10; “metafictional” imaginings 112; and Modest Imagined Seeing 109; with paintings 137; processing of truth 122 – 3; and projection 141; self-other differentiation 143; use of heuristics 123 visual effects 194 – 5 visuality condition 82 Vogler, Elizabeth (character) 140 Wachowski, Andrew 184 Wachowski, Laurence 184 Waking Life (Linklater) 78, 168, 169, 198n1 Walton, Kendall 5, 62, 81, 82, 149, 155, 187 Warhol, Andy 186, 193 Wartenberg, Thomas 11, 185, 193, 198n4, 199n7 Ways of Seeing (Berger) 197, 199n14 Weissmuller, Johnny 58, 60 Welles, Orson 138, 176, 178 Whale, James 122 Wilder, Billy 100, 161 Williams, Bernard 172 – 4 Wilson, George 5, 84 – 5, 108, 109 – 11, 115, 116 Winged Migration (Perrin, Cluzaud, and Debats) 48 Witt, Michael 186 Wittgenstein, Ludwig 70, 74n10 The Wizard of Oz (Fleming) 103 Wollheim, Richard 40, 72, 137 Woman I (de Kooning) 41 written texts, as philosophy 167 X-Phi (experimental philosophy) 183