The Philosophy of Motion Pictures

Foundations o f the Philosophy o f the Arts

Series Editor: Philip Alperson, Temple University

The Foundations of the Philosophy of the Arts series is designed to provide

a comprehensive but flexible series of concise texts addressing both
fundamental general questions about art as well as questions about the
several arts (literature, film, music, painting, etc.) and the various kinds
and dimensions of artistic practice.

A consistent approach across the series provides a crisp, contemporary

introduction to the main topics in each area of the arts, written in a clear
and accessible style that provides a responsible, comprehensive, and
informative account of the relevant issues, reflecting classic and recent
work in the field. Books in the series are written by a truly distinguished
roster of philosophers with international renown.

1. The Philosophy of Art, Stephen Davies

2. The Philosophy of Motion Pictures, Noel Carroll

Forthcoming:
The Philosophy of Literature, Peter Lamarque
The Philosophy of Music, Philip Alperson
Black is Beautiful: A Philosophy of Black Aesthetics, Paul Taylor
The Philosophy of
Motion Pictures

Noel Carroll

jfk Blackwell
* C r Publishing
© 2008 by Noel Carroll

BLACKWELL PUBLISHING
350 Main Street, Maiden, MA 02148-5020, USA
9600 Garsington Road, Oxford OX4 2DQ, UK
550 Swanston Street, Carlton, Victoria 3053, Australia

The right of Noel Carroll to be identified as the author of this work has been asserted in accordance
with the UK Copyright, Designs, and Patents Act 1988.

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or
transmitted, in any form or by any means, electronic, mechanical, photocopying, recording or
otherwise, except as permitted by the UK Copyright, Designs, and Patents Act 1988, without the
prior permission of the publisher.

Designations used by companies to distinguish their products are often claimed as trademarks. All »
brand names and product names used in this book are trade names, service marks, trademarks, or
registered trademarks of their respective owners. The publisher is not associated with any product or
vendor mentioned in this book.

This publication is designed to provide accurate and authoritative information in regard to the subject
matter covered. It is sold on the understanding that the pubHsher is not engaged in rendering
professional services. If professional advice or other expert assistance is required, the services of a
competent professional should be sought.

First published 2008 by Blackwell Publishing Ltd

1 2008

Library of Congress Cataloging-in-?ublication Data

Carroll, Noel, 1947-

The philosophy of motion pictures / Noel Carroll.
p. cm. — (Foundations of the philosophy of the arts)
Includes bibliographical references and index.
ISBN 978-1-4051-2024-1 (hardcover : alk. paper) — ISBN 978-1-4051-2025-8 (pbk. : alk. paper)
1. Motion pictures—Philosophy. I. Title.
PN1995.C3574 2008
791.4301—dc22
2007019837

A catalogue record for this title is available from the British Library.

Set in 11.5/13pt Perpetua

by SPi Publisher Services, Pondicherry, India
Printed and bound in Singapore
by C.O.S. Printers Pte Ltd

The publisher's policy is to use permanent paper from mills that operate a sustainable forestry policy,
and which has been manufactured from pulp processed using acid-free and elementary chlorine-free
practices. Furthermore, the publisher ensures that the text paper and cover board used have met
acceptable environmental accreditation standards.

For further information on

Blackwell Publishing, visit our website at
www.blackwellpublishing.com
 To Loretta and Maureen
for taking care of my brothers
 Contents

Acknowledgments viii

Introduction: From Film Theory to the Philosophy

of the Moving Image 1

1 Film as Art 7

2 Medium Specificity 35

3 What Is Cinema? 53

4 The Moving Picture — the Shot 80

5 Moving Images — Cinematic Sequencing and Narration 116

6 Affect and the Moving Image 147

7 Evaluation 192

Select Bibliography 227

Index 233
 Acknowledgments

The author would like to acknowledge the helpful comments, criticisms,

and suggestions offered in the preparation of parts of this book by
Philip Alperson, Susan Feagin, Margaret Moore, Jonathan Frome, Vitor
Moura, Jinhee Choi, Murray Smith, Gregory Currie, Aaron Smuts, Tom
Wartenberg, Cynthia Freeland, Annette Michelson, Jeff Dean/ George
Wilson, Elisa Galgut, Ward Jones, Amy Coplan, Patrick Keating, and
Deborah Knight. They helped make this a better book. I'm the one who
made it worse.
Introduction
From Film Theory to
the Philosophy of the
Moving Image

Though the philosophy of the motion picture — or, as I prefer to say, the
moving image — began early in the twentieth century, perhaps arguably with
the publication in 1916 of The Photoplay: A Psychological Study by Hugo
Munsterberg (a Harvard professor of philosophy and psychology in the
department of William James), the philosophy of motion pictures did not
become a thriving sub-field of philosophy until quite recently. Although
Ludwig Wittgenstein enjoyed movies and attended them often — he especially
liked westerns — he did not philosophize about them. But as of late, the
discussion of movies by philosophers has become quite literally volurninous.
Why?
At least two factors may account for this, one demographic and the
other intellectual.
The demographic consideration is this: for the philosophy of motion
pictures to take root in any serious way, a substantial cadre of philosophers
steeped in motion pictures was necessary in order for a deep and informed
philosophical conversation to be sustained. Historically speaking, that
condition did not begin to be satisfied sufficiently until the late 1960s
and 1970s. By then there was at least one generation of philosophers who
had grown up going to the movies in their neighborhood playhouses, and
also a second generation who had access, through television, to a wide
selection of the history of their own national and/or regional cinema

See for example the bibliography assembled by Jinhee Choi in Philosophy of Film
and Motion Pictures, edited by Noel Carroll and Jinhee Choi (Oxford: Blackwell
Publishing, 2006).
2 FROM THEORY TO PHILOSOPHY

traditions. Thus, toward the end of the twentieth century there were —
suddenly — enough philosophers with enough knowledge about motion
pictures for rich and wide-ranging philosophical debates to begin and for
positions to be refined dialectically.
The demographic situation that I've just described, of course, not only
explains the emergence of the field of the philosophy of motion pictures.
It also accounts for the evolution of cinema studies (or moving image
studies, or just media studies) as a rapidly expanding academic enterprise.
However, though cinema scholars initially followed in the footsteps of the
major film theorists (such as Rudolf Arnheim, Sergei Eisenstein, V. I. Pudovin,
Andre Bazin, and Siegfried Kracauer, among others), by the 1980s cinema
studies, like other branches of the humanities, took what has come to be
called "the cultural or social turn." That is, academics in cinema studies
decided to reorient their field in the direction of what came to be known as
"cultural studies." And in doing so, they left in mid-air many of the discussions
of a lot of the issues that had perplexed earlier film theorists.
Intellectually, a vacuum appeared. And as Richard Allen, a former
chairperson of the New York University Department of Cinema Studies,
has pointed out, philosophers stepped into that gap. In effect, the professors
of cinema studies have ceded what was once a central part of their field to
philosophers of the moving image.
Perhaps needless to say, the philosophical appropriation of many of the
topics of the earlier film theorists is by no means a matter of an alien
colonization. For traditional film theory was always mixed through and
through with philosophy. For example, to take a position on whether film
is or is not art presupposes a philosophy of art. Film theorists also helped
themselves to theses from many other branches of philosophy as well.
Philosophy was never far from the thinking of classical film theorists. So, in
this respect, the philosophy of the moving image is a legitimate heir to film
theory, and not a usurper.
Many of the topics in this book — especially in terms of the questions
asked — reflect the legacy of traditional film theory for the contemporary
philosophy of the moving image. The first chapter addresses the question of
whether or not film can be art. This is undoubtedly the question that got
film theory and the philosophy of the motion pictures rolling in the early
decades of the twentieth century. As we shall see, the issue has been
revived of late due to some recent, highly sophisticated theories about
the nature of photography. As in the past, showing that film can be an art
forces us to look at and think closely about the nature of our object of
study. In this way, meeting the charge that film is somehow precluded from
 FROM THEORY TO PHILOSOPHY 3

the order of art in fact becomes an opportunity to deepen our understanding

of the moving image.
The second chapter concerns what can be called the "medium specificity
thesis." This is the view that the artistic exploration of cinematic possibi-
lities must follow the implicit directives of the medium in which those
possibilities are realized. The precise medium that figures in debates like
this is typically film — that is, photographically or celluloid based motion
pictures. The claim is that the nature of this medium has normative
consequences with regard to that which moviemakers should pursue
and avoid in their artistic endeavors. Quite simply, it is argued, they should
strive to be cinematic and shun being un-cinematic (a condition that is often
also equated with being theatrical). Because this was an article of faith for
so long in the history of film and because the position appeals so seductively
to common sense, it is worth a chapter to scrutinize the medium specificity
hypothesis in depth.
Chapter 3 focuses on the question, "What is cinema?" The title, of
course, comes from the legendary collection of essays by Andre Bazin, one
of the most renowned theorists in the history of the moving image.
It probably goes without saying that nearly every major film theorist has
organized his or her thinking about cinema around this question.
In this chapter, I defend the notion that cinema is best understood in
terms of the category of the moving image. "The moving image?" you
might ask: "In contrast to what?" The short answer is: in contrast to film —
that is, to be more explicit, in contrast to celluloidTmounted, photographi-
cally based film. I will argue that our object of study here is more fruitfully
conceptualized under the broader category of the moving image than it is
under the rubric offilm, narrowly construed.
Film, properly so called, was undoubtedly the most important early
implementation of the moving image (a.k.a. movies), but the impression of
movement — including moving pictures and moving stories — can be
realized in many other media including kinetoscopes, video, broadcast
TV, GGI, and technologies not yet even imagined. Of course, ordinary
folks don't haggle over whether a videocassette is a movie or not. And
neither, I will argue, should philosophers.
This, of course, is a conceptual point. It is not my intention to initiate a
crusade for linguistic reform. That would be quixotic. In everyday speech,
many use the labelfilm to designate things that are really the product of other
media. For example, they may refer to a high-definition video as a film.
This is rather like the use of the name "Coke" for any cola, or "Xerox"
for any copying machine, or "Levi" for all jeans, or "zipper" for every slide
4 FROM THEORY TO PHILOSOPHY

fastener. In these cases, the names of the earliest, most popular entrants to
a field get used — in an inaccurate way, strictly speaking — to refer to their
successors and even their competitors. Because of this tendency, we under-
stand why sometimes digital cinematography will get called film, though it
does not involve the use of film (i.e., the use of a filmstrip). There is little
damage here in the daily course of events. Nevertheless, as we shall see, it
can and does cause philosophical mischief.
Chapter 4 follows the discussion of the nature of cinema with an analysis
of the nature of the cinematic image, construed as a single shot. Obviously,
these two topics are related, if only because throughout the history of
motion pictures, the temptation has endured to treat movies as if they were
equivalent to photographs, where photographs, in turn, are conceived of as
modeling single shots.* That is, many have attempted to extrapolate the
nature of cinema tout court from the nature of the photographic shot. Thus,
it is imperative for the philosopher of the moving image to get straight
about the nature of the shot.
Of course, typical motion pictures, excluding experiments like Andy
WarhoFs Empire, are usually more than one shot in length. Shots are
characteristically strung together in cinematic sequences, usually by
means of editing. Chapter 5 examines prevailing structures of cinematic
sequencing from a functional point of view. In this regard, one might see
the analysis here as returning to an exploration of the terrain that was of the
greatest interest to the montage theorists of the Soviet period.
Moreover, in composing the image series in a motion picture, one not
only standardly combines shots to construct sequences but then also joins
sequences to build whole movies. Consequently, in the second part of
chapter 5 we turn to the most common way of connecting sequences to
make popular, mass-market movies — a process that we call erotetic narra-
tion, that is, a method of generating stories by means of questions the
narrator implicitly promises to answer.
Just as chapter 5 revisits, with a difference, the concerns of the monta-
gists, so chapter 6 also tackles a subject near and dear to the heart of Sergei
Eisenstein — the way in which cinema addresses feeling. Unlike Eisenstein,
however, in this chapter I will take advantage of recent refinements in the
philosophy of mind and cognitive science in order to appreciate the wide
gamut of ways in which movies can engage our affective reactions. I will try

For example, we call an afternoon photographing fashion models "a shoot"

and a successful photo a "nice shot."
 FROM THEORY TO PHILOSOPHY 5

to be precise where Eisenstein was often impressionistic. Nevertheless, at

the end of the chapter, I will attempt to argue that at least one of
Eisenstein's insights into the mechanics of influencing audience affect was
spot-on. These audience responses are what we can call mirror reflexes;
and Eisenstein was dead right about their significance.
The last chapter in the book focuses on the evaluation of motion
pictures. Though many might be tempted to maintain that they are wholly
a subjective affair, I will try to demonstrate that quite often movie
evaluations can be shown to be rational and objective. Often, this relies
upon determining the category in which the movie under discussion is to
be correctly classified. Since more than one category comes into play when
evaluating the range of available motion pictures, the position defended in
our last chapter is called the pluralistic-category approach.
The pluralistic-category approach contrasts sharply with traditional
approaches which attempt to identify a single category — often called the
cinematic — into which all movies allegedly fall and in accordance with which all
motion pictures can be evaluated. The pluralistic-category approach, instead,
accepts that there are many categories of motion pictures — from comedies
and splatter films to travelogues and onto instructional videos about install-
ing software or wearing condoms — and that, inasmuch as specimens of
these different categories are designed to fulfill different functions, they call
forth different criteria of evaluation.
In the past, the theory (or philosophy) of cinema was often pursued in a
very top-down manner. One identified the essence of cinema — usually
understood in terms of photographic film — and then attempted to deduce
accounts of every other feature of film on the basis of that essence.
The conception of the moving image championed in this book is much
looser. Although I attempt to define the moving image, I do so in a way
that remains wide open not only to the media in which moving images
may be realized, but also in terms of the purposes moving images may
legitimately serve. Our characterizations of the elements of the moving
image — the shot, the sequence, the erotetic narrative, and its modes of
affective address — are not deduced from first principles. Rather, we
proceed from topic to topic in a piecemeal fashion.
Thus, the end product is in nowise as unified as the philosophies of
our very distinguished predecessors in film theory. Instead, our results are
pluralistic. Nevertheless, that appears to be where the argument leads us.
Of course, whether or not that is really so is for you to decide.
Therefore, read on.
b FROM THEORY TO PHILOSOPHY

Suggested Reading

There are several anthologies that may be useful for readers interested in
exploring the field of the philosophy of the motion pictures. A very serviceable
introductory text is The Philosophy of Film: Introductory Text and Readings, edited by
Thomas E. Wartenberg and Angela Curran (Oxford: Blackwell Publishing, 2005).
A slightly more advanced textbook is The Philosophy of Film and Motion Pictures,
edited by Noel Carroll and Jinhee Choi (Oxford: Blackwell Publishing, 2006).
Also valuable are any of the many editions of Oxford University Press's Film Theory
and Criticism, edited by Marshall Cohen, Gerald Mast, and later Leo Braudy. For
philosophical purposes, the earlier editions are to be preferred to the later ones.
Two anthologies that present a wide selection of topics in the discipline of the
philosophy of motion pictures are Philosophy and Film, edited by Cynthia Freeland
and Thomas E. Wartenberg (New York: Routledge, 1995), and Film Theory and
Philosophy, edited by Richard Allen and Murray Smith (Oxford: Oxford University
Press, 1997). The journal Film and Philosophy, edited by Daniel Shaw, also publishes
essays in this area of inquiry on a regular basis.
Although the present volume is concerned with the philosophy of motion
pictures, many philosophers are also interested in the prospect of philosophy in
motion pictures. This involves the interpretation of specific motion pictures
as illustrations of — and sometimes even as original proposals of — philosophical
themes. An example of this approach might be a reading of Groundhog Day as a
version of Nietzsche's myth of the eternal return. Many of the articles in the journal
Film and Philosophy are in this genre. An important recent collection devoted to the
topic of philosophy in film is Thinking Through Cinema: Film as Philosophy, edited by
Murray Smith and Thomas E. Wartenberg (Oxford: Blackwell Publishing, 2006).
This book derives from the special issue of the Journal of Aesthetics and Art Criticism
64:1 (Winter 2006).
Chapter One
Film as A r t

The philosophy of the motion picture was born over the issue of whether or not
film can be art. The question here, of course, was not whether allfilmsare art —
surely raw surveillance footage is not. Rather the question was whether some
films could be art. At the beginning of the twentieth century, the lurking
suspicion in many quarters was that, since films are photographic, they are
somehow precluded from the order of art. That is, photography as a medium —
and film as an extension of photography — lacked the capacity to create art,
properly so called. Thus, no films could possibly be artworks, since photo-
graphic media, whether still or moving, by their very nature, are incapable of
producing art. This debate, moreover, was (and remains) a philosophical one,
because it presupposes views about what is required of an object, if it is to
be legitimately classified or categorized as an instance of the concept of art.
To contemporary ears, the contention that no films can be art undoubt-
edly sounds bizarre insofar as some films — such as Citizen Kane — number
among our paradigms of twentieth-century art. We are sure that some
films are art because we think we've seen a number of straightforward
cases; in fact, we believe we've seen quite a few. We even, call some films —
such as Robert Bresson's Mouchet, Federico Fellini's #^?, and Hou Hsiao
Hsien's Millennium Mambo — art films. But things were far less obvious when
film first arrived on the cultural scene. For in the first decade of the
twentieth century there were not that many, if any, agreed-upon master-
pieces of cinema for the friend of cinema to cite.
However, the problem went deeper than merely the lack of available
evidence. Skeptics concerning the possibility of film art thought that they
had prior reasons that conclusively estabHshed that no film could be an
artwork. Those reasons had to do with the supposition that film was
essentially photographic in nature. Moreover, cinematography, on their
view, was merely photography that moved. So, since these skeptics were
convinced that photographs, given their essential nature, could not be art,
they therefore surmised that moving photographs couldn't be either.
8 FILM AS ART

That is, the argument that films could not be art rested upon antecedent
arguments — arguments that had been voiced throughout the nineteenth
century — which maintained that photography could not be art. That the
first moving pictures were films — the very name of which betrays their
photographic (celluloid) basis — encouraged skeptics who contemplated the
prospects for film art to extrapolate their reservations about the possibility
of photographic art to the products of the new technological medium of
motion pictures. The skeptics did so just because they believed that motion
pictures were nothing more than moving photographs. Thus, in this chapter,
in order to discuss the possibility of film art, we will have to spend a very
great deal of time talking about photography.

Against Photography
So why, from the outset, did skeptics challenge the artistic credentials
of photographs? Even in the nineteenth century were there not already
photographs — by people like Julia Cameron Mitchell, for example — which
were undeniably artworks? However, for often subtle reasons, many
skeptics were not prepared to grant this.
Skeptics about the potential for an art of photography begin by taking
note of the fact that photographs are mechanical productions. Photographs
are the causal consequences of a series of physio-chemical processes — the
exposure of silver haloids to light. A photograph was the sheer physical
output of the operation of brute laws of nature. Thus, the skeptics concluded
that photography precluded the creative* imaginative, subjective, expressive
contribution of the photographer. Photographic images, according to the
skeptic, were the automatic product of a machine, not of a mind. Press
a button and you get a picture. But art is not made thusly. Art requires an
artist who expresses herself through her work and who imposes form or
style upon her materials. Yet that cannot be achieved by a machine slavishly
grinding its way through a sequence of physical states.
According to the philosopher, Benedetto Croce, "if photography be not
quite art, that is precisely because the element of nature in it remains more
or less unconquered and ineradicable" in a way that blocks the transmission
of artistic intuitions and points of view. The visual artist gives form to his
subject, grouping its elements in a way predicated upon securing a specific
effect. But the photograph allegedly does nothing more than mechanically
reproduce the formless flow of reality as it passes before the lens of
the camera.
 FILM AS ART 9

This view of photography, moreover, was confirmed for many by the

earliest film screenings, such as the newsreels of the Lumiere Brothers in
France. Called actualites, these one-minute shorts captured street scenes
from home and abroad, travelogue footage, domestic affairs, folkways,
historic events, everyday details, and the like. These films appeared to be
little more than documentary records of whatever flitted before the cameras
of the itinerant Lumiere photographers. People wandered in and out of the
frame. Audience fascination was with the reproduction of reality in all its
shapeless bustle. Putatively, it was not as though some comment or feeling
on the part of the photographers regarding their subjects emerged from
the screen. What there was to see and wonder at was arguably nothing more
than the simulacra of reality mechanically reproduced with neither the
intervention of a subjective artistic interpretation nor formal invention.
On the one hand, the skeptic argued that inasmuch as photography was
mechanical, it foreclosed the opportunity for creativity — both expressive
and formal — on the part of the artist. An objective physical process, it left
no space for the subjective influence of the artist. Photography was simply
a mechanism. Moreover, the skeptic also drew attention to the kind of
mechanism it is. It is a machine for automatically reproducing whatever
finds its way in front of the camera lens. In this regard it is nothing but a
recording device. When the camera is pointed at a street scene, it mechani-
cally records reality. And, the skeptic maintains, just as reality comes
without expressivity or form, so does a mere mechanical recording of it.
But what of the case where what the motion picture camera is trained
upon is not an everyday street scene, but a dramatic enactment of a fiction?
Might not that be an artwork? After all, aren't the actors expressing their
thoughts and feelings about the characters they are playing? However, the
skeptic rejects this possibility, again by stressing that photography, moving
or otherwise, is essentially a process of exact recording and, therefore,
incapable of being art in its own right.
Consider the case of the museum-shop postcard of Munch's painting The
Scream. It is not an artwork in its own right, though The Scream is. The
postcard is at best the photographic reproduction of an artwork. It is not an
artwork itself. Likewise, a CD of Bizet's Carmen is not an artwork itself; it
is the recording of one. The postcard and the CD give us access to already
pre-existing artworks, but they are not artworks themselves. They are
recording media, not artistic media.
The photograph of The Scream does not make art; it only preserves art.
Photographs and, by extension, films are recordings; they are rather like
time capsules — temporal containers — that convey artistic achievements
10 FILM AS ART

that are far away in space and/or time from us. But just as the jar in which
the caviar is sold is not the delicacy itself, neither is the photographic
recording of the artwork — whether of a painting or a dramatic fiction — the
work of art. No one would mistake the photographic plates in an art history
book for the works of art they showcase. Furthermore, this argument can be
applied to dramatic films. If a filmed drama is thought to be an artwork, then
it is not the film that is the pertinent artwork. Rather it is the drama —
enacted before the camera and preserved on film — that is the artwork
proper. Fiction films are not artworks under their own steam; they are at
best slavish, mechanical recordings of theatrical or dramatic artworks — that
is, recordings of artworks staged in front of the camera.
For decades, many contended that the German expressionist film The
Cabinet of Dr. Caligari was not really an example of cinematic art, but only
the mechanical recording of a stage play. Yes, the sets are expressive —
expressive of a kind of existential claustrophobia. But this, it was said, is
the product of the work of the set designers; the filmmaker simply
photographed the antecedently expressive sets that stood before the camera.
Similarly, the acting is expressive. But that is the contribution of the
performers. The filmmakers merely mechanically recorded that which the
actors creatively invented.
Whatever artistry is attributable to Caligari, the skeptic protests, belongs
to the enactment of the narrative and the design of the stage sets which were
automatically imprinted on the film stock as the camera routinely cranked
on. Surely, the skeptic charges, exposing film is not the work of an artist.
If there is art in Caligari, then it is due to the work of the writers, the actors,
and the set designers and not to the people running the cameras. If
the director, Robert Wiene, added art to the proceedings, it is as a theatrical
director of actors and not as a cinematic director of cameras that he makes
his mark. The film as such simply preserves an already existing dramatic
accomplishment for posterity. It does not make art itself; at best it makes
available art from elsewhere in space and/or time. The Cabinet of Dr. Caligari
is just like the postcard of Munch's Scream, except that it moves.
Because photography, still or moving, was regarded as the mere
mechanical recording or copying of reality, its potential for creating art
was denied. For it seems reasonable to presuppose that art requires the
addition of artistic expression and/or formal articulation to its subject
matter. And a mere copy adds nothing. Unlike a painting of a street scene,
the Lumiere actualite reproduces mechanically the look of said reality sans
artistic invention. On the other hand, where the subject of a film is a
fiction, such as The Cabinet of Dr. Caligari, then, allegedly, the art is in the
 FILM AS ART 11

dramatic enactment of the story as it played in front of the camera. The art
is not in the camerawork — not in the film as film. In contrast to theater,
which creates artistic fictions, film at best records their performance.
To summarize: if a film is a nonfiction — say, a Lumiere newsreel — it cannot
be art, since art, properly so called, affords the scope for artistic expression
and/or formal invention. Machine reproductions do not afford the scope for
artistic expression and/or formal invention, since they are, by definition,
automatic mechanical transcriptions of how things look, So insofar as non-
fiction films are machine reproductions, they cannot be artworks.
Nor, on the other hand, argues the skeptic, are fiction films art. Again,
the problem is that film is essentially a recording medium. Where a fiction
film is said to be art, the skeptic maintains that this confuses the recording
of something artistic — namely, the performance of a drama — with an
artwork in its own right. If Charlie Chaplin created art when in The Gold
Rush he made us see his boot as a turkey dinner, it was his pantomime
before the camera and not the filming of it that belonged to the Muses. But
in that case, there is no art of the film as such; there is only theater art — in
Chaplin's case mime — captured on film.
Of course, many remained unconvinced by these skeptical arguments
against photography and then film. And in the process of resisting the
skeptic's case, the philosophy of film took flight. For in addressing the skeptic's
allegations that films could not be art on a priori grounds, the friends of film
began to clarify the ways in which films could qualify as authentic candidates
for the status of artworks.
Some theorists, like Rudolf Arnheim, pointed out that in spite of their
mechanical dimensions, both photography and film were not perfect
replicas of that which they represented and that, in virtue of the ways in
which they diverged from being perfect mechanical recordings, photo-
graphs and films could be expressive. A low-angle shot, for example, can
portray a typewriter as massive, thereby commenting upon it. In his The
General Line, Sergei Eisenstein uses shots like this to underscore the
oppressiveness of the bureaucracy that employs such typewriters. Similarly,
a filmmaker can use a distorting lens to make a point; laughing faces can be
made to appear hideous by means of a wide-angle lens, thereby unmasking
the cruelty that can lurk underneath a smile.
Moreover, as silent film theorists frequently emphasized, film is not
reducible to moving photography. Editing is at least as important to film as
we know it. But since film editing can rearrange the spatio-temporal
continuum, including the sequence of events in a play, a film need not
be a mere slavish recording of anything — of either an actual everyday event
12 FILM AS ART

or the theatrical performance of a fictional one. Editing gives the film-

maker the capacity to rearrange reality creatively as well as to body forth
imaginatively alternative fictional worlds — worlds of works of art.
Considerations like these have by now won the day — hardly anyone
presently questions whether or not the motion picture is an artform.
Indeed, some maintain that it was the artform of the twentieth century.
Nevertheless, the arguments of the skeptics should not be dismissed as
nothing but cranky conservatism. The skeptics recognized certain pro-
found features of film whose possible implications enthusiasts sometimes
failed to appreciate. Thus, one can learn about film from the skeptics, even
in the process of trying to refute them. So, for the rest of this chapter, I will
be taking a long look at the skeptic's case against the possibility of film art
both in order to see how it goes wrong and what it gets right.

The Skeptical Argument

The skeptic's leading argument against the possibility of film art has three
major movements. It goes like this:

1 Photography is not art.

2 A film is at best a photographic recording of a work of dramatic art.
3 Therefore, film itself is not art.

This skeptical argument, in turn, is supported by three further arguments,

each of which is intended to bolster the first premise of the leading argument.
For convenience sake, let us call these three supporting arguments respec-
tively: (1) the causation argument; (2) the control argument; and (3) the
aesthetic interest argument. Typically the skeptic does not appear to have
additional arguments to reinforce the second premise of the lead argument.
This premise is sometimes treated as if it were absolutely unobjectionable.
This, however, is not so, and we will return to this lacuna after reviewing the
three arguments in behalf of the first premise of the skeptic's master argument.

The Sheer Physical Causation Argument

Crucial for the skeptic's case against film is the premise that photography
is not an art. One way in which the skeptic supports this is by means of
the sheer physical causation argument (henceforth, usually just called "the
 FILM AS ART 13

causation argument"). The skeptic presupposes that anything worthy

of being classified as art is necessarily an expression of thought — the
expression of an idea, a feeling, an attitude, a point of view, or the
articulation of a formal conception. If the snowflakes falling in my drive-
way assemble themselves in such a way that the results look exactly like one
of Rodin's statues of Balzac, the skeptic would argue (and I think that most
of us would agree) that that thing in my driveway is not a work of art.
Unlike Rodin's statue of Balzac, there is no thinking, expression, or formal
invention behind our putative snow "sculpture." Indeed, does it even make
sense to call it a sculpture, since, given the case as we've described it, there
was no sculpting involved. And, in any event, it is not the product of a
sentient human agency. It is a natural event, the result of a mindless
sequence of causes and effects, the outcome, albeit freakish, of the sheer
operation of physical laws. That it resembles Rodin's Balzac, or, for that
matter, Balzac, is an accident.
True, things that look like my snowdrift are usually the product of a
mind — but not always. There are shadows cast by the mountains of the
moon that from a certain angle are said to resemble a face. But such
anomalies are not even representations, let alone artworks, the skeptics
argue, since they are not the products of mental acts. They are not the
result of human intentions, but of the sheer movement of matter. They do
not express thought, since they do not express anything. If expression
involves bringing something that is inside, such as a thought, outside, these
stochastic flukes have nothing inside to manifest outside. They are com-
pletely the product of the random operation of physics, not psychology.
They involve -no thinker; they express no thinking; they are neither
portraits nor artworks properly so called.
But what does this have to do with photography, including moving
photography? A photograph can occur sans human agency. Imagine a
camera falling off the back of a truck in such a way that it starts to click
off perfectly focused snapshots. These images would not meet the skeptic's
criterion for art status, since they express no thoughts. How could they?
No one is taking the picture.
Likewise airport surveillance cameras do not require operators; they
may be fully automated. That is why, in the normal course of things, we do
not count that footage as art. For, once again, it is not plausible to suppose
that any thought is expressed by an automated surveillance camera.
Perhaps an artist can take such footage and re-edit it in such a way that
it counts as art. But that is another story. Raw surveillance footage is not
art, if only because it is mindless. So again: that a photographic image of x,
14 FILM AS ART

moving or otherwise, enables us to recognize x in it does not entail that the

image of x is an intentional product. Thus, we might also call this argument
the absence of intentionality argument.
A photograph need not be the product of an intentional agent; it may be
the consequence of a concatenation of sheer physical events. For example,
we might imagine certain improbable but nevertheless logically possible
instances of photographs that express no thoughts. Think of a cavern in the
recesses of which there is a puddle of photographic salts; overhead there is
a tiny crack in the ceiling of the cave that admits a ray of light. This light, in
turn, functions like a pin-hole camera and fixes the outline of a neighboring
volcano on the floor of the cave. A randomly occurring, "natural" photo-
graph like this would not require a human photographer nor a human
intention. And for that reason, we would not count an image such as this as
an artistic representation of a volcano, no matter how nicely it appeared to
resemble one. For, since it lacks a photographerj it expresses no thought
about the volcano.
What these examples are intended to show is that photography does not
necessarily require a camera operator. There could be cases of things that
we would be perfectly willing to call photographs — i.e., genuine products
of the process of photography — that are the outcome of a sequence of
physical causes without any human intervention. Stripped down to its essence
then, a photograph does not require human input, and, consequently, need
not involve the expression of human thought or the implementation of human
intentions. The expression of thought, in other words, is not an essential
feature of photography. Therefore, with respect to its essential nature,
photography is not art.
In this regard, what might be called the essential photograph differs
profoundly from the essential painting, which, for the purposes of argu-
ment, we may take to be paradigmatic of art status. All paintings putatively
have the property of intentionality which means, according to the skeptic,
that paintings are (1) about something because (2) the painter intends this.
A painting is intentional both in the sense that it is directed at something
and in the sense that it is a vehicle for the thought of its maker (such as a
comment on or interpretation of whatever the painting is about). But
essential photographs are not intentional in either of these senses. They
can be the result of completely causal processes lacking any intentional/
mental mediation. Nor is the appearance of the volcano in our fanciful
example of the aforesaid "natural" photograph about anything; it is just a
happenstance collision of events with no more content than a hghtning
bolt. Such an essential photograph could not express the thought that the
 FILM AS ART 15

volcano is awesome, as a painting of a volcano might, since an essential

photograph of the sort imagined lacks an author to bear the thought, whereas
a genuine painting, according to the skeptic, requires a sentient artificer by
definition.
So the sheer physical causation argument ultimately boils down to this:

1 If something is art, then it must involve the expression of thought.

2 If something involves the expression of thought, it requires a mental
dimension, also known as intentionality.
3 Intentionality is not an essential feature of photography.
4 Therefore, photography (including moving photography) is not essentially
or necessarily art.

But if photography is not essentially art, then photography — whether still

or moving — is not art qua photography, i.e., not in virtue of being
photography,
Furthermore, whereas a painting can depict imaginary things, a
photograph allegedly always delivers up the appearance of something that
quite literally stood before the camera. Photographs present the specta-
tor with what philosophers call referentially transparent contexts — in
other words, in the typical case, the photograph P is the effect of cause G
(a volcano, for instance) in such a way that the existence of P allows us
to infer the existence of G (a volcano or, at least, something that looks like
a volcano).
Paintings, in contrast* are referentially opaque. From the painting of
a satyr, you are not advised to infer the existence of a satyr. Paintings
can be about whatever painters imagine; their subjects do not necessarily
exist. This too is a feature of intentionality, sometimes called intentional
inexistence. On the other hand, that which gives rise to a photographic
image must have existed and the photograph gives one inferential grounds
to affirm it. Thus, this is another way in which the photograph lacks
intentionality.
But painting is a representational artform in large measure because
it involves intentionality — which feature of painting is intimately related
to its capacity authorially to express thoughts concerning its subjects.
Instead, photography, seen in terms of its minimal or essential nature,
produces its images by means of mechanical causation rather than by means
of intentionality. Consequently, photography as such — photography qua
photography — is not a representational artform and neither is film, since
film is basically to be understood as merely moving photography.
16 FILM AS ART

The Control Argument

In addition to the causation argument, the skeptic may also bring forward the
control argument to elaborate the case against photographic art. According to
the skeptic, an artist working in a medium requires control over the
medium. Why? If art is the expression of thought, then in order to express
one's thought clearly, you need to have control over every detail of the
medium. If one lacks control, wayward effects may intrude into the results,
undermining the thought or the feeling one intends to express. To articulate
clearly one's thoughts, one must have a handle on the elements that go into
presenting such thoughts. Ideally, it is said, every element in a genuine
artwork either serves or should be capable of serving the expression of the
artist's thoughts. To express a thought or a feeling is to clarify it; to clarify
a thought or a feeling requires the capacity to hone every element that makes
the thought or feeling mor e definite, and the capacity to expel or modify
any element that might render it ambiguous. That is, art as the expression
or articulation of thought or feeling requires consummate control over the
relevant medium.
Painting evinces the requisite degree of control, since everything in a
painting is ostensibly there because the painter intentionally put it there.
Also, as we have seen, the painter — again in virtue of intentionality — can
clarify her ideas or emotions by imagining whatever she needs to sharpen
her thoughts, not restrained by what is. However, the photographer,
including the cinematographer, supposedly has nothing approaching these
resources. The photographer/cinematographer is at the mercy of what is
in a way that it seems a painter never is.
The camera is akin to a mirror; the lens captures whatever looms before
it, whether or not the camera operator is aware of the details she records
a n d / o r whether she intends to photograph them. For this reason, many
historical spectacles are marred by the intrusion of unwanted anachronisms.
A telephone pole, for example, may appear in the background of a medieval
romance to the embarrassment of the camera crew just because it was a
detail of the actual mise en scene* albeit not one the cinematographer noticed
or desired to record. However, since photography is a causal process, if
something stands before the camera, the physical procedure guarantees, all
things being equal, that it will appear in the image, no matter what the
cinematographer intended.

Tlie scenographic layout — what is next to what and in front of what else, etc.
 FILM AS ART 17

Were a telephone pole to appear in the background of a painting of

what otherwise looked to be a medieval scene, we would suppose the
painter intentionally drew it there and then go on to interpret what
the painter meant by doing so. But when comparable anachronisms erupt
in historical films, as they do with unnerving frequency, we infer that they
arrived there because things got out of control. Since we know that the
photographic process is automatic, abiding blindly by the laws of chemistry
and optics, we understand that things appear, to the chagrin of filmmakers,
in motion picture images that were never meant to be there.
There are so many things in front of the motion picture camera that may
make an inadvertent debut in the final print of a film. Indeed, there are
typically far too many details — large and small — for the cinematographer
to keep track of. Many surprises appear uninvited in the finished film and
the filmmaker just has to learn to live with them. On the other hand,
everything in a painting is putatively there because that is what the painter
intended. The painter is not surprised by what she finds in her painting,
because she put it there. What would we make of a painter who, looking at
her medieval scene, asked: "How did that telephone pole get there?" The
painter has ultimate control over what inhabits her canvas.
Not so the photographer/cinematographer. Details appear in the photo-
graphic image, whether still or moving, just because they were in the
camera's visual field even if no one on the film crew took heed of their
presence. Furthermore, there are always an indefinitely large number of
potential gaffs lying in wait for the unwary cinematographer. In this regard,
the painter must be said to exert far more control than either the still or
moving picture photographer.
Since paintings are intentionally produced, painters enjoy a degree of
control such that there is nothing whose appearance in her pictures takes
her aback. But photographers are often startled by what they find in their
photos, because the causal process they initiate evolves in utter obliviousness
of whatever the photographer believes, desires, or intends.
The photographer not only falls far short of the level of control the
painter imposes over what can show up in her image; the photographer
also lacks the painter's imaginative freedom to picture whatever she will —
absolutely unrestrained by what is — inasmuch as the photographer can only
present what can be placed before a camera. The painter can picture a
werewolf for us; the filmmaker can only give us a photographic picture of
a man got up in a werewolf outfit. The painter, in this regard, has greater
freedom than the photographer, which, in turn, entails greater control
over shaping her subject matter and the expression of her thiiiking about it.
18 FILM AS ART

Because art involves the expression of thought, only media that afford a
high level of control are suitable for the purposes of artmaking. The
clarification of thought requires a degree of plasticity. To get across an
idea or a feeling clearly you may have to depart from the way the world is —
either by adding to it or by subtracting from it. Photographic media,
however, are inhospitable in this regard. They are tied to what is and they
incorporate details, like inadvertent telephone poles, that are beyond the
photographer's ken. Unlike painting, photography is not an artform; it lacks
the degree of control evinced by an unquestionable artform like painting. And
for the same reasons, moving photography suffers the same liabilities as does
still photography. So, once again, we must conclude that film is not an art.

The Aesthetic Interest Argument

A third skeptical argument to the conclusion that photography/cinemato-
graphy is not art can be labeled the aesthetic interest argument. An aesthetic
interest is an interest that we take in something for its own sake, that is,
because of the kind of thing it is. We take an aesthetic interest in the novel
Remembrance of Things Past when we are preoccupied by the kind of thing it
essentially is — an expression of thought — rather than as a set of books the
color of the covers of which contrasts appealingly with the drapes in our
living room.
According to the skeptic, photography cannot command aesthetic interest
because a photograph is strictly analogous to a mirror. When we are
interested in a mirror the object of our interest is whatever is reflected
there and not the mirror itself. When one brushes off one's jacket in front of
the looking glass, one is concerned with whether there is any telltale lint
anywhere. We look into the glass in order to learn something about our
jacket; our interest is in our jacket, not in the mirror reflection for its own
sake. We are only interested in the mirror reflection because of what it tells
us about something else, namely our jacket. Thus, the interest one typically
has in a mirror is not an aesthetic interest.
Moreover, the skeptic argues that what is true of mirrors is also true of
photographs. We are interested in them for what they show — historical
events, atrocities, family weddings, dead relatives, and so on — and not for
their own sake. It is the object in the photograph and not the photograph as
an object that engages our attention.
With respect to the art of painting, it is the painting as object (indeed, as
art object) that commands our attention and it does so because it is an
 FILM AS ART 19

expression of thought. Mirrors are not expressions of thought; they are

optical appliances whose images are produced naturally, not intentionally.
They relay appearances to us mechanically. There is no point in taking an
interest in them in the way that we care about paintings. In fact, the only
way to take an interest in the images in mirrors, if our interest is not
scientific, is to be occupied by what they show us — like the police cruiser
with the flashing blue light in our rear-view mirror.
Likewise, the skeptic argues that photos record appearances; they do
not convey thoughts. We are interested in a photograph of x — say of
Osama Bin Laden — because we are interested in seeing the way he looks.
We are not interested in the photograph for itself, but only as a mechanism
that transmits appearances. The interest is not aesthetic, since it is not an
interest in the kind of thing the photograph is, but rather an interest in the
person whose appearance it delivers.
Another way to make the skeptic's point here is to say that the skeptic
is claiming that photographs, properly so called, are transparent. We have
already seen that photographs are referentially transparent in contrast
to the referential opacity characteristic of the productions of intention-
ality. That is^ there is a causal connection between the photo and the
object that gives rise to it such that the photo supplies grounds for
inferring the existence of the object in the photo. My appearance in
the surveillance footage supplies evidence that I was in the bank on the
night of the robbery. All things being equal, the jury will legitimately
infer it was so. But in addition to providing this inferential warrantj
skeptics are also maintaining that photography as it is usually employed —
what has been called "straight photography" — is also transparent in the
sense that it provides us with indirect perceptual access to the object
it represents. Just as seeing an object through a rear-view mirror is a special,
indirect way of seeing the police cruiser behind me, so a photograph is a
special way of seeing an object;
That is, we see objects through photographs just as I see objects through
mirrors. In this way, photographs, like mirrors, are perceptually transpar-
ent. But, the skeptic adds, if photographs are perceptually transparent,
then it is not the photograph in its own right that preoccupies us. It is
the object to which the photograph gives us perceptual access that interests
us, either for its own sake, or otherwise. In either case, our interest is not
an aesthetic interest. We are not interested in the photograph as an object;
we are not interested in the photograph as an artwork; we are interested
in it because of what it shows and not how it shows it. Moreover, if a
photograph cannot support aesthetic interest, then why suppose it is an
20 FILM AS ART

artwork? For, the skeptic presumes, an artwork is just the kind of thing
that enjoins aesthetic interest — that is, an artwork is the kind of thing
in which we are interested because of the kind of thing (the kind of object)
it is (rather than because of the kind of object to which it affords us
perceptual access).
Suppose that the object of a photograph is something very beautiful —
like a flowering rose — of the sort whose appearance we are typically said to
value for its own sake. The, skeptic maintains that the putative aesthetic
interest here is the beautiful rose. Our attention is not drawn to the
photograph itself — as it would be were it a genuine artistic object — but
to the appearance of the flower which the photograph serves up transpar-
ently and mechanically. It is not, so to speak, a beautiful photograph but
a photograph of a beautiful thing. And it is the beautiful thing in the'
photograph that is the object of our aesthetic interest i properly speaking.
Likewise, when the movie presents us with a beautiful vista, it is the vista
that is the object of our aesthetic interest, not the movie that allegedly
merely records it.
The Soviet filmmaker Dziga Vertov called film the "microscope and
telescope of time. ,, Just as perceptual apparatuses like microscopes
and telescopes enable us to penetrate very small and very large spaces,
photographic processes enable us to "see into" the past, to "bridge" temporal
distances at a glance. But, equally, just as in the normal course of affairs neither
the microscope nor the telescope is the object of our interest, unless we are
optical technicians or repairmen, neither, the skeptic maintains, is the photo-
graphic or cinematographic image. With respect to these things, we are meant
to see "through" them; they are transparent.
To sum up: we are not interested in the photograph — as we are
interested in genuine art objects — for the sake of the object it is. The
photograph does not sustain aesthetic interest on its own. It is not an
instance of genuine representational art. Consequently, if film, as its name
indicates, is basically, essentially photographic in nature, then it is not a
representational artform in its own right either. If a given film appears to
elicit aesthetic interest — as The Cabinet of Dr. Caligari does — that is because
it is a transparent photographic record of a dramatic (theatrical) represen-
tation. It is the dramatic representation itself— as it was staged and enacted
in front of the motion picture camera — that grips our interest, just as it is
Groucho Marx's monologues and not the moving picture recording of
them of them that is the locus of our aesthetic interest in Animal Crackers.
To suspect otherwise is to confuse the box that contains the cookies for the
cookies themselves.
 FILM AS ART 21

Responding to the Skeptics

These three arguments — the causation argument, the control argument, and
the aesthetic interest argument — are the basis for the skeptic's brief against
the possibility that some films are artworks. These arguments, however
benighted they may appear, call our attention to intriguing features of film,
even if the skeptic does not always draw the right conclusions from his
observations. The arguments often sound convincing, but are they decisive?
Let us review them critically in order to locate their potential shortcomings.

Round Two: The Causation Argument

First, let us reconsider the causation argument. The skeptic correctly
points out that the production of a photograph, properly so called, does
not necessarily require the intentional contribution of an agent. A photo-
graph, still or moving, could result from a series of physical events with
no intentional input. A malfunctioning camera might switch on when
dropped, or, more fantastically, photographs might appear in nature as
the result of a sequence of statistically improbable but logical possible
events as discussed earlier in the case of our natural "cave" photo of the
volcano. These cases suggest that intentionality is not an essential feature of
photos. That is, there can be photos stripped of intentionahty altogether
which are nevertheless still appropriately classified as photos. But if there
is no intentionality involved, they cannot be expressions of thought, for
they are bereft of thought content (something they are about). Thus, what
might be called essential or minimal photographs are not necessarily art-
works, where to be an artwork requires, ex hypothesis being an expression
of thought.
So there are conceivable photographs that are not artworks. But> granting
that there may be such a case, what about actual photographs rather than
merely conceivable ones? When we admire the photos of Edward Steichen
as works of art, we are talking about actually existing photographs and
not just conceivable ones. Moreover, in the cases of the actually existing
photographs that we classify as artworks, do we not believe, with good
cause, that the intentions of the relevant photographer are expressed in his
images? Steichen's photos convey thoughts and feelings about the content
toward which they draw our attention and we believe they are undoubtedly
intended to do precisely that by their maker.
22 FILM AS ART

As Walter Benjamin noted, Atget's photographs, for example, express

thoughts about Parisian cityscapes. The pertinent photographs, in other
words, possess intentionality in the ways the skeptic mandates, and said
photographs should, consequently, count as works of art in accordance
with the terms the skeptic's argument dictates. What difference does it
make to the artistic status of actually existing photographs that express
thought, as Atget's do, that there are some other conceivable photographs
that do not?
That some photographs express thoughts is enough to establish that
some photographs are art. The fact that there may be conceivable photo-
graphs, like those we have explored, that do not express thought at best
shows that not all photographs must necessarily be art. But, on the one
hand, no one wants to prove that; even the greatest fan of photography will
concede that not every photograph is or must be art. Who thinks the
photos on standard automobile licenses are art?
On the other hand, the skeptic has to prove more than just the
proposition that some photos are not or may not be art. The skeptic is
committed to proving that necessarily no photograph is art nor can it be.
But how will the skeptic show that? What the skeptic, invoking the notion
of an essential or minimal photograph, has demonstrated is that intention-
ality is not necessarily involved in producing something that we would be
willing to categorize as a photograph. But it does not follow logically from
the proposition that intentionality is not necessarily involved in producing
some photographs that necessarily intentionality is not involved in the
production of any photographs. The fact that a surveillance camera at a
traffic light records a scofflaw automatically — without the intervention of
any intention — does not entail that intentionality and intentions do not
come into play productively when I select the subject of my snapshot,
choose the lens, film stock and aperture setting that I desire, and adjust the
lighting for my purposes at the same time as I prepare to execute my
photograph. To imagine that the lack of intentionality in the case of the
surveillance camera must imply a lack of intentionality in the shoot I've just
described is a logical error (called a modal fallacy because it is an error in
alethic modal logic, i.e., the logic of necessity and possibility).
From the fact that mere photographic status does not necessarily require
intentionality, it does not follow that anything legitimately classified as
photography must necessarily lack intentionality. Though the skeptic appears
to presume this inference, it is a non sequitur. Technically, it is a fallacy
of alethic modal logic to attempt to move from (1) photography is not
necessarily an intentional product to (2) necessarily photography is not an
 FILM AS ART 23

intentional product. But the skeptic seems to attempt, erroneously, to

make precisely this inferential leap.
Against the skeptic, we can argue that even if certain conceivable photos
lack intentionality and are not artworks for that very reason, this does not
even remotely suggest that all actual photos are equally compromised. For
as we have just seen, if we did not know already, some actual photos can
and do have intentional content and can be and have been intentionally
produced, thereby satisfying even the skeptic's criteria for art status. And if
still photographs can be representational artworks on these grounds, so can
films, even films considered as merely essentially moving photographs.
Perhaps the skeptic will argue that anything short of what we've called
an essential or minimal film should not count as an authentic specimen of
the medium. But when adjudicating the question of whether some photos
are artworks, why would we grant greater weight to how they might have
been made as opposed to how they were actually made?
Moreover, the skeptic cannot plausibly argue that if causation is implicated
in the production of a work; then that precludes the kind of intentionaHty
requisite for art status. For were that so, virtually no medium could be said to
be capable of producing art, since almost all of them possess an meliminable
causal dimension. Consider: even a paintbrush is a tool; once an artist touches
his brush to the canvas, a causal process is set in motion; the paint flows at a rate
determined by chemical law. The sculptor is beholden to physics. The organ,
of course, is a machine. But in none of these cases do the facts that there are
causal processes involved preclude the expression of thought. So, if in count-
less cases of already acknowledged artistic media their causal or mechanical
aspects do not foreclose the possibility of expressing thought, then why does
the skeptic suppose matters to stand differently with photography and film?

The Control Argument Revisited

The skeptic emphasizes that the expression of thought — for him, the sine qua
non of art status — requires control. If there is no control over the medium in
question, the skeptic contends, then one is simply not expressing one's own
thought. This seems fairly persuasive. For if there is no control whatsoever-
then it is doubtful that any thought is being expressed at all. Furthermore,
if you believe that there is an essential connection between the expression
of thought and the possession of style, then if one lacks control over one's
medium, what one produces lacks style, with the upshot: no control, no
style, no art. Such is the control argument.
24 FILM AS ART

An initial response to the control argument must be that it is Utopian.

If what is required for art status is total control of a medium, then we will
be compelled to discount as artforms most of the practices that we esteem
as such. Most artists have to negotiate compromises with their medium —
to adjust to the material at hand. A theater director will have to cast her
drama with the available actors; these may not be and often are not exactly
how she imagined the characters upon her first reading of the script. But
she will make do. A director does not have total control; she has to work
within parameters set by others and by circumstances.
Likewise a choreographer has just these dancers and just these spaces.
Whatever dreams she had for her dance must be brought down to earth, so
to say. Virtually every particular artistic element comes with limitations.
Even the writer has to work within the confines of a specific language* The
composer has to work with the array of existing instruments and the
musician has to work with the peculiarities of his tool. Architects and
sculptors are evidently very constrained physically in terms of the amount
of leeway they have for exerting their will on their materials, while even
the painter will have to live with the resources to which she has access —
such as the available varieties of paint — and their attendant properties.
Though critics may speak of this or that artist as controlling every aspect of
her creation, this is, in the main, journalistic hyperbole — a feel-good
fantasy for the Sunday arts and leisure section.
Few artforms, if any, realize the fantasy of total control. Are we to say,
then, that no medium is capable of producing art? Surely a more
reasonable tack is to lower our sights — to require only that a medium-
permit its practitioners sufficient control, rather than total control. But this
adjustment then reframes the terms of the debate — to wit: does the film
medium allow enough control for the filmmaker to project intentional
content — i.e., to express thoughts and feelings and/or to structure formally
his subject matter?
We can start to formulate a positive response to this by reminding
ourselves of some of the points that we made in the process of refuting the
causation argument. We noted that the photographer has a number of
variables at her disposal that enable her to articulate and, thereby, express
her thoughts. She can select her subject matter, for instance. Certainly it is
not random statistical coincidence that explains the undeniable consistency
in the work of a Diane Arbus. Also, the photographer may choose the type
of camera, lens, film stock, aperture setting, and lighting effects that
she determines to be suitable for whatever thoughts and/or feelings
she means to express. And, in addition, the photographer may set up the
 FILM AS ART 25

shot — staging it, establishing the right subject-to-camera distance, electing

the props, costumes, and make-up, and, indeed, often the very models who
then strike the pose in this regalia.
Frequently, the photographer makes a series of photos of a given subject,
settling ultimately on the one from an array of alternatives that most realizes
her conception. And once she shoots the film, she can then go on to clarify
more specifically her expression by deploying a gamut of laboratory processes,
including printing on various papers, cropping, air-brushing, and so forth.
At each step in the creation of a photograph, there are choices that can
be made, each of which will inflect the expression of thought and/or
feeling embodied in the picture. Opting for one of these alternatives rather
than another — a wide-angle lens rather than a long lens — may make a
difference that expresses with greater precision the content and/or qualities
of the thoughts and/or feelings that the photographer endeavors to convey.
Should not all this count as enough control of the medium to warrant
regarding photography as a representational artform?
One reason to think so is that there is a striking inventory of correlations
between many of the acknowledged variables for modifying paintings
expressively arid those that obtain with respect to photography, including
choice of subject, lighting, scale, framing, and so on. Moreover, we already
know that variations along these expressive dimensions have proven rich
enough to differentiate the thoughts and feelings that an August Sandier
intends to express from those of a Robert Mapplethorpe.
Whether a medium grants sufficient control to be categorized as a
representational art can be put to the test by gauging if there exist enough
alternative variables of articulation for us to identify distinctive stylistic or
expressive profiles. That we can isolate stylistic and expressive contrasts
between photos by Cindy Sherman and Walker Evans and then go on to
associate these with the expression of different thoughts strongly supports
the conjecture that photography possesses the degree of control sufficient
to qualify as a medium capable of artmaking.
The style or form of a work of art is the ensemble of choices that
function to realize its point or purpose — for example, the expression of
thought and/or affect. Control is required in order to possess a style and
to express thoughts. The level of control is a function of having at one's
disposal a range of articulatory options such that the choice of one of these
strategies with respect to the pertinent variable rather than another makes
for stylistic and/or expressive differences; for it is choice from a field of
alternative elements with contrasting qualities and associations that clarifies
thought and particularizes style.
26 FILM AS ART

Photography possesses multiple expressive channels, replete with alter-

native, contrasting strategies of construction at each level of articulation.
This wealth of options affords the creative wherewithal for the photo-
grapher to implement her intentions by strategic decisions. Furthermore,
such has proven to be more than enough control for artistic purposes, since
it has enabled us to locate the divergent "authorial intentionalities" of a
Man Ray, a Nan Goldin, a Weegie, and a Moholy-Nagy.
The cinematographic image, moreover, offers even more levers of
stylistic control and choice than does the still photographic image. For
example, camera movement allows the cameraman the opportunity to
explore the array, bracketing this or that detail for special attention. Of
course, neither medium grants absolute control. But, as indicated, the real
argument is over the issue of whether the relevant medium permits
sufficient control. And for the reasons just rehearsed, we can insist against
the skeptic that photography and film appear to satisfy the requisite level of
control justifiably expected of an artform, representational or otherwise.

Back to the Aesthetic-Interest Argument

Plato demeaned the seriousness of mimetic painting and poetry by analogiz-
ing them to mirrors. Similarly, the skeptic disparages the pretensions of
photography, still and moving, to the status of art by alleging that cameras
are little more than mirrors artlessly reflecting whatever wanders into their
purview. For if photos and films are like mirrors, then they no more express
thoughts than does the glass on the front of your bathroom medicine cabinet.
A mirror trades in mere appearances rather than intentionality. Thus, if
photography and film are strictly analogous to mirrors, then we do not care
about them for their own sake; we do not take an aesthetic interest in them.
Rather, we care about them or take an interest in them for the sake of the
objects they show us. I do not care about the curling, sepia snapshot of my
beloved grandmother as a photograph but as a reminder of Nanny.
The capacity of film and photography to afford a window into the past
has been called "photographic transparency." The notion of photographic
transparency provides the basis for the skeptic's assertion that photography
and film are not artforms, representational or otherwise, because they are
not capable of commanding our attention to the kind of objects they are in
their own right. Consequently, neither photography nor film can sustain
aesthetic interest. For if they are transparent* then we see through them
and our interest lands on the appearance of whatever they are images of.
 FILM AS ART 27

If the photograph is of President Bush's inauguration, then my interest lies

in his inauguration. It supposedly makes no sense for me to be interested in
the photograph as such, since it is transparent — i.e., we see through it.
But there seems to be a dubious presupposition underlying the skeptic's
polemic, namely, that if a photograph, still or moving* is transparent, then
it is invisible. Here "invisible" means something like seeing an object through
a photograph is just like seeing the object in question face to face. That is, the
mediating agency of the photograph goes undetected by the human eye, and,
since we allegedly don't see it, we don't take an interest in it, where it is the
photograph as such — the photograph as an object of attention in its own
right and not merely the image of some other object.
Yet even if we grant that photographs are transparent, it would not
follow that they are invisible. In fact, not even all mirrors are invisible in
the way the skeptic suggests. When I look into one of those convex mirrors
that enables one to see around corners in parking lots, I may see through it
to the car that is racing toward me; however, I also see the distortion in the
mirror. Such a mirror is hardly invisible even if it is transparent; transparency
does not entail invisibility. But the skeptic's aesthetic interest argument
against photography only gains traction if photos are invisible; alleging they
are simply transparent is not enough. Moreover, the case for the invisibility
of photographic mediation is even more unlikely than the case for the
invisibility of mirror mediation.
Throughout most of the history of photography, both still and moving,
most photographs were black and white. This feature of photos was
scarcely invisible. Looking at a photograph of the blood-spattered body
of the Archduke Ferdinand was not just like seeing a dead body staring you
in the face. It was obvious to everyone that the absence of color was a
feature of the photographic mediating process.
Needless to say, one could take an interest in this aspect of a certain
photograph — noticing whether the blacks and whites contrasted starkly
and brutally, or whether they were subtly mapped along the grey scale.
Moreover, a photographer could intentionally design a photograph in such
a way as to reward attention to photographic features like this. That is,
photographers could make photos that could support taking an aesthetic
interest in such objects in their own right. In short, there is no reason to
suspect that the photographer cannot meet the skeptic's challenge on the
skeptic's own terms.
Photography may be transparent, but even if it is, it may still be
interesting because of the ways in which the medium displays views of
objects that depart, in various respects, from the ways in which those
28 FILM AS ART

objects would appear to "photographically unassisted" or "normal" vision.

The photograph, even on the transparency thesis, need not be thought to
disappear; the image need not collapse into the object of which it is an
image. Thus, it is possible to take an interest in a photographic object for its
own sake — an aesthetic interest, that is, in a photographic art object. Taking
an aesthetic interest in a photograph as an object in its own right, moreover,
will be especially apposite where the photograph has been designed to
encourage such an interest.
If we are willing to say that we see objects through photographs, then
that kind of seeing, like seeing through a mirror, is indirect,\ rather than
direct. Furthermore, indirect "photographic seeing" is different in kind
from direct seeing — seeing that involves a close encounter of the third
kind, a face-to-face brush with the object of attention. Photographs, for
example, lift the object out of its natural context and reframe or bracket it,
often enabling us to see the object afresh in ways denied to us in routine
face-to-face encounters with it. Photographs freeze the object in a moment,
potentially disclosing aspects of it that might be otherwise typically occluded
in everyday life. And the object pictured in a photograph is absent from the
viewer, inviting an opportunity for scrutiny that in "real life," where the
object is present to the viewer, might be dangerous, callous, insensitive,
impolite, or confusing. Moreover, it is this feature of absence that can make
nostalgia an appropriate emotional response to the someone or something
depicted in the photograph, whereas nostalgia would be an absurd reaction
were they "in our face."
In virtue of the differences between photographic seeing and face-to-face
seeing, features of the photographic medium can defamiliarize the objects
whose appearances they convey. Defamiliarization, in turn, may serve as an
artistic strategy for expressing thoughts about the relevant objects, persons,
events, actions, and places captured on film. Even if photographs are said
to be transparent, their departure from face-to-face or direct seeing leaves
open the possibility of artistic defamiliarization and, thence, the expression
of thought and feeling. Furthermore, if a photograph by means of features of
the photographic medium expresses a thought and/or emotion about its
objects, then we may take an interest in the photograph qua photograph —
that is, we may take an aesthetic interest in the way in which the photograph
at hand mobilizes characteristic features of its medium to express thought.
Our concern, in other words, need not be presumed to be absorbed
solely in the appearance of the object as it has been relayed to us by the
photograph. The defamiliarizing resources of the photographic medium
can be harnessed to express thoughts about the objects portrayed and this
 FILM AS ART 29

process, in consequence, can give rise to ruminating aesthetically about the

way in which the kind of object a photograph is has given birth to a
thought. That photographs may defamiliarize opens the possibility for them
to express thoughts about the object they depict, contrary to what the
skeptic assumes about the alleged transparency of photography. Moreover,
the process of expressing thoughts not only gives us a renewed interest
in the object upon which the photograph comments, but an interest in the
photograph itself as viewers ask themselves how the use of various features
of the medium has managed to engender thought.
Like the still photograph, the motion picture shot has an estimable
repertoire of devices that diverge from the direct face-to-face perception
of objects. These include the close-up, high^ and low-angle shots, lenses of
various types, filters, traveling shots, alternating camera distances, and so
forth. Each of these devices makes possible the expression of thought and/or
feeling about the objects presented in the shots. If such shots are transparent,
they can nevertheless be expressive, since there is no coherent view of
transparency that implies that the photographic image, still or moving, is
plausibly indiscernible from that of which it is an image. But that is what
the aesthetic interest argument that the skeptic advances would appear to
require. And that is just unconvincing.
The skeptic argues that an aesthetic interest is an interest that we take in
an object because of the kind of thing it is. Photographs do not afford
aesthetic interest because whatever interest we have in photographs, still or
moving, collapses into the interest we have in the objects of which they are
photographs. Two things need to be said about this.
First, as has already been emphasized, we may take an interest in the
photograph qua photograph in virtue of the ways in which it has exploited
the photographic (or cinematographic) medium to express a thought or
feeling.
However, we may also wish to challenge the skeptic's conception of an
aesthetic interest. For the skeptic, it seems that an aesthetic interest in an
object must be about the kind of object it is; so it must be an interest in the
modes and materials that comprise the object. An aesthetic interest in an
oil painting must be one that takes notice of how oil paint has been used to
achieve a certain effect. This seems to imply that an aesthetic interest in an
artwork requires insider knowledge about properties of the modes and
materials of the artform in which the object is created.
But this seems to be an overly stipulative conception of what counts as
an aesthetic interest. Surely, I have an aesthetic interest in a film if I am
absorbed by its expressive properties. For example, I think many viewers
30 FILM AS ART

had legitimate aesthetic interests in the "hot chocolate" number in Polar

Express, taken as they were with its infectious exuberance and good spirits.
And yet I suspect that few viewers, including myself, have much of an
understanding of how it was made. However, even though we might not be
able to be interested in the object in terms of the immensely technical
achievement it is, our interests were still aesthetic in the usual sense of
that term.
Moreover, the relevance of this for the prospects of taking an aesthetic
interest in photography should be clear-cut. Photography, still and moving,
can manifest expressive properties of objects that might otherwise go
unremarked, especially in direct, face-to-face encounters with said objects
in quotidian experience. People are very often absorbed in these qualities
as revealed by a master photographer or cinematographef, like Gregg
Toland. Thus, it is false to suggest, that photographs, whether still or
moving, cannot reward aesthetic attention and sustain aesthetic interest.

What about the Skeptic's Second Premise?

Thus far we have been bickering with the first premise of the skeptic's
master argument. This has been a fairly intricate exercise because the
skeptic supports this opening move with an elaborate set of considerations.
But the skeptic's case also depends upon the truth of the second premise in
the argument — namely, that film is at best a photograph (a moving
photograph) of a dramatic representation. Without this premise, the skeptic
cannot extrapolate what he alleges to be the limitations of photography to
the case of film. That is, without this second premise the argument that film
cannot be art falters.
Luckily sketching our reservations about the second premise in the
skeptic's argument need not be as complex as unhorsing his defense of
the first premise, since the skeptic presents little by way of motivating the
claim that films are nothing more than moving photographs of dramatic
representations. The skeptic appears to regard this as self-evident, which it
truly extraordinary, since it seems so palpably mistaken.
Films are not mere photographs of dramatic representations for the
painfully obvious reason that films are not merely photographs. Photo-
graphy, that is, is not the only constituent element that comprises film. In
addition, there is, among other things, most notably, also editing. So even
if all the arguments against photography had won the day, those objections
could not be transferred to the case against film, since the edited dimension
 FILM AS ART 31

of cinema can express thoughts and feelings, is not transparent, can

facilitate a more than sufficient level of artistic control, and can abet a
locus of aesthetic interest, both in the skeptic's narrow sense and in the
broader, more ordinary sense. The Odessa Steps sequence of Eisensteins
film Potemkin expresses a viewpoint about the event that it is implausible to
suppose any stationary-camera position on the actual event in 1905 could
have delivered. Perhaps skeptics will carp that editing is not really film art,
but something else. However, to deny that editing is a characteristic
element of the motion picture begs the question. But it also invites the
question: if not film, what else?
Nor can the rich repertoire of techniques available to the contemporary
filmmaker for coordinating the audio with the visual components of film be
reduced to the photographic aspect of the medium. When the horror-
suspense filmmaker of The Descent introduces frightening, off-screen
sounds, how is the emotional anxiety she is provoking simply attributable
to photography. The frisson would not occur without the sound. And the
sound design of a movie is patently not part of its photographic dimension,
Film, in sum, has many more dimensions than the photographic one.
So even if photography/cinematography was as sullied artistically as the
skeptic insinuates, it would still not follow that cinema is comparably
incapacitated as a player in the art game.
But another problem with the skeptic's second premise is that it
presupposes that the image track of film is exclusively photographic.
Undoubtedly, in the main, for most of film history, this has been true
of most movies. But it is not necessarily true of motion picture images,
as one realizes when one contemplates the development of new moving
picture technologies, such as computer-generated mattes. As seen in Sky
Captain and the World of Tomorrow, not to mention recent installments of
Star Wars and Lord of the Rings, whole cities, armies, and air fleets can
be created without primary recourse to the mediation of traditional
photography/cinematography. These things can be concocted out of thin
air, digitally fabricated, if you will. Indeed; in the future, entire feature-
length motion pictures are likely to be constructed in the computer. I say
that this is likely if only because it is so economically attractive. Producers
will not have to drag casts and crews to expensive locations, and trouble-
some, highly paid actors can be retired in favor of digitally synthesized
ones. I can see the producers smiling already.
Again, the skeptic may say that these are not really films, since they are
not celluloid-based. But I doubt that most people will hesitate to call
moving picture narratives that look like the movies of yesteryear film. They
32 FILM AS ART

will continue to call them film for the same reason they continue to call
blue jeans "Levis," copiers "Xeroxes," and cola "Coke." Film was one of
the first names we had for this sort of thing and it will stick, just as we
continue to call those things "zippers" after their inventor.*
Of course, although it might be more accurate to call the optical
creations made possible by computer technology and whatever new inven-
tions come down the pike in the future "moving images," most people,
I predict, will continue indiscriminately to call motion picture narratives,
however they are produced, films in honor of the first medium that made
these moving spectacles popular.
But it is important to remember that photographic film is not the only
delivery system for what we call film (a.k.a. motion pictures) in the
broader sense. And since some of these delivery systems need not employ
photographic film in any way, it is false to allege that a dramatic film is
a photograph of a dramatic representation. It could be a computer-generated
representation in its own right. Moreover, this has always been a logical
possibility for the film medium, even if the mass production of such work
awaits the future.
Furthermore, it should be evident that computer-generated imagery
faces none of the challenges that the skeptic levels at straight photography.
Such imagery may mirror nothing* nor can it always be said to be
transparent because sometimes it creates its own objects. What appears
on screen is as much under the control of the CGI specialist as what appears on
the canvas is under the control of the painter. Nor can there be any question
of whether computer-generated imagery can express thoughts. It can do so
in the same ways that a painting does.
At this point* the skeptic may reject these claims on the grounds
that computer-generated imagery and anything else like it that is invented
in the future is not really film, but, perhaps, animation or even painting.
But at that point, the skeptic is in danger of exiting a genuine philosophical
debate about the evolving practice of motion picture making and indulging
in, at best, an antiquarian and primarily verbal debate about some of its
earliest stages.

For example, I doubt anyone, including both ordinary and informed viewers,
will hesitate to call the last section of Kiarostami's 1997 Taste of Cherry "film" just
because it was shot on video. "Film" has just become another way of saying
"moving image," irrespective of the provenance of its material process of production,
Or so I conjecture.
 FILM AS ART 33

In short, the second premise of the skeptic's master argument is as

troubled as the first one: Consequently, at this point in the dialectic, it
seems fair to say that the skeptics have failed to carry their point. Thus, we
may continue to share the common-sense conviction that (some) films
(some motion pictures or moving images) are art. The burden of proving
otherwise belongs to the skeptic.
In sum, throughout this chapter we have been engaged in addressing the
arguments of skeptics who contend that art cannot be made by means of
film. Though this debate, from the outset, may not have seemed particularly
alive to many readers, it is to be hoped that in considering the skeptic s brief
a number of interesting and heuristically valuable claims about photography
and photographically based cinematography have been brought to light —
such as their alleged transparency. As we will see in ensuing chapters, these
claims will continue to figure in other debates concerning the moving image.
However, until further notice, the debate about whether film, or, m o r e
precisely, the moving image, is art is closed. Motion pictures can be art.
It is important to establish this unequivocally from the outset, since I doubt
that we would have any interest at all in developing a philosophy of motion
pictures were it not the case that this particular technology (or set of
technologies) had evolved into a genuine artistic practice.

Suggested Reading

If some of the topics in this chapter have intrigued you and stimulated your
interest in pursuing them in greater depth, let us offer some further readings for
your consideration. If the skeptical position about the possibility of an art of film
appears worth a second look, by far the most sophisticated version of it can
be found in "Photography and Representation," by Roger Scruton in his book
The Aesthetic Understanding (South Bend, IN: St. Augustine Press, 1998), 119-^-8.
A discussion of the historical background of the prejudice against photography
and then cinematography as art can be found in the first chapter of Philosophical
Problems of Classical Film Theory by Noel Carroll (Princeton, NJ: Princeton University
Press, 1988). A particularly effective and extremely historically influential answer
to the traditional charges that neither photography nor film is art is Film as Art
by the indomitable Rudolf Arnheim (Berkeley, CA: University of California
Press, 1956).
Powerful responses to Roger Scruton's arguments include Berys Gaut, "Cinematic
Art," Journal of Aesthetics and Art Criticism 60:4 (Fall 2002), 299-312, and Dominic
Mclver Lopes, "The Aesthetics of Photographic Transparency," Mind 112:447
Quly 2003), 433-48.
34 FILM AS ART

The idea of photographic transparency is developed in Patrick Maynard,

"Drawing and Shooting: Causality in Depiction," Journal of Aesthetics and Art
Criticism 44:2 (Winter 1985/6), 115-29, and Kendall Walton, "Transparent
Pictures: On the Nature of Photographic Realism," Critical Inquiry 11:2
(Dec. 1984), 246-77. Kendall Walton has defended criticisms of his view in the
article, "On Pictures and Photographs: Objections Answered," in Film Theory and
Philosophy, edited by Richard Allen and Murray Smith (Oxford: Oxford University
Press, 1997).
Of classical film theorists, the author who comes closest to articulating the
"transparency" view is Ande Bazin. See his "The Ontology of the Photographic
Image," in What Is Cinema?', vol. 1, trans. Hugh Gray (Berkeley, CAf University
of California Press, 1967). Consider also Stanley Gavell, The World Viewed
(Cambridge, MA: Harvard University Press, 1971, 1974, 1979).
Chapter Two
Medium Specificity

The topic of this chapter is the doctrine of medium specificity or, as it is

sometimes better known, the problem of the purity of the medium — the
view that artists should be true to the medium in which they work, or, in
other words, they should not pursue effects that belong to some other
medium. Needless to say, the medium that concerns us is the one that
is usually labeled alternatively film, cinema, or the movies. Traditionally,
purists with regard to film are particularly anxious that filmmakers not
dilute the cinematic medium with that of theater.

What is a Medium?
A medium is something that mediates. The word itself derives from the
Latin word medius, which means "middle." The medium is a middleman,
a go-between betwixt one thing and another. In the arts, it is that which
brings the artist's conception to the audience, or, more exactly, it is
generally what makes the artist's vision physically manifest for reception.
With respect to the debates to be canvassed in this chapter, the notion of
media is to be understood primarily in terms of physical media.
Media, on this construal, are (1) the materials (the stuff) out of which
works are made and/or (2) the physical instruments employed to shape
or to otherwise fashion those materials. The medium of painting, in the
material sense, is paint, including watercolor, pastels, acrylic, gouache,
tempera, oil, and so forth. The medium of painting in the instrument sense
includes brushes, palette knives, air guns, fingers, etc. With respect to
cinema, the celluloid-based, photographic filmstrip, coated with silver
haloids, is — or at least was for many years — the most obvious material
medium; the film camera is a physical instrumentality that gives shape
to the emulsion on the filmstrip in a manner roughly analogous to the way
in which the brush gives form — or informs — the oil paint on the canvas.
36 MEDIUM SPECIFICITY

Editing, the literal piecing together of discrete shots, is also a physical

means by which films are composed, much as collage is a physical technique
in fine art.
(Here it is important to emphasize that, for the most part, given the
nature of the debate regarding cinema as it has evolved, we shall be dis-
cussing the physical media from which artworks are constructed and
not what might be call the artistic media — that is to say, the conventions,
such as the reversal of fortune in tragedy, which are employed in the
production of certain genres of artworks. The reason for this is simply that
conventions cross media, materially or instrumentally construed.)

Medium Specificity
The idea of medium specificity is that each artform has its own specific
medium — as painting has paint, and film has film. These media, to a great
extent, individuate artforms. Furthermore, it is believed, the media of
specific artforms possess a definite range of effects — things they can do
especially well and other things that they cannot do, or, at least, that they
cannot do as well as the media that correlate with some other artform or
artforms. For example, it is frequently said that, in virtue of its medium
(especially editing), film represents physical action — such as chases, races,
and intergalactic dog-fights — exceedingly well, but intricate speeches less
well, notably in contrast with theater.
According to the doctrine of medium specificity, each artform, given
its physical medium — in either the material and/or instrumental sense —
has a range of representational, expressive, and/or formal capacities. Some
of these may overlap with the representational, expressive, and formal
capabilities associated with the media of other artforms. However, some of
these possibilities are said to be distinctive of the medium that serves to
identify or to individuate the artform in question. By "distinctive" is meant
that certain of the effects of the relevant medium are managed both
(1) better than the other things it (the medium) does and (2) better than
said effects are managed by the media possessed by any other artform. The
medium-specificity thesis recommends that artists exploit the distinctive
possibilities of the medium in which they ply their trade and that they
abjure the effects that are discharged better or equally well by the media
of other artforms.
Because the doctrine of medium specificity advises the artist to exploit
all and only those effects that are distinctive, in the preceding sense, to her
 MEDIUM SPECIFICITY 37

medium, the doctrine is a form of purism. It instructs artists to operate

within the pristine boundaries of their chosen medium and warns them
not to flirt with counterfeiting the characteristic effects of the media that
belong to other artforms.
An artist must be true to her medium. For example, the sculptor should
not attempt to imitate the effects of bronze in plastic. In German, this
doctrine is sometimes referred to as Materialgerechtigkeit, which translates
as "doing justice to the nature of the material." That is, the medium has
a distinctive nature with specific capacities — ones that differentiate it
from the media of other artforms — and it is precisely these possibilities
that the artist ought to explore rather than the effects that are distinctive
of the media that identify or individuate other artforms. For example,
metal engravings should not attempt to fake the grainy look of woodwork.
Artforms are distinguished in terms of their physical media. Paint is
different from film. Artforms should, normatively speaking, be true to
the possibilities governed by the physical media that serve as the basis of
what individuates them:
Why?

Considerations on Behalf of Medium Specificity

One suggestion on behalf of medium specificity is ostensibly prudential:
namely that trafficking in the distinctive effects that belong to the medium
of another artform is a prescription for failure. A blizzard, for example, is
an apt subject for painting, but not for sculpture. Painting can capture the
lightness of snow; stone cannot shed its own heaviness. A painting can
portray individual snowflakes; a sculpture made of a white block of stone
will look more like a giant, milky ice cube than it will look like a flurry of
snowflakes. If a sculptor trespasses into the domain of painting, so the story
goes, he courts disaster.
This way of thinking underwrites the traditional account of the failure
of early talking pictures. The charge is that these movies were artistic
catastrophes exactly because they were nothing more than canned theater —
that is, theater in a film can. Actors awkwardly declaimed their lines, rigidly
arrested in space so as not to wander out of range of the microphone;
the camera was equally static, too large to move because it was housed in a
heavy, trailer-like, soundproof structure.
Producers bought up theater properties — actors, writers, and scripts — in
order to give the audience something to hear. But, the writing was crafted
38 MEDIUM SPECIFICITY

for the stage, and the actors, accustomed to reciting it there, were scarcely
prone to movement. Nor did these films exploit the fluid possibilities
of changing camera positions. The cinematic grace of the silent film, which
emphasized the putatively unique art of motion pictures, was betrayed for
a mess of words.
That a great many early talkies were artistically wooden and aestheti-
cally stilted no one disputes. However, the proponent of medium specifi-
city adds that they were failures precisely because they eschewed the
"natural" inclinations of the cinema — which exploits movement in
every dimension (hence, moving pictures) — while simultaneously and un-
advisedly these early talkies aspired to the condition of theater. As such,
they fell stillborn from the projector and the putative reason for their
failure is that they violated the dictates of medium specificity; they were
impure to their medium. And the medium bit back.
The doctrine of medium specificity is a very bold idea. In its most
radical version, it dares to predict which artistic endeavors will succeed
and which will fail in a given artform. Only those strategies that exploit
the distinctive possibilities of the pertinent medium have any prospect of
success, while those which ignore or neglect the demands of their medium
do so at their own peril — they will either fail, or, at least, will not go very
well. In this regard, the doctrine, supposing it to be true, is extremely
attractive, because it supplies the artist with general guidelines about
what will fare nicely in her medium and what will supposedly go badly.
The doctrine of medium specificity tells the artist what lines of exploration
are promising and what directions should be shunned.
The doctrine is also seductive for critics-at-large and ordinary audience
members, as we shall see in the last chapter of this book, because it gives
them the means to account for why some films fly and others flop. Those
that excel are cinematic — that is, they engage in and exploit the distinctive
properties of the medium. Those that are insolvent can often be explained
away on the grounds that they have failed to take advantage of the special
resources and distinctive capabilities of film, often by stumbling into the
realm of the medium of an adjacent artform (usually theater). Thus,
someone might claim that the doctrine of medium specificity explains
why Hitchcock's film Psycho is superior to his The Paradine Case. The latter,
with all its palaver, crossed over into the domain of theater, whereas
Psycho — remember the shower-sequence montage — is pure cinema
(cutting, that is to say, editing, every inch of the way).
The doctrine of medium specificity also appears to have common sense
on its side. It just seems intuitively right. Who would deny that an artist
 MEDIUM SPECIFICITY 39

should be knowledgeable concerning her medium? If the filmmaker is

shooting a scene in the dark, she should know the speed of her film
stock, lest nothing show up on the screen. She should not attempt to
make the medium do what it cannot do. It would be absurd to try to
shoot at night with film stock that is too slow to do the job. Isn't that
obvious?
Clearly, the artist must know her medium. She should be savvy enough
to use the right tool for the right job. Using a piledriver to open a sardine
can is likely to go awry. Similarly, it is argued that using cinema to render
a story dominated by ornate and complicated language is a recipe for
ennui; for, in order to enable the audience to follow such dialogue, the
camera will have to hold on to it rather than taking the risk of distracting
us away from it by shifting the camera's position in a gesture that could
call attention to itself. But the result of this caution will be uncinematic;
it will not do justice to the distinctive fluidity of cinema — its potential
not only for character movement but for moving (shifting) camera set-ups
as well. The film will feel stultifying; and the audience will squirm
claustrophobically.
Moreover, the proponent of medium specificity adds, the reason for
this is that our filmmaker is attempting to use the medium to do something
for which the medium is not suited. She has opted for the wrong tool
for the job. If this is the kind of script that interests her, then she should
look toward Broadway, not Hollywood. It is just common sense that you
should not try to make an instrument do what it cannot do. You don't use
a machete to cut a diamond. Don't enlist film to do theater's work, nor
vice versa.

The Legacy of Lessing

The idea of medium specificity was popularized in the eighteenth century,
especially by Gotthold Ephraim Lessing in his 1776 treatise entitled
Laocoon. In that book, Lessing compares, among other things, painting
and poetry, and, he argues that, insofar as the basic elements of painting —
daubs of paint — lie side by side, painting is naturally best suited to
represent static states of affairs, like landscapes and frozen historical
moments; whereas, since the basic building blocks in poetry — namely,
words — follow each other sequentially in time, poetry, or literature in
general, is the best means for representing events, i.e., things that unfold
over time. In other words, the very structure of the medium putatively
40 MEDIUM SPECIFICITY

is such that in painting spatial arrangements are the ideal subject, while in
poetry actions and events are.
Furthermore, artists invite trouble if they fail to respect the nature
of their medium. Lessing worries that the statue after which his book is
named may be flawed just because it attempts to depict an event that is
too action-packed for sculpture — Laocoon and his sons struggling violently
against the coiling serpent about to devour them. Such strenuous move-
ment supposedly looks strange in stone; the subject is not right for the
physical medium.
Early lovers of film were predictably drawn to the idea-of medium
specificity, because it gave them a framework in which to defend the
possibility of film as art. It was simply a question of showing that there
was some range of aesthetic results that film achieved best in virtue of
its physical medium. Commentators often stressed editing as the essence
of the film medium and pointed to all the kinds of things that film could
represent with incomparable facility by means of editing that, in effect,
were plausibly thought to be unequaled in neighboring artforms. Devices
like crosscutting, for instance, permitted opportunities for the parallel
development of simultaneous events virtually unrivaled by theater as
conventionally staged.
But camera movement too can deliver distinctively cinematic effects.
For instance, the follower of medium specificity might argue, at the end
of Sansho the Bailiff, Mizoguchi's slowly rising camera, moving from the
characters to the vista of the beach and sky overhead, secures its emotional
punctuation in a way that is practically inconceivable by conventional
theatrical means.
Lessing emphasized the suitability of a medium for representing certain
kinds of things more perfectly than other media. For example, media
whose basic elements follow each other in time supposedly represent
temporal succession more perfectly than media whose basic elements are
not related temporally. However, proponents of the art of cinema were —
more often than not — concerned with the ways in which film departed
from the perfect representation of its subject matter. This difference from
Lessing, moreover, can be readily explained once we recall the arguments
against the possibility of film art rehearsed in the previous chapter.
For, as we saw there, the worry was that if film were merely the
perfect representation (recording) of whatever it represented, film could
not be art at all, or, if it seemed to be, that was only because it was really
a perfect representation/recording of some work of theater art. Thus,
where Lessing was preoccupied with matching the properties of the
 MEDIUM SPECIFICITY 41

medium with what it represents, the defender of film art purposefully

zeroes in on the ways in which the inherent properties of the film medium
result in deviations from the perfect replication of whatever it represents.
The film lover argues in this way, as we have seen, in order to secure a
space for artistic expression in the teeth of skeptical arguments about
slavish, mechanical recording.
In the days of the silent film, the film theorist pointed out the way in
which the film image of the talking woman departed from normal experi-
ence and then went on to suggest how this very silence might be used for
expressive effect; for example, the actress would have to "speak" with her
eyes. Likewise, the absence of sound effects impelled the filmmaker to find
creative ways to communicate things such as the noise of a gunshot — for
instance, by showing scores of frightened birds taking wing.
Because of the charge, discussed at length in the last chapter, that film
was precluded from the ranks of the fine arts inasmuch as it was nothing
but a recording machine, early defenders of film were at pains to exhibit
and to delineate all the ways in which film did not record impeccably
but also diverged from what it portrayed in ways that were potentially
expressive and creative.
Were cinematic representation truly invisible, then there would puta-
tively be no way for the filmmaker to express — that is, to make manifest or
to mark — his thoughts and feelings regarding whatever it was he was
representing. The object would look no different than it does in its
everyday context where it dwells expressionlessly in reality. For expression
to gain a foothold, there must be some difference between the object
and its representation; the mediation cannot be invisible — leastwise, there
is no way of signaling the addition of expression.
That is why twentieth-century expressionists have so often favored
strategies of distortion. Distortion has the potential to make expression
tangible. Similarly, the filmmaker uses the differences and divergences
between the filmic mediation of what it represents and the normal
perception thereof in order to render her expressive viewpoint upon the
matter before her perceptible.
Film editing had special importance in this regard, since it clearly
involves going way beyond mere recording; it is a means for creatively
and assertively rearranging whatever events and objects the camera took
in and, therefore, it affords wide latitude for artistic invention and
creation. But cinematography itself already also has ample resources to
enable the filmmaker to go beyond mere recording and to defamiliarize
her subject matter. As we saw in the last chapter, the filmmaker may
42 MEDIUM SPECIFICITY

frame an object in a way that lifts it out of its ordinary context, enabling
us to see it afresh — to apprehend certain of its properties that may have
gone heretofore unnoticed in the normal course of perception. Likewise,
freeze frames, slow and accelerated motion, high- and low-angle shots, as
well as close shots and camera movement cannot be dismissed as nothing
but the simulacra of an act of seeing with one's own eyes. They depart
from reality noticeably and can, therefore, function as potential expressive
devices.
The friends of film embraced the idea of medium specificity in the
first instance because it supplied them with a clear-cut way in.which to
defend the artistic promise of the medium of cinema by demonstrating
the kinds of artistic effects that its distinctive medium could deliver which,
at the same time, did not have ready equivalents in the distinctive medium
of other arts, as always, notably theater. Medium specificity — not to put
too fine a point on the matter — was a conceptual framework made for
enfranchising media artistically. But the proponents of film did not just
dragoon medium specificity for the purpose of legitimatizing film art.
They also bought into the further normative commitments of the medium-
specificity program — the notion that being pure to the medium is the
litmus test of success and failure as an artwork.
The Soviet montagists, enthusiasts of the belief that editing is the essence
of cinema, hailed those films as cinematic successes that exploited cutting
and castigated those that did not as retrograde. Films such as Potemkin,
Strike, By the Law, Fragment of Empire, Mother The End of St. Petersburg, Earth,
and Arsenal were rated as masterpieces, whereas Russian films with long
takes, for example Father Sergius, were dismissed as artistic fossils. Editing
could not only reshape the spatio-temporal continuum; it could also
be deployed to express attitudes — as Sergei Eisenstein does in his
film October when he juxtaposes a shot of Kerensky against a statue of
Napoleon.
In the late 1920s, a kind of international film style developed which
reached a high point in works such as F. W. Murnau's Sunrise, Joe May's
Asphalt, and Josef von Sternberg's Last Command, that combined stylish
editing, sophisticated camera movement, special effects, assertive lighting
and strikingly composed camera set-ups; these films celebrated empha-
tically the potential of the medium to depart from mere recording in
order to defy the suspicions of the skeptic about the proposition that
film could be art; and, in turn, cinematic extravaganzas of this sort were
rewarded by medium-specificity advocates, like Rudolf Arnheim, with the
highest aesthetic accolades.
 MEDIUM SPECIFICITY 43

Cinematic artifice was treated as the high road to art. The influence
of the idea of medium specificity was so entrenched that one could still
hear its undertones as late as the 1960s, when a film like Ingmar Bergman's
Winter Light might be ranked artistically beneath Alfred Hitchcock's To
Catch a Thief on the grounds that Hitchcock exploited the distinctive
potential of cinema — Hitchcock's films were cinematic, whereas Winter
Light was not. Moreover, it was often argued that Hitchcock was the
greater artist, for this very reason. And to the extent that that conviction
is still widespread among cinephiles, the doctrine of medium specificity
continues to have clout.
But what is the argument that gets us from a defense of the possibility
of film art to the conclusion that motion pictures that exploit the dis-
tinctive capacities of the film medium (however those are to be specified)
are better artworks than those that do not? This normative conclusion
is quite spectacular when one comes to think about it. For it would seem to
entail that a film like Tarzan the Apeman is a better work of art than Louis
Malle's My Dinner with Andre, since the latter is arguably essentially a work
of theater, an evening of talk, talk, talk only minimally relieved by the
rhythm of the shot-counter-shot; whereas the jungle picture revels in the
spatio-temporal liberty of the medium, representing Tarzan swinging
across miles, vine by vine. And yet, many would feel uncomfortable
with the judgment that the Tarzan film is a greater artistic achievement
than Louis Malle's film. How can the medium-specificity advocate defend
this conclusion?

The Medium-Specificity Argument

The argument begins by reminding us that film can be art precisely
because, in both its cinematographic and its montage components, it can
depart from the slavish, mechanical recording of that which stands before
the camera and/or from what we would see if we encountered whatever
stood before the camera face to face. Film is not bound to record —
perfectly, automatically, and without deviation — anything, neither nature
nor the products of other artforms. Film can express the thoughts and
feelings of filmmakers because of certain distinctive features of the
medium, ones that dp not simply mirror whatever the camera is aimed at.
But if it is the exploitation of these features of the medium that makes
the films in question art — specifically film art (as opposed to, say, theater
art) — then certainly those films that possess the pertinent features (let us
44 MEDIUM SPECIFICITY

call them cinematic features) must be better as film (qua film) than films
that lack those features.
Why so? With respect to the art of film, aren't films that are artworks
better than films that are not art? That is, if a film qualifies as a work of
film art, then, all things being equal, it would appear to stand to reason
that it is better qua film art than a work that does not qualify as a work of
film art. If Helen qualifies for the long-distance swimming team after a series
of fair trials and Jane doesn't, then, all things being equal, Helen is a better
long-distance swimmer than Jane. Similarly, a film that qualifies as a work of
art is a better work oifilm art than a film that does not so qualify.
But recall: putatively films are works oifilm art only if they exploit
those features of the medium that transcend mere recording. Those
features are cinematic ones. So a cinematic film is better as a work of
film art than an uncinematic film. Let us test this hypothesis by examining a
pair of examples.
First, consider the Oceana Roll sequence in Charlie Chaplin's The Gold
Rush. This is the scene in which the Tramp, in a dream, entertains Georgia
and her friends on New Year's Eve by sticking two forks into two bread
rolls in such a way that calls to mind feet, which "feet" the Tramp then sets
to gigging through his graceful mime. The camera just stands by and takes
in this glorious performance by Chaplin. Let us agree, for the sake of
argument, that this is merely the recording of a stage pantomime. It is
not particularly cinematic, since it does not transcend mere recording.
On the other hand, the chase sequence of the Keystone Kops comedy
Lizzies of the Field is unquestionably cinematic; at the very least, it is an
exercise in editing. According to the medium-specificity theorist, the chase
sequence of Lizzies of the Field is a better work of film art than the Oceana
Roll routine from The Gold Rush.
Perhaps many can live with this conclusion. However, the strongest
version of the medium-specificity thesis goes even farther. As art (not
just film art, but art as such), the chase sequence is superior to the Oceana
Roll sequence. That is, the thesis holds that, from
(a) If a film qualifies as a work of film art, then, all things being equal, it is
better qua film art than a work that does not qualify as a work of film art
we are entitled to infer that
(b) If an artwork qualifies as an artwork in the medium in which it was
created then, all things being equal, it is a better artwork qua art than
an artwork that does not.
 MEDIUM SPECIFICITY 45

But what justifies the move from (a) to (b)?

The reasoning may go something like this: any artist aspires to expertise
in a specific medium — a painter in paint, for example. What a painter
knows about as an artist is how to express feelings and ideas, including
ideas of form, by exploiting the distinctive advantages of the sensuous
medium of paint. She might be another kind of artist in addition to being a
painter. She might also be a drummer. But as a drummer, what she knows
is how to express herself by means of drums. In short, an artist is always
an artist in some medium.
An artist achieves excellence as the kind of artist she is by mastering
the distinctive features of the medium that make hers the artform it is
and not some other artform. Artistic accomplishment, in other words, is
always relative to the medium in question. To be a great artist is to be
someone accomplished in marshalling the distinctive capacities of one's
medium to great effect. That is what artistic skill amounts to. A genuine
artist, as such, is a mistress (or master) of her medium.
Furthermore, to say that one artwork qua artwork is better than another
is to say that it is a greater artistic accomplishment — a greater accom-
plishment by an artist — than another. But this is to say that the artist being
commended has done her — somewhat specialized — job better than
someone else. Clearly an artist whose work qualifies as an artwork in the
medium in which it has been created has done a better job in her medium
than the artist whose work does not so qualify. But since artistry is always
relative to a medium, then the artist who excels in her medium is better
as an artist than one who does not. Her artistic accomplishment as such
is greater. Therefore, her artwork is better.
With regard to film, then, a film that qualifies as a work of film art —
which it does by the artist engaging the distinctive features of the film
medium which transcend mere recording — achieves a greater level of
artistic accomplishment than a film that does not, since artistry as such
is always relative to mastery of a medium. Or, the chase sequence in Lizzies
of the Field is a greater artistic accomplishment than the Oceana Roll
sequence from The Gold Rush.

Resisting Medium Specificity

This is a conclusion that few will find palatable. Though the Keystone Kops
chase is a perfectly respectable piece of film comedy, few would accord
it anywhere near the admiration they believe the Chaplin turn deserves.
46 MEDIUM SPECIFICITY

I wonder whether, if confronted with a Sophie's choice with regard to these

two pieces of film, there are any film historians who would plump for the
chase over the Oceana Roll geste? Chaplin's mime is more subtle, better
observed, more imaginative, original, and more thematically dense than
the simple, progressively deteriorating car chase that Mack Sennett's police
force offers us. Though Sennett's gags are eminently serviceable ones,
Chaplin's conception is possibly profound. An argument that concludes
that Lizzies of the Field, or a part thereof, outranks even this moment of
The Gold Rush is automatically suspect.
What appears to carry the medium-specificity advocate's argument
over the finish line is the presupposition that "better as an artwork"
means "a higher level of achievement by the artist in handling the medium
in which the work is created than some other work in the medium." But
what warrants this presupposition? It seems to be the idea that an artist as
such just is someone who handles the distinctive medium of her artform.
But an artist is not someone who just handles her medium; she does it
to some effect. Surely her accomplishment as an artist is to be gauged by
the quality of her effects. If Chaplin's effects are superior to Sennett's, what
difference should it make that in some sense Sennett relied on distinctive
features of the film medium that transcend recording and Chaplin did not?
The Oceana Roll sequence is excellent on film, even if it would also
have been excellent on stage. The sequence is no less excellent for being
on film. When we speak of artistic excellence as such, what warrants
requiring that in addition to being fine the work must also be pure to its
medium? Indeed, isn't defining "artistic excellence" or "better as an
artwork" in the preceding manner merely begging the question in favor
of the medium-specificity argument?
Nor is it compelling to say that an artist must always be the master of
his medium. For surely there are artists whom we recognize to be better
at manipulating their medium than some others, but who are nevertheless
counted to be the lesser artists in our all-things-considered judgments.
There were armies of hack Hollywood directors who handled the medium
better than Marcel Pagnol but whose films were never the artistic equal
of his Cesar. Surely artistic excellence need not be tied to the medium.
Even if a film that engages the medium more directly is better as film art,
it is not evident that it follows that it is a better artwork qua art. For the
quality of the artist as such and the artwork as such depends qn the overall
excellence of the output.
We know that there can be excellent artworks on film that medium-
specificity advocates will maintain are not excellent works of film art.
 MEDIUM SPECIFICITY 47

If only for the purpose of the argument, let us grant them that, in a certain
sense, this is true. Nevertheless, it does not follow from this concession
that these are not excellent artworks, indeed, excellent artworks that may
even be superior to works more beholden to the allegedly distinctive capacities
of the medium. The medium-specificity advocate must tell us why consider-
ations of the medium should be overriding in our all-things-considered
judgments of said artworks as artworks. Why must considerations of the
purity of the medium trump considerations of the excellence of the effect of
the work, irrespective of the constraints of media?
The defender of medium specificity really has no convincing reason for
valuing the purity df the medium over the excellence of the artistic effect.
Both artists and artworks would appear to be validly assessed in terms of
the excellence of the effects they achieve and not in terms of whether their
means of securing of those effects are pure. The requirement of purity
here seems to fetishize the medium.
This is an especially anomalous position for the defender of medium
specificity to land in. It is as if she has forgotten what a medium is! For a
medium is the kind of thing that is instrumental!/ valuable. If a medium is
valuable, it is as the means to something else. Isn't this the very definition
of a medium? The media that are associated with artforms are presumably
valuable ultimately as means to artistic excellence. Thus, it would appear
virtually self-contradictory for the medium-specificity theorist to maintain
that respecting the medium is more important than the achievement of
excellence. Doesn't that seem to reverse the means^ends relationship
here? In demanding the purity of the medium, the medium-specificity
proponent acts as though the medium were valuable for its own sake or
intrinsically valuable, rather than being only instrumentally valuable. But
that flies in the face of what it means to be a medium in the first place.
Earlier the integrity of the medium was defended by analogizing it to
tools. But tools are instrumentally valuable, not valuable for their own
sake. If you can achieve your end by using a hammer in a non-standard
way — say, propping a door open with it — then its effectiveness is what
warrants your use of it. If holding the door open is what you want, the
observation that that is not what hammers are most suited for is beside
the point. We do not respect tools for themselves, but for what they get
done. Who values a screwdriver for its own sake? If the media associated
with artforms are really analogous to tools, we should treat them likewise —
that is to say, as only instrumentally valuable.
Undoubtedly, the friend of medium specificity will challenge the pre-
ceding observations. She may begin by claiming that our objections are
48 MEDIUM SPECIFICITY

based on imagining cases that will never occur, namely cases where there
is authentic artistic excellence despite the artist's neglect of the distinctive
capacities of the medium. But the defender of medium specificity is too
confident here. There are such cases, as exemplified by the Oceana Roll
sequence from Chaplin's The Gold Rush. Indeed, it might be argued that
throughout his career, Chaplin was never a very cinematic director by the
lights of medium-specificity theorists of a certain vintage, but nevertheless,
on everyone else's hit list, he was one of the greatest artists ever to have
worked in film.
At this point in the debate, the proponent of medium specificity is
likely to say that Chaplin's work was never really all that good. But this
will run afoul of received wisdom. Moreover, if the medium-specificity
theorist says that Chaplin's work was not very good because Chaplin did
not exploit the medium, then the theorist will again be accused of begging
the question. The medium-specificity thesis often seems attractive because
it looks like a powerful empirical theory that promises to predict what
will work or not in cinema.* However, no one has ever succeeded fielding
a mediums-specificity hypothesis of any degree of detail that has not been
immediately accosted by counterexamples.
Typically medium-specificity theorists attempt to fend off these counter-
examples by denying that the masterpieces in question are genuinely
estimable due to their compromised employment of the medium. But that
merely assumes as a premise what should be the conclusion of the argument.
It is an attempt to transform an empirical statement into a definition of what
is good in cinema.
But how, the believer in medium specificity will demand, can it be that
an artist who ignores what is distinctive about the medium can accomplish
anything? Earlier it was emphasized that you cannot make a medium do
that which it is beyond its realm of possibility. That is true, but only
because it is a tautology — by definition, if something is impossible for x,

Another way in which to see that being true to the medium — a.k.a. being
cinematic — is not a guarantee of quality is to recall that most of the films and
TV programs we see will pass the medium-specificity muster. But a great many
of them are not very good at all; in fact many are bad. So being, as they say,
cinematic, is not a reliable predictor of goodness. Moreover, interestingly enough,
the reason they are bad often has to do with factors they share with other media — for
example, they are badly plotted or acted, or poorly written. So, purity is not a key
to artistic success, just as impurity is not a sure road to aesthetic perdition. Hence,
we have yet another nail to hammer into the coffin of the medium-specificity thesis.
 MEDIUM SPECIFICITY 49

then x cannot do it. Yet it was not impossible for Chaplin to make the
Oceana Roll sequence in the way he did. Had it been, he would not have
been able to make it. The camera didn't explode when he mimed before
it; the footage did not ignite when projected in theaters. In no logical or
physical sense of impossibility was the Oceana Roll sequence impossible.
In fact, there would be no cause for the medium-specificity theorist
to warn artists away from putatively impure uses of the medium, if said
uses were truly impossible options. N o artist can execute anything that is
genuinely impossible. If the proponent of medium specificity is afraid of
people doing the impossible, she should stop fretting and find some other
crusade to rally.
Common sense is naturally attracted to the medium-specificity position
because of the previously mentioned analogy between media and tools. Use
hacksaws for cutting through steel; use jack-hammers for busting concrete.
Similarly, use film for representing fast-moving action that occurs over
large distances; use theater for presenting densely packed speeches with
elaborate verbal textures.
Behind the tool analogy, of course, there is some presiding notion of
efficiency. Use the right tool for the right job. But the media of the various

By the way, it should be added that it is a mistake to think that elaborate

verbal textures have no place in cinema. For, at the very least, the juxtaposition
of complex verbiage with imagery can be immensely stimulating, as shown by,
among other motion pictures, Yvonne Rainer's Journeys to Berlin, Peter Greenaway s
Prosperous Books, Hans-Jurgen Syberberg's Our Hitler, and Jean-Luc Godard's L'Histoire
du cinema. Moreover, even medium purists will have to accept these examples, since
audio-visual juxtaposition is an inherent possibility of the sound film (though, of
course, not a possibility beyond the reach of theater).
Another problem for the medium-specificity thesis may be that sometimes
the apparently uncinematic choice may be the right artistic choice for a specific
motion picture project. In Dogville, director Lars von Trier opts for highly
theatrical sets, sets lacking walls, as one might find in a production of Our Town.
That the movie is shot on a sound stage is pretty evident. And yet this highly
theatrical style — indeed, this style that even alludes to theater — serves von Trier's
artistic purposes perfectly. The absence of walls enables us to "see through" the
town in both senses of the phrase; von Trier wants to get across the idea that we
are seeing the town stripped of its facades, figuratively speaking, and the stagecraft
literalizes this. Likewise, von Trier intends to present this specimen of Americana
as a matter of roles, and the allusion to theater through the sets helps establish
this conception. Were von Trier to forswear theatricality here in order to pay his
dues to the cinematic, he would sacrifice the precise figuration that expresses
50 MEDIUM SPECIFICITY

artforms are not tools with fixed uses. Indeed, it is generally up to artists
to discover uses for their media rather than the other way around. Film-
makers discovered the artistic potential of the technology of film; the
medium did not come with a set of artmaking instructions
So it is not clear that the tool analogy supports the dicta of medium
specificity. Perhaps there is an underlying suspicion that there is something
wasteful about using one medium to do what another medium can do. But
what exactly was wasted by Chaplin doing the Oceana Roll pantomime on
film rather than in the theater — the opportunity to do it in the theater?
But what if he had done it onstage and on screen? That neither sounds
like something "wasteful," in the ordinary sense of the term, nor does
there seem anything else wrong with it. The world would be a richer place
for having many performances of the Oceana Roll, even across media.
Andj in any event, the notion that waste is to be avoided in art is an
odd one. Artists do not eschew procedures because they are too time-
consuming, or materials because they are too dear. Efficiency, in short, is
not a standard to which we often appeal in art; no one chides James Joyce
as a slacker for dragging his pen in the production of Finnegans Wake.
Excellence of effect is typically all that counts.
In this chapter we have looked at the doctrine of medium specificity.
This doctrine grows naturally out of the strategy for establishing the
possibility of film art by asserting that the film medium is not a perfect
recording mechanism but rather that it deviates from the invisible, mirror-
like transmission of images of its referents in ways that make expression
possible (and formal invention apparent). For the proponent of medium
specificity surmises that film art emerges precisely by exploiting those
medium-specific deviations. Historically, the doctrine of medium specifi-
city has been of immense heuristic value, because it urges filmmakers and
film viewers to pay very close attention to many of the most important
variables whose manipulation have made the, cinema such a powerful
channel of expression. Medium specificity is the kind of "noble lie" that
can get people to see motion pictures with greater accuracy and sensitivity.
But the doctrine of medium specificity is very ambitious. It attempts
to advance an a priori account of what will succeed and fail artistically on
film. Films that fail to exploit the distinctive — a.k.a. cinematic — potential

his thoughts and feelings about American society. Similarly, the film The History
Boys is highly theatrical in a way that is appropriate to its subject — acting,
performance, and striking the pose in the pursuit of academic achievement.
 MEDIUM SPECIFICITY 51

of the medium, perhaps by being essentially theatrical — are going to be as

artistically dismal as the early films d'art; on the other hand, films that
embrace the distinctive possibiHties of the medium, like Godard's Married
Woman, can enter the charmed circle of winning artworks.
We have attempted to show that the normative reach of the doctrine of
medium specificity is indefensible. This is not to deny that an artist should
understand her medium and its properties. Indeed, it is precisely because
Abbas Kiarostami appreciates the difference between film and video that
he can, for expressive purposes, exploit their difference in the contrast he
draws between them in the body of A Taste of Cherry versus its coda.
Our only intention in this chapter has been to reject the suggestion
that these observations in any way provide generalizable, a priori dictates
of what the artist should or should not do and/or that one can predict
systematically what will fail or succeed as art on the basis of some
conception of engaging the distinctive features of the medium of any
artform, including, especially, film.
Medium specificity may seem like an old-fashioned idea, one that finds
itself at home in the early days of film, but which is unlikely to command
much of a following in the era of postmodernism in which boundaries
between everything — artforms and their media included — are supposedly
evaporating. However, medium specificity is an idea that never goes away
and is always standing in the flies ready to make a comeback. That is why
it is worth studying at the length we have spent upon it in this chapter.
For sometime soon, probably someone will latch on to it in order to
defend the possibility of computer art and to recommend the direction
it should take. That is, I predict that there is likely to be someone in the
near future who will claim that what is distinctive about the medium of
computer art is, let's say, interactivity, and that only computer artworks
that .exploit interactivity will be artistically worthwhile. If a view like this
has not already found its way into cyberspace, it will. And at that point we
will need to reopen the debate about the intellectual credentials of the
doctrine of medium specificity, invoking, it is to be hoped, the case of film
as a cautionary tale.

Suggested Reading

For a historical overview of the medium-specificity idea in cinema, see Noel

Carroll, Philosophical Problems of Classical Film Theory (Princeton, NJ: Princeton
University Press, 1988). For summary criticism of the position, see "Forget the
52 MEDIUM SPECIFICITY

Medium!," in Engaging the Moving Image by Noel Carroll (New Haven, CT: Yale
University Press, 2003).
Of proponents of the medium-specificity view, the most powerful and thor-
oughly argued statement is Rudolf Arnheim's in his book Film as Art (Berkeley, CA:
University of California Press, 1956). Most of the other canonical film theorists are
also committed to the medium-specificity idea, or a variation thereof. Some classic
texts in this tradition include: Hugo Munsterberg, The Photoplay: A Psychological
Study (New York: Dover, 1916); Bela Balasz, Theory of Film (New York: Dover,
1952); V. I. Pudovkin, Film Acting and Film Technique (London: Vision Press, 1958);
Sergei Eisenstein, Selected Works (London: BFI Publishing/Indiana University
Press, 1988); Andre Bazin, What Is Cinema?y vols. 1 and 2, (Berkeley, CA:
University of California Press, 1967); and Siegfried Kracauer, Theory of Film
(Oxford: Oxford University Press, 1960).
An extremely perceptive survey and sophisticated extension of this viewpoint
is Victor Perkins* Film as Film: Understanding and Judging Movies (Baltimore, MD:
Penguin Books, 1972). This book will be discussed in the final chapter of this
text but deserves mention now as one of the most astute accounts of the medium-
specificity debates.
Modified defenses of medium specificity can be found in Edward Sankowski,
"Uniqueness Arguments and Artist's Actions," Journal of Aesthetics and Art Criticism
38:1 (Fall 1979), 61-74, and Donald Crawford, "The Uniqueness of the Medium,"
The Personalist 51 (Autumn 1970), 447—69. Comment upon these.articles is offered
by Noel Carroll in "Medium Specificity Arguments and the Self-Consciously
Invented Arts: Film, Video, and Photography," in his collection Theorizing the Moving
Image (Cambridge: Cambridge University Press, 1996). Exchanges on the issue
of medium specificity can be found in the following: between Alan Goldman and
Noel Carroll in Film and Philosophy 5—6 (2002), and between Murray Smith and Noel
Carroll in Film Studies 8 (Summer 2006).
For an excellent, critical review of the film—theater debate, see Susan Sontag,
"Film and Theater," in her collection Styles of Radical Will (New York: Farrar,,
Straus & Giroux, 1966).
Chapter Three
What Is Cinema?

The question, "What is cinema?" (which we derive from an unequivocally

marvelous collection of essays by the distinguished film theorist and critic
Andre Bazin) is an ontological one. "Ontology" is the study of being. The
ontology of cinema is an inquiry into the being of cinema, or, to put it less
awkwardly, it is an inquiry into the kind of being, the kind of thing, cinema
is. To what category does it belong, or, in other,words, under which
concept do we classify it?
Asked what iron is, we say "a mineral." That is the category of which it
is a member. Likewise, philosophers want to ascertain the concept or
concepts under which cinema falls. For instance, is it essentially photo-
graphy, is it art, is it both, or is it neither? What is the nature of cinema?
The ontology of cinema interrogates the mode of existence of cinema —
its manner of existing (of being). It aspires to establish the kind of thing
cinema is essentially. Another way of framing the ontologist's question is to
ask "What is the essence of cinema?"
The philosophy of any practice attempts to clarify the concepts that are
indispensable to carrying on that practice. The notion of a number is
fundamental to mathematics; so the philosophy of mathematics tries to
define number, asking whether a number is something real or only a logical
fiction. Similarly, since the concept of a law is the sine qua non for the practice
of the law, the philosopher of law speculates about the nature of law — is a law
essentially merely an injunction promulgated by the appropriate people
(such as the members of a parliament) in accordance with the correct
procedures (such as voting), or does a genuine law derive from something
much deeper, like a universal moral code? Of course, law is not the only
concept that is essential to the practice of law. Concepts like intention,
voluntariness, and guilt are also necessary to our legal practices. Conse-
quently, they too fall under the purview of the philosophy of law.
With regard to the cinema, there are a plethora of concepts that are
essential to the practices of making, interpreting, and evaluating motion
54 WHAT IS CINEMA?

pictures. We will analyze a number of them in this book. But before plunging
ahead, we need first to denominate the boundaries of our investigation.
We need to answer the question "What is cinema?"

Defining Cinema
The issue of defining cinema arose within the first two decades of the
advent of moving pictures. One reason for this was that cinema was a new
thing under the sun and, therefore, it needed to be fitted into our scheme
of things — where it is, the task of philosophy, in the intellectual division of
labor, to situate items in our biggest picture of the way things are.
But there were other reasons why commentators became engaged with
defining the essence of cinema. Some, as we saw in chapter 1, believed that
a proper analysis of cinema would reveal that it could not be art, properly
so called, given its photographic provenance, whereas others were invested
in demonstrating that, to the contrary, an accurate analysis of the nature of
cinema would show that art was within its reach. Thus, both sides of the
debate needed an account of the essence of cinema in order to make their case.
Moreover, as we saw in chapter 2, many of the friends of film — those
who characterized cinema as possessing inherent artistic capacities — also
thought that their analysis of the film medium would provide them with
the wherewithal to tell the good films from the bad ones.
Of course, the pressure to prove that moving pictures can be art is, by
now, far behind us. Furthermore, as we saw in the previous chapter, we
have reason to believe that a descriptive definition of film will not provide
us with a reliable set of normative guidelines to the filmically good, the
bad, and the ugly. Nevertheless, even if isolating the essence of cinema
holds forth no promise of disclosing the secrets of cinematic success,
curiosity still lingers about the place of cinema on our overall conceptual
map. How does cinema stand in relation to things like theater and painting?
We should at least know what we are talking about when we say we are
going to the movies, lest we find ourselves at the ballet. For such reasons,
we need a definition or analysis of cinema.
When philosophers attempt to produce what they variously call a
definition, a theory, or an analysis of x (whatever x is), they usually have
in mind something very specific — what is often called an essential or real
definition. What is a real definition? The short answer is: a definition or
analysis of some concept in terms of a set of necessary conditions that are
conjointly sufficient for the application of said concept. But what are those?
 WHAT IS CINEMA? 55

A necessary condition states a property that something must possess in

order to be a member of the category under analysis. For example, in order
to be a nun in the Roman Catholic Church, one must be a woman. It is a
necessary condition or a necessary criterion for the sisterhood that the
candidate be a woman. That is, someone is a nun only if she is a woman.
There are several other necessary conditions for sisterhood, including that
one be a Catholic, that one take the requisite vows, and that one be inducted
by the proper authorities. Taken together — that is to say conjointly — these
necessary conditions are sufficient for classifying a candidate as a nun; they
add up, so to speak, to counting the woman in question as a sister. That is, if
someone is a woman, a Catholic, has taken the requisite vows, and has been
inducted by the proper authorities then she is a nun.
You can usually tell if a philosopher is trying to offer you a real definition
by his use of the phrase "if and only if." For example, "x is a bachelor if
and only if (1) x is male and (2) x is unmarried" is a real definition, an
analysis in terms of two necessary conditions (signaled by the locution "only
if") that are conjointly sufficient (signaled by the word "if"). The necessary
conditions tell us what properties a candidate must have in order to be an
instance of the concept under analysis; the sufficient conditions mark off
how members of the relevant class of things differ from the members of
other classes — how, for example, bachelors differ from husbands, on the
one hand, and spinsters, on the other.
Generally^ when asked a question of the form "What is x?" the philoso-
pher will initially attempt to provide a conceptual analysis or real definition
of x in terms of necessary conditions that are conjointly sufficient. In this
chapter, we are facing the question "What is cinema?" Let us see whether
or not we can answer that question in terms of a real definition.

Seeing x Versus Seeing a Motion Picture of x

As we have learned, one of the earliest proposals for defining cinema was
the idea that cinema is essentially nothing but moving photography, a
process of sheer mechanical recording. With this definition comes the
prejudice that cinema cannot be art, since the mere mechanical recording
of something putatively leaves no opportunity for the artist to add or to
express her thoughts and feelings regarding whatever she has filmed.
The suggestion here was that whatever you see on film is exactly what
you would see in a face-to-face encounter with the object that stood before
the camera. Seeing a motion picture of x, say a stone falling, in other
56 WHAT IS CINEMA?

words, is, on this conception, putatively exactly like seeing the stone
falling in front of you, where your view is not mediated by cinema.
Consequently, just as a falling stone in nature literally expresses nothing,
so a moving picture image of one expresses nothing. The moving picture
image is rather like a mirror — it simply delivers up the way things look and
nothing more. What you see on film, what you see in a mirror, and what
you see in a close encounter of the third kind with whatever gave rise to
the aforesaid cinematic image supposedly are perceptually equivalent.
Seeing a cinematic image, according to this theory, is precisely like seeing
whatever it is an image of in the pro-filmic world — the world as it is
arrayed in front of the camera.
That is, perceiving a cinematic image is virtually the same as seeing its
pro-filmic referent; the cinematic image itself is, in a manner of speaking,
diaphanous. Thus, what we might call "cinematic seeing" is just like normal
perception. The normal perception of something and its perception via
cinema are in nowise appreciably different.
This view, as we have already seen, is false. There are quite a few
differences between seeing an object in nature and seeing an object on film.
For example, for many years, most of the objects seen on film were black
and white, whereas their referents were in living color. Undoubtedly,
cinematic color is still often perceptually discernible from color in nature.
However, perhaps color technology could be improved to the point where
any differences in hue that the naked eye might detect could be erased.
Indeed, let us imagine that motion picture technology advances to the
point where moving picture images of objects and events are virtually
indiscernible by normal viewers from the things of which they are images.
Suppose, for instance, that so many scan lines are added to high-definition
television images that the unaided eye cannot confidently cut the difference
between a medium shot of Samuel L. Jackson and Samuel L. Jackson in the
flesh. Imagine, in other words, that the image was as persuasive as a sculpture
by Duane Hanson. Is there nevertheless some fundamental or essential differ-
ence between seeing a moving picture image of x and a normal perceptual
encounter with x in, so to speak, nature?

Detached Displays

One such difference is the fact that in a face-to-face perceptual encounter

with an object, I am typically able to orient my body toward that object
within the very spatial continuum that I inhabit physically. If I see a
 WHAT IS CINEMA? 57

Mercedes Benz parked across the street, I can point my body in its direction
and walk to it. Normal perception enables me to navigate my way in the
space where I live toward the objects I see in the distance. Normal veridical
perception, that is, comes with egocentric information about where I stand
in relation to what I see. But this characteristic feature of normal perception
contrasts radically with our perceptual encounters with movies.
For when I see an image of the Kremlin in cinema, I cannot, on the basis
of the spatial information available intrinsically in that image, orient myself
to that precinct of Moscow, nor, if it is a movie set rather than the real
thing, can I point myself in the direction of the locale where the motion
picture was shot. If we are, in a manner of speaking, attached to the arrays
that surround us visually in our lives off screen, then cinematic images are
detached displays — they display persons, places, things, actions, and events
that are phenomenologically detached, in the preceding sense, from the
actual space of our bodies.
Suppose that I am watching Casablanca and what I see onscreen is Rick's
bar. I can, of course, orient myself to the screen; I can point myself in its
direction. The screen is part of my phenomenologically attached array. But
I cannot, on the basis of the spatial information intrinsic to the shot
onscreen, orient myself reliably toward Rick's bar — toward the spatial
coordinates of that structure as it existed some time in the early 1940s in
California (nor could I point my body, by means of the information in the
image, toward, or away from the fictional locale of the film in North
Africa). Simply by looking at the image on the screeni I would not be
able to walk, drive, swim, fly, or otherwise move toward the site of Rick's
bar — toward the remnants, if any, of the set on some sound stage in LA.
The image itself would not tell me how to get to the set, presuming that
part, if any, of it still remains, nor the way in which to get to the place in
the world where, if it no longer exists, it once did. For the space, so to say,
where the set of Rick's bar is located is disconnected phenomenologically
from the space in which I dwell bodily as a physical being.
Some readers might say that when they see a movie image of Big Ben in
London, then they are confident that they will be able to point their body
in right direction. But if this is so, it is because these viewers already know
in which direction London lies. It is not information intrinsic to the image
on the screen that gives them their marching orders. It is the extrinsic
knowledge of geography which they antecedently possess that sets them
straight.
Ordinarily, our sense of where we are depends upon our sense of balance
and our kinesthetic feelings. What we see is integrated with these cues in
58 WHAT IS CINEMA?

such a way as to yield a sense of where we are situated. But if we call what we
see on the silver screen a "view," then it is a disembodied view. When I see a
visual expanse, like Rick's bar, onscreen, I have no sense of where the
portrayed space really is in relation to my body. It is detached. I cannot relate
myself reliably to it spatially; it is not attached spatially to my life-world.
In the past, I think, film theorists have attempted to get at the essence of
cinema by talking obscurely about how film engenders a peculiar sort of
absence, or even a unique play of presence and absence. For the most part,
this way of speaking has been unfortunately vague, even murky. However,
perhaps what these theorists were trying to get at by the notion of absence
is the way in which the cinematic display is discontinuous from the space
we inhabit. What is present to us is a display, but the referent of the display
is discontinuous with, or detached or absent from, the space in which our
eyes can plot the way to our destination. Thus, the play of presence and
absence that intrigued film theorists might be better characterized by
saying — in a more positive and descriptively substantial way — that, in
cinema, the image or image series is a detached display.
This feature of the cinematic image, moreover, appears universal; it
pertains to every moving picture representation. There are no exceptions.
Therefore, we may begin to answer our animating question — "What is
cinema?" — by means of a first approximation: "x is an instance of cinema
only if it, x, is a detached display (or a series of detached displays)." That is,
being a detached display is a necessary condition for being an instance of
cinema.

Moving Images
That the cinema is a detached display draws a distinction between cinematic
arrays and the normal perception of the visible "attached" arrays that give
rise to them. However, it does not draw a distinction between near
neighbors of cinema, like painting and photography, on the one hand, and^
theater, on the other. Quite clearly, these artforms are also detached
displays. For neither from a painting of Elsinore nor from a stage set of
Hamlet can one point ones body toward that castle on the shore of Denmark.
So what distinguishes cinema from still pictures and theater? Let's start with
still pictures — with paintings, photographs, and the like.
Imagine a painting of a cartoon character, say Manga, by some
postmodern artist. Now imagine alongside of it a film loop of the same
cartoon character striking precisely the same pose. Like the painting of the
 WHAT IS CINEMA? 59

cartoon character, the image of Manga on the filmstrip remains frozen.

Both images are the same size. The image on the screen is a high-definition
one of the most technologically sophisticated sort, and there is no flicker
effect. Let us suppose, then, that the two images — the painting and the film
image — are indiscernible to the naked eye. No one can see the difference.
But is there nevertheless a difference and, if so, what is it?
If you were confronted by these two images and you were told that the
one on the left was a film projection and that the one on the right was a
painting, it would make sense for you to keep watching the one on the left,
at least on first viewing, to see whether the cartoon character ensconced
there starts to move, whereas anticipating that Manga might move in the
painting would be absurd. For movement is a possibility in movies but not
in painting. That is why in ordinary language we call film images moving
pictures or motion pictures. One of the earliest labels for film was bio — as
in "biograph" — which meant life and signaled animation. On the other
hand, paintings, photographs, slides, and so forth are called still pictures,
pictures which, by definition, do not move — indeed, pictures for which
the very possibility of movement is ontologieally foreclosed.
The difference between a painting and a cinematic image, then, is
movement. This much is perhaps already built into common speech in
locutions like the moving picture show. But we need to be careful about the
way in which we exploit this clue from ordinary language. It is not simply
the case that cinematic pictures move and still pictures do not. For there
are a number of films in which there is no movement, such as Nagisa
Oshima's Band ofNinjas (a film of a comic strip), Michael Snow's One Second
in Montreal (a film of photos) and his So Is This (a film of sentences), Hollis
Frampton's Poetic Justice (a film of a shooting script), Godard and Gorin's
Letter to Jane (another film of photos), and Takahiko Iimura's 1 in 10 (a, film
of addition and subtraction tables).
A perhaps better-known example than any of these is Chris Marker's La
Jetee, a film of almost no movement whose time-travel narrative is told
primarily through the projection of still photographs. Of course, there is
one movement in Marker's film; however, it should be easy for you to
imagine a film just like La Jetee but with no movement whatsoever.
Thus, it would be wrong to require movement as a necessary condition
for membership in the category of cinema, since a film like the reimagined
La Jetee is indisputably a motion picture. Rather, it is more accurate to
propose that something is an instance of cinema only if it belongs to the
class of things for which creating the impression of movement is a technical
possibility.
60 WHAT IS CINEMA?

However, some might balk at this adjustment, resisting the impulse to

accommodate motionless films into the corpus of cinema. It might be
charged that the prospect of movies without movement is at least oxymoro-
nic (motionless movies! Holy Cow!), and perhaps even self-contradictory.
Such experiments are little more than slide shows mounted on celluloid,
maybe merely for the purpose of assuring the most efficient mode of
projection.
Yet there is a profound difference between a shot of a character from
our imagined movie version of La Jetee and a slide image of that character
abstracted from the film. For as long as you know that what you are
watching is a film, even a film of what appears to be a photograph, it is
always justifiable to entertain the possibility that the image might move. On
the other hand, if you know that you are looking at a slide, then it is
categorically impossible that the image might move. If you know that what
you are looking at is a slide and you understand what a slide is, then it is
irrational to suppose that the image could move because it is conceptually
impossible.
Movement in a slide would require a miracle; movement in a cinematic
image is an artistic choice which is always technically available. Though
most frequently exploited, it can also be forgone for artistic effect. Before
Band ofNinjas concludes — that is, until the last image flickers through the
projection gate — the viewer may reasonably presume, if she knows she is
watching a film, that there may yet be movement onscreen. For movement
is a permanent possibility in cinema. But if she knows she is looking at a
slide, then it would be absurd for her to entertain even the possibility that
it might burst into movement. It would be irrational of her to do so, of
course, because if it is a slide and she knows what a slide is, then she must
know that it is beyond the capacity of a slide to support movement within
the boundary of its frame.
Furthermore, the difference between slides and moving images applies
across the board to the distinction between every species of still picturing —
including paintings, drawings, still photos and the like — and every sort of.
moving picture — including videos, mutoscopes, movies, computer-generated
animation, and so forth. When it comes to still pictures, one commits a
category error if one expects movement. It is, by definition, self-contradictory
for still pictures to move. That is why they are called still pictures. That
is why it is conceptually confused to look at what one understands to be
a painting with the expectation that it might move. But it is eminently
rational — and never unreasonable — to stay on the lookout for movement in
a film because of the kind of thing — a motion picture or a movie — that a
 WHAT IS CINEMA? 61

film is. Even with a static film, like Poetic Justice, it makes sense to wonder
whether there will be movement up until the last inch of film has run its
course.
With a film like Poetic Justice, it is an intelligible question to ask why the
filmmaker, Hollis Frampton, made a static film, since he had movement as
a genuine stylistic option at his disposal. But it makes no sense to ask why
Holbein forswore literal movement in his painting The Ambassadors. Unlike
Frampton, he had no other alternative. Asking why Holbein's ambassadors
don't move is like asking why spiders don't sing Tosca.
Of course, once one has seen a static film from beginning to end, it is no
longer justifiable to anticipate possible movement in repeated viewings,
unless you suspect that the film has been tampered with since the time of
your initial viewing. On first viewing, it is reasonable, or at least not
irrational, to be open to the possibility that movement will appear onscreen
up until the final credits; on second, third, and subsequent viewings, such
anticipation is out of place. However, on first viewings, one can never be
sure that the film is entirely still till it is over. And this is what makes it
plausible to continue watching for movement throughout the first viewing
of static films. But to await movement from what one knows to be a painting
or a slide is conceptually absurd.
Why categorize static films as cinema rather than as slides or as some
other sort of still picture show? Because, as already suggested, stasis is
stylistic choice dn static films. It is an option that contributes to the artistic
effect of a film. It is something whose significance the audience contem-
plates when trying to make sense of a film. Encountering Lajetee, a natural
interpretive question is, "Why did Chris Marker choose to make a movie
out of still photographs?" Perhaps the answer has to do with his attempt to
simulate the fragmentary nature of memory. But, in any event, note that
this approach makes sense with a static film, whereas it would not with a
slide show, every image of which, of necessity, is a still photograph. That is,
one crucial reason to countenance the notion of static films and to modify
our definition of cinema to assimilate them is an aesthetic reason, viz.,
stasis is a cinematic variable of potential significance in our interpretive and
appreciative responses to a motion picture, whereas stasis is an unalterable,
and therefore mute and meaningless, constant in a slide show.
With an individual shot, such as the one that concludes Lo Wei's movie
The Chinese Connection and which shows Bruce Lee, frozen mid-air deliver-
ing a flying kick, the arrest of movement exercises a cinematic option with
expressive potential. In this particular instance, freezing the frame, among
other things, serves to memorialize Lee and his character — to turn him
62 WHAT IS CINEMA?

into the stuff of legend at the height of his prowess and glory. But just as
the choice of stasis can have an expressive effect by eschewing motion at the
level of the individual shot, so too an entire film may accrue meaning by
staying still, whereas a painting or a slide cannot exploit this possibiUty
precisely because its only option is to be motionless.
It is informative to say that a film is static; it alerts a prospective viewer
to a pertinent lever of stylistic articulation in the work. Contrariwise,
there is no point in saying of a painting that it is a literally still painting. It could
not have been otherwise. To call a painting or a slide a still painting or a still
slide is redundant.
One can imagine a slide of a procession and a cinematic freeze frame of
the exact same moment in a parade. The two images may, in effect, be
perceptually indiscernible. And yet they are metaphysically different.
Moreover, this difference translates into a difference in the epistemic states
of the respective viewers of the slide versus the cinematic freeze frame.
With the motion picture, the anticipation that the freeze frame may start
to move is reasonable or, at least, conceptually permissible; but with still
pictures such as slides it is never conceptually permissible. The reason for
this is quite clear. Film belongs to the class of things where movement is a
technical possibiUty while paintings, slides, and the like belong to a class of
things that are, by definition, still.
Ordinary language broadly informs us about one of the necessary condi-
tions of cinema by referring to it collectively as "the motion pictures" But
the wisdom in this common way of speaking needs to be unpacked a bit. It is
not the case that every film or every film image imparts an impression of
movement. As we have seen, there are static films. However, static films
belong to the class of things where the possibility of movement is always
technically available in such a way that stasis is a styUstic variable in films in a
way that it cannot be with respect to still pictures. Perhaps this is one of the
reasons that "moving pictures" or "motion pictures" serve as better labels
for the practice of cinema than does "film," since these phrases advertise
what may be the most pertinent essential feature of the artform, namely, the ?
possibiUty of movement.
Of course, the category of moving pictures is somewhat broader than that
which has traditionally been discussed by film theorists, since it would
include such things as video and computer imaging as well as celluloid-
based cinematography. Indeed, if we could produce moving images sono-
graphically, those too should count as motion pictures. Moreover, this
expansion of the class of objects under consideration to moving pictures
in general is, in my opinion, theoretically advisable, since I predict that one
 WHAT IS CINEMA? 63

day the history of what we now call cinema and the history of video, TV,
computer-generated imaging, and whatever comes next will all be thought
of as a piece.
That is, by defining cinema as belonging to the class of the moving
picture, we are characterizing cinematic artifacts in terms of their function —
rather than their physical basis — which function is, in short, to impart the
impression of movement. This function can be implemented in an indefinite
number of media — celluloid-based film, most obviously, but also video,
broadcast TV, handmade cartoon flip books, CGI, and, truth be told, who
knows what successive generations will invent? Furthermore, the capacity
to impart movement to pictures is also connected to an even narrower
function that, for many, is the central cultural rationale for cinema — its
ability to narrate visually by means of pictures that move.
Nevertheless, there is at least one limitation in calling the relevant
artform moving pictures. For the term "picture" implies the sort of inten-
tional visual artifact in which one recognizes the depiction of objects,
persons, and events simply by looking, rather than by reading or decipher-
ing or decoding. That is, we recognize what pictures are pictures of
through the operation of certain subpersonal perceptual routines that are
automatic and bred in the bone. But not all of the images found in cinema
are of this sort. Many films and videos are abstract or nonfigurative or
nonrepresentational. Think of the films of figures like Viking Eggeling,
Marcel Duchamp, and some of the work of Stan Brakhage. These works
may be comprised of non-recognizable shapes and purely visual structures.
Thus, rather than speaking of moving pictures, it may be more advisable
to talk of moving images. For the term image covers both pictures and
abstractions — Hokusai and Kandinsky; Zhang Yimou and Jordan Belson.
That is, the concept of the moving image — like that of the still image — is
more comprehensive than that of the moving or motion picture. For
whether the image is pictorial or abstract is less relevant for the purposes
of classification than that it is moving imagery in the sense of possessing the
technical possibility of movement.

Types, Tokens, and Templates

By adverting to the possibility of movement, we have been able to draw a
boundary between cinema, on the one hand, and painting, drawing, still
photography, and so forth, on the other hand. And we have added another
necessary condition to our characterization of cinema: x belongs to the
64 WHAT IS CINEMA?

category of cinema (1) only if x is a detached display (or a series of

detached displays), and (2) only if x belongs to the class of things from
which the impression of movement is technically possible. However, these
two conditions are not jointly sufficient to differentiate film from its near
neighbor theater.
For a theatrical performance is also a detached display. Watching a
theatrical performance of A Streetcar Named Desire, we cannot orient our
bodies — on the basis of the visual information onstage — in the direction of
New Orleans. The space in the play is not the space of my body. It is not
true, as I look on, that Julius Caesar dies fifteen feet in front of me, even if
I am sitting in the first row. Nor can I point my body toward the forum
in ancient Rome on the basis of the information available intrinsically in
the theatrical image before me. Moreover, quite clearly, theater, as cinema,
affords the possibility of movement.
Cinema and theater are in many ways alike. Cinema, as it is most
commonly practiced, is an art of dramatic fiction, telling stories through
the play of actors. In many ways, cinema was a successor of theater. Even
the language of cinema echoes that of theater. We show movies in
"theaters," and movie showings are sometimes still called "performances";
at the turn of the century, motion pictures were called "photoplays." Many
movie emporiums were housed in refurbished "legit" theaters where some
continued to employ rising and falling curtains to signal the beginning and
end of the screening.
Indeed, it is hard to deny that cinema was a rival to live theater,
providing a cheap substitute whose economies of scale eventually came
to dwarf the grandest spectacles on stage. This rivalry, moreover, as we
have seen, fueled some of the earhest debates about film, with the lovers of
live theater impugning the artistic credentials of film, calling it nothing but
the slavish mechanical reproduction of live dramas,
And even today, when the battle between cinema and theater has abated,
the affinities between the two remain deep. Arguably it remains the case
that the best training for film actors and film directors is theater. Film
acting and directing, of course, involve skills different from live theater.
Nevertheless, the basics of building character and orchestrating ensemble
playing are still best learnt in drama schools.
Furthermore, in addition to these historical correlations, there is also
a deep metaphysical bond between theater and cinema. Both are what
philosophers classify as type-artforms or multiple-instance artforms. That
is, one possible division amongst the arts is between artforms that involve
unique singular objects, like typical, one-of-a-kind paintings, and those arts
 WHAT IS CINEMA? 6S

that involve multiple copies of the same artwork. The novel, for example,
is a multiple-instance artform. There are probably millions of copies of the
novel Ivanhoe.
Multiple-instance artforms, in turn, can be analyzed in terms of the
type—token relationship. Just as the dime in my pocket is a token of the type
designed by the US Treasury, so my copy of Ivanhoe is a token of the
type authored by Sir Walter Scott in the nineteenth century. My token of
Ivanhoe grants me access to the type — the artwork— that Sir Walter Scott
created. Ivanhoe, moreover, is not an ordinary material object. It is not
Scott's original manuscript, which itself is a token of the novel. For the
original manuscript could be lost or destroyed and we would still say
that Ivanhoe exists. In fact, every copy of Ivanhoe could be destroyed and it
could continue to exist, perhaps as human memory traces — a possibility
illustrated by Ray Bradbury's novel Fahrenheit 451 (which was adapted for
the screen by Francois Truffaut).
Along with novels, plays and films also fall into the category of multiple-
instance or type-artforms. Tonights performance at 8 p.m. of the play
Romance is a token of the comedy created by David Mamet, just as tonight's
performance (showing, screening) at 8 p.m. of Raging Bull is a token of
the movie-type directed by Martin Scorsese. Thus theater and cinema are
brethren ontologically as well as historically.
On the basis of the type/token distinction theatrical performances
and movie performances do not look very different. In both cases, the
performance in question is a token of a type. Does this imply that there
really is no deep difference between theatrical performances and motion
picture performances? Does it suggest that there is no significant ontological
divide between theater and cinema?
No. For, though the simple type /token distinction is useful as far as it
goes, it does not go far enough. Theatrical performances and motion
picture performances may, indeed, both justifiably be called tokens; how-
ever, the ways in which these tokens are generated from their respective types
diverge profoundly. And therein lies a clue to the ontological distinction
between theater and cinema. Specifically: live theatrical, token performances
are generated by interpretations, whereas token performances of motion
pictures are generated by templates.
Moreover, this yields a perhaps unexpected artistic distinction between
the two kinds of performance. Theatrical performances are artworks in
their own right and, as such, are subject to artistic evaluation, but the
performance of a motion picture (a motion picture showing) is neither an
artwork nor a reasonable candidate for artistic evaluation.
66 WHAT IS CINEMA?

The motion picture performance — a film showing or screening — is

generated from a template. Standardly, this was a film print, but in recent
decades it might be a videotape, a laser disk, a DVD, or an instantiated
computer program. It is through these templates that we gain access to the
original designs of the creators of the motion picture in question.
These templates are tokens. Each one of them can be assigned a location
in space and time; each one of them can be destroyed. But the film — say,
Chungking Express by Wan Gar Wai — is not destroyed when any of the prints
are destroyed. One might think that the master or negative is privileged.
But the negative of Murnau's Nosferatu was destroyed as the result of a court
order in response to charges that the film plagiarized Dracula, yet Nosferatu
survives and will continue to do so as long as its design is retrievable either
through duplicate prints, DVDs, a collection of photos of all of the frames,
or as a digital program on disk, or tape, or paper, or otherwise.
To get to a token performance of a movie — tonight's screening of Ozu's
/ Was Born But . . . — we require a template which is itself a token of the
film type. The film performance — the projection or showing of/ Was Born
But . . . — is a token of the type created by Ozu, which token conveys
Ozu's masterpiece, the type, to us, the spectators.
This account contrasts sharply with the itinerary from the play type to
the token performance of the play. Plays have as tokens both objects
(the script) and performances (the enactment of the drama). That is,
when considered as a literary work, a token of The Bacchae is a graphic
text of the same order as my copy of the novel The Devils. But considered
from the viewpoint of live theater a token of The Bacchae is a performance
which occurs at a specific time and place.
Moreover, unlike the movie performance (the performance of the
movie), the theatrical performance is not generated by a template. It is
generated by an interpretation. For, when considered from the perspective
of a live theatrical performance, the play by Euripides is much like a recipe
to be filled in by other artists including the director, the actors, the set and
lighting designers, the costumers, and so forth.
This interpretation or set of interpretations is a conception of the play
and it is this conception of the play that governs its performance from night
to night. The interpretation may be performed in different theaters and on
different nights. It may be revived after a hiatus. For the interpretation is a
type, a type which mediates between the play-text type by Euripides and
its token performance at such and such a place and time. Thus, whereas
token performances of a motion picture are generated by templates, token
performances of a play type are generated by a type — an interpretation
 WHAT IS GINEMA? 67

of the play type. The connection between the play type and its token
performance is negotiated by a conception of the play, itself a type, thereby
indicating that an interpretation is a type within a type. What gets us from
the play to a token performance is not a template, which is a token, but an
interpretation that is a type.
Furthermore, and of the utmost importance: an interpretation is an affair
of intentional or mental causation. The token performance of The Bacchae, in
other words, is mind-generated. The token performance of the movie, on
the other hand, is mechanically (and/or electronically) generated; it is the
product primarily of physical causation. Perhaps this distinction is already
built into the phrase "live theater." But it pays to be pedantic here. Live
theater is to be understood in contrast to mechanical movies (cinema) where
what makes the former living is that it is the product of mental or intentional
mediation via an interpretation or series of interpretations.
So one difference between the performance of a play and the performance
of a motion picture is that the former is generated by an interpretation and
the latter is generated by a template. Moreover, this difference is connected
to another very interesting distinction* namely, that performances of plays
are artworks in their own right and can be aesthetically evaluated as such,
whereas performances of films and videos are not artworks. Nor does it
make sense to evaluate them as such. A film may be projected out of focus or
the video tracking may be badly adjusted, but these are not artistic failures.
They are mechanical or electrical glitches. That is, a film projectionist may
be mechanically challenged but he is not artistically incompetent. Artistic
competence does not even enter the picture.
In the theater, the play, the interpretation, and the performance are each
discrete arenas of artistic achievement. It is to be hoped, of course, that
they will be integrated. And in the best of all cases they are. Nevertheless,
we recognize that these are separable strata of artistry. We realize that
they can come apart. We often speak of a good play interpreted blandly, of
a mediocre play interpreted ingeniously and performed brilliantly, and
every other combination thereof. But this manner of speaking, of course,
presupposes that we regard the play, the interpretation, and the perform-
ance as separate levels of artistic achievement — even where the play is
written by someone who directs it and acts in it as well. The play by the
playwright is one artwork which is then interpreted like a recipe or set of
instructions by a director and others in the process of producing another
artwork — the performance of the play.
In this regard, theater, like music and dance, can be called a two-tiered
artform. In music, at least since the advent of notation, there is the score — the
68 WHAT IS CINEMA?

artwork created by the composer; and then there is also the performance —
the score as played by a group of musicians. Similarly, there is the choreo-
graphy of the dance which is one artwork; and then there is choreography
as performed by the troupe which constitutes a discrete artwork. And
in theater, there is the play — if not of the well-made sort then, at least,
some kind of performance plan — and subsequently its execution (or
executions).
But cinema is not a two-tiered artform. There is one, and only one,
artwork in cinema — the finished film or video. In this regard, though a
multiple-instance artwork, a finished film or video is like a painting and
unlike a theatrical performance, which, in a manner of speaking, finishes
the play-text, or an interpretation thereof, by enacting it. It is that
enactment that deserves to be treated as an artwork in its own right.
But the performance of a motion picture — a film showing — is not an
artwork in its own right. With two-tiered artforms it makes sense to say
that last night's performance was better than tonight's. For example, we
might say of a ballet that the corps danced more crisply in yesterday's
rendition but is a bit sloppy today. But we cannot talk that way about films.
What would we make of someone who said Gregory Peck's portrayal of
Atticus Finch in To Kill a Mockingbird was riveting last week but a bit shabby
last night? The suggestion strains intelligibility.
Our practices with respect to theater and motion pictures are very
different. If* in theater, the play-text type is a recipe or sketch that the
director and/or the actors interpret, and, furthermore, if the recipe and
the interpretation can be regarded as different though related artworks, in
motion pictures the recipe and the interpretations are constituents of the
self-same integral artwork. When the playwright creates a play, we appre-
ciate it independently of what its theatrical interpreters make of it. We can
and should read a play-text as we read a novel.
But in the world of motion pictures, as we know it, scenarios are not
read like plays and novels. For example, the voters at the Academy Awards
do not read the scripts of nominees when they vote for the best screenplay;
they watch the movie in which it is a contributing creative ingredient. The
only people who read screenplays, apart from actors and directors, are people
who are trying to learn how to write them or the people who teach people to
write them or film scholars engaged in historical and/or critical research.
In contrast to the theater, where the recipe and the interpretation are two
different artworks, in cinema the recipe and its interpretation are presented
together in one indissoluble package. The scenarios of most motion
pictures, like Duel, are not literature, whereas the typical play-text, save
 WHAT IS CINEMA? 69

perhaps for comedy skits, lays claim to literary status — that is, the status of
being an artwork in its own right.
Movie scripts are nothing like the play-text of Hedda Gabler. Often, as in
the case of Sunset Boulevard, shooting begins without a completed script;
pages arrive on the set day by day, just in time to start the next scene.
Moreover, the script is usually reworked and reworked, often as the movie
is being shot, and even after. The director changes it, the actors change it
(often simply by paraphrasing the screenwriter's lines their own way), then
the writers have second thoughts, and the producer intervenes — sooner or
later everyone puts in their two bits, and so on. These are not the products
of the uncompromising modernist genius a la Becket. The movie scripts
that come to us in published form are not the scripts used during the
production process — which were quite frequently being constantly altered
as the movie was being made — but are m o r e of the nature of transcriptions
or records of the words that finally wound up being said in the finished
product. They are not the recipes that were used to launch an interpret-
ation. They are documents of what got made — less recipes, and m o r e parts
of the dish that was served. That is, once again, they are ontologically
ingredients in the motion pictures with which they are associated rather
than being independent artworks.
Let us now also consider another difference between theater and
cinema, namely, the putative contrast between how we regard the relation
of movie a c t o p to their characters and how we understand the relation of
stage actors to theirs. Frequently, people appear to hold the conviction that
many different actors can play Khlestakov and the performance will still
be a token of the play-type The Inspector General, but that it would not be
the film-type Shadow of a Doubt if someone other than Joseph Gotten played
Uncle Charlie. The reason for this is that Joseph Gotten's performance
of Uncle Charlie — his interpretation of the role — in concert with
the direction of Alfred Hitchcock is ostensibly a non-detachable constituent
of the movie. It would not be a screening (a token performance) of Shadow
of a Doubt if this particular ingredient were missing. Cotten's interpretation
of Uncle Charlie has been etched onto celluloid indelibly. That is, speaking
less metaphorically, the interpretation by the thespians in the case of
film is not separable from the film-type in the way that such interpretations
are separable from the play-type. W h e n new actors are cast in what are
called movie "remakes," we count the results as a new film-type; b u t for
all the cast changes that Hamlet has undergone in the course of history,
there is still the one and only play-type, the one created by William
Shakespeare.
70 WHAT IS CINEMA?

Whereas film performances — performances of the film — are generated

from templates which are tokens, play performances are generated from
interpretation types. Thus, where film performances in the relevant sense
are counterfactually dependent on certain electrical, chemical, mechanical,
and otherwise routine processes and procedures, play performances are
counterfactually dependent upon the beliefs, intentions and judgments of
people — actors, directors, lighting designers, make-up artists, and so on.
Though, in modern Western theater, there is typically an overarching
directorial interpretation of the playwright's recipe, the realization of
the token performance on any given night depends on the continuous
interpretation of that play, given the special exigencies of the unique
performance situation. It is because of the contribution that interpretation
qua intentional action makes in the production of the performance that
the performance warrants artistic appreciation, whereas the performance
of a film — a film showing — warrants no artistic appreciation, since it is
simply a function of the physical mechanisms engaging the template in the
right way.
A successful motion picture performance — the projection of a film or
the running of a videocassette or a DVD — does not command aesthetic
appreciation. We do not applaud projectionists as we do violinists or
pianists. We are likely to complain and perhaps to demand our money
back if the film emulsifies in the projector beam, but that is a technical
failure, not an artistic failure. If it were an artistic failure, wouldn't we
expect the audience to cheer when the film doesn't burn? But they don't.
For the happy motion picture performance merely depends upon operating
the apparatus as it was designed to be engaged, and, since that involves
no more than often quite minimal mechanical know-how, running the
template through the machine is not regarded as an artistic achievement.
On the other hand, a successful theatrical performance involves a token
interpretation of an interpretation type, and, inasmuch as that depends
upon intentional factors, such as artistic understanding and judgment, it is
an apposite object of aesthetic appreciation.
Furthermore, if the preceding is correct, then we may conjecture that a
major difference between motion picture (or moving image) performances
and theatrical performances is that the latter are artworks and the former
are not. Thus, performances of motion pictures are not objects of artistic
assessment, whereas theatrical performances are. Another way to state this
conclusion is that, in one sense, cinema is not a performing art — i.e., an
artform whose token performances themselves are artworks — in contrast
to theater, which is a performing artform in the requisite sense.
 WHAT IS CINEMA? 71

This, of course, sounds somewhat counterintuitive. Undoubtedly

what is so disturbing about this suggestion is that cinema artists — such as
actors — appear to do precisely the same sorts of things that the stage actors,
whom we call performing artists, do. Stage actors act; film actors act. If the
former are performing artists, so too are the latter. So we seem to be denying
the obvious.
Let us grant this much. However, the obvious may be more complex
than it appears at first glance. When we go to a performance — a showing —
of Seven Samurai, we are not viewing a performance by Toshiro Mifune. If
we say that it is a performance by Mifune that is on the screen, then we must
immediately qualify our statement: it is not a performance by Mifune, but
the one-and-only one; it is the same every night. Mifune's acting is an
inseparable element of Kurosawa's Seven Samurai. Mifune's acting is not a
performance out of many in the sense that we can track its vicissitudes from
night to night. There can be no variation from showing to showing. A film
performer is not a live performer.
As Walter Benjamin pointed out in his often quoted essay, "The Work of
Art in the Age of Mechanical Reproduction," the stage actor can adjust his
performance to;the living audience before him through his interpretation.
If the audience seems to appreciate broad humor, she might emphasize the
sexual innuendo of certain lines. But the motion picture actor's interpret-
ation is fixed on the template for good.
Returning to the example of Mifune, his artistic contribution is done by
the time his acting is stored on the reel of celluloid and shipped off in the
film can. Whatever alternative interpretations Mifune essayed in his role in
Seven Samurai were lost on the cutting-room floor. Mifune's definitive inter-
pretation of his role was invented before the first performance of the film,
Mifune's interpretations of each scene in Seven Samurai were integrated
and edited into the final product of the motion picture-type before it ever
reached any distribution site. By the time those objects reached the movie
theaters, there was nothing left for Mifune or any other performing artist
to do. It was just a matter of a technician running a template through a
machine by, as they say, the numbers.
In sum, there are two striking differences between the performance of a
motion picture and the performance of a play. One is that the play
performance is generated by an interpretation that is a type, whereas,
with respect to movies, the moving image performance is generated by a
template that is a token. The second difference is that the performance
itself of a play is typically an artwork in its own right, while the token
performance of a motion picture is not.
72 WHAT IS CINEMA?

Moreover, the first of these contrasts helps us explain the second. For it
is insofar as the performance of the moving image is generated merely by
engaging the template mechanically and/or electrically that it is not to be
counted as an artistic performance, properly so called, whereas a perform-
ance generated by an interpretation or a set of interpretations is. These two
features of motion pictures are enough to differentiate performances of
moving images from performances of plays, and, furthermore, the two
differentiae under consideration apply to all films, videos, and the like,
whether the types themselves are artworks or not.
Thus far, we have identified four necessary conditions for the moving
image. Something x is a moving image only if (1) x is a detached display;
(2) x belongs to the class of things from which the promotion of the impression
of movement is possible; (3) performance tokens of x are generated by
templates that are tokens; and (4) performance tokens of x are not artworks
in their own right. Moreover, these criteria supply us with the conceptual
wherewithal to discriminate instances of the moving image from its neighbors
in the worlds of painting and theater.

Two-Dimensionality
Nevertheless, our formula so far is not completely adequate. For example.,
it is vulnerable to at least one sort of counterexample. Consider what
might be called moving sculptures of the sort that are exemplified by
the moving figurines on various antique clocks; as the clock chimes noon, a
door swings open and a toy grenadier figure marches across the top of the
time piece. Similarly, there are the moving sculptures that one meets at the
various Disneylands; such as the mechanical effigies of Abraham Lincoln
intoning the Gettysburg Address like, well, clockwork.
Inasmuch as these sorts of figures are multiples, they meet the condi-
tions stated above for membership in the category of the moving image, as
do the statuettes of ballerinas that once cavorted on the tops of music
boxes. All these examples are detached displays in the same sense that the
locale of Burnham Wood on a theater stage is literally discontinuous with
the space we inhabit. These ballerinas are not dancing in our space, but
some ideal, ethereal space.
Moreover, the aforesaid sculptures, we are assuming, are manufactured
from a template and come in multiples. And the mechanical movement
in the semblance of a pirouette of the ballerina — her performance, if
you will — is not an artwork in its own right. Yet, although she satisfies the
 WHAT IS CINEMA? 73

criteria, clearly, our ballerina does not seem to be the sort of thing that we
have in mind when we speak of motion pictures or the moving image.
In order to exclude cases like this, let us propose a fifth condition for
membership in the regime of the moving image. Let us require additionally
that, in order to be a moving image, x must also be a two-dimensional
array. So x is a moving image only if (5) x is two-dimensional. Perhaps, it
might seem unnecessary to supplement the formula above with this
emendation, since it might be argued that two-dimensionality is already
entailed by the fact that we are talking about motion pictures and moving
images which are by definition two-dimensional. However, even if this is
true when it comes to pictures, it helps to make it explicit that the
pertinent images, when speaking of moving images, are two-dimensional.
Here, it may be fair to ask why two-dimensionality wasn't introduced earlier,
since it might appear that that would have given us a neat way in which to
cleave motion pictures from theater. But appearances can be deceptive, since
there is, in fact, theater that is two-dimensional, for example the shadow-
puppet shows of Bah, Java, and China. In order to count them as theater
rather than moving pictures, we require recourse to the notion that token
performances of motion pictures, in particular, and moving images, more
generally, are generated by means of templates that are themselves tokens.
So, x is a moving image if and only if (1) x is a detached display or a series
thereof; (2) x belongs to the class of things from which the promotion of the
impression of movement is technically possible; (3) performance tokens of x
are generated by templates that are themselves tokens; (4) performance
tokens of x are not artworks in their own right; and (5) x is two-dimensional.
Notice that each of these five conditions is alleged to be necessary and to be
conjointly sufficient.
However, as soon as the claim of sufficiency is issued, the worry may arise
that the preceding formula is too exclusive. Suppose that it became possible to
remake motion pictures, like the HBO television series Rome, in the round by
means of holography. Imagine that we could project a scene of mortal combat
in the Coliseum three-dimensionally with the audience seated around the
virtual arena like ancient Romans. Would not such a spectacle be rightfully
categorized as a moving image? And yet it will not pass muster according
to our definition, since a holographic remake of Rome is not two-dimensional:
for we can measure our legionnaire-gladiators up, down, and around.
There are at least two strategies for dealing with this case. On the one hand,
we may argue that it is an illusion that our holograph is three-dimensional.
It is produced by an ensemble of two-dimensional projections, and if you try
to strike our legionnaire-gladiators, your hand will pass right through them.
74 WHAT IS CINEMA?

But it is not clear that the three-dimensionality of the holograph is an illusory

property rather than an objective, response-dependent property. It is true that,
though our legionnaire-gladiators have girth, they have no bulk. They have no
weight, but they do have height, width, and depth. But since it is not evident
that solidity is really a necessary condition for three-dimensionality, it may not
be an option to claim that a holographic Rome is actually two-dimensional.
On the other hand, perhaps we should bite the bullet on this one. We can
argue that just because the holograph is three-dimensional, it is not a moving
image. Remember that the criterion of two-dimensionality was introduced
precisely in order to distinguish what we call moving images from moving
sculptures. Sculptures are differentiated from pictures exactly in terms of
three-dimensionality versus two-dimensionality. Since the holographic Rome
is three-dimensional, then, it is natural to classify it as a moving sculpture.
Is this merely an ad hoc solution — one that basically invokes a new
category in order to evade an embarrassing theoretical predicament (other-
wise known as a counterexample)? Not really. Many classic sculptures have
moving parts. Many fountains involve sculptural ensembles, for example, in
which moving water stands for moving water — the great rivers of the world,
for instance or, less exaltedly, as can be observed in Brussels, urine issuing
from a young man. Moreover, there are abstract moving sculptures, such as
Alexander Calder's mobiles.
Thus, the category of moving sculpture has not been made up on the
spot. There are such things. Why should not moving holographs be placed
in this category, if they are really three-dimensional? Indeed, for an
example of such a sculpture, which I think it is natural to think of as a
kind of sculpture, albeit light sculpture, see the interactive holograph of
Benjamin Franklin at the Lights of Liberty in Philadelphia, which was
designed for the celebration of Franklin's 300th anniversary.
Someone might object to this proposal on the grounds that sculptures
are not only three-dimensional, but solid, and our holographs are not. But
Robert Morris's sculpture Steam, the very name of which tells you what
you need to know about it, will no more resist your hand swinging through
it than our holographic Rome. And, if you maintain that Steam still has some
particles of substance, whereas our legionnaire-gladiators do not, then
recall many of Dan Flavin's sculptural installations, such as Alternating Pink
and Gold. For these sculptural installations are not only comprised of
fluorescent rods but also the light they throw off, and the way it illuminates
the surrounding walls. If Flavin can create still sculptures with light, then
there should be no objection in principle to moving sculptures concocted
of the same stuff.
 WHAT IS CINEMA? 75

In addition to the anxiety that our characterization of the moving image

is too narrow, there is also the suspicion that it is too broad — that it
includes what it should not. For example, photographically mass-produced
flip books meet all of our criteria for membership in the category of the
moving image. Consider, for instance, the upper right-hand corners of
Arlene Croce's wonderful The Fred Astaire and Ginger Rogers Book. There you
will find still photographs of Astaire and Rogers dancing. If you flick the
pages quickly enough, as you are patently intended to do, you can animate
the dancers. What is in operation here, of course, is the principle behind a
children's flip book. And although the third condition of the preceding
theory of the moving image excludes handmade flip books of the kind we
once made in grammar school, photographically manufactured or other-
wise mechanico-chemically reproduced ones would satisfy that condition;
furthermore, if the page-turning were automated, the token performance
of the animation would clearly not count as art.
Similarly, Muybridge-type photographs of horses, brought to "life" by
the nineteenth-century device known as the zoetrope, fit our formula.
But these too are not the sorts of things that people usually have in mind
when they talk about motion pictures. If I told you that in repayment for a
favor I was going to take you to see a motion picture, and I led you to the
library and handed you a flip book, you would probably feel cheated.
Admittedly, neither a zoetrope nor a flip book is a paradigmatic or
prototypical instance of our concept of the movies. But I see no reason not
to include them in the category of the moving image. Indeed, there are at
least two reasons, apart from our theory, to do so. First, the label moving
image, as understood in ordinary language, suits them perfectly. Whatever
else would we call these things but moving pictures or moving images? The
name naturally fits them. Furthermore, they rely upon the same principles
as do standard film images and, in fact, these phenomena are intimately
related historically to the development of the motion picture film images we
see projected in our local cineplexes. They were, so to speak, the movies
before our prototypical movies. So let us embrace them as instances, albeit
non-paradigmatic ones, of the category of the moving image.

What about a Simpler Theory?

Some philosophers will be unhappy with the resolution, arguing that it

strays too far away from our ordinary concept of motion pictures as we
currently apply it. In order to block the inclusion of flip books, zoetropes,
76 WHAT IS CINEMA?

and the like, they may propose that membership in the category requires
that the candidate be projected onto a screen. One philosopher, Alan
Goldman, for example, defines something as a moving image if and only
if it is (1) an image capable of movement (2) that is mechanically projected
on a screen. Since this conception of the moving image is far simpler than
the one we have been evolving throughout this chapter, we can call it "the
simplified definition." Moreover, the second condition of the simplified
definition — that a moving image be projected onto a screen — ejects flip
books and zoetropes from the realm of the moving image. Shouldn't that
count in favor of the simplified view? However, on the down side, it also
excludes too many other items that surely belong to the category.
The earliest films produced by the Edison Corporation were not
projected onto screens but were viewed as kinetoscopes — that is, not
screened but viewed in boxes into which customers peered one at a time
and which are sometimes referred to as "peep shows." Often these film
shorts were recordings of vaudeville turns, like Annie Oakley's sharp-
shooting. It was not until the Lumiere Brothers popularized film projection
that the screening of films became the order of the day. But these early
Edison films were surely moving images, even though they were never
screened. A definition that does not classify them as moving images is
clearly inadequate. Moreover, these viewing devices remained popular in
amusement parks and in adult pornography emporiums until the advent of
the videocassette. Like the Edison shorts, these films too were undeniably
moving images. Indeed, until quite recently, films were usually cut on
machines like movieolas and Steembeck editing tables. That is, editors viewed
the raw footage in the same way in which the original audiences viewed the
products of the Edison company. But who would deny that these editors
were watching moving images? Isn't that what they bequeathed to (movie)
theater viewers?
In response to these objections, the proponent of the simplified definition
might counter that his definition of the moving image is meant to capture
current usage and, therefore, is not threatened by counterexamples from the
past. I am not sure that this is a compelling response, since current users will
still have to classify things like the Edison productions, and I suspect they
will categorize them as moving images, or, at least, as motion pictures. But,
in any event, even if it were permissible somehow to bracket all these
historical counterexamples, there is a remaining conundrum for the simpli-
fied definition, since there are quite a large number of ostensibly legitimate
candidates of contemporary vintage for the category of the moving image
which the simplified definition simply can't accommodate.
 WHAT IS CINEMA? 77

Consider, for example, videocassettes and DVDs. These are not

projected onto a screen but, rather, are beamed directly at viewers. By
the way, the same is true of everything on television. Aren't all these things
moving images? If— since they are not projected — the simplified definition
rejects them, then at least two difficulties arise. First, there is the obvious
problem that pre-theoretically one is disposed to think that virtually
everything on TV should count as belonging to the category of the moving
image. But second, denying this leads to inconsistency. If I see There's
Something About Mary in the movie theater, the simplified definition assigns
it to the class of moving images. Yet if I watch it on broadcast TV or a
videocassette or a DVD, it is not. Thus, the simplified definition invites
paradox and should be discounted for that very reason.
At this point in the debate, the defender of the simplified definition is
apt to object that televisions have screens, and, consequently, none of the
alleged counterexamples concocted in the preceding paragraph really hit
their mark. But this is not true. We look directly into TV images; they are
not reflected off a screen. Standardly, scholars distinguish between looking
at TV and looking at film in this way. Why would the friend of the
simplified view think otherwise?
Perhaps the proponent of the simplified view has in mind what most
people refer to as the "TV screen" — the thing that we clean with Windex?
But this is somewhat of a misnomer, since the image is projected through
this piece of glass and not onto it. Moreover, this glass is not a necessary
feature of the TV delivery mechanism. You can remove this glass and still
receive a moving image. If you don't believe me, try it.
Alternatively, by screen, the defender of the simplified view may be
thinking of the phosphor screen in the cathode ray tube. This lies between
the shadow mask and the front glass of the picture tube. It is coated with a
phosphor strip that glows when different colored beams of light hit it. But
the phosphor screen is part of the picture tube — which is to say that it is
part of the projector mechanism. It is not a screen onto which the image
is projected. It is a device for generating the image. It is part of what
projects the image — it is part of the picture tube — and not something,
such as a screen, that receives the image. For typically there is not screen
onto which the picture tube throws the image.
This is not to deny that televisions cannot be outfitted with mechanisms
to project their images onto bare walls and screens. But usually they are
not so equipped. The possibility of projection onto a screen is not a
necessary feature of televisions. In television — whether the image comes
to us via a videocassette, a DVD, or a broadcast — we look directly into
78 WHAT IS CINEMA?

the projector mechanism rather than at a screen onto which something is

projected. A screen is not a necessary condition for a televisual image. Thus,
inasmuch as televisual images are paradigmatic instances of the moving
image, the simplified definition does not even suit contemporary usage of
the concept.
Of course, the defender of the simplified view may claim that all she
means by a screen is that upon which you see something else or through
which you see something else. But this is clearly too broad. For it would
fail to exclude certain theatrical inventions from the order of the moving
image. There are, of course, the shadow-puppet plays mentioned earlier.
And, in addition, if a play-actors stage a scene or an entire play behind a
back-lit scrim, that will fall under the simplified definition of the moving
image, though this seems to be a deeply counterintuitive result.
Philosophers in general prefer simpler theories over more elaborate
ones. That is the attraction of the simplified theory. However, accuracy
and comprehensiveness are even more important than simplicity. The
simplified definition is n o t as accurate as its m o r e complex competitor.
It errs in excluding too many phenomena — from Edison shorts to DVDs
of The Incredibles — while also including too much — for example, shadow-
puppet plays. Consequently, unless a simplified theory can be crafted to
avoid these shortcomings, for the duration we will regard something as
an instance of the moving image if and only if (1) it is a detached display
or a series thereof; (2) it belongs to the class of things from which
the production of the impression of movement is technically possible;
(3) performance tokens of it are generated by templates which are tokens;
(4) performance tokens of it are not artworks in their own right; and (5) it
is a two-dimensional array.
To answer the question with which we began — what is cinema? — we can
now say that cinema is, at least, necessarily, the practice of making moving
images. Admittedly, that's not, as Descartes might say, the whole ball of
wax. But isn't it a pretty helpful start?

Suggested Reading

The quest to define cinema has been a perennial one. All of the film theorists cited at
the end of the last chapter have their own version of an answer to the question "What
is cinema?" See those references as a guide to reading your way through the maze.
An approximation of the theory offered in this chapter is also available in
Noel Carroll, "Defining the Moving Image," in his Theorizing the Moving Image
 WHAT IS CINEMA? 79

(Cambridge: Cambridge University Press, 1996), though the present version is

bolder and better defended than the earlier one.
The idea of detached displays or disembodied arrays benefits greatly from Francis E.
Sparshott's speculations, though establishing this condition as a necessary one for
cinema is not the primary purpose of Sparshott's article. See Sparshott's "Vision
and Dream in the Cinema," Philosophic Exchange 1 (Summer 1971), 111—22.
Our emphasis upon moving pictures here is very deeply indebted to Arthur
Danto's magnificent article, "Moving Pictures," Quarterly Review of Film Studies 4:1
(Winter 1979), 1-21.
The discussion of tokens, types, and templates with regard to art begins with
Richard Wollheim's Art and its Objects (Cambridge: Cambridge University Press,
1980), especially sections 35—8. In this chapter, we are working out of Wollheim's
framework.
Walter Benjamin's contrast between film acting and stage acting can be found in
his "The Work of Art in the Age of Mechanical Reproduction," in Illuminations,
edited by Hannah Arendt and translated by Harry Zorn (New York: Schoeken
Books, 1968).
The simplified definition of motion pictures can be found in Alan Goldman,
"Specificity, Popularity, and Engagement in the Moving Image," Film and Philosophy
5:6 (2002), 93-9.
Chapter Four
The Moving
Picture — the Shot

Movies are typically comprised of a series of moving pictures, called

shots, which themselves are animated as a succession of still pictures,
called frames, pass before the projector beam. A shot is characteristi-
cally an uninterrupted camera take — one with no spatial or temporal
discontinuities that are detectable by the human eye.
These shots can be very near to their subjects — as in a close-up of Gong
Li's face or of the barrel of a gun — or they can be further away, for
example: a medium shot of a man, from the knees up, waiting for a bus;
or a long shot of an ocean liner steaming into port; or an overhead shot
of chorines distributing their limbs decoratively into kaleidoscopic figures
in a musical number by Busby Berkeley. Shots may also involve moving
points of view as the camera tracks toward or zooms in on its subject or
away from it, or travels next to it, as in an exciting car chase.
These shots are, so to speak, the building blocks of traditional cinema.
The question to be pursued in this chapter is how best to understand
these moving picture shots, ranging from close shots of Tony Soprano to
long shots of the spaceship Enterprise, (Note that in this chapter our concern
is with moving pictures and not with, more broadly, moving images. That is,
we will not attempt to analyze cinematic abstractions here.)
Motion picture shots are representations — pictorial representations of
houses, horses, and Hondas. But how it is it that they succeed in conveying
their content to audiences? There seem to be at least these four leading
alternatives: (1) moving picture shots are illusions of that of which they
are pictures (the illusion thesis); (2) shots are transparencies, providing us
with indirect perceptual access to whatever gave rise to them (the trans-
parency thesis); (3) shots are coded symbols, somewhat like words (the
 THE MOVING PICTURE 81

code thesis); or (4) shots are recognitional prompts, naturally generative

symbols that engage the subpersonal perceptual processes that we employ
daily for object recognition in such a way that a motion picture shot of, say, a
dog will elicit our apprehension of a dog in the picture without any subtending
operations of decoding, reading, or inferring (the recognition thesis).
Let us proceed through these options one by one. However, for the
sake of full disclosure, I should say out front that my money is on the
recognition thesis.

The Illusion Thesis

The first alternative — that the motion picture shot is an illusion — is
probably the best known one; I suspect that most readers have at least
heard this idea before. It involves the contention that the image on screen
is so convincing that one is seduced or deluded into the illusion that the
objects and events that gave rise to it are actually before us. Sometimes
this or that cinematic process, like Cinerama or Imax, is promoted by
declaring that it delivers the illusion of reality. Or a critic may recommend
a motion picture by saying that it makes you feel that you are really there —
there in the world of the movie.
The notion that the moving picture shot is an illusion appeared early on
in the history of cinema. One of the earliest films screened by the Lumiere
Brothers showed a train arriving at a station. Allegedly, as the image on
screen chugged toward the camera recording it, viewers scrambled to
get out of the path of the engine. Supposedly they felt in mortal danger.
Whether this event ever really happened or is merely the hyperbole of
some publicist, we may never know for sure (though I frankly admit that
I am very skeptical about the likelihood that this report is accurate).
Nevertheless, the mantra that a moving picture is an illusion of reality is
often repeated; it was the prevailing article of faith of film theorists in the
1970s and 1980s.
What we may call the illusion thesis can be framed in at least two ways:
(1) that the moving image is a cognitive or epistemic illusion; or (2) that it
is a perceptual illusion. To say that a moving picture shot is a cognitive or
epistemic illusion maintains that it involves the provocation of a belief
on the part of the viewer. Specifically, it involves engendering a false
belief in the viewer — namely, that one is literally in the presence of the
object or event pictured on screen.
82 THE MOVING PICTURE

Imagine that the image on screen is Big Jake. According to the cognitive
or epistemic version of the illusion thesis, as Big Jake saunters toward the
camera, I believe that Big Jake is actually sauntering toward me. If Big Jake
reaches for his Winchester and points it in my direction, then I fear for my
life. For, putatively, I believe that Big Jake is in the same space that I am in
and that his bullets can cut me to the quick.
The perceptual version of the illusion thesis can be illustrated by that
perennial philosophical example — the straight stick that appears bent in
water. As you know, if you take a straight stick and dip it into a running
stream, it will appear bent to you. That is, the stick will look bent to you,
even though you realize that it is straight. This is what we call an optical
illusion. Of course, since you know that the stick is straight, you believe
it is straight, despite its bent appearance. You are not deceived by the look
of the stick. Unlike a case of cognitive or epistemic illusion, you do not
form a false belief about the stick. Nevertheless, it does look perceptually
as though it were bent.
Another famous example of a perceptual illusion is the Miller-Lyre
Illusion. This involves two lines of equal length. But on the tips of the
two ends of the bottom line there are what look like arrowheads pointing
away from the line, while on the top line the arrowheads point into the
line. The effect of this arrangement is that the top line appears longer than
the lower line, though they are both precisely the same length. To prove
this you may take a ruler to the two lines. This will establish that they
are equal, but still one cannot shake the impression that they look to be of
different lengths.
Even though you come to believe that the two lines are equal, that belief
will not change the way in which the lines appear to you. The perception of the
inequality of the lines is what philosophers call cognitively impenetrable — i.e.,
knowing (and believing) that the lines are equal will not change the fact that
they look unequal to you. The perceptual illusion here is impervious to our
beliefs. We know cognitively that the lines are the same, but we cannot but see
the top line as longer than the bottom line.
Cognitive illusions involve beliefs, indeed false beliefs. We are deceived
by cognitive or epistemic illusions. Perceptual illusions need not involve
being deceived at all. We may know (and, therefore, believe) that a stick
in water is straight and yet it still seems stubbornly bent to us. What we
know and believe does not change the perceptual impression the stimulus
provokes in us. Believing otherwise — for example, that the two lines in the
Miller—Lyre example are equal — has no impact on the fact that to our eyes
they remain obdurately unequal.
 THE MOVING PICTURE 83

The cognitive or epistemic version of the illusion thesis is not very

promising. On this view, moving picture shots succeed in conveying
whatever they are pictures of by instilling the illusion that whatever they
represent on screen is literally present to us. We are, for example, while
watching Finding Neverland, somehow convinced, deceptively, of course,
that we are looking on as James Barrie — in the flesh, so to say — composes
Peter Pan. Sometimes this phenomenon is described as a suspension of
disbelief. But this is simply a roundabout way of saying that, since we do
not disbelieve Barrie is in front of us, we believe that he is. Yet this
hypothesis strains credulity.
Among other things, beliefs are mental states that we posit in order
to explain people's behavior. Hortense announces that she is thirsty and
she reaches for the refrigerator door. We explain her action of reaching
for the door by postulating that Hortense believes that there is thirst-
quenching liquid in the refrigerator. Herbert reads his assignment; we
explain this by indicating that Herbert believes by reading his assignment,
he will pass the test tomorrow. Beliefs are connected to behavior, since
beliefs make behaviors happen.
But consider the ramification of this for the cognitive or epistemic
version of the illusion theory. Putatively, we believe that the objects and
events in the motion picture are present to us. But what if the film is King
Kong? If we truly believed that a fifty-foot tall, very angry gorilla was within
arm's length of us, would we sit by idly and wait for him to wrench us
from our seats and do — God knows what — to us? No. If we believed Kong
was within stomping distance, we'd start running away from him as quickly
as possible. Or imagine that we are watching The Wild Bunch. Bullets are
flying everywhere. But we are casually munching our cheese nachos. Is that
how people behave who believe that they are in the middle of a fire fight?
Or, to consider less violent prospects, let us suppose that we are
watching two young lovers in the process of removing their clothes and
fondling each other. In real life (as opposed to reel life), if we came upon
such a situation and believed we were inadvertent witnesses to a tryst, I
suspect most of us would turn away, perhaps in embarrassment. But we
look on shamelessly at scenes like this on film. Consequently, how likely
is it that we believe that we are actually in the presence of this couple?
Not very. For how else would one explain our very different behavior
when we do not believe the lovers are only on screen?
These behavioral anomalies imply the overwhelming improbability of
the conjecture that we are cognitively deceived into believing we are
actually involved in a close encounter of the third kind with the folks
84 THE MOVING PICTURE

pictured in movie shots. Our behavior in the movie theater or, for that
matter, in our living rooms, is not the sort we would expect from
people who believe that King Kong is in the neighborhood, or that they
are pinned down by gunfire, or that they have just blundered into a love
nest. Given the connection between beliefs and behavior, our behaviors at
the movies indicate that we are not in the thrall of any cognitive or
epistemic illusion.
Apart from the aforesaid kinds of behavioral anomalies, there are
further aspects of the movie-viewing experience that make it positively
unlikely that we are deluded into believing that what is pictured on screen
is "in our face." Often we have to see through various surface interfer-
ences in order to attend to a movie — we must look past the scratches, the
dirt on the filmstrip, and the hair in the projector gate, or we must see past
the overexposure of the footage in order to secure visual access to the
story. But if we are knowingly "seeing through" the image, we cannot be
under the misapprehension that what is pictured is somehow really a living
presence that is close up and personal, so to speak. For such a mental state
would be self-contradictory.
Likewise, sometimes the screen is soiled. Perhaps an unhappy patron
threw a glass of Coke at it eons ago and the negligent theater owner never got
around to cleaning it off. That splotch is also something that we need to see
through. But how can we see through this surface interference and believe in
the reality of what is pictured at the same time? That is, movie viewing
appears to involve an awareness of the screen, if we know to "see through"
these various surface incongruities. But, at the same time, the cognitive
illusion thesis has us disbelieving in the screen and believing in the real
presence of what is shown. However, this supposes us to be in a contradict-
ory mental state — simultaneously believing and not believing in the presence
of the screen — a mental state that borders dubiously upon psychosis.
Moreover, often when appreciating a motion picture we say "How
very lifelike." But we would hardly call a movie lifeMe, if we thought it
was, as the cognitive illusion thesis requires, real life. Sitting in our office,
suffering through our daily grind, we never say "How lifelike." We're
more apt to say "That's life."
Similarly, we wouldn't call a movie realistic if we thought it was real.
When caught in a traffic jam, we don't say "How realistic." If anything we
say "I can't believe this is real (or really happening)." Thus, it seems that
certain of the ways in which we articulate our appreciation of cinema entail
that we cannot hold the beliefs that the epistemic version of the illusion
thesis purports we do.
 THE MOVING PICTURE 85

Furthermore, if we were seduced by these alleged cognitive illusions,

we would be hard put to render coherently what we experience under
their sway with the rest of our experience. How do I explain to myself
when watching Gandhi how my home entertainment center in the base-
ment of my condo in Philadelphia now somehow abuts onto the Ganges?
When a moving picture shot is taken from the ceiling, am I deluded
into believing that I am hanging from the rafters? How did I get there?
How do I get down? Or recall my watching The Wild Bunch. How is it that
I emerged from that welter of whizzing bullets unscathed?
Or, even more unintelligibly, suppose I am watching the episode of Six
Feet Under in which it is true in the fictional world that he inhabits that
Nate buries his wife Lisa in the desert unseen by anyone. Yet, if 1 am caught
in the undertow of a cognitive illusion, don't I believe that I am a witness
to the event? So how can I also believe that no one saw the burial? Again,
the cognitive illusion theory would seem to have the extremely counter-
intuitive consequence of leading normal viewers into webs of belief so
contradictory that said viewers would appear to be virtually psychotic.
The cognitive or epistemic variant of the illusion thesis is fraught with
problems. But perhaps the perceptual version of illusionism is not so
troubled. According to the perceptual illusion thesis, we are not necessar-
ily deceived by what we see on screen. Nevertheless, we find what we
see irresistibly, perceptually compelling. We do not believe Big Jake is in
front of us; it only seems to be or looks that way. But is it really true that
the objects and events pictured in motion pictures strike us perceptually
as really being present? If the movie is black and white, how is this possible?
People in broad daylight don't look black and white to us. Moreover, the
film image is odorless. When we perceive the world, we not only see and
hear objects, we also smell them.
Perceiving motion picture images typically diverges in a large number
of ways from normal perception; it hardly strikes us as though we are
perceiving these cinematic images in exactly the same way we would perceive
their referents outside the cinema. It is almost always possible to perceive
the difference between a motion picture image of a person and the living,
breathing being.
If Big Jake were standing in front of me, and I changed my position
relative to him, different aspects of Big Jake would come into view. For
example, if I moved to my right, more of his left side would occupy my
field of vision than was visible previously, just as more of his right side
would become proportionately occluded. Psychologists call these "edge
phenomena"; that is, I start to see around the edges of things when I alter
86 THE MOVING PICTURE

my perspective. But there are no edge phenomena with respect to a static

motion picture shot. The shot is taken from a single point of view or
monocular station point. If I move from the right aisle of the movie theater
to the opposite aisle, there will be no change in what I see of Big Jake. I will
not start to see around his other side. In other words, I will not perceive
what I would perceive were I moving relative to a real person who was
actually present to me. The perceptual experience will be radically differ-
ent; it will not feel, even if I am not able to put my finger on why precisely,
as though I am seeing someone in the flesh.
Additionally, the muscles in our eyes register a phenomenological
difference between the flickering image on the screen as an object that
inhabits the same physical space in which we are situated, versus the virtual
space portrayed in the picture. So, once again, the film image does not
seem — does not even feel — literally co-present to me. My body, so to speak,
rejects this impression.
Furthermore, many of the objections brought against the cognitive
or epistemic version of illusionism can be recycled against perceptual
illusionism. I must see through surface anomahes on the screen; but
there are no surface anomalies like that in reality. Reality doesn't look as
though it has been scratched. And what of moving picture shots that
employ a distorting lens? People's anatomies do not come squashed or
elongated as they frequently do in film; nor is soft focus a recurrent feature
of our perceptual world.
Those early film theorists, like Rudolf Arnheim, were right about
the myriad ways in which film departs from normal perceptual experience.
They were also right that filmmakers often use these deviations as oppor-
tunities for expression which they justifiably expected viewers to notice or,
at least, to sense just because the relevant film images diverge from the way
in which things normally look. But if this sort of appreciative response to
film is plausible — and it gives every indication that it is — then perceptual
illusionism cannot be true. Film shots do not strike us perceptually as
though the objects they picture are literally in front of us. Moving picture
shots are not perceptual illusions of what they present. They just don't
look like the real thing.
Even if at some point in the future it becomes technologically feasible to
produce motion picture shots that look just like what they are pictures of,
this possibility would not entail that that is how we have typically experi-
enced moving picture shots up until now. For the vast majority of moving
picture shots do not look like their referents. They have not fooled and will
not fool the eye. Even if, one day, there will be shots of Big Jake that look
 THE MOVING PICTURE 87

as though the bruiser is in front of us, we have not met those shots so far.
Consequently, what perceptual illusionism offers by way of understanding
the operation of moving picture shots fails to characterize shots as we now
know them.
Nevertheless, the defender of perceptual illusionism may argue that
these objections do not show that there is absolutely no place for the
notion of illusion in our account of the moving picture shot. Even if Big
Jake does not strike us perceptually as being in front of us in the flesh,
there still is an element of the image that is implacably illusory, namely:
that something is moving on screen. For, it may be argued, nothing is really
moving on screen in the movie theater. That is, the appearance of
movement is an illusion — specifically a perceptual illusion.
The impression of movement in film is standardly generated by a
succession of still pictures projected, nowadays, typically at the rate of
twenty-four per second. These still pictures are frames, and they are said
not to move. This is not exactly right, since they do move through the gate
of the projector. Of course, we do not perceive these frames or their
movement. The projector puts projected images of these frames on the
screen for us to look at. These images move too — they move on and off
the screen — but that is not something our perceptual apparatus registers.
Instead, we perceive shots, not frames, or rather projected shots.
Moreover, these projected shots don't move, in the sense that they don't
jump off the screen and fly around the movie theater. Rather, parts of the
projected shots — projected people, vehicles, animals, etc. — give the
impression of moving. Furthermore, the friend of perceptual illusionism
maintains> those impressions are illusions, since what is actually on screen
are still images.
That is, what is really on screen is a succession of still pictures; but we
cannot help but see movement up there. Movement in film is a perceptual
illusion. Therefore, one cannot discount the illusion thesis entirely when it
comes to understanding how the moving picture shot operates.
Parts of the shot — a car, for example — appear to move. Comparable
parts of the frame don't move, or, if they do, because they wobble a bit,
this is not usually detectable by the human eye. So, the parts of the frame
are still, whereas the corresponding parts of projected shots give the
impression of movement, that is, in shots where there is movement (as
opposed to freeze frames). Perceptual illusionism says that this impression
of movement or apparent motion is an illusion.
I think that, at first blush, perceptual illusionism with respect to motion
picture movement appeals to common sense. But might not there be some
88 THE MOVING PICTURE

way in which the movement is not illusory? Or, in other words, might
there be a way in which we might be prepared to say that the movement
within the shot is real, or, at least, objective?
The argument that so readily convinces common sense that motion picture
movement is illusory rests upon setting up a forced choice: either the
movement in the motion picture shot is metaphysically real or it is illusory.
Here, for something to be metaphysically real means: the way in which things
stand from the view from nowhere or, if that notion is oxymoronic, then how
ultimate reality appears to an omniscient being, like God, who is not limited
to what can be detected by human sensory apparatus. This being sees tables
and chairs in their ultimate physical reality as swarms of sub-atomic particles;
this being sees film frames click by one at a time. Moreover, if something
is not metaphysically real in this sense, then it is illusory.
But are these the only alternatives that we should consider? Insofar as
common sense starts with this disjunction, common sense sees no other
option but to consign movement within the shot to the rank of an illusion.
Yet might we not think differently if there were more options at our
disposal? Needless to say, that will depend upon what those alternatives are
and how compelling they appear.
One alternative, not canvassed so far, is that the motion is real in the
sense that it is real, response-dependent motion. That is, the movement
within the shot is an objective appearance. A standard motion picture shot
will yield the impression of movement to creatures constituted with
perceptual systems like ours. Creatures with our perceptual capabilities,
including not only other humans, but the members of other species as well,
will unfailingly perceive motion on screen when the pertinent frames are
projected at certain speeds. This will not be a subjective matter in the
sense that some people will see the motion and others won't, depending
upon their taste. The design of the motion picture apparatus engages the
"design" of our perceptual apparatus in a way that the detection of
movement on screen obtains consistently and regularly for normally
sighted, human viewers. That is why we say that the appearance of motion
is objective in the sense of being inter-subjective; it is uniform across sighted,
human percipients. Therefore, it seems very misleading to call it an
illusion, since everyone will see it that way. Indeed, if a sighted human
being reported sincerely that he did not see movement when exposed to
standard motion picture projection, we might suspect him of being the
victim of some sort of perceptual malfunction.
Nor is this convergence only observable among humans. There is a 1994
videotape entitled Kitty Safari which was produced in order to make
 THE MOVING PICTURE 89

household cats purr with delight as they watch assorted birds, squirrels,
rats, mice, and guppies cavort before the camera. Humans are not the only
animals that see movement in motion pictures; felines do as well. In fact,
in one installment of Americas Funniest Videos, a cat perched on top of a
television attempted to snatch the racing automobiles zooming toward the
camera as they seemed to converge on lower left-hand corner of the
screen. Cats, like humans, perceive movement in motion picture arrays.
Furthermore, the motion effect imparted from movies is so uniform that
were someone to deny that they saw movement on screen as the guns of
Navarone blow skyward, we would suspect that that person were suffering
some sort of delusion — an illusion or hallucination of stillness perhaps.
If it does not make sense to say that what everyone sees is illusory then,
equally, it admittedly does not seem right to say that the parts of the shot
are literally moving from the perspective of ultimate reality. The move-
ment is not metaphysically (or even physically) real. But it may be real —
that is to say objective — nevertheless; the movement may be a real,
response-dependent property of the shot. But what does that mean?
Color is a real, response-dependent property. A property is a real,
response-dependent one if it can be detected by normal viewers in
appropriately conditioned circumstances. Color is like this. We say that
the American post box really is blue, in the response-dependent sense, if it
looks blue to normal viewers in standard lighting; it is not green in
standard lighting, nor would we say it was green if we suffused it in a
yellow light or if the viewers were jaundiced.
Arguably, cinematic motion is real in a comparable, response-dependent
way. That the post box is blue is not an illusion. Blueness is a real,
response-dependent property of the mailbox. Analogously, cinematic
movement is real — real, response-dependent movement — and not illusory
movement. Just as we detect the real, response-dependent property of
blueness in the post box, so the movement we detect on screen is real,
albeit in a response-dependent way.
Perhaps this position may gain support from considerations like this: if,
while watching To Kill a Mockingbird, one spectator asks another spectator
"Did Atticus Finch just cross the courtroom?" the pertinent answer will be
"Yes" rather than "There was no movement, just a succession of projected,
still frames"; just as when we are asked "Is the fire engine red?", the
answer is "Yes" rather than "No, there are just colorless photons over
there." Color and cinematic movement are both objective appearances, i.e.,
real, response-dependent properties that are inter-subjectively verifiable,
though only detectable via certain perceptual systems.
90 THE MOVING PICTURE

It is not an illusion that the desk upon which I now write is solid. It is
true that at the subatomic level there are gaps between the particles that
comprise the desk. That may dispose us to say that the desk is not solid —
gapless — in the sense of ultimate metaphysical reality. But the solidity of
the desk is not illusory. Anyone who thought they could pass their hand
through it would be counted as a madman. Rather, the desk appears solid,
objectively, for those of us possessed of certain perceptual apparatus.
Similarly, we say that the epidermis of an ant is solid, though under a
powerful microscope we would see that it is perforated with numerous
spiracles through which it takes in air. It is not illusion to say that the ant's
back is solid; it is an objective property of the ant's epidermis detectable by
creatures with perceptual apparatuses like ours. So, by analogous argu-
mentation, cinematic movement is not an illusion. It is an objective
appearance, inter-subjectively available to anyone constituted perceptually
in the way we are. Cinematic movement is a real, response-dependent
property of the shot.
That is, our alternatives in this discussion are not limited conceptually to
a choice between cinematic movement being metaphysically real and its
being illusory. There is also the option that is an objective appearance —
real, that is, in a response-dependent way. And if cinematic movement
is real in this way, then it is not an illusion.
Nevertheless, this argument against relegating cinematic movement to
the status of an illusion may strike some readers as proving far too much.
They may fear that this notion of response-dependent reality is far too
liberal. It seems to threaten to banish entirely the very category of
perceptual illusion. Will we say of the Miller-Lyre Illusion, for example,
that it is not an illusion, because the apparently greater length of the top
line is objective for creatures with our perceptual capacities — that its
apparently greater length is a real, response-dependent property of the
line? Gr that the bentness of the stick in the stream is a real, response-
dependent property of the stick? Or, for a fresh example, that waves have
the real, response-dependent property of coming toward us, though
speaking in terms of what is happening physically, water molecules are
merely being displaced sequentially?
The notion of real, response-dependent properties, it may be argued,
will make all these illusions disappear. But we cannot dispense with the
category of perceptual illusions, since there are such things under the sun.
The invocation of real, response-dependent properties, however, does
not evaporate the category of perceptual illusions. An important distinc-
tion between the solidity of the epidermis of the ant and the bentness of the
 THE MOVING PICTURE 91

stick in water is that we can discover the illusory nature of the latter by using
nothing more than our normal perceptual capacities and ordinary or every-
day or conventional procedures, ones readily within our reach without
resort to special equipment or tests. That is, we can lift the stick out of
the water and, using our ordinary powers of sight, confirm that it is straight;
or we can bend down and feel that it is not bent while it is still below the
waterline. In other words, we can locate the illusion within the realm of our
own senses. On the other hand, in order to see the air holes on the back of an
ant, we need special equipment — a high-powered microscope, for example.
So if we can detect the pertinent, deviant phenomena by employing
our ordinary perceptual capabilities and conventional means, then it is a
perceptual illusion. To determine that the lines are equal in the Miller-Lyre
configuration, we need only cover the arrowheads with our hands, or apply
some ready-to-hand measuring device, like a piece of string, to the lines.
We can establish that the wave is not a single block of water rushing at us
by floating something in the water and then noting that it stays bobbing
in place when the wave moves past it. Thus, these two phenomena are
perceptual illusions, as is the bent stick in water. But establishing that the
desk is composed of atoms and the void, or that the redness of the fire engine
is an affair of colorless photons requires far more extraordinary means.
Given this way of marking off perceptual illusions from realj response-
dependent properties, in which category does cinematic motion belong? If
it is a perceptual illusion, we should be able to discover this using our normal
sensory capabilities and ordinary, household paraphernalia. Can we?
In order to assert that the cinematic movement is an illusion, it will not
suffice to take the celluloid filmstrip in your hand and to notice that it is
composed 6f static frames. For it is not the film unengaged by the
projector that imparts the impression of movement; you have to be able
to detect the static frames on the screen while the filmstrip is moving. How
would you do that, and would it involve extraordinary means or not?
It may be suggested that the way to establish that there is no movement
on the screen is to slow the projector down to a speed where the spectator
sees the projected frames one at a time in all their stillness. But this is not a
conventional way of viewing films. Moreover, it turns out that it is not, as
a matter of fact, a readily available way. Most projectors do not have the
capacity to slow down the projection rate to the point where the spectator
can see one frame at a time. In all probability, almost all, if not all, of the
extant movie projectors in theaters today lack this capacity. Moreover,
the projectors that have the capability to run alternately at silent speed and
sound speed do not run slowly enough to extinguish the impression of
92 THE MOVING PICTURE

movement, but only to change the apparent velocity of the moving objects
on screen.
Of course, certain projectors have been designed that can exhibit the
filmstrip frame by frame — for example, analytic projectors like the Athena
of yesteryear. But is resorting to such devices a conventional way of
checking the projected film for actual movement? Rather, such projectors
seem to be like microscopes — they are ways of going beyond what can
be discovered by the naked eye. They are expressly produced to pierce
through response-dependent appearances. We would not say that the
epidermis of the ant is not solid just because if we put it under a suitably
adjusted microscope, we would see air holes. There is a perfectly respect-
able way in which it is appropriate to say that the ant's body is solid.
Shouldn't we adopt the same stance with respect to cinematic movement?
Moreover, the same kind of argument might be made if it was suggested
that we examine the film on a movieola or and editing table, like a
Steembeck; these are specialized instruments, like microscopes, and not
the sorts of things to which most viewers have access.
So one reason to deny that the cinematic movement in the film shot is
illusory is that it is an objective appearance; the motion, that is, is a real,
response-dependent property of the shot. But another reason to deny the
illusion thesis is that calling filmic movement illusory invites paradox.
Why? Well suppose, pace the preceding discussion, that we could slow
ordinary film projectors down to the point where one could see that the
cinematic movement was caused by the passage of still frames through the
gate. Imagine we slowed down the projection of Nashville in just this way.
We could then say that the movement in the film Nashville is illusory. But
now suppose we transfer Nashville to a videocassette or a DVD. These
systems do not engender the impression of movement frame by frame, nor
can the manner in which they impart motion be decomposed in a way that
we can grasp using the naked eye. Consequently, we will be forced to the
patently contradictory conclusion that the impression of motion in the film
Nashville is illusory, but the perceived movement in the videocassette
version and the DVD is not.
Also, there is another impression caused by film projection that may be
pertinent here. Call it the impression of the continuous image. As each
frame is pulled away from the projector lamp and replaced by another,
there is an interval of darkness. But on screen we see the shot as a
continuous picture, though for every second of screening time there are
so many blank screens. Moreover, even if there is some flickering, this
impression will not be extinguished by slowing the projector or even
 THE MOVING PICTURE 93

moving frame by frame. Given the inextinguishabihty of the impression

of the continuous image in the shot, the proponent of perceptual illusion-
ism will have to accept the impression of continuity in the shot as a real,
response-dependent property of the shot. But doesn't consistency then
demand that we have the same position with respect to the appearance
of movement in the shot that we have with respect to its appearance of
continuity?
Undoubtedly one of the reasons that the defender of the perceptual
illusion of movement feels compelled to deny that cinematic motion is real
is that he presupposes that ordinary real movement from A to B involves
passing through every point between them. But as Big Jake makes his way
across the screen, he does not move through every point on the streets of
Laredo; there are gaps. So the movement is not ordinary movement, but
an illusion.
However, is it really the case that ordinary movement must be continu-
ous in this way? Recall the phenomenon of shadows. Positioned appropri-
ately with regard to a light source, we are happy to say that my shadow
moves when I move from one side of the courtyard to the other. Now
imagine that I pass behind a tree for a moment. My shadow will moment-
arily cease, only to reappear as I clear the tree. Yet we will regard the
movement of my shadow as continuous. We will not say there are two
shadows here, each with its own discrete spate of movement. Even though
there is a gap in the space the shadow traverses, we do not hesitate to
regard the movement as ordinary. Furthermore, when we remember that
the film image is in many ways an affair of shadows, we should be ready to
treat it on a par with the way we regard the ordinary movement of shadows
which, among other things, tolerates gaps. If shadow movement is not
necessarily illusory, neither need be cinematic motion.

The Transparency Thesis

We have already encountered the transparency thesis in the first chapter of
this book. It is the view that the cinematic image is transparent. Putatively
we see through the image to the object of which it is an image. In this
regard, the cinematic image is analogous to telescopes, microscopes,
periscopes, and to those parking-lot mirrors that enable you to see around
corners. When we look through devices like these, we say that we see the
objects -to which these instruments give us access. We see distant planets
through telescopes; molecules through microscopes; oncoming traffic
94 THE MOVING PICTURE

through parking-lot mirrors; and enemy vessels through periscopes. These

devices augment our visual powers. They are, in this respect, prosthetic
devices. They enable us literally to see things we could not see otherwise.
Moreover, we do not regard these prosthetic devices as delivering repre-
sentations of the things they show us; we say that we see the very things
themselves when we look into these instruments.
When I am sitting high up in the balcony and I look through my opera
glasses at the soprano, I see the soprano, not a representation of the
soprano. Devices such as opera glasses, and the instruments mentioned
above, enhance the range of my vision. They do not produce representa-
tions of what they reveal, but put us in visual contact with the real thing.
These devices are not, in principle, different from the corrective lenses that
we wear to rectify near-sightedness or far sightedness. They are prostheses.
They help us to overcome visual shortcomings and to see what is otherwise
visually unavailable.
But if we are willing to speak this way about microscopes and telescopes
as prosthetic devices, why not extend this idea to photography and
cinematography? Cameras also put us in visual contact with things that
are remote — temporally remote that is — things from the past, A snapshot
enables me to see my childhood sweetheart; the filmstrip grants me visual
access to the performance of a silent film star who died before I was born.
The telescope puts us in visual contact with a distant galaxy; a camera does
the same thing with respect to the distant past.
The claim that we can see into the past may seem like voodoo to you.
But it shouldn't. Since the light from far off galaxies takes so long to reach
us, it can be said that some of the stars that we see in the here and now had
actually collapsed into black holes centuries ago. As we watch the glow
stream out of them, we are quite literally seeing into the past. If we can see
into the past with a telescope, there should be no impediment to saying
that cameras enable us to do likewise.
The argument for the transparency thesis exploits a slippery slope. If a
periscope enables us to see over the wall into the adjacent courtyard, why
not say that a video set-up does as well? Here one feels like saying "But
we don't actually see the courtyard on the video monitor." But what do we
mean by "actually see"?
One thing that it means is that our perception is counterfactually
dependent upon the visible properties of the objects of our perception — i.e.y
had the visible properties of those objects been different, then
our perceptions would have been different accordingly. There is a causal
chain of physical events between the objects of our perception and our
 THE MOVING PICTURE 95

perceptions such that if the starting point in that network were otherwise,
then our perceptions would vary in concert. For instance, I see the redness
of the tomato because the tomato is red, but had the tomato been a green
one, what I would see would be green. And had the object been a tuba,
rather than a tomato, then, all things remaining constant, what I would have
seen would have been a tuba.
Likewise, when I look through a periscope, what I see is counter-
factually dependent upon the objects that give rise to my perception.
I see an ocean liner, because that is what occupies the pertinent visual
arc; had it been a battle cruiser, a battle cruiser is what the periscope
would deliver up to me. It is because devices like periscopes and binoculars
preserve this sort of counterfactual dependency that we are prepared to
say that we see through them to the very objects to which they grant us
visual access.
These instruments boost our visual powers. They are on a continuum
with unaided sight, in part because they sustain the same land of relations
of counterfactual dependence with the objects of sight that natural per-
ception does. That is, what we see through them would have been different
had the visible properties they were aimed at been different. The causal
chain of physical events involved in looking through a pair of binoculars
may add a step when contrasted to unaided vision. But the step is not one
that is different in kind from what one finds in unaugmented perception.
It remains a causal process that preserves the relevant visual relations of
counterfactual dependency. The prostheses and the perception they abet
blend together seamlessly.
But is the situation so different when it comes to cinematography?
Cinematic "vision" and unaided vision are as alike as the "vision" available
through a pair of binoculars is analogous to unaided perception. Counter-
factual dependency obtains equally. We expect a cinematic image of x to
present the visible properties of x in such a way that had the visible
properties of .x been different, the image of x would differ correspondingly.
For instance, we expect, all things being equal, the cinematic image of
a white church to be white, though if, counterfactually, the church had
been black, then we would have expected the cinematic depiction of it
to be black.
This, of course, correlates both with prosthetically unassisted vision
and with the enhanced vision available through the optical instruments
discussed thus far. So if we are comfortable with saying we see through
binoculars because of the way in which they preserve the counterfactual
dependency of the visual output on the visual input, then why should we
96 THE MOVING PICTURE

balk at saying precisely the same thing about cinematic images which possess
comparable powers? Since the same kind of physico-causal processes are
involved in seeing through a telescope as are engaged in the delivery of a
cinematic image to an auditorium, we have no reason, in principle, to treat
the one as different from the other. We see the extinct star through the
telescope; we see the assassination of JFK through the Zapruder footage.
In order to "see through" a cinematic shot, it is a necessary condition
that the photographic process put us in contact with the object by purely
mechanical means of the sort that guarantees the preservation of the
pertinent relations of counterfactual dependency. However, the proponent
of the transparency thesis does not think that this is enough to establish that
the shot is transparent. Why not? Well, imagine a computer that is capable
of scanning a visual array of simple geometrical shapes and of then printing
out a verbal description of it. When a square appears in the array, the
computer prints out "square," and so forth. However, no one supposes
that in looking at the word "square" on the printout we are seeing through
the paper and ink to the very square that the computer detected.
A description is not a transparent picture for the simple reason that it is
not a picture at all. So what must be added to the concept of transparency
in order to forestall the inclusion of computer-generated descriptions of
the previous sort from the category?
One way to begin to do this is to think about the ways in which we
might be confused by pictures versus the way in which we can be confused
by verbal descriptions. When reading, for example, we might confuse the
word sun for son, since the lettering is so very similar. This kind of error is
even more likely if we are reading very quickly. However, when we are out
in the world viewing objects in nature it would seem virtually impossible
(discounting Apollo) to mistake a male offspring for the celestial body.
On the other hand, when it comes to seeing objects out in the world,
it is quite possible to mistake the back of a garage for the back of a house,
though even when we are tired or reading quickly, it seems highly
improbable that one could mistake the word garage for the word house.
What accounts for these differences?
One very plausible hypothesis is that confusions between objects in the
case of seeing in nature are rooted in genuine or real similarities between
the objects in question, whereas confusions involving words are caused
by similarities in the lettering which, in one sense, are perfectly arbitrary.
Consequently, we may hypothesize that seeing through a photographic
process obtains only where confusion over the object in the photographic or
cinematic image is a function of genuine similarity relations between the
 THE MOVING PICTURE 97

pertinent cinematic referents. Descriptions, even if mechanically gener-

ated, do not provoke visual confusion on the basis of the real similarities
between the objects they depict, but only confusion in terms of lettering,
which is arbitrary. Authentic transparent presentations, in contrast, traffic
in real or natural similarities, whereas descriptions do not.
So, x is a transparent image if and only if (1) x puts us in mechanical
contact with its object in such a way that the visible output in x is counter-
factually dependent upon its visible input, and (2) x preserves genuine or real
similarity relations between the objects and events presented by it. Further-
more, cinematic images are transparent in this sense. Cinematic images are
like telescopes into the past. We see through them to the sound stages of
yesterday.
The first thing to note about this proposal is that even if we accept the
notion that photographic images are transparent, it would not follow that
all cinematic images are, since, as we have noted earlier, not all cinematic
images are based in photography. Computer-generated mattes, like those
in Sin City, and other comparable digital effects have figured increasingly in
recent motion pictures. Because computerization permits greater control
over the image and lower costs when it comes to mounting spectacles, we
are likely to see more and more CGI rather than less. But a great many of
these images cannot be said to be transparent. You cannot see through
many of them to the objects that gave rise to them for the simple reason
that there are no such objects. The images were fabricated entirely within a
computer. Consider the army conjured up by CGI in The Mummy. Likewise
much of the background rockface in 300 was built in 3D geometry,
employing Autodesk Maya. So, one objection to the transparency thesis
is that it is not a general account of every cinematic shot, but only of those
created by what we might call straight photographic means.
But suppose this much is conceded. The question still remains about
whether the transparency thesis helps us understand all those cinematic
shots that were produced photographically. In all probability, most of the
cinematic shots to date fall into that category. Does the transparency thesis
enhance our understanding of them?
An initial problem with the transparency thesis is that, as it has been
articulated so far, it is not clear that everything that meets the two
conditions stated above is something that we would be prepared to say
we see through to its object. The definition, that is, is not sufficient to pick
out so-called transparent pictures.
Consider a case like this one. Imagine a fully automated, enclosed,
railway yard, one that runs itself and has been insulated from the intrusion
98 THE MOVING PICTURE

of living creatures. Now suppose we build a point-by-point model of

the railway yard that links every square inch of it to a super computer,
so that every change in the surface of the railway yard registers a change in
the model. If an engine switches tracks in the railway yard, then a facsimile
engine switches tracks in exactly the same way in the model. As the
original engine rusts, so does its counterpart in the model. And so on.
The model, however, is in Argentina, while its source is in Samoa, though
looking at the model provides no clue to the location of the original.
Following the preceding account of transparency, we are forced to the
bizarre conclusion that the model gives us transparent perceptual access to
the railway yard — that we can literally see through the model in Argentina
to its source in Samoa. For the model is counterfactually dependent upon
the visible properties of the original with which it is in mechanical contact.
Every change we observe in the model mirrors precisely comparable
changes in the railroad yard in Samoa. Moreover, where we might tend
to confuse objects — like water tanks and gasoline tanks — in the original
railway yard, we will also tend to confuse objects in the model, because the
model preserves real similarity relations between things.
But surely this conclusion is unacceptable. Were we to rebuild the
model in Samoa in such a way that it stood between us and the original
and thus occluded the original in our visual field, we would not say we saw
through it, but rather that it blocked our vision. You can't see through
solid objects, can you?
What this example may show is that we do not yet have a full account
of what it would take for something to be a transparent presentation of x,
Therefore, until that uncertainty is resolved, and transparency is adequately
defined, we cannot really determine whether we see through, in the
requisite sense, photos and cinematic images to their objects.
However, the proponent of the transparency thesis might argue that
even if he has not provided a complete defiirition of seeing through, he has
brought our attention to a continuum of cases where it seems appropriate,
each step of the way, to speak literally of seeing. Thus, he argues, the
burden of proof belongs to those of us who find it hard to believe that we
see through cinematic shots to their objects.
The transparency theorist starts by comparing ordinary seeing and
seeing with a corrective lens. Both strike us unequivocally as seeing. But
if seeing through a corrective lens warrants being called seeing, why stand
on ceremony when it comes to telescopes, binoculars, and microscopes?
Moreover, if these prostheses are so regarded, why not also treat peri-
scopes and mirrors in the same fashion? And if we see through mirrors,
 THE MOVING PICTURE 99

there seems no reason not to speak in the same way about photographic
and cinematic images which, in effect, function rather like celluloid mirrors.
Putatively, seeing through cinematic images is merely a straightforward
expansion of the notion of seeing through binoculars.
That is, starting with ordinary seeing, we seem to encounter no
principled difference in kind as we make our way through an inventory
of cases all the way to the cinematic image. We find ourselves sliding down
the transparency theorist's slippery slope. Is there any way to break our
fall? In other words, less metaphorically, can we adduce any principled
differences between the items on this apparent continuum such that we
can distinguish between ordinary seeing and seeing with certain prostheses,
on the one hand, from our perceptual engagement with the contents of
cinematic shots on the other? That is the challenge with which the
transparency theorist confronts us.
Yet, despite the transparency theorist's suggestion, there is at least one
difference between ordinary seeing and certain forms of seeing through
prostheses, on the one hand, and the alleged seeing through photographs
and cinematic images on the other. We encountered it in the preceding
chapter where we noted that something counted as a moving image only if
it was a detached display. Moreover, there is a striking disparity between
seeing x and seeing a detached display of x. When one sees x, one can point
one's body toward x, but in seeing a detached display of x, one cannot.
Ordinary perception carries egocentric information about the relation of
my body to the object of perception. But photos and cinematic images do
not afford comparable egocentric information. Thus, the everyday case of
seeing x in front of me and seeing a detached display of x should not be
treated as instances of the same kind of perceptual experience. For I can
orient myself physically to x in the spatial-temporal continuum I inhabit in
cases of ordinary seeing, but not with cases of seeing a detached display of x.
If it entered my visual field, I could find my way to that railroad yard in
Samoa; but if I saw a detached display of it — like the aforesaid model
in Argentina — I would have no clues about which way to turn to get to
the original.
Egocentric information about my bodily relation to the objects of
perception is also available in many of the cases where I employ prosthetic
aids. I can point my body in the direction of the star when I use a telescope;
toward the microbes when I look into the microscope; at the parking
space, when I peer into the rear-view mirror in my car; at the ingenue
when i adjust my opera glasses; at the aircraft carrier when I see through a
periscope. And so on. In all of these cases, these prosthetic instruments can
100 THE MOVING PICTURE

direct me to the objects they deliver to us. But I could not hope find my
way to the White House by looking at a photographic postcard or a
newsreel of it. For they are detached displays and, all things being equalj
a detached display does not provide perceptually the sort of egocentric
information about its object that I need to get from point A to point B.
If I look through a pair of binoculars at a brace of horses racing furiously,
the visual array I obtain, though magnified, is still connected to my own
body in such a way that I would be able to walk reliably toward the finish
line if asked. When I use the binoculars, I can still point myself accurately
toward the finish line. My bodily orientation with respect to the things
that I perceive is preserved. But someone who does not already know
where St. Petersburg is cannot point themselves reliably in the direction of
the Winter Palace by watching Eisenstein's October.
Ex hypothesis we do not literally see x, unless we can orient reliably
ourselves spatially toward x by, for example, turning our body, if asked, in
the direction of x. We can achieve this in the case of ordinary seeing and
in many cases of prosthetically empowered perception. It is not the case
that we can do this with respect to cinematic images, save in unusual
conditions, as when we view a video in the self-same space the video
depicts. The possession of versus the lack of egocentric information is the
distinction we need to thwart the slippery slope argument introduced to us
by the transparency thesis. Since a cinematic image of Fido lacks egocentric
information about the relation of my body to Fido, we do not see Fido
through it in the same way in which we see Fido in the flesh when he is
prancing before us on the lawn.
The transparency theorist may attempt to meet this challenge with two
challenges of his own. First, consider the sight of a rose situated ten miles
away but delivered to us by a relay of hundreds of mirrors arranged after
the fashion of a labyrinth so complex that, by peering into the mirror in
front of us, it is impossible to find our way to the rose reflected in it.
Ostensibly we have agreed that we see through mirrors. And yet this
particular arrangement of mirrors does not enable us to point our bodies
toward the rose. Therefore, the transparency theorist argues, we must
agree that the preservation of egocentric information is not a necessary
condition for seeing.
Furthermore, the transparency theorist invites us to consider a second
case. Again we see a rose before us. But we have also been apprised that
it is very likely that it is being presented to us through a maze of mirrors
like that described in the previous paragraph. Thus, we may come to doubt
that we do know how to find our way to the rose. However, we have been
 THE MOVING PICTURE 101

misled. The rose is actually right in front of us. We do not believe that it is
in front of us; we doubt we can walk toward it. But surely we are seeing
the rose, even if we do not believe we are seeing it. We are seeing it, even
if we do not believe that we know how to move in its direction. This too
indicates that knowing how to orient one's body toward an object is not a
logical requirement for seeing the object.
What both of these examples are supposed to show is that egocentric
information is not a genuine criterion for either seeing x or seeing through
j to x. Thus, the transparency theorist argues, the putative lack of
egocentric information cannot be invoked as the reason why we cannot
be said to see through photographic and cinematographic images to the
objects at which they are aimed.
However, it is not obvious that we must concede the case of the
labyrinth of mirrors to the transparency theorist. If every time I try to
walk toward the rose I bump my head against a mirror, it is not evident
that I am really seeing the rose. Agreeing that some arrangements of
mirrors enable us to see through them does not force us to say that we
see through every imaginable relay of mirrors. This position, moreover, is
not arbitrary, so long as we have a principled way of distinguishing between
the two cases. Do we? Of course — the mirrors we see through are the ones
through which we can retrieve reliable egocentric information; the ones
we do not see through are the mirror arrangements through which we
cannot. There is no inconsistency here. If the transparency theorist alleges
that there is, he begs the question.
But what of the second example — the case where the rose that
dominates our visual field is really directly in front of us but where we
doubt it because we are entertaining the suspicion that a Cartesian-type
deceiver has surrounded us with a baroque maze of misleading mirrors?
Certainly we are seeing the rose, even though, if asked, we might say we
don?t believe we can point or otherwise orient our bodies toward it.
Nevertheless, it pays to notice that there is a difference between being
able reliably to point our bodies toward the rose and believing that one can
reliably point one's body toward the rose. The transparency theorist's
example focuses on the doxastic state (the belief state) of the percipient.
But even if, given the peculiar circumstances of the thought experiment,
the percipient does not believe that she can reliably point her body toward
the rose, she may still be able to do so. I may not think or believe that I can
lift a fifty-pound barbell over my head, but I may, in spite of what I believe,
be able to do so. Similarly, I may not believe that I can walk reHably toward
the rose, but, if instructed to give it my best shot, in the situation as
102 THE MOVING PICTURE

outlined by the transparency theorist, I will be able to do so reliably,

because actually directly seeing the rose — the very perceptual state I am
in — is one that reliably tracks egocentric information about my spatial
relationship to the object of perception.
That is, we can say, along with the transparency theorist, that in his
second example the percipient is seeing the rose, even if she is not confident
about where it is located in space, just because the percipient has access to
the kind of egocentric information that, if push comes to shove, will enable
her behaviorally to find her way to the rose. Her doxastic state notwith-
standing, she is able to do so behaviorally precisely because genuine seeing
carries egocentric information. The transparency theorist's case is not a
counterexample that refutes the necessity of egocentric information to
authentic seeing, but rather an example that reconfirms it. For whether or
not the percipient believes she can successfully point her body toward the
rose, if she is genuinely seeing it, then she can so orient herself reliably
toward it, since genuine perception gives her access, whether she believes
it or not, to the pertinent egocentric information.
One objection to the hypothesis that genuine seeing requires the ability
to reliably orient oneself spatially in the direction of the object of percep-
tion might be that this has the counterintuitive consequence that disabled
persons, deprived of the capacity for locomotion cannot see, for they
cannot move their bodies in any direction. Consequently, we need to adjust
our claim so that we assert that genuine seeing only occurs in persons who,
if they either possess or were to possess the relevant movement capacities,
would be able to reliably orient their bodies in the direction of the relevant
objects of perception. Since this criterion is behavioral rather than dox-
astic, it is not threatened by the transparency theorist's second example.
Genuine seeing, including genuine seeing-through, then, is a process
that gives us reliable egocentric information about the relationship of our
body to the objects in the space we ii|habit. Cinematic shots do not afford
this sort of information. Thus, we do not literally see through them.
Therefore, the transparency thesis appears unacceptable.

The Code Thesis

In the era of the silent film, cinema was often analogized to language.
D. W. Griffith, for example, claimed to derive certain of his techniques
from the writing of Charles Dickens. The most ambitious early versions of
this kind of claim were undoubtedly articulated by the Soviet montagists.
 THE MOVING PICTURE 103

For example, one of the leading theoreticians and filmmakers of that school
of cinema, V. I. Pudovkin, argued that "[Film] editing is the language of the
film director. Just as in Hving speech^ so one may say in editing: there is a
word — the exposed piece of film, the image; a phrase — the combination of
these pieces."
That is, according to Pudovkin, the cinematic shot should be understood as
something strictly word-like. Perhaps we see a shot of the muzzle of a cannon;
this is equivalent to the word "cannon." Then we are shown a close-up of a
fuse being lit; this is equivalent to " fires." The two shots are spliced
together and we get "The cannon fires."
But even if this somewhat contrived example seems persuasive, it is hard
to imagine this account being extrapolated to the operation of shots in
general. For shots do not typically function as words. Characteristically,
shots carry an amount of information that it would take sentences and even
paragraphs to articulate in language; this is an instance of the old truism
that "a picture is worth a thousand words." Indeed, moving pictures are
probably most often worth even more words than still pictures. And, in any
event, clearly a feature-length film, like Russian Ark, composed of a single
shot cannot be equated with a word, even if we turned off the voice track.
Except in very special situations, such as those in which shots are being
manipulated to advance cinematic similes (e.g., from Eisenstein's October:
"Kerensky is like Napoleon"), moving picture shots are not the equivalent
of words, especially if the test of being word-like is that they be summari-
zable by means of a single word. It was only for the sake of the example
that we "translated" the igniting of the fuse above as "fires." It would be
more natural to characterize such a shot with a sentence.
But, be that as it may, the content of most shots cannot be readily
captured by one word. So it would appear unlikely that most shots are
strictly equivalent or even analogous to words. Moreover^ if it is argued
that the capacity to be summed up in a word is not the true test of shots
being word-like, then the burden of providing a method for ascertaining
whether a shot is functioning as a word falls to Pudovkin and his followers.
Unfortunately, to date, they have not met this challenge.
In addition, another problem is that words are abstract in a manner
quite unlike moving picture shots, which are concrete. The word "cannon"
need not call any particular cannon to mind, whereas a shot of a cannon is
always of a particular cannon — the 155 mm howitzer with a flat right front
tire and the initials "NC" scratched on the inside of its barrel, for example.
So, shots are arrestingly different from words given their inevitable
concreteness, which property, in turn, is connected to the fact that they
104 THE MOVING PICTURE

are naturally described by many words, not one. Related to this and
further supporting the disanalogy between cinematic imagery and language
is that the neural bases of imagery and language are different, suggesting
that we are dealing with distinct communicative systems.
Furthermore, the relation of a word to its referent is arbitrary. We
use the word "dog" in order to denote canines, but it is easy to imagine
that we might have introduced the word "cat," had it not been otherwise
employed, to do the job. O n the other hand, that a well-made shot of a dog
depicts a dog is not arbitrary; a competent shot of a cat, all things being
equal, cannot represent a dog cinematically. The shot of a dog represents a
dog, to a large extent, because of a nonarbitrary perceptual relation between
certain cinematic processes and certain automatic, human, visual processes —
ones to be discussed at greater length anon. But for the moment, suffice it to
say that a shot is not a word, nor is it particularly word-like because, among
other things, its relation to its referent is not, on the face of it, word-like
in terms of arbitrariness. That is, whereas the relation of a word to its
referent is conventional, it is not merely a matter of convention that the
shots of Lassie in the old TV series depict a dog, indeed a collie (rather than,
say, an ostrich).
We have argued that a cinematic shot is not a word because, among
other things, it would generally appear to carry m o r e information than a
word does. At the very least, it must usually be described by more than
one word. Shots do not typically evoke single words, save in certain cases
of highly controlled figuration, as one finds most famously in constructivist
montages. If shots evoke linguistic responses at all, it is more plausible to
predict that they would be descriptions of at least the order of sentences.
But it will do no good to attempt to repair the code thesis by saying that
a shot is not a word, but rather a sentence.
For cinematic shots are neither sentences nor are they sentence-like.
Sentences are composed of atomic particles — words — that are minimal
units. But shots cannot be decomposed into minimal units like verbs and
nouns. Shots do not have discrete subjects and predicates. In a shot of
the cannon firing, which part of the picture is the subject and which part
is the predicate?
Shots do not contain grammatical units at all. A shot is not structurally
analogous to a sentence such as "This is a tall man." N o t only does it lack
analogues for "this," "is," and "a," but the subject of the shot is not the
result of the law-like combination of the acontextually definably meaningful
units "tall" and "man." Rather the shot of a man shows all his properties at
once and nondetachably, so to speak. You can't break the down the moving
 THE MOVING PICTURE 105

picture of the tall man into its component parts in the way that you can
analyze the sentence into minimal units such as "tall" and "man." Moreover,
if negation is a natural part of language, then film cannot be a normal
specimen of language, since it would appear to lack the means to say "no" in
its putative vocabulary. Showing a picture of Napoleon is scarcely the
equivalent of saying that "Napoleon is not tall."
There is nothing in the shot that corresponds to words. In sentences,
meaning is made by the application of certain recursive rules to the atoms
of meaning — the words — that comprise the sentences. That is, sentences
abide by grammar in order to be well formed. However, moving picture
shots are not grammatical, or even ungrammatical. In part, this is because
they are not comprised of minimal units of meaning — like words — that can
then be assembled in accordance with certain rules.
But in addition there are no rules about what can be combined with
what in a movie shot. Whatever can be represented pictorially is kosher;
there are no ungrammatical pictures. There may be pictures of physically
impossible objects and events — like Dumbo or Mighty Mouse in flight.
However, these are impossible pictures, not ungrammatical ones. Even the
picture of a monstrous entity — say a head with an arm sprouting from it, as
one might find in a film by Brian Yunza, is only a picture of something
biologically ill formed rather than a grammatically ill-formed picture.
In this regard, a shot is more like a representational painting than it is
like a sentence.
Sentences are meaningful in virtue of being built up, via recursive
operations, from acontextually defined, meaningful, detachable units,
like nouns and verbs, adjectives and adverbs. But shots lack comparably
discrete building blocks. The red hair on screen is not made up of two
discrete components red and hair, combined in virtue of certain semantic
rules, but is just red-hair-all-at-once, so to say.
Furthermore, the arm of a woman in a film shot is not like the word
"arm" in a sentence, because the word "arm" bears an arbitrary relation to
its referent, whereas the arm in a film image does not bear an arbitrary
relation to its referent. Nor is the perceived biological relation of the
woman's arm to her torso a conventional one in the shot of a film, though
the grammatical relation of the word "arm" to the word "torso" in the
sentence "The woman's arm was connected to her torso" is utterly
conventional, its sense governed by its grammar.
Moving picture shots also lack the panoply of other structures available
to natural languages like English. Sentences come in a variety of moods, but
how, without resort to the soundtrack or titles, does one articulate an
106 THE MOVING PICTURE

interrogative shot? Declarative sentences can be active or passive. They

have voice. But was our earlier example of the cannon firing active ("The
cannon fires") or passive ("The cannon is being fired")?
Confronted by these sorts of objections, the proponent of the code
thesis may be moved to abandon the strong thesis that shots are strict
analogues of words or even sentences, but, nevertheless, maintain that they
are somewhat language-like in the sense that the communication achieved by
means of motion picture shots is dependent on conventional codes. Umber to
Eco, a leading semiologist, for example, has suggested that single-shot
images — of\ say, the Eiffel Tower — are best understood as conventional
signs. Commenting upon a shot from Fritz Lang's classic film M, Eco says
"on the basis of some conventions the addressee [the viewer] understands
that those traces upon the screen send us back to the content 'a girl holding
a balloon/"
The idea here seems to be that the relation between a shot and what
it is a shot of is analogous to a dictionary-like correlation. In the process
of viewing moving pictures, we acquire the vocabulary of cinema piece
by piece, correlation by correlation. There is a code here, though unfor-
tunately the associational structure of the code has never been spelt out to
any degree by semiologists, nor, for that matter, have they even elaborated
adequately their operating conception of a code.
For semiologists such as like Eco, understanding images must be learned.
What must be learned are conventions that link the image in the shot — "the
traces on the screen" — to its referents. Film shots are code-like insofar as
they are somehow a matter of conventional associations. Comprehending
a shot is allegedly a process of decoding it, though the units of the code
and their combinatory principles are lamentably left rather obscure by
proponents of the code thesis.
Nevertheless, the code theorist believes that a condition for compre-
hending what a shot stands for is putatively knowledge of a certain set of
codes or conventions, perhaps involving dictionary-like correlations.
Moreover, these codes and conventions are said to be culturally specific;
they could have been otherwise. Thus, the semiologist alleges that they are
arbitrary.
Indeed, on this view comprehending a single shot is akin to reading or,
at least, deciphering or decoding. Films are decoded by audiences who have
learned the conventions of cinema, including conventions for translating or
deciphering a single shot. We identify the shot of the man as such in the
same way that we associate or correlate the word "man" with "adult
male." We employ a code that we have previously learnt.
 THE MOVING PICTURE 107

But the suggestion that the way in which we learn a language is on a par
with the way in which we become conversant with a pictorial idiom is
immensely dubious. Suppose we call the ancient Egyptian "frontal eye"
approach to pictorial representation a code. If it is a code, then we acquire
command of it in a way that is radically different from the way in which we
would achieve mastery of spoken Egyptian during the reign of King Tut.
Once you encounter one or two pictures in the "frontal eye" style, you
can readily recognize the kind of thing represented by any other picture in
that tradition. But if we were to learn the meaning of one or two spoken
words of ancient Egyptian, we would not be any wiser about the rest of
the language.
This is not to say that there are no pictorial conventions in the frontal eye
style. That we interpret the fact that the pharaoh is so many times larger than
other folk in the picture as a measure of his relative social superiority is a
learned convention. However, that we recognize the picture of the pharaoh
as a picture of a man is not something we need to learn to do. If we can
recognize m e n "in nature," so to speak, we can recognize pictures of men.
That is, once we see one or two ancient Egyptian pictures, we can then
go on to identify what virtually every other Egyptian picture is a picture
of — at least with respect to the general kind of thing it is a picture of. But
were we to hear one or two ancient Egyptian words and learn what they
mean, we would not, on the basis of that experience, be able to grasp the
meaning of any other Egyptian words, n o t even the words that denote
different natural kinds.
The way in which we acquire knowledge of a code, like a natural
language, and the way in which we acquire mastery of a pictorial system
differ extremely. Ex hypothesis the reason is because the underlying pro-
cesses are different. Arbitrary codes are associational and, therefore, must
be acquired correlation by correlation over a protracted period of time,
whereas pictorial practices can be grasped almost immediately — or, at
least, with astounding alacrity in comparison to language learning — and, in
a manner of speaking, "all in one shot." Indeed, the empirical evidence is
that the underlying brain-processing systems that negotiate language and
pictures are different.
This contention is borne out by the research of two psychologists —
Julian Hochberg and Virginia Brooks. This couple raised one of their
children in an environment effectively devoid of pictures of any sort,
including drawings, photographs, and, of course, motion pictures. The
child had no picture-books, nor did he have access to television; the labels
with visual graphics on them were removed from all the household
108 THE MOVING PICTURE

products in his vicinity. Moreover, the child was never tutored in any way
about associations between words and pictures. He was never read a story
from a picture-book. And yet at the age of 2 or so, when the child's
vocabulary was sufficiently developed, he was exposed to a selection of line
drawings and black and white photographs and he was asked to identify
what they were pictures of. He succeeded in doing so in most cases. He did
so at rates that were scarcely random.
In this case, there was no question of the child's learning or being
instructed in codes or conventions for decoding or reading pictorial
images, including moving picture shots. The child was able to recognize
what the pictures were pictures of simply by looking. The code hypothesis
fares badly with a case like this one, since there is virtually no process of
learning detectable here.
Picture comprehension, including motion picture comprehension,
appears to require no special training or instruction. It seems to come
naturally. That is, one hypothesis that is superior to the code hypothesis
for explaining the way in which we come to comprehend what a picture is a
picture of is that this comprehension occurs via the activation a natural
recognitional capacity — one that arises in tandem with our capacity to
recognize objects but which is also triggered by things that are appreciably
similar to the visual properties, such as the outline contours, of said objects.
No special training is required to understand what a picture — including
a moving picture shot — pictures, so long as one is already familiar with the
kind of thing being depicted by the image in question. One can identify
what a typical shot of x is a moving picture of by looking, if one can identify
x in "real life," even if one has never previously seen a shot of x. The fact
that this can be done without any significant training period indicates that
nothing like the acquisition of a code is involved in accessing the meaning
of moving picture shots. Indeed, the communication here seems to happen
in the absence of a code. Therefore, the thesis that the shot is a coded
signal seems unlikely. This is not to say that the shot is not a symbol. Rather
it is a symbol that communicates in large measure without reference to a
code, a vocabulary, or a set of combinatory conventions.

The Recognitional Prompt Thesis

This hypothesis emerges almost unavoidably from a consideration of the
shortcomings of the code hypothesis. The code hypothesis would appear to
require a learning period during which the code — whether a natural
 THE MOVING PICTURE 109

language or something else — is acquired. This seems to follow from the

allegation that the codes relevant to understanding shots are arbitrary. How
else would one achieve access to an arbitrary code except by learning its
basic elements and rules piece by piece? But even if some pictorial styles
may require an interval of time upon first exposure in order to be grasped,
that interval is infinitesimal compared to the chunk of time required to
begin to master a language or similarly complex code.
Moreover, once the percipient has — quite quickly — gotten the hang of a
pictorial style through one or a few of its symbols, she can pretty much
effortlessly apprehend everything else represented in that style. But from
one or a few words, one has hardly started to understand the language.
Arbitrary codes, especially ones with large vocabularies, take a long time to
learn. The velocity with which we acquire mastery of pictorial styles — such
as the Japanese "floating eye" style or that of Renaissance perspective — is
astounding compared to the amount of time that we expend in absorbing
a complex, arbitrary code. Moving picture shots, like other pictures, are
grasped virtually at a glance with little or no tutelage. They are so different
from codes that there is no reason to suspect they belong to the same
psychological category.
Learning a complex code, like a language, involves an extended period
of training. It takes a great deal of instruction, including trial, error, and
correction. But the subject in the previously cited experiment by Hochberg
and Brooks manifested the capacity for pictorial understanding in "one fell
swoop." A child does not assimilate the meaning of a word, let alone a series
of words or any other complex code, like that. Thus, it is unlikely that
anything akin to learning a code, like a language, is operative in mastering
the comprehension of a moving picture shot.
Instead, a more attractive hypothesis, given the data, is that humans
acquire the capacity to recognize pictures, including moving pictures,
naturally, rather than conventionally, at the same time that they acquire
the capacity to recognize the objects that pictures represent. Pictorial
recognition, including the recognition of moving picture shots, is not a
skill that needs to be taught. Pictorial comprehension is a natural recogni-
tional capacity that comes with object recognition. When we see a picture,
including a moving picture, of something with which we are already
familiar, the same recognitional processes are thrown in gear as when we
see that kind of thing "in the wild." Exactly what convergent features of
the objects in the pictures and the corresponding objects outside of the
pictures serve to trigger this response is a question for psychological
inquiry. Philosophically, it is enough to say that the motion picture shot
110 THE MOVING PICTURE

is a natural recognitional prompt — a stimulus that engages something bred in

the bone to apprehend what it is a picture of simply by looking.
This, however, is not a reversion to the illusion thesis. The hypothesis is
that the stimulus imparted by moving picture shots, taken from typical
angles, activates the same recognitional capacities that the objects pictured
would engage in everyday, "off-screen" life; but, at the same time, we also,
by dint of the same faculties, recognize the screen as an independent object
in our environment. That is, we recognize both the objects pictured and the
screen. Pace the illusion thesis, we do not take the objects pictured in the
shot to be objects in the space of our immediate physical environment.
Indeed, we detect this difference phenomenologically in our own bodies,
inasmuch as the muscles in our eyes register the distance between the
location of the picture as the surface of an object — the screen — that inhabits
the very physical space in which we are situated, and the objects in the
virtual space portrayed inside the moving picture.
Nor is the recognitional prompt thesis the same as the transparency
thesis. For on the former view all that is claimed is that we recognize the
objects and events in the shot, not that we see through the moving picture
to the objects and events that gave rise to them. It is not, in other words,
that the original objects are engaging our perceptual systems through a
prosthetic device, but rather that there is something about the picture that
naturally arouses our powers of object recognition in virtue of which we
are enabled to identify what the picture is a picture of. What that something
is, we depend ultimately upon perceptual psychologists to define. But what
we can say now is that moving picture shots naturally prompt our percep-
tual system to recognize in them the kinds of objects and events that we
apprehend visually in everyday life.
Like the recognitional prompt thesis, both the illusion thesis and the
transparency thesis presuppose that we mobilize, without any appreciable
prior training, our ordinary recognitional capacities in order to identify
what we see in moving picture shots. In that, the three theses are in
agreement. But according the recognitional prompt hypothesis, we do not
take what we see to be the very object we recognize either mistakenly, as
the illusion thesis would have it, or veridically, as the transparency thesis
maintains. Rather, a motion picture shot of a gazelle causes us to see or to
recognize a gazelle in it; it naturally generates both the recognition of the
screen and of the things depicted by the moving shadows reflected upon it.
The hypothesis that motion picture shots are those whose depicta
are recognized naturally by untutored spectators — spectators who have
not been trained in some specialized processes of reading, decoding,
 THE MOVING PICTURE 111

or inferring — simply by looking also fits nicely with the historical fact of the
notably rapid, cross-cultural dissemination of the movies. Motion pictures
became the dominant artform of the twentieth century and have evolved
into an omnipresent channel of communication worldwide because the
basic symbols in the practice — the moving picture shots — are accessible
virtually automatically to everyone from urban centers in North America
to rural communities in Eurasia.
Undoubtedly, these assertions fly in the face of certain anecdotes,
repeated endlessly in anthropology classes, about the incomprehension
with which tribal peoples greeted pictures in the Western tradition.
However, the first thing to emphasize in this regard is that the systematic
evidence about the responses of pictorially inexperienced, tribal peoples to
pictures is overwhelmingly on the side of the hypothesis that they do,
without instruction, recognize what the pictures which are presented to
them depict.
Furthermore, in those cases where subjects manifested some trouble in
identifying the referents of the alien pictures, it seems likely that their
initial difficulties had more to do with misunderstanding what they were
being asked to do, rather than with an inability to recognize the depicta of
the relevant pictures. For example, when pictures were presented to the
Me'en tribe of Ethiopia, they sniffed them instead of looking at them,
because the pictures reminded them of a kind of cloth material with which
they were already familiar. They did not look at the pictures at all.
However, when their attention was directed to the task of identifying the
objects being pictured, the vast majority of the respondents had no
problems comprehending the pictures.
That is, their putative inability to understand the pictures had to do with
their lack of familiarity with the point of picturing and not with an inability
to recognize what the pictures were of, since, once the practice of
picturing was explained to them, they had no problems with offering the
pertinent identifications. Moreover, since they were able to do so with
virtually no preparation, it is not plausible to suppose that they had — on
the spot — digested some complex body of conventions.
Recognition rather than convention is the fundamental key to pictorial
comprehension. To recognize something, of course, requires that what is
presented to us be something or, at least, a type of something, with which we
are already familiar. It is a matter of re-cognizing. So, pictorial recognition
is a process of realizing what is depicted in a picture where it is something —
or an instance of a type of thing — that is already known to us simply by
looking at (rather than by reading or decoding) the picture.
112 THE MOVING PICTURE

Moving picture shots are, of course, typically pictures. Thus, these

building blocks of movies are such that what they depict is recognized by
untutored spectators simply by looking. Spectators do not comprehend the
reference of single shots by means of reading, decoding, or inferring from
so-called traces — as semiologieal proponents of the code thesis intimate.
Rather, the process proceeds by engaging object-recognitional capacities
that are hardwired.
The single shot is the basic unit of film communication in the sense
that, without the single shot, other levers of cinematic communication,
like film editing, would remain mute. Without single-shot units, there
would be no editing. The single shot is typically, in a manner of speaking,
the sine qua non of filmic communication. Or, in other words, if there were
no single shots to be comprehended, there would be nothing for editing
to do — nothing for it to organize. So, in this sense, the moving picture shot
is the basic element of cinematic communication.
Moreover, the single shot is standardly a picture, and pictures are the
sorts of symbol whose depicta are grasped by natural recognitional capaci-
ties. That these capacities are natural, of course, implies that they are
comprehensible at the level of what they portray to most perceptually
competent human beings, virtually automatically. Humans around the
world can recognize what a picture — including a motion picture image —
is a picture of, simply by looking. And this entails that motion picture shots
are accessible to nearly every human being without special training in any
code or convention. Needless to say, this is why motion pictures are so
eminently exportable globally. This is why motion pictures are every-
where. This is part of the reason that moving pictures are so important.
A benefit of the recognitional prompt theory of the moving picture shot,
then, is that it can contribute to an explanation of the widespread success
of the motion picture as a means of art and communication. Whereas the
code thesis leaves us with the mystery of when spectators ever learned the
correlations between Eco's vaguely described traces and their meanings,
there is no enigma remaining when we hypothesize that the basic commu-
nicative function of shots can be discharged without the mediation of codes
or meaning conventions. Of course, the illusion thesis and the transparency
thesis also have this advantage over the code thesis. But the recognitional
prompt thesis does not have the other liabilities of those alternative
conceptions of the motion picture shot.
On the recognitional prompt thesis, we recognize what a shot is a
picture of by looking; our inbred capacity for recognizing objects swings
into action. Where the image is fictional, we then go on to imagine that the
 THE MOVING PICTURE 113

kind of thing we recognize — say, a young man — is an agent, such as the

protagonist, in the story. Or maybe we recognize the facade of an ante-
bellum mansion in the neoclassical style of the old South and imagine, as
the filmmakers mandate we do, that this is Tara. If these images are then
dubbed appropriately for the pertinent non-English-speaking audiences,
these fictions, all things being equal, will be immediately accessible to
audiences everywhere in large part because the pictures that are commu-
nicating the story can be recognized by one and all sighted humans.
In this chapter we have reviewed four hypotheses concerning the nature of
the moving picture shot — that shots are illusions, transparencies, coded
symbols, or natural recognitional prompts. The illusion thesis founders in
large measure because we do not behave as though we believe that the events
on screen are literally before us; we do not flee, for example, from the
onslaught of swarms of raptors. It may even be argued that the movement
we detect pn screen is objective — in the way that color is objective — rather
than illusory.
The transparency thesis proposes that moving picture shots are pros-
thetic instruments through which we gain perceptual access to objects and
events in the past. This cannot be a necessary feature of all motion picture
shots, since some shots, such as CGI mattes (which are not produced by
means of straight photography) need have no referent from the past to
show us. The claim of the transparency thesis that we see through moving
picture shots may also be challenged on the grounds that even shots
produced by straight photography fail to provide the sort of egocentric
information characteristic of genuine seeing and seeing-through.
The code thesis maintains that the moving picture shot is an arbitrary
symbol of the sort found in a natural language. But moving picture shots
lack the structure of words and/or sentences; nor do we secure our
understanding of them as we do when gaining entry to a complex vocabu-
lary of arbitrary symbols. Rather, we appear to get what motion picture
shots communicate virtually immediately. This suggests that they trigger
some natural recognitional capacity of which the most likely suspect is our
inborn capacity for object recognition. Therefore, it seems very probable
that the moving picture shot is best conceptualized as a natural recogni-
tional prompt — a stimulus that communicates its content to spectators
who apprehend it simply by looking at the moving picture.
The natural recognitional prompt hypothesis, furthermore, has the added
attraction of offering an important part of the explanation of the way in
which movies managed to become such a widespread and often transcultural
form of art, entertainment, and communication. For the basic symbols
114 THE MOVING PICTURE

in the practice — the moving picture shots — can be grasped at a certain

level of understanding by everyone with standard-issue human perceptual
equipment.
Another advantage of the review of the motion picture shot in this
chapter is that often whole conceptions of cinema are based upon extrapo-
lations from analyses of the nature of the shot. From the illusion theory of
shots, some have jumped to the illusion theory of cinema — that is, to the
theory that whole films are illusions. Some (though not the philosophers
who developed the transparency thesis) allege that all mainstream cinema
strives to the condition of transparency. And others have claimed that
entire films are coded or semiotic systems. O n e side benefit of this chapter
is that some of the premises that ground these larger extrapolations have
been undermined.
O n the positive side of the ledger, we also have a naturalistic account
of the moving picture shot. Let us see if we can develop further theories
in this vein for other aspects of motion pictures.

Suggested Reading

The illusion theory undoubtedly derives ultimately from Plato's discussion of

pictorial representation in Book X of his Republic. That account has been so
influential that it has been applied virtually without hesitation to cinema. A great
many theorists simply extrapolate, with scant modification, from Plato the view
that the motion picture image is an illusion and go on from there. A sophisticated
version of the illusion theory, however, can be found in Richard Allen, Projecting
Illusion (New York: Cambridge University Press, 1995).
The illusion theory is critically discussed in Gregory Currie *s fine book Image
and Mind: Film, Philosophy and Cognitive Science (New York: Cambridge University
Press, 1995). Currie, among other things, defends the view that the impression of
movement in motion pictures is not illusory. Currie's view, in turn, is rejected by
Andrew Kania in "The Illusion of Realism in Film," British Journal of Aesthetics 42:3
(July 2002), 243-58.
The transparency theory is suggested, if only obscurely, in Andre Bazin's What Is
Cinema?', vol. 1 (Berkeley, CA: University of California Press, 1967). A version
also seems to be advanced by Stanley Cavell in The World Viewed (New York:
Viking, 1972). Recent defenders of transparency with respect to photography
include: Kendall Walton, "Transparent Pictures: On the Nature of Photographic
Realism," Critical Inquiry 11:2 (Dec. 1984), 246-77; and Patrick Maynard,
"Drawing and Shooting: Causality in Depiction," Journal of Aesthetics and Art
Criticism 44:2 (Winter 1985/6), 115-29.
 THE MOVING PICTURE 115

Critics of the transparency theory include: Nigel Warburton, "Seeing Through

'Seeing Through' Photographs," Ratio new series 1 (1988), 64-77; Gregory
Currie, "Photography, Painting, and Perception," Journal ofAesthetics and Art Criticism
49:1 (Winter 1991), 23-9; and Noel Carroll, "Defining the Moving Image," in his
Theorizing the Moving Image (New York: Cambridge University Press, 1996). Kendall
Walton has responded to critics of the transparency theory in his "On Pictures and
Photographs," in Film Theory and Philosophy, edited by Richard Allen and Murray
Smith (Oxford: Oxford University Press, 1997). Recent important articles on the
transparency thesis include Aaron Meskin and Jonathan Cohen, "On the Epistemic
Value of Photographs," Journal of Aesthetics and Art Criticism 62:2 ( Spring 2004),
197—210, and Aaron Meskin and Jonathan Cohen, "Photographs as Evidence," in
Photography and Philosophy, edited by Scott Walden (Oxford: Blackwell Publishing,
forthcoming).
Regarding the film-shot-as-linguistic theory, the quotation from V. I. Pudovkin
comes from his Film Technique and Film Acting (London: Vision Press, 1958).
Umberto Eco develops his view of the code thesis in his article, "On the Contri-
bution of Film to Semiotics," in Film Theory and Film Criticism, edited by Gerald Mast
and Marshall Cohen, 2nd edn. (Oxford: Oxford University Press, 1979). Eco's
position is criticized by Gilbert Harmon in his article "Eco-location," in the same
volume. See also Sol Worth, "Pictures Can't Say Ain't," in Film Culture, edited by
Sari Thomas (Meteuchen, NJ: Scarecrow Press, 1982). The code/language thesis is
criticized at length in Gregory Currie's superb "The Long Goodbye: The Imaginary
Language of Film," British Journal of Aesthetics 33:3 (July 1993), 207-19.
The recognitional prompt thesis was introduced by Noel Carroll in his "Address
to m e Heathen," October 23 (Winter 1982/3), 89-163 and "The Power of Movies,"
Daedalus 114:1 (Winter 1985), 79—103. Under the rubric of natural generativity,
a similar, more intricately defended view of picturing can be found in Flint Schier's
Deeper Into Pictures (Cambridge: Cambridge University Press, 1986). In his Image and
Mind, Gregory Currie also aligns himself with a naturalist account.
For a report of the empirical findings of Julian Hochberg and Virginia Brooks, see
their "Pictorial Recognition as an Unlearned Ability?" American Journal of Psychology
75:4 (Dec. 1962), 624-8. On cross-cultural pictorial recognition, see Jan
B. Deregowski, Illusions, Patterns, and Pictures: A Cross-Cultural Perspective (New
York: Academic Press, 1980), and Paul Messaris, Visual Literacy: Image, Mind and
Reality (Boulder, CO: Westview Press, 1994), esp. 60-4.
Chapter Five
Moving Images —
Cinematic Sequencing
and Narration

In the previous chapter, we discussed the single motion picture shot, with
primary emphasis on the single shot as taken from a static camera position.
However, most motion pictures are comprised of more than one shot.
Typically shots are combined into sequences. They are added together in
order to perform certain functions. Paradigmatically, that function has
been to narrate a fiction. But there are other possible functions as well,
including: to depict an environment, as in a travelogue or a nature film; to
make an argument, as in the case of Fahrenheit 9/11; to draw a comparison
or contrast; to make a comment; to arouse an emotion; and so forth.
This list, of course, is not exhaustive, nor are the items on it mutually
exclusive.
Cinematic sequences are typically assembled by means of film editing.
Discrete shots, taken from different camera positions, are spliced together
in order to discharge one or another of the preceding functions. The scene
of a man becoming aware that he is surrounded, for example, might be
narrated cinematically by taking a close-up of his face, registering the
emotion of surprise, and then cutting back to a long shot that reveals
him to be surrounded by hostile forces of whom he has just become
cognizant.
However, a sequence like this one can also be built up within a single
shot via camera movement — the image can go from the close-up to the
long shot by moving or tracking the camera backwards, thus achieving an
effect that may be functionally equivalent (in terms of the information
conveyed) to cutting between two different camera positions, but while
continuously recording the space traversed between the view from close
up and the more distant one. This is sometimes called a sequence shot.
 MOVING IMAGES 117

In editing, the space between the close-up camera position and the long
view is excised. Hence, the idea of a "cut": something has been cut out.
In camera movements, which were once called "traveling shots," the
"space between" is exhibited.
Furthermore, the preceding sequence of images — from a close camera
view to a far camera view — might be secured by action within the camera, as
when one zooms out using a telephoto lens. Whether by editing, camera
movement, or lens movement, cinematic sequencing — the coherent shifting
of discernibly discrete camera views — is a fundamental feature of most
motion pictures, save single-shot films made from a fixed camera position as
in the case of many Lumiere actualites.
This chapter is divided into two parts. First, we will be concerned with
understanding cinematic sequencing — that is, with how changing camera
positions facilitate motion picture communication. Though editing is a
primary example of this, we will also consider the ways in which camera
movement and the deployment of various types of camera lenses can also
contribute to this effect. Then, after examining how sequences are con-
structed, in the second part of the chapter we will focus on some of the
ways in which sequences are combined, most notably in terms of the most
common form of movie narrative.

1 Cinematic Sequencing
Cinematic sequences are composed of views taken from different vantage
points whether by means of editing together shots from discrete camera
positions or by means of camera movements that shift our perspective on
the action, or via the adjustment of camera lenses that have functionally
equivalent consequences. By way of sequences, moviemakers communicate
action, events, and states of affairs as well as often the attitudes and
opinions they take toward the circumstances they portray. But how is
communication possible through a disparate array of views?
We have already encountered a putative answer to this question. It is
perhaps one of the earliest accounts, one popularized by the Soviet
montage school of film theory. We heard a clarion statement of this view
in the last chapter where V. I. Pudovkin declared: "[Film] editing is the
language of the film director. Just as in living speech, so one may say in
editing: there is a word: the exposed piece of film, the image [the shot]; a
phrase — the combination of these pieces." Call this the "film language
hypothesis." How plausible is it?
118 MOVING IMAGES

The film language hypothesis

Proponents of the film language hypothesis are struck by certain very
broad analogies between sentences, on the one hand, and cinematic
sequences, especially edited arrays, on the other hand. A sentence is
putatively composed word by word, the combination of meaningful
units resulting in a larger meaningful complex. Similarly, a shot chain is
composed of independently sense-bearing shots which when combined
yield larger meanings: the shot of the man, followed by the shot of the
barrel of a rifle firing, followed by the shot of the man falling commu-
nicates the narrative that the man has been shot. Of course, the story
could also be told by moving the camera from the man to the rifle and
back, or by zooming, in a comparable manner, between these viewpoints.
Thus, it appears that sentences and cinematic sequences communicate —
that is, they make sense — by combining smaller, meaningful segments into
larger ones.*
Sentences are meaningful in virtue of being built up, via recursive
operations, from smaller meaningful units like nouns and verbs. This
feature of sentences is called compositionality. Linguistic meaning relies
upon syntax, a set of rules for composing meaningful wholes, a.ka.
sentences, out of words. A language has a finite vocabulary of arbitrary
symbols that can, in principle, be organized by recursive, grammatical
operations into an infinite number of sentences.
Butj as we saw in the previous chapter, shots, save in extraordinary
circumstances (e.g., shots of street signs), are rarely word-like. Shots may
signify, but they are typically nothing like words. There is no vocabulary of
cinema; no finite dictionary of motion pictures. There are as many motion
pictures as there are things and combinations thereof to photograph from
an indefinitely large number of camera positions.
The nose of a woman in a motion picture shot is very different from the
word "nose" in a sentence, because the word "nose" bears an arbitrary
relation to its referent, whereas the woman's nose in the film has a natural

Though it may be easier to think of the individual shots in edited arrays as

discrete units akin to words, it is far more difficult to know how we could parse
different moments of a camera movement as discrete units. Perhaps this should
count as a reason to abandon the film language hypothesis from the get-go, viz., it
cannot handle all cinematic sequences, notably functionally equivalent ones, in the
same way. Indeed, maybe it cannot handle some of them at all, specifically camera
movements.
 MOVING IMAGES 119

relation to its referent. Moreover, the picture of the woman's nose is

not decomposable into smaller units as the phrase "the woman's nose"
is decomposable into "the," "woman's," "nose." They are all of a piece, so
to speak; they are not atomic particles of a larger, recursively assembled
meaning-compound, as is a sentence. So a meaningful shot is not con-
structed or composed in the manner of a meaningful sentence. The shot is
not a minimal unit of meaning in the same way that a word is, and>
furthermore, for the same reason, neither are strings of shots built up
after the fashion of sentences. There are no comparable minimal units of
meaning in motion pictures, so cinematic sequences are not of a piece with
sentences.
Shots, whether one by one or in combinations, are not appropriately
equated with words. Consequently, film lacks the kind of discrete building
blocks that make linguistic communication possible. Even if cinema had the
kind of recursive rules that languages possess, motion pictures standardly
would have nothing for them to operate upon, lacking as they do anything
that functions like words. Motion pictures, pace Pudovkin, have no
proper semantics — no discrete, arbitrary, recombinable, minimal mean-
ingful units — for a syntax to organize. There is no vocabulary of film.
So, there is no compositionality.
But not only is there no semantics of cinema; there is no grammar
either. Pudovkin seems to think that film is a combinatory system exactly
akin to a language. Shots are supposedly assembled by editing in the way
words are organized into sentences. This suggests that he thinks film, or at
least film editing, possesses something like a grammar, i.e., recursively
applicable rules for ascertaining what is and is not well formed. But cinema
possesses no such rules. So, in addition to being without anything like a
requisitely structured vocabulary, motion pictures, as we know them, are
also without syntax.
Film instructors, teaching courses with titles like "The Language of
Sight and Sound," or "The Grammar of Cinema/' may go on and on about
rules — such as "Don't cross the one hundred and eighty degree line"
(because it will flip the direction of movement on screen) or "Don't edit
together shots from virtually the same but somewhat discernibly different
camera positions" (because such a cut; called "a jump cut," will cause the
appearance of an abrupt tremor in the image). But, in truth* these are rules
of thumb, not grammatical rules. The test of whether a stretch of cine-
matic sequencing is successful is whether or not it works — whether or not
it (communicates, whether or not an audience can follow it, whether or
not it achieves the effect the image maker intends. For every alleged
120 MOVING IMAGES

"grammatical" rule of filmmaking, one can come up with cinematically

distinguished counterexamples.
John Ford, for instance, violated the one hundred and eighty degree rule
in Stagecoach and yet the scene in which this occurs is perfectly intelligible
to spectators. No one has ever complained that it was ill formed. They
understand that the stagecoach is moving in the same direction after the
cut; they do not think it makes a hairpin volte-face. Audiences are able to
do this, moreover, because this understanding best coheres with their
conception of the rest of the narrative.
Grossing the one hundred and eighty degree line is sometimes used, as
David Bordwell points out, to give a movie a realistic, rough-edged look,
as in the case of Homicide: Life on the Street, or it may be used to arouse a
feeling of tumult in audiences, as in the case of 8 Mile. In Half Nelson, the
camera crosses the one hundred and eighty degree line when the character
Drey is riding her bicycle in order to highlight the dramatic tension and
transition she is undergoing. In a film about how opposites bring about
change, there is a related scene later in the film where Drey is riding her
bike in one screen direction and her teacher is driving in the opposite
direction. Though not strictly a matter of crossing the one hundred and
eighty degree line, the editing here shows how opposing screen directions
can be used to mark thematic points, a potential exploited in the earlier
scene just mentioned in which the one hundred and eighty degree line is
literally crossed.
Similarly, Jean-Luc Godard employs jump cuts in a number of
films, including Breathless, where they function as a sort of expressive
punctuation, signaling the improvisatory temper of the film and its film-
maker. Viewers do not interpret these edits as rents in the fabric of the
universe, but as signs of authorial exuberance. The jump cuts are perfectly
intelligible to viewers, who understand them as the fingerprints of the auteur.
Nor is the jump cut merely the gesture of a single moment of motion picture
making. The use of jump cuts for expressive purposes appears in contem-
porary movies such as Ridley Scott's Matchstick Man.
In language, we can distinguish between a string of symbols being intelli-
gible and its being well formed (or grammatical). We can say that we
understand a young child's ungrammatical utterances and that we can interpret
garbled telephone conversations with missing parts, even if neither consists
of complete sentences. For there is a distinction between intelligibility and
grammaticality in language. But with motion pictures, intelligibility is the
whole story.
 MOVING IMAGES 121

Moreover, insofar as cinema is an artform, motion picture makers need

not regard any rule as inviolable. In principle, any of the so-called rules of
thumb of cinema can be forsaken, so long as the edited array is successful —
intelligible to the relevant spectators. Cinematic sequencing, though it
possesses certain tried-and-true strategies, lacks a specific syntax, and,
therefore, there is no grammar of film in the strict sense of the term,
despite the fact that you might pick up a book entitled The Grammar of Film
in a used book store.
Just as we speak of the art of poetry rather than the grammar of poetry, so
we should think of cinematic sequencing more in terms of art rather than
syntax. We do not chide a poetic effect if it departs from grammar, so long
as it is effective. Think of the work of e. e. cummings. Similarly, when it
comes to the art of cinematic sequencing, everything is a matter of artistic
effect, since there are no genuine grammatical rules in this vicinity to
speak of.
When confronted with an edited array in a movie — say a shot of
one man walking, cut against the shot of another man walking — we infer the
significance of the shot chain on the basis of what makes sense given the
context (which is generally narrative). If we know, given the rest of
the story, that the two characters are in different countries, we infer that
they are both taking their walks at approximately the same time of day.
But, if we know that they are due for a showdown on the streets of Dodge
City, then we infer that they are probably in the same locale, heading
toward each other. The shot chains may look exactly alike, but we will
infer different meanings depending on the best way of making it cohere
with whatever else we know about the rest of the fiction.
But, it must be stressed that we have no rules to which to resort in
cases like this. Rather, we depend upon holistically interpreting the array of
shots in context; we do not read the cinematic sequence. The notion
of "reading a film" is at best a metaphor, and a misleading one at that.
We process the flow of information delivered to us by cinematic sequenr
cing through an iterated series of hypotheses to the best explanation where
our abiding concern is the search for coherence. This involves something
quite different than the exercise of rudimentary reading skills. Thus,
inasmuch as a certain sort of inference, rather than anything resembling
basic reading, underwrites cinematic sequencing, and especially editing,
cinematic sequencing does not appear to be linguistic. So, once again, as
we concluded in the previous chapter, the film language hypothesis is best
rejected.
122 MOVING IMAGES

The attention management hypothesis

But if cinematic sequencing does not communicate after the manner of a
language, how does it do so? An alternative hypothesis is that it does this by
what might be called attention management. Motion picture makers
communicate to spectators by controlling their attention. Through cine-
matic sequencing, the moviemakers select what the viewer sees as well as
the order in which she sees it along with the duration of what she sees and
its scale, The moving picture maker articulates her own intentions to the
audience by guiding our attention. We attend to what the filmmaker
displays, from the angle the moviemaker elects, at a screen size propor-
tionate to her choosing, for as long a time — or as short a time — as she
deems appropriate. Then we fit these details together with the story (or
the argument, etc.) that we have been prompted to construct so far from
what we have been shown.
The motion picture maker determines the object* the acuity, and
cadence of our attention as well as its precise place in the relevant
succession of images. The moviemaker both holds our attention and shapes
it by controlling the objects of our perception in addition to the order,
pace, and scale at which we peruse them. The constant shifting of camera
views draws our attention to the screen insofar as our perceptual system is
keyed to change and movement. And once our attention is riveted to the
array, the alternative, successive framings of the situation guide us through
it in accordance with a highly deliberate pathway.
The motion picture maker communicates with us by means of the
aforesaid factors — such as the scale and duration of the camera view —
whose modulation not only tells us what it is important to see, but
does so in an ordered succession that alerts us contextually to the
significance of the relevant items in the emerging configuration, whether
the configuration be that of a narrative or an argument or something
else. That is, a very fundamental level of cinematic communication
involves the motion picture maker articulating her intentions by directing
and molding our attention in terms of cinematic sequencing — showing
us this rather than that in a certain order, and at a certain velocity
and scale.
Needless to say, manipulating just these variables is not the whole story
of cinematic communication. There are many other important contribut-
ing factors — including, as already noted, the audience's sense of the
overarching direction of the motion picture, whether in terms of a story,
an argument, or some poetic vision. Undoubtedly, we use our emerging
 MOVING IMAGES 123

sense of the whole, generally abetted by language (dialogue, voice-over,

subtitles, and/or intertitles), to contextualize the array of details that the
shifting camera viewpoints in the cinematic sequencing bring to our
attention. However, for the narrative or argument to emerge in the course
of a motion picture, the audience must typically be shown images in an
appropriate order, cadence, and scale, namely ones that are pertinent to
the communication that the motion picture maker intends to deliver.
Moreover, in general, what the image track brings to the viewer's attention
should ideally reinforce or be otherwise coordinate with the dialogue and/or
voice-over, unless what is intended is ironic, comic, or a gesture of
avant-garde subversion.
The construction of the cinematic sequence — the flow of disparate
images (or shifting camera views) — is executed by the movie editor, if it is
a matter of cutting, or by the cinematographer, if it is a matter of camera
movement or lens movement. Such construction is typically done in
conjunction with the motion picture director. Working in concert^ the
motion picture production team fixes the cinematic sequence, determining
what will be shown, in what order, at what pace, and at what screen scale.
Where successful, these cinematic sequences strike us as intelligible com-
munications not because — as the film language hypothesis would have it —
we have learnt some sort of grammar which we use to render them
coherent. Rather, to a substantial degree, the cinematic sequencing engages
features of our cognitive and perceptual make-up that were already in place
before our first exposure to moving pictures and mobilizes them to make
sense of the array.
For example, in order to underline the importance of an element in the
storyworld — say a gun in a mystery in which it figures as a valuable clue —
the filmmaker moves in for a close-up. In this way, we know to look at the
gun; indeed, we have no real perceptual alternative. And we know
that the gun is important* not as the result of some language-like code,
but because (1) it is the only thing to see and (2) its scale has been
magnified to the point that it occupies the whole screen; it has been writ
large, so to speak. Nor is the emphatic scale of the gun as a signal of its
narrative weight an arbitrary association. It is rooted in perception where
there is a natural correlation between bigness and importance that can be
observed cross-culturally at rates of occurrence that are far from random.
That is, our disposition to look at the gun in the close-up is not
something we need to learn. It happens automatically. First, it is the
only thing visible. Moreover, that making something large is a signal of
importance — a way of visually announcing "Look here now" — is no more
124 MOVING IMAGES

of an arbitrary communicative strategy than is raising one's voice for

emphasis. We use both techniques in communicating with infants. That
these techniques work strongly suggests that humans are innately suscep-
tible to their influence — that we do not acquire them as we do linguistic
codes. Indeed, the application of techniques like this would appear prior to
the acquisition of any particular language just because they are among the
strategies that facilitate language acquisition.
The preceding close shot signals that the gun is important. But why?
Perhaps the next shot pulls back to show us a detective and his partner
leaning over the gun at a crime scene. This second shot reframes the
situation in order to lead us to consider the relevance of the discovery of
the gun to the investigation. Refraining the elements of the situation in this
way — in this case, by placing the gun in a larger gestalt or narrative
configuration — strongly inclines the spectator to ponder and then probably
to infer the role of the gun in the larger narrative context. The shifting
camera position achieves this by directing one's attention from the gun
back to the ongoing investigation. Next, when the detective notes some-
thing special about the gun — perhaps the initials on its handle — the camera
cuts or otherwise moves in to a closer view in order to fill our attention
with this telltale clue.
In this example, the cinematic sequence moves from close shot to broad
shot and then back to a closer view, perhaps one taken at a slightly different
angle than the first one. By constantly refraining the scene, the filmmakers
assure, by exploiting our natural perceptual tendencies, that we will be
attending just where we need to be attending in order to follow what is
going on — in order to be apprised of the detail or details that we require in
order to be prepared to make the inferences necessary to move us forward
as the story moves forward. Or at least this is what transpires in the
standard case, as opposed to intentionally disjunctive cases, such as Le Chien
Andalou. That is, typically the filmmaker selects various details and shows
them from various angles and distances, choosing camera positions in such
a way that the audience member is attending precisely to what she needs to
consider so as to keep track of what is happening in the story or the
argument. We can call this process variable framing.

Variable framing
Cinematic sequences are built through variable framing which exploits our
natural perceptual dispositions in order to guide our attention to where the
motion picture maker wants it to be — usually to whatever is important to
 MOVING IMAGES 125

the unfolding of the action or argument. That is, it is through variable

framing that the moviemaker manages the audience's attention.
So far examples of variable framing involve shifting from close shots to
broader shots. However, variable framing also occurs when the camera
changes its position on the action without changing its distance from it
appreciably. For example, in the celebrated mortar sequence in Buster
Keaton's The General, there is a medium shot of the cannon lowering its
muzzle, taken from in front of the artillery piece, and then there is a shot
at roughly the same distance from behind the gun, underscoring the
importance of this state of affairs in the narrative — namely, that the mortar
is aimed dead-on at our hero.
Variable framing is a matter of changing camera positioning in order for
the motion picture maker to ensure that the audience is attending to the
action exactly in the way the moviemaker intends. This is a lever of control
over the audience's attention that has virtually no parallel in theater. In the
theater, the audience, for the most part, needs to guide its own attention as
to where to look. This is why theatergoing requires the acquisition of certain
skills, including disciplined attention, that are virtually unnecessary when
encountering the typical movie. In theater, one must finally find one's own
way through the spectacle. One must make one's own perceptual pathway
through all the finery and the stage flats and the extras milling about, not
only to fix on the heroine, but also not to be distracted by everything else
there is to see. It takes a deliberate effort, especially for those unaccustomed
to theater. In motion pictures, even for neophytes, this generally happens
virtually automatically due, in large part, to variable framing.
How does variable framing work?
Variable framing is fundamentally a matter of changing the position of
the camera on the action. Typically, changing the position of the camera
provides the filmmaker with three distinct levers for directing or influen-
cing the audience's vector of attention. These three levers — the variables of
variable framing — can be labeled: indexing, bracketing, scaling.
As we have seen, editing, camera movement, and "lens movement" are
the best-known mechanical bases for variable framing (though methods
which look just like these can be contrived via animation, whether manual
or digital). These techniques, in turn, provide the motion picture team
with three formal instruments for influencing or directing the audience's
attention. These formal variables, as already noted, can be called indexing,
bracketing, and scaling respectively.
Indexing occurs automatically when the camera is brought closer to its
subject, whether by means of a camera movement, a cut, or a zoom. It
126 MOVING IMAGES

functions like the gesture of pointing. One might think of it as a form of

cinematic ostension. When the camera moves, that indicates that the
spectator should be looking in the direction the camera is headed, if the
camera's movement is being recorded, or in the direction toward which
the camera is aimed, if the shot has been presented by means of a cut.
In Casablanca, there is a scene where Elsa visits Rick in order to plead
with him to sell her a set of exit visas. The relevant shot begins with Rick
opening the door to his office screen right, and then the camera pans
screen left to show us Elsa by the window. As the camera travels from left
to right, our eyes follow it irresistibly. Its movement points us in the
direction we should be looking. Indexing is not an arbitrary, language-like
convention for where we should look. It is keyed to the way in which our
perceptual system works. It is a technique that manages or guides our
attention automatically, even naturally.
In addition to indexing, the filmmaker also directs our attention by
bracketing what we see. That is, as the camera is moved toward an object,
a character or an event, it effectively screens out everything beyond the
perimeter of the frame. By moving a camera toward an object, either by
means of editing or camera movement^ the filmmaker indicates that what is
important at this moment is what is on the screen. The frame acts as a
bracket. Everything outside the frame has been bracketed from attention.
On the one hand, bracketing is exclusionary. What does not fall inside
the cinematic bracket, in effect, should not concern the audience's atten-
tion. But bracketing is also inclusive. The bracket indicates that what is
important for the audience to consider is what falls inside the bracket —
that which is visible within the cinematic frame. Furthermore, that we look
inside the bracket, rather than outside it, is hardly an arbitrary, language-
like convention. It does not need to be learned. For we have no alternative
but to abide by the placement of the bracket. Bracketing is not an
unmotivated element in an artificial code. Rather, the technique is tailored
by its very nature to compel viewers, endowed with standard-issue per-
ceptual equipment, to attend to just what the director intends, as signaled
by variables like bracketing.
In terms of cinematic sequencing, most editing and camera movement
function by employing the exclusionary and the inclusive dimensions of
bracketing simultaneously — both to eliminate distracting detail and to
focalize attention on the center of the bracketed area. Thus, bracketing
both filters out irrelevancies and spotlights what is significant.
However, in some cases (indeed, in enough cases that this option should
be regarded as a standard deviation), a motion picture maker may
 MOVING IMAGES 127

deliberately place the bracket in such a way that what may be most significant
for the narrative is left offscreen, perhaps in an adjacent space. This is often
done for emotional effect. Thus, the approach of the murderer is frequently
hinted at onscreen, but the killer is "hidden" outside the bracket, as is the
case in early scenes in Fritz Lang's classic M. We hear the child-killer, but we
don't see him. This raises a certain frisson in the spectator and keeps us
riveted to the screen in the expectation that sooner or later we will get to see
the murderer — that what is offscreen will come onscreen.
Moving the camera forward not only indexes and brackets certain
details rather than others. It also changes the scale of what the audience
is looking at. In Hitchcock's Notorious, there is a famous crane shot that
begins on a second-floor landing and edges over behind a chandelier before
it swoops down toward the character of Alicia Huberman, finally landing
on her hand, where she has secreted the cellar key that she plans to give to
the American agent, Devlin. As the camera moves in on Alicia, it not only
points to her and removes surrounding, distracting detail from the most
significant element of the shot, but it also makes the key larger. That is, the
key literally occupies more screen space at the end of the shot than at
earlier moments. As it becomes visible, it "grows" in scale, taking up more
screen space and more retinal space to boot. Moreover, the increase in scale
alerts the audience not only to the presence but also to the importance of the
key in the scene.
Although someone might be tempted to say that it is a convention that
visual largeness should come to mark importance, it is difficult to agree
that this is purely arbitrary; Would it be equally effective psychologically to
make what is important in a scene small? But then how could a director be
assured that the audience would take note of it? By means of scaling, the
director fills the audience's awareness with whatever she deems appropriate.
Likewise, scaling enables the director to make what is less significant to the
narrative smaller and, therefore, a less likely object of audience attention.
This is not merely a matter of marking an item as less important. It is a way
of making it less probable that the audience will be looking at it or that
they may be able to see it. Scaling may be a "conventional" device in that
it is frequently used in film; but it is not conventional in the sense of
the arbitrariness that we speak of with relation to language, since, given the
human perceptual make-up, it is not arbitrary that we attend to the objects
that dominate our visual field nor that we are less likely to attend to
objects that are not prominent.
Scaling, bracketing, and indexing are three different ways at the director's
disposal to make objects and events, characters, and actions salient to the
128 MOVING IMAGES

audience. To a large extent, through these means, the director manages or

guides the audience's attention from moment to moment to where it is
most appropriately fixated, given the purposes of the ongoing narrative or
argument.
Moreover, these devices help the filmmaker modulate the audience's
attention over time. First the audience is shown a medium long shot of a
scene and then the camera moves in, and, by means of indexing, scaling,
and bracketing, certain details are selected out for special attention. Or,
the sequence may start with a closer shot and then the shot is changed to a
broader view, encouraging the audience to think of the detail presented in
the close shot in terms of its relevance to the larger gestalt encompassed by
the altered bracket — for example, first we see a close-up of the simmering
fuse on the end of a stick of dynamite, indexed, bracketed, and made large-
scale, and then we see the larger context — the targeted bridge exploding —
in the broader shot that follows.
Constant reframing permits the director to control not only what we
see, but when we see it in terms of the sequence of views that is most
advantageous for our (usually inferential) processing of the story or the
argument that she is developing. Variable framing, that is, is the primary
means by which a director directs. And what does a director direct? The
audience's attention.
Typically, scaling, bracketing, and indexing come into play whenever the
camera is moved. However, they need not always occur coincidentally.
One may change the bracket without changing scaling — for example, when
a camera follows a character but maintains her screen dimensions. Or one
may index a scene by moving the camera forward slightly without appre-
ciably altering the scale or the bracket of the shot. Or one may emphasize
the changing bracket in a shot, as in the famous trolley-car ride in Sunrise,
without indexing anything in particular, or introducing notable scale
changes. Nevertheless, scaling, bracketing, and indexing are generally
employed in concert and, when so employed, they afford a very powerful
means through which the motion picture maker controls the audience's
attention, delivering to them with great clarity what they should be
thinking about with respect to the larger evolving direction — whether
narrative or otherwise — of the motion picture.
In the typical motion picture, these means are placed in the service of
the ongoing narrative. Bracketing, scaling, and indexing are generally
deployed in such a way that the first item or configuration of items in the
shot of which the audience is induced to take note is that which is most
pertinent to following the story as it is emerging. These devices make
 MOVING IMAGES 129

certain narrative details — characters, objects, actions and parts thereof —

salient; they make us cognizant of what is (ideally) maximally significant for
us to attend to in the story at the moment the shot appears. And then we
are presented with new narrative information in the next shot, which has
also been structured in terms of the principle that thefirst item or group of
items that grabs our attention again be the one that is most relevant
to tracking the progress of the narrative. Furthermore, as a consequence
of this coordination of narrative purposes with visual structure, most
typical movies are not only easy to follow; they seem perspicuous through
and through.
Of course, for various reasons, a motion picture maker may deviate
from the principle that the first item the audience is prompted to see be
the one that is the most significant for the narrative. Frequently, a film-
maker, perhaps for expressive effect or emotional impact, may lead the
audience's eye so that the most significant narrative detail is not evident
immediately, but only after a second look. For example, we might see a
jittery character first, and, only after first taking him in, are we prompted
to notice that in the background there is a man with a knife standing
ominously behind him. This kind of standard deviation itself, however,
owes much of its forcefulness to the fact that the basic approach to visual
narration in motion pictures is to show us the most diegetically significant
item in a shot first — that is, to coordinate all the visual strategies available
to the director, including bracketing, scaling, and indexing, in such a way
that the most important narrative item is the first thing upon which our
attention battens. Nor is this principle a grammatical rule, since the art of
storytelling is often enhanced by initially diverting the audience's attention.
Furthermore, if the film is not a narrative, but, for example, an
argumentative nonfiction film, the same principles of cinematic sequencing
generally apply. Even avant-garde films may employ these principles,
though it can often be difficult to divine the overarching point of directing
our attention to this or that item in motion pictures of this sort. And, of
course, some avant-garde films may subvert these principles for the very
purpose of drawing our attention to them. Indeed, such exercises may
accrue a large portion of their fascination by baring the devices, such as
variable framing, that make mainstream motion pictures so accessible to
viewers with little or no prior exposure to them.
If we compare the stage drama to the ordinary movie drama> we might
say metaphorically, that, in terms of attention, the narrative in the motion
picture has been visually predigested for us. In theater, we must ultimately
locate and select the objects of our attention; we have to find the objects
130 MOVING IMAGES

that are most relevant to the ongoing narrative. But with movies, a great
deal more of that work is done for us by the variable framing than is done
for us in the staging of the typical play, housed, as it usually is, under the
proscenium arch with its rigidly fixed dimensions. Think of how much
more difficult it was to take in the first plays you saw in comparison with
the first movies you encountered.
With motion pictures, the right standpoint on the action is delivered
automatically to the viewer most of the time, especially in those movies
made for mass distribution. That is why, as suggested previously, motion
picture viewing involves less effort than most theater viewing. It is also
why movies may strike viewers with little experience of live theater as
more readily intelligible than staged dramas, inasmuch as each cinematic
moment is such that it shows in bold reHef what is necessary to see in order
to grasp the emerging story.
Of course, variable framing does not operate without substantial input
from the spectator. Not only must the spectator follow the perceptual
trajectory that is mapped out by the variable framing, but she must also
assimilate what she sees, in combination with what she knows about the
evolving story, in order to make inferences about what is going on. In Fritz
Lang's You Only Live Once, there is a scene about which the spectator infers
that heroine is on the verge of suicide. This is never stated outright.
Rather, the viewer has to put together the details — that the variable
framing selects out and makes salient — with what is happening in the
storyworld. As the clock tells us that the time of her husband's execution is
at hand, we ask ourselves why we are being so emphatically shown the wife
filling a glass with water and then adding an unidentified white powder to
it. Suicide, we surmise. Variable framing leads the viewer by highlighting
the ingredients we need to take into account in order to infer what is going
on; but generally it remains up to the viewer herself to assemble these cues
by making the desired or proponed inferences.
Here we see that although the variable framing makes certain inferences
likely — indeed, often virtually unavoidable — it still usually depends upon
the viewer to arrive at the hypothesis that completes the thought the
filmmaker intends to convey. Furthermore, what is interesting about the
way in which the viewer puts this information together is that it does not,
for the most part, rely on special cinematic and/or language-like codes and
conventions, but instead upon the same sort of inferential and interpret-
ational processes — including abduction, inference to the best explanation,
and practical reasoning — that we employ to negotiate situations in every-
day life.
 MOVING IMAGES 131

Think about what goes on in your mind when you chance upon a crime
scene in the course of your daily affairs. First, you begin to notice that
there are a number of cops milling about; then you observe other details —
police barriers, photographers, and the like. Gradually you colligate these
facts under the concept of a crime scene. But isn't it the case that a
comparable mode of thinking is thrown into gear when Lang presents the
wife's near-suicide to us in You Only Live Once?
In order to process a film narrative, our default assumption is that we
should fill it in with the kinds of beliefs and inferential strategies that
we employ in everyday life. Of course, this is only a default assumption. It
may need to be waived under certain conditions. Specific genres, for
example, mandate frequent departures from the reality principle in order
to make sense of them — for instance, with respect to vampire films, we are
meant to assume that something that looks very like a man, such as Count
Dracula, is nevertheless impervious to bullets. Or a particular film may
stipulate that something that is physically impossible is a fact of life in its
storyworld — perhaps that one can travel through a black hole and come out
the other side intact and alive.
In other cases, the cultural or historical origins of the motion picture
may dispose us to modify our assumptions about the storyworld in order to
understand what a motion picture maker, whose beliefs diverge from our
own, intends to convey. However, we usually begin computing the narra-
tive of a motion picture by deploying our own working cognitive stock,
unless alerted to do otherwise or unless our presumptions are evidently
culturally or historically inapposite. Undoubtedly we do this because it is
our default assumption in any of our dealings with our conspecifics. And
though we often do need to adjust our interpretive assumptions to
exceptional circumstances like those just canvassed, it is surprising the
extent to which the default assumption is almost as reliable with respect to
many, many aspects in movies as it is in life.
That much of the reasoning that we mobilize while following most
motion pictures is continuous with ordinary, everyday reasoning — within
the context of an ongoing narrative — helps explain how it is possible for
films to travel so easily internationally: for The Last Samurai to be a hit in
Tokyo and for Fists of Fury to flourish among audiences from Harlem. For, it
is not necessary to learn a special code to follow film narratives. To a large
degree, motion pictures can be understood by means of ordinary reasoning
and the recruitment of our standing concepts and beliefs, including folk-
psychological beliefs about the recurrent scenarios of human motivation
and action.
132 MOVING IMAGES

With respect to mass-market movies, the variable framing of cinematic

sequencing serves up hmpidly the details the audience needs to input
into its standing cognitive and emotive routines which we then weave
into a coherent fit with what we already know has occurred in the motion
picture and perhaps with what we anticipate about the future of the story-
world. The variable framing operates automatically as a way of directing our
attention to what is pertinent, and then those details are assimilated to what
we already know and expect of the narrative primarily by means of engaging
cognitive and emotive structures that are already in place.
Cinematic sequencing is a process of attention management that
employs variable framing to keep the audience focused upon what it
needs to know and when it needs to know it in order to follow the
emerging point or points of the motion picture. Where the motion picture
is a narrative, the cinematic sequencing feeds the spectator the inputs
she needs in order to perform the continuing battery of inferences
requisite for tracking the story. On the attention management account,
cinematic sequencing relies upon activating innate perceptual responses
by means of variable framing, which framing, in turn, governs the flow of
attention naturally rather than by way of language-like conventions.
Furthermore, this hypothesis fits remarkably well with our hypothesis in
the last chapter regarding the status of the single cinematic image as a
natural recognitional prompt instead of as something akin to a linguistic
symbol. Earlier we argued that one of the primary advantages of movies as
a mass artform derives from the fact that their basic symbol — the cinematic
image — is accessible to every cognitively and perceptually competent
human being; if, in a close encounter of the third kind, you can recognize
at least the type of object that the cinematography portrays, then you
will be able, all things being equal, to apprehend what the cinematic image
is about. This is an important source of the power of movies. That is,
movies are accessible because in large measure they recruit our natural
capacities.
But this is true not only of the way in which the cinematic image
engages perception; it is also true of how cinematic sequencing, by
means of variable framing, shapes our attention. Via variable framing,
the cinematic sequencing has us attending precisely, though automatically,
to the elements in the motion picture array that the moviemaker intends
us to be processing at exactly the moment she ordains. This happens
automatically because of the way in which the variable framing
naturally orchestrates our attention. Moreover, the fact that the standard
 MOVING IMAGES 133

usages of variable framing tap into our natural perceptual dispositions

accounts for the way in which most people follow the average motion
picture with great ease and accuracy. Both with respect to the cinematic
image and cinematic sequencing, mass-market motion pictures, on film
and TV, are, to an appreciable extent, popular and almost universal in
their appeal because they mesh so well with certain human capabilities
rather than being dependent primarily upon communication on the basis
of completely arbitrary conventions or codes.

2 Cinematic Narration

Cinematic scenes and sequences are parts of larger totalities. These scenes
and sequences fit into these larger constellations in terms of the functions
they perform for the given motion picture work as a whole. The
sequences, that is, are subordinate to the purposes of the overarching
work and, therefore, the principles that connect the scenes and sequences
therein are likewise dependent upon the superordinate aims of the overall
motion picture.
Furthermore, motion pictures, as noted earlier, can have a diversity of
agendas. A great many tell stories; some make arguments. Others, such as
Man With a Movie Camera by Dziga Vertov, are involved in extended
explorations of comparisons and contrasts. Still others, such as Cocteau's
Blood of a Poet, portray the inner life of dreams and visions or seek to shock,
as in the case of 11Age d'or, the surrealist masterpiece by Bufiuel and Dali.
With different organizing concerns come different structures for connect-
ing scenes and sequences. What motivates scene linkages in Anticipation of
the Night is different than what makes the scenes and sequences of Sorry
Wrong Number hang together.
Needless to say, an introductory book like this one is not the place to
pursue all the various structures, relative to the disparate programs, that
can be deployed to organize the diverse kinds of motion pictures as
internally coherent entities. However, it may be fruitful to look at one
kind of structure for coordinating sequences, and the variable framing that
organizes them, within certain of the larger structures that render entire
motion pictures coherent. The specific larger structure that I have in mind
is that which determines the shape of the vast majority of popular, mass-
market motion pictures — what many affectionately call movies (whether
they appear on screen or on television).
134 MOVING IMAGES

Erotetic narration
This basic structure of movie narration is an instance of what can be called
erotetic narration. It will be profitable to examine it at some length for
two reasons: first, in order to illustrate how the principles that govern
variable framing cohere or segue with the principles that hierarchically
control the narrative enterprise at large at the level of the entire individual
motion picture work; and second, because the issue of the nature of basic
movie narration is of philosophical interest in and of itself.
How does the typical movie tell its story? First, in contrast, for
example, to a soap opera, the typical movie tells its story in one uninter-
rupted sitting. It begins and then comes to an end, usually within two to
four or so hours after it begins. On the other hand, it is the fervent wish of
the makers of soap operas that their sagas continue on and on. If movies
follow Aristotle's advice with respect to having beginnings, middles, and
ends, soap operas have, as the critic Dennis Porter nicely remarks,
"indefinitely expandable middles."
Moreover, the difference here is not simply that the movie stops
somewhere whereas the soap opera, ideally, plows on perpetually. The
end of a motion picture narrative has an aura of consummation about it; it
yields a quite definite impression of completeness. In contrast, the soap
opera is, by its nature, incomplete. Even if General Hospital were about to
go off the air, it is probably by now practically impossible for a soap opera
like this to tie up all of its loose ends — that is, to resolve all of its
exfoliating plot lines. But ideally we expect a movie to "wrap everything
up." "That's a wrap," the director shouts. In short, a movie is not supposed
merely to come to a halt; a movie is meant to have what theorists call
closure.

Closure
The notion of closure is connected to the sense of finality with which a
piece of music, a poem, a story, or a typical movie concludes. It is the
impression that exactly the point where the work does end is just the right
point. To have gone beyond that point would have been an error. It would
have been to have gone too far. But to have stopped before that point
would also have been a mistake. It would have been too abrupt. Closure is
a matter of concluding rather than simply stopping or ceasing or running
out of steam or crashing. When a filmmaker effectuates closure, then we
feel that there is nothing remaining for her to do. There is nothing left to
 MOVING IMAGES 135

recount. Closure involves a feeling of completeness with respect to the

motion picture. When the last image flickers off the screen, there is no more
of the story to tell, nor has anything that needed to be told been left unsaid.
Or that, at least, is what ideally the viewer feels and is intended to feel.
But what accounts for closure? As already observed, we often say with
respect to a successful movie narrative that "everything has been wrapped
up." But what is meant by saying that a movie has "wrapped things up"?
A tantalizing suggestion can be found in David Hume's essay "Of Tragedy,"
where he points out: "Had you any intention to move a person extremely
by the narration of any event, the best method of increasing its effect
would be artfully to delay informing him of it, and first to excite his
curiosity and impatience before you let him into the secret." For example,
when early on the narrator of Martin Anus's novel House of Meetings makes
a point of saying that he has a letter in his pocket that he hasn't opened for
twenty-two years, the reader automatically wants to know why and also
what is in the letter, and she is willing to stay with the story until she finds
out. Call this Hume's principle. And although Hume is offering this advice
to playwrights, it has also become the creed of screenwriters.
On Hume's view, an immensely effective way of holding on to the
attention of a viewer involves presenting the audience with a chain of events
about whose outcome her curiosity has been perked — about which she wants
to know what happens next — but then hold off informing her. For example,
arouse in an audience the desire to know whether the secret agent will escape
the dungeon and destroy the paralyzing nerve-gas machine, or whether the
evil genius will victoriously carry out his plans to enslave the world — and then
withhold the answer to that question until the last reel of the film.
One can test the reliability of this technique — as well as the hypothesis
that curiosity is the glue that keeps the audience stuck to the screen — by
shutting the projector down just before the final reel unravels. The
audience will jump to its collective feet and demand to know whether
and how the secret agent saved the world. For those are the questions that
had been organizing their viewing activity.
What Hume calls a "secret" we may more prosaically refer to as a
question or a set of questions to which the audience expects answers. And
using this suggestion, we can propose an account of how closure is secured,
or, more colloquially, how the filmmaker "wraps things up." To wit: the
filmmaker wraps the story up when she has answered all the questions that
have stoked the audience's curiosity. Those questions, needless to say, do
not come from nowhere. They have been planted by the filmmaker in a
way that ideally grabs hold of the attention of the intended audience. These
136 MOVING IMAGES

questions won't relax their grip on us as the story moves forward. Closure
transpires when all the questions that have been saliently posed by the
narrative get answered. It is the point at which the audience can presume
that everyone lived happily ever after and leave it at that.
The film Totosi revolves around the question of whether the infant who has
been accidentally kidnapped by Totosi will be returned to his parents. When
we learn the answer to that question, the film is over. We do not ask whether
the child got into Harvard, since that is not a question the film encouraged us
to ask. The impression of completeness that arrives with closure derives
from our sense that all our pressing questions have been answered.
Moreover, this account of closure gives us an important clue to the
basics of the way in which plots in movies are constructed. They are
constructed by generating questions about the storyworld of the movie
which the movie then goes on to answer. At one level, the plot is a network
of events and states of affairs held together by the cement of causation. Yet,
at another level — namely, the level of rhetorical address — a typical movie
narrative is a network of questions and answers, where the questions are
self-generated but then finally resolved.
In order to pith the narrative structure of a typical movie, it is useful to
begin at the end. Ascertain what questions are being answered at the
conclusion of the film, and then work your way backwards to the scenes
and sequences where those questions were introduced, partially answered,
or otherwise sustained, refined, transformed, mutated, and so forth. What
results is the skeleton of the plot with the animating questions functioning
rather like the joints.
Some scenes and sequences set out the conditions necessary for certain
questions to take hold. These scenes themselves, including most of intro-
ductory or establishing scenes in a movie, generally involve answering the
standing questions that we have when presented with a description or
depiction of any set of circumstances: where the action is set, and when,
who these people are and how they are related, what they want, who they
are for and against, why they behave thus and such, and so forth.
Some scenes evoke questions; others answer said questions directly.
Still other scenes or sequences sustain earlier questions: the failure to
apprehend the escaped prisoner leaves us still asking whether he will be
caught in a subsequent scene. Sometimes our questions are incompletely
answered: we learn that the culprit limps, but we still do not know who he is.
One question may be answered in a way that introduces a new question
or set of questions. The question of whether or not King Kong can be
stopped is answered when he is gassed on Skull Island, but this raises the
 MOVING IMAGES 137

question of what will happen once he is shipped to New York. Likewise,

after the shower scene in Psycho, the question of whether Marion can put
her life back together is replaced by the questions of whether her murder
will be discovered and her murderer apprehended.
Not all the questions and answers that belong in the unifying network of
a movie are of the same order. Some questions orchestrate our attention to
the emerging story from virtually one end to the other; others organize
large parts of the tale, but not the tale in its entirety, and others are of a
still smaller gauge. Questions that structure an entire text, or, at least most
of it, we can call presiding macro-questions. Two of the presiding macro-
questions of Moby-Dick are "Will Ahab and his crew ever find the white
whale?" and "If they do, will they be able to destroy it?"
When the Pequod encounters other ships that have knowledge of the
whereabouts of the white whale, those scenes contribute to keeping us
bound to the story; they sustain the presiding macro-questions. When we
finally are shown the deadly confrontation and see that all of the crew, save
one, of the Pequod have perished, the story is over, closured. We do not ask
whether Ishmael will open a dry goods store in Gloucester, since the film
(and the novel) have not encouraged us to ask anything about Ishmael that
is not germane to the hunt for the white whale.
Of course, movie narratives are driven from moment to moment by
more than just presiding macro-questions. In the pursuit of some over-
arching goal, the protagonist may confront a local obstacle — for example,
in order to save Christendom from the machinations of Satan, the heroine
may have to retrieve a vial of holy water, perched precariously on a ledge
in a cavern and protected by gargoyles. Will she be able to reach the perch
on which the holy water rests or not? The solution to this problem is
connected to the protagonist's larger project as a means to an end.
Questions of this scale, which govern the structure of one or two
sequences, we may call micro-questions. They are questions whose answer
will eventually contribute to answering a presiding macro-question, but
which do not, on their own, answer the presiding macro-question directly
and completely.
Between the presiding macro-questions and the micro-questions
there may be questions that organize large portions of the movie but
which do not organize the plot as a whole. Because of their scale we can
call them macro-questions, but not presiding macro-questions. In Psycho,
the question of whether Marion will be caught organizes a number of
scenes, but it is not ultimately one of the presiding macro-questions
of the film.
138 MOVING IMAGES

Clearly, there is a hierarchical relation between the questions posed by

an ongoing movie narrative. Micro-questions are generally subordinate
to the presiding macro-questions, and closure obtains when all of the
presiding macro-questions and all or most of the micro-questions that
are relevant to settling the macro-^questions have been answered. More-
over, since it is the function of scenes and sequences to keep the audience
bound to this circuit of questions and answers, the variable framing in each
sequence will generally operate in the typical movie in such a way that the
element or elements that first arrest our attention in successive shots are
those that are most pertinent to raising, sustaining, refining, transforming,
or answering the questions that are driving the scene.

The problems I solutions model

An alternative to the question/answer model for conceptualizing the
erotetic narrative invokes the vocabulary of problems and solutions. Closure,
in this way of speakings occurs when the protagonists have solved all the
problems with which the movie has burdened them — for example, when
the cowboy in white has gunned down the villain and all his henchmen and,
just in the nick of time, swept the heroine off the railroad tracks, thereby
saving her from being crushed by the oncoming locomotive. Obviously, the
problem/solution model applies to a very large number of eases. Yet, the
question/answer model is arguably superior for several reasons.
First, the question/answer framework will apply to every case where
the problem/solution model works, since we can always ask: Will the
protagonist solve her problems or not? However^ there are also cases
where the question/answer model fits, but where the problem /solution
model seems strained. Does it really make sense to regard the question of
whether two people will fall in love a problem? Certainly their falling
in love could involve problems — obstacles, like parents and rivals, to be
overcome, and so on. But it need not. Though all problems may be translated
into questions, it is not evident that all questions can be translated into
problems.* Thus, the question/answer model seems more comprehensive
than the problem/solution model.

In Volver, the neighbor wants to know whether or not her mother is dead. It
seems more natural to describe her request for information as a question rather
than a problem. Moreover, it becomes a question whose answer we discover is
material to securing closure in Volver.
 MOVING IMAGES 139

Furthermore, is the problem/solution model suited to deal with cases

where the problem goes unsolved — where, for example, the aliens do
inherit the Earth as in the 1978 remake of Invasion of the Body Snatchers? That
is, once again, the question/answer model appears more comprehensive
than the problem/solution model, and, for that reason, is to be preferred.

Flashbacks and Jlashforwards

It is a fact about storytelling that the time of the telling need not
correspond to the time of the tale. This is as true of movies as it is of
any other kind of narrative. Though the story is shown in a linear fashion in
real time with the opening credits preceding the end credits, the events
that comprise the story need not be ordered consecutively in that way.
There can be flashbacks and flashforwards.
The temporal order in which events are disclosed in a movie can diverge
from the order in which they occurred in the storyworld. For example, in
Sunset Blvd. we learn that Joe Gillis has been murdered before we learn
who killed him and why. Nevertheless, the temporal ordering of the telling
of the tale is generally not divorced from the erotetic structure of the
movie. For flashbacks and flashforwards typically play a role in raising,
sustaining, or answering questions. The flashback in Sunset Blvd. explains
how Gillis died. Likewise the flashbacks at the end of The Illusionist,
intercut with the inspector's realization that he, along with everyone
else, has been tricked, answer our questions about how Eisenheim's scam
was executed.
Flashforwards also typically fit into the erotetic structure of narration.
The burning motorcycle in Easy Rider prompts us to ask who has died and
how, just as the flashforward in Don't Look Now raises the question of whose
funeral we are seeing. And these retrospectively directed questions keep us
glued to their respective narratives.

How can movies ask questions?

It may seem strange to claim that movie narratives proceed by posing
questions. After all, if it makes any sense to attribute grammatical mood to
motion picture footage (and it probably doesn't), most motion picture
footage, without any further, stipulated conventions, would have to be
classified> if at all, as declarative or indicative. Movie footage, minus
dialogue or inter titles, does not ask questions or issue commands. So
does it make sense to say that movies ask questions?
140 MOVING IMAGES

On behalf of the question /answer model, the first thing to notice is that
questions can and quite frequently do arise from declarative sentences.
When we are told that the mayor has just eluded abduction, it is normal to
ask who attempted to kidnap him and why. Likewise, when we receive
information through media other than language, questions frequently
ensue. When we see a photograph of our parents standing next to people
we don't recognize, we ask who they are.
Question formation on the basis of received information is a natural
thought process. It plays a role in every kind of inquiry. With respect to
erotetic movie narration, we use the questions evoked by the narrative
to organize and to keep track of the representations of events and states of
affairs that the story presents to us. A young farmer's wife is raped by
marauders. He returns home and finds her dead. We ask ourselves what he
will do. He takes a fancy pair of revolvers from their hiding place and
straps them on. We surmise that he is bent on revenge. And this then
arouses the question of whether he will succeed, fail, or possibly relent —
questions that structure our viewing of what is to come.
The evocation of such questions is a matter of a natural thought
process. Though we are speaking of questions here, we are not reverting
to a version of the film language model. As we have already argued
strenuously, there is no grammar of film. In ordinary life when we see
our front door ajar, we naturally ask who opened it. Similarly, in a movie,
when we are shown that the lock on the door of the heroine's house has
been forced, we want to know who opened it and for what reason. It is
not that the movie has a syntactical structure isomorphic to the inter-
rogative in language. It is the narrative that raises the question — not some
celluloid sentence. For, narrative events and states of affairs in fictional
movies can provoke questions, just as events and states of affairs in
everyday life can do so.
In order to understand how the events and states of affairs represented
by the scenes and sequences of narrative movies can evoke questions in
viewers, it pays to remember that many of these events and states of affairs
stand in various causal relations to one another, albeit causal relations of
different strengths. Some of the earlier events in the storyworld are only
necessary causal conditions to the later developments; so rather than being
fully deterministic of future plot developments, they instead open a range
of possible outcomes. For example, once the bridge is mined, it either
will or will not explode as the enemy crosses it; but we do not know
which state will obtain. Thus, we quite naturally come to wonder what
will happen.
 MOVING IMAGES 141

The very insufficiency of many of the causal connections in most

movie narratives incites our curiosity about what will happen. For, though
unpredictable, an outcome is presumed to be in the offing. And, it is
natural to ask, if only tacitly, what it shall be.
It is primarily the causal inputs, broadly construed, in erotetic narratives
that raise the presiding macro-questions and the pressing micro-questions
whose answers secure closure in typical movies. A character forms an
intention, has a desire, a need, a purpose, a goal, or a plan; or she has a
commitment or an ideal, or makes a promise. These motivational states
comprise part of the causal conditions of her action. They also generate
questions about whether her intentions, desires, motives, purposes, goals
and/or plans will be realized. Can she live up to her commitments, ideals,
and promises? Certain obstacles, challenges, or problems erupt. These too
structure the causal context of her action. Can she meet them or will they
defeat her?
A robbery motivates a chase; a kidnap, a rescue attempt; an injury,
a desire for revenge. Will the thieves be caught, the abductee recovered,
and the insult repaired? Someone finds himself in a desperate financial
situation — his bicycle has been stolen and he cannot afford a new one. But
he needs a new one in order to find work to support his family. What
will he do?
Characters form beliefs about the situations in which they find them-
selves. They interpret their circumstances. Othello, egged on by Iago,
comes to believe that Desdemona has been unfaithful. Recognizing Othello's
belief, we ask: "What will he do?" "Will he kill her??
So far, these examples have exploited the mental states of fictional
characters as the relevant causal inputs that frequently generate the kinds
of questions whose answers will eventually bring on closure. However,
there is no reason to restrict the pertinent causal inputs to subjective
states. In the film March of the Penguins, the rapidly falling temperature abets
the presiding macro-question of whether the eponymous penguins and
their progeny will survive.
Of course, not only causal inputs may raise macro-questions, but also
effects. Presented forcefully with certain effects, like the crop circles in
Signs, we will expect answers about their origins. Thus, in addition to
causal inputs, presiding or large-scale macro-questions can be launched
by puzzling events. This, of course, is the way of classical detective films;
no sooner is the murder discovered than we ask: Whodunit? In short, any
question that arises from the causal nexus of the story can function as an
ingredient in the erotetic structure of the movie. For example, Citizen
142 MOVING IMAGES

Kane begins with the question of what "rosebud" portends, while Mildred
Peirce starts off with the question of who killed the title character's
husband.
Moreover, it is the role of the variable framing from shot to shot to
make salient the element or elements in the sequence that animate the
network of questions and answers that compel our attention. For example,
the close-up toward the end of The Illusionist that indexes, scales, and
brackets Eisenheim taking, the gems from the blade of the Archduke's
sword answers our question about the way in which the heir-apparent
has been framed. In turn, the sequence in which this scene is embedded
functions to answer the lingering question the spectator has about the
earlier scene in which Eisenheim and one of his confederates were hatching
a plot. Whatever came of that plot? That is what the flashback involving the
inspector in the railroad station discloses.
Of course, not all motion pictures are narrated in this manner. Not even
all movies are full-fledged, erotetic narratives. For, though they raise
questions, the weekly installments of the movie serials of yesteryear and
the daily episodes of the soap operas of today studiously refrain from
answering all of their presiding macro-questions in order to guarantee
that viewers will return for the next round of plotting.
In contrast to American horror movies, which try to tie up all the loose
ends (if only with forays into fantastic physics), recent Asian horror flicks —
like Ringu by Hideo Nakata and Ju-on by Takashi Shimizu — leave the super-
natural forces not only at large but also scarcely explained — the refusal to
answer the viewers' questions thereby engendering a kind of metaphysical
unease about the mystery of things that is far more chilling and far less
reassuring than Stateside ghost films where the wraiths get sent back to hell
in the course of our learning how they came to be here in the first place.
Or, on a higher plane, one might eschew closure for the purpose of
scoring a more elevated thematic point, as Kurosawa does in Rashomon,
where, by not divulging what happened from the vantage of an omniscient
narrator, the director encourages the viewer to entertain perspectivism as
the best interpretation of the film and, perhaps, of interpersonal relations
in general as well. Some films, such as L'Avventura by Michelangelo
Antonioni, may not answer the questions they put in motion in order to
suggest that not all our questions have answers — that existential meaning-
lessness is a genuine possibility.
Similarly Nashville by Robert Altman defies closure — its ending coming
from nowhere — in order to suggest the utter unintelligibility of America in
the 1970s. That is, narrative incompleteness = chaos.
 MOVING IMAGES 143

The scenes in Satyricon are not connected by a nexus of questions and

answers, but that, of course, is a leading factor in provoking the alien
quality that Fellini sought when he created what he called this "science
fiction" film of the past. Likewise, Orson Welles's The Trial abruptly
shifts scenes in order to inflict a sense of arbitrariness on the viewer
which is intended to parallel the plight of the protagonist trapped in an
incomprehensible legal system.
The question/answer structure, moreover, may not only be loosened for
expressive effect. It may also be slackened for the sake of realism. Often
realist films, like Fellini's Amarcord, are narrated episodically; scenes are not
linked erotetically but by principles of rough temporal contiguity and often
geographical propinquity. This type of narration often has as its aim the
desire to impart a holistic impression of a specific milieu by itemizing or
layering details of life in a certain culture or subculture at a given time.
Furthermore, modernist films, such as Last Year at Marienbad and India
Song, literally defy the erotetic model. For they are all questions with no
answers. In some cases, it is the very point of avant-garde films like these to
lay bare reflexively the structures that make the normal narrative motion
picture experience what it is. These moving pictures are examples of meta-
cinema; they are films about films. Last Year at Marienbad, for instance,
provokes a range of questions — such as whether certain of the characters
were at Marienbad during the preceding season — but refuses to answer
them in order to reveal to the apperceptive viewer the degree to which
motion picture narration proceeds by engaging our expectations (a.k.a.
questions). That is, by frustrating our expectations in these matters, the
plot arouses in the viewer the recognition that we are accustomed to
having all of our pressing questions about the storyworld answered before
the lights go up in the auditorium. By subverting those expectations, we
are brought to an awareness of the way in which these expectations enable
us to follow and organize the motion picture as it unfolds.
So again: not all moving pictures, nor even every movie, employ
erotetic narration. But a great many do. In fact, I would hazard to guess
that by far the vast majority of feature-length, theatrically released motion
pictures (not to mention TV dramas and comedies) deploy this form of
narration. Indeed, it is probably one of the major calling cards of popular
cinema — one of the central factors that make mass-market movies so
attractive for so many. For, when erotetic narration is wedded to variable
framing in such a way that, in a given shot, the first item or group of items
that grabs the viewer's eyes is whatever is relevant to supporting the
pertinent question/answer network, the resulting visual narrative has a
144 MOVING IMAGES

level of intelligibility and uncluttered clarity rarely rivaled in daily life.

Neurologists have begun to establish the brain's preference for clarity. In
this regard, movies are appetizing brain food.
The storyworld gives the impression of being significant through and
through. From a narrative point of view, everything that is brought to
our attention is explained. The fiction has a completeness and compre-
hensibility that far exceeds ordinary experience. Thus, the wish for the
meaningfulness that so often escapes us in life can be recuperated at the
movies due to the coordination of the variable framing within sequences
with the erotetic narration across sequences.
Moreover, the notion of erotetic narration not only gives us insight into
the power of movies to hold mass audiences in their grip. It also may
suggest important clues to the operation of other sorts of plot structures.
Recall our previous examination of some of the effects that can be incurred
when motion pictures forgo erotetic narration. In some cases, the result
involved disposing the audience to embark upon interpretations — thematic
interpretations in the case of Rashomon; reflexive ones with respect to Last Year
at Marienbad. It was also noted that loosening the erotetic structure of a motion
picture narrative could also result in either expressive or realistic effects.
In these cases, the erotetic structure seems to capture something like
one of our default expectations when it comes to narratives of all kinds —
that they are underwritten by the promise of answers to our questions,
even if those answers are generally unpredictable ones. But when that
default expectation is not realized by the story, we are inclined to search
for other sources of motivation for the instance at hand, including inter-
pretive, expressive, and realistic significance.
That is, where we have grounds to suspect that the erotetic irregularity
in the movie is not a mistake, but is intentional, we treat it as a heuristic
invitation or prompt to search for some other-than-narrative significance.
In short, the failure to secure erotetic closure is what in certain cases can
make other-than-narrative dimensions of signification possible. That is, the
erotetic narrative is the plot structure upon which many seemingly more
sophisticated narrative structures play their variations and build.

Suggested Reading

Titles claiming that cinema is a kind of language are readily cited. For example,
consider Raymond Spottiswode, A Grammar of the Film: Basic Film Techniques
(London, 1935). And, as indicated in earlier chapters, one source for the idea
 MOVING IMAGES 145

of a language of film editing is V. I. Pudovkin's Film Acting and Film Technique

(London: Vision Press, 1958). In order to be fair to Pudovkin, however, it should
be acknowledged that he did not work out this proposal in genuine technical
detail. His was more of the nature of a programmatic suggestion. Indeed, the same
may be said of most of the theorists who have floated this idea, including many
semioticians.
Nevertheless, Brian Henderson does appear strictly committed to such a
project in his Critique of Film Theory (New York: Dutton, 1980). Henderson's
approach is criticized by George Wilson in "Morals for Method," in Philosophy of
Film and Motion Pictures^ edited by Noel Carroll and Jinhee Choi (Oxford: Blackwell
Publishing, 2006).
John M. Carroll, employing the Chomskyan framework of transformational
grammar, has attempted to work out the editing-as-language notion technically.
See Carroll, Towards a Structural Psychology of Cinema (The Hague: Mouton
Publishers, 1980). A speech-act approach to pictures in general (including motion
pictures) is suggested in Soren Kjorup, "George Inness and the Battle of Hastings, or
Doing Things with Pictures," The Monist 58 (Apr. 1974), 216-35, and id., "Pictorial
Speech Acts," Erkenntis 12:1 (Jan. 1978), 55—71. As the publication dates of the last
three items indicate, the cinema-as-language view is not as influential today as it was
in the latter part of the twentieth century.
Philosophical criticism of the approach include Gilbert Harman, "Semiotics of
the Cinema: Metz and Wollen," Quarterly Review of Film and Video Studies 2:1 (1977),
15—24. As mentioned in a previous chapter, Gregory Currie has challenged the
cinema/language couplet in his "The Long Goodbye: The Imaginary Language of
Film," British Journal of Aesthetics 33:3 (July 1993), 207-19.
The approach to cinematic sequencing as a matter of attention management
is developed in Noel Carroll, "Film, Attention, Communication: a Naturalistic
Approach," in Engaging the Moving Image (New Haven: Yale University Press, 2003).
The quotation from David Hume with respect to narrating comes from his "Of
Tragedy," in David Hume: Selected Essays (Oxford: Oxford University Press, 1993).
An impressive overview of narrative structures in the cinema is David Bordwell,
Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985).
A "sequel" to that book — which brings the discussion up to date with an examination
of contemporary narrative practices in the commercial cinema — is David Bordwell,
The Way Hollywood Tells It: Story and Style in Modern Movies (Berkeley, CA: University of
California Press, 2006). The notion of erotetic narration is defended at greater length
in Noel Carroll, "Narrative Closure," Philosophical Studies (formcoming).
A current topic of very lively philosophical exchange regarding cinematic
narration — which is not discussed in this chapter — is whether or not all motion
pictures have fictional narrators. The view that all fictions, including motion picture
fictions, have fictional narrators is advanced by Seymour Chatman in Coming to Terms
(Ithaca, NY: Cornell University Press, 1990), and by Kendall Walton in Mimesis as
Make-Believe (Cambridge, MA: Harvard University Press, 1990). This claim is
146 MOVING IMAGES

rejected by David Bordwell in Narration in the Fiction Film and by Gregory Currie in
his Image and Mind.
The debate is deftly summarized by George Wilson in his "Le Grand Imagier
Steps Out: The Primitive Basis of Film Narration," Philosophical Topics 25:1 (Spring
1997), 295—318. Wilson, himself, comes out in favor of the thesis that all fiction
motion pictures have fictional narrators. This position is challenged in: Berys
Gaut, "The Movies: Cinematic Narration," in The Blackwell Guide to Aesthetics,
edited by Peter Kivy (Oxford: Blackwell Pubhshing, 2004); Andrew Kania,
"Against the Ubiquity of Fictional Narrators," Journal of Aesthetics and Art Criticism
63:1 (Winter 2005), 47-54; and Noel Carroll, "Film Narrative/Narration:
Introduction," in The Philosophy of Film and Motion Pictures, ed. Carroll and Choi.
Though this particular debate is not canvassed in this book, it is well worth a
look. One very exciting place to go in order to continue your exploration of
the philosophy of motion pictures, then, might be to start by checking out the
references in the preceding two paragraphs.
Chapter Six
Affect and the Moving
Image

There is an undeniable relationship between moving images and our affective

life — that is, our life of feeling. Many mass-market motion picture genres
derive their very labels from the types of affects or feelings they are designed
to provoke — suspense, horror, mystery, weepies (a.k.a. tear-jerkers), and
thrills (thrillers). In addition to being motion pictures, movies also generally
aspire to being e-motion pictures; they not only yield the impression of
movement, they also move us.
Indeed, one of the primary reasons that the kind of moving fictions
we affectionately call movies are so popular is undoubtedly their potential
to arouse us affectively. After a day of numbing routine, we crave some
excitement — a jolt of suspense perhaps, or horror, or fear, or a good cry,
or maybe some robust laughter. Movie fictions facilitate experiencing these
affective conditions imaginatively, that is, without having to pay the price
that these states usually exact — as tears of sadness typically require some
genuine loss.
Rather, fictional motion pictures give the system a hearty workout — our
affects are enlivened — though we are not the worse for it, as we would be
if we had to confront the sort of perils that typically engender horror or
suspense. Even if there can be motion pictures that do not massage the
emotions — such as surveillance footage of empty parking lots — energizing
our affective system is so central to the enterprise of the movies that no
philosophical account of the moving image would be complete without
discussing it.
Of course, motion pictures may not only be valued for the affective
calisthenics they promote. Some motion pictures may be esteemed for
engendering feelings in us for peoples and causes that heretofore left us
unmoved — such as the plight of exploited factory workers in so-called
148 AFFECT AND THE MOVING IMAGE

developing countries. Furthermore, certain motion pictures can even

afford us self-knowledge, by awakening feelings in us we never knew
we had and enabling us to examine them apperceptively. But whether
for the sake of bracing affective stimulation or for the purpose of
expanding our emotional reach — either socially or introspectively —
the link between the moving image and our affective life is one of its
major draws.
However, it is not the case that moving pictures engage our affective
reactions in some simple, unitary fashion; motion pictures engender
feeling-responses in us in a number of different ways. This is not only
because movies possess several different channels of affective address, but
also because our repertoire of affective responses is itself variegated. That is,
the motion picture can elicit a range of affects not only because it has a battery
of diverse triggers at its disposal, but also because there is a multiplicity of
affects available for stimulation. These affects include reflexes, phobias, affect
programs, cognitive emotions, and moods. "Affect" is our name for this
entire domain; others might prefer to call it the "emotions," but for reasons
that will emerge in what follows, I reserve that label for only certain species
within this genus.
It is the purpose of this chapter to explore, to varying degrees,
a diversity of phenomena in the realm of affect in relation to various
structures of the moving image. Because there are a number of different
affects and a number of different structures at issue, I will attempt to
impose some order here by dividing the chapter into two parts. These
parts correspond to a crude and ultimately artificial distinction between
two levels of motion picture address.
Part 1 pertains to the affective reactions of audiences to the depictive
array at large, while part 2 zooms in more specifically upon the range of
audience responses to fictional characters, notably to protagonists and
antagonists. Or, to put the division more perspicuously, part 2 is about
our affective relationships to fictional characters, and part 1 canvasses the
broader category of that plus everything else. Part 2 is not opposed to part 1,
but rather is more focused on one of the perhaps most important levers
movies have upon audience affect.

* And, perhaps needless to say, other motion pictures can be particularly

dangerous for stoking morally inappropriate hatreds between people.
 AFFECT AND THE MOVING IMAGE 149

1 Audience and Affect

The gamut of affect
By "affect," I am referring to felt bodily states — states that involve feelings
or sensations. The compass of affect is broad, comprising, among other
things, hard-wired reflex reactions, like the startle response, sensations
(including pleasure, pain, and sexual arousal), phobias, desires, various
occurrent, feeling-toned mental states — such as fear, anger, and jealousy —
and moods. Through various strategies and devices, motion pictures
possess the capacity to kindle and even to inflame a number of these states.
In this section, we will begin to examine a selection of these capacities in
relation to the kinds of affective experiences that they are suited to enlist.
Through the manipulation of sound and image, the creators of moving
images can induce changes in our bodies. The plummeting camera in This Is
Cinerama caused roller-coaster chills in the stomachs of audiences, while
some of the abrupt cuts in Bullitt made our bellies flip. Moreover, moving-
image arrays have the power to activate the involuntary and automatic
reflexes of audiences at a subcognitive, or cognitively impenetrable, level
of response. For example, loud noises — either recorded effects or musical
sounds — can elicit instinctual reactions from spectators.
A sudden explosion — even if expected — is apt to make the viewer flinch.
The response is cognitively impenetrable in the sense that, despite the fact
that you know that you are not in danger, your body will respond otherwise
and prepare you affectively for flight or some other self-protective behavior.
Adrenaline will rush into our veins as we are primed for action. Further-
more, this "high" can be enjoyable — can be savored — if there really is no
real danger in the vicinity.
This kind of response is called the "startle response." It is an innate feature
of the human organism. All things being equal, the adrenaline rush elicited by
a fiction is pleasurable — as it is in the case of fireworks (as opposed to
battlefield bombardments). This is why so many movies, come seasonal
vacation periods, are tantamount to fireworks displays; they literally excite
spectators by triggering inborn perceptual/hormonal mechanisms for the
purpose of throwing audiences into heightened affective states. So many
genres are devoted to this enterprise — from disaster films to war films
(whether on earth, or in outer space, or in some fantasy realm, like Narnia,
or in the distant future, like The Time Machine or Planet of the Apes).
150 AFFECT AND THE MOVING IMAGE

These reflex states, of course, are not only tripped by loud noises.
Sudden movements toward the camera or of the camera or rapid move-
ments laterally across the screen can put the body on high alert. The maw
of the giant shark surges forward and we start backwards. The movie
screen is a rich phenomenal field in terms of variables like size, altitude,
and speed, which have the capability to draw forth intense, feeling-tuned,
automatic responses from the bodies of viewers, as do the variations in the
loudness and cadence on the soundtrack.
This is why so many movies are "action-packed." The relentless move-
ment in movies provokes a level of inner commotion that is experienced
positively. In this way, the movie can keep the audience percolating with
affect from the pre-title scene to the end credits. And that is the aspiration
of the many summer action spectacles which subscribe, according to some
French critics, to the "Boom-Boom" theory of filmmaking.
In addition to reflexes, the standard-issue human organism comes
equipped with certain broadly shared phobias. Fear of insects and snakes,
for example. Moviemakers exploit these in all sorts of ways; the number of
films named after their presiding bug is legion, from Killer Bees to The Spider
(a.k.a. Earth Versus the Spider). Ditto snakes — as in the case of Anaconda.
Sometimes our phobias are titillated by making these creatures enormous —
for example, The Black Scorpion — but also by marshalling swarming masses
of them together as in the case of the army ants in Byron Haskin's The
Naked Jungle. O r one can have them both larger than life and swarming, as
Peter Jackson does in the scene in the pit in his remake of King Kong
Horror fictions, of course, specialize in phobic creatures. O u r instinctual
aversion to dead and decaying things is exploited by monsters from A to Z
(for zombies). In these examples, we relish the shudder they invite, since we
are in no danger of being eaten alive or infected. Again, the heightened affect
comes cost-free. But horror fictions are not the only movies that traffic in
tantalizing phobic reactions. An instinctual fear of heights plays a role in
many action genres; that is probably why so many films feature mountain
climbing, airline disasters, roof-top chases, and so on.
Alfred Hitchcock was identified by Francois Truffaut as a director who
overtly strives to tickle phobic responses. In North by Northwest he is said
to have experimented with agoraphobia (fear of open spaces) in the crop-
dusting sequence, while Vertigo is, of course, named after a phobia, one
which the film attempts to simulate cinematically. The recent film Snakes on
a Plane exploits several phobias at once by releasing venomous vipers in the
cabin of an airplane in flight — thereby racking up fear of reptiles, heights,
claustrophobia, and fear of flying in one shot.
 AFFECT AND THE M O V I N G IMAGE 151

Perhaps, little needs to be said about the ways in which motion pictures
can arouse sexual feelings; we will leave that research to the reader.

Emotions
The affective responses reviewed so far are somewhat rigid. They are dedicated—
they issue a fixed response to a very specific kind of stimulus. This is not to
discount their importance to either self-preservation or moviegoing. They
were evolved to protect us in environments fraught with danger. However,
where there is no danger, as is typically the case in most movie theaters,
inciting them, as already indicated, can be a source of great pleasure.
However, in addition to these somewhat primitive responses, the body has
affective resources that are more discerning — smarter, if you will — in their
activation. Whereas the startle response warns the organism on the occasion
of a loud noise, these affective systems — which we will call the emotions
proper — can detect danger not only in a resounding explosion, but also in a
whispered threat or in one's spouse's overly attentive laughter to the attractive
stranger.
These resources size up the situations and things that give rise to them
and elect differential reactions to the aforesaid stimuli on the basis of
antecedent computations, whether immediately prior or after some interval.
That is, these computations may occur on contact at the initial level of
perception or they may be processed cognitively in the forecourts of the
mind, either tacitly or consciously. They may engage the frontal cortex of
the brain or they may bypass it and may be relayed directly to our behavioral-
response centers. It is the function of these affect systems to evaluate the
circumstances before us in terms of certain recurring existential-human
themes — like loss — and to prepare us to react appropriately.
For example, if the stimulus is appraised to be harmful, the organism is
primed to fight or flee. This response, of course, is what we call fear.
Likewise, if the situation that confronts us is one in which we perceive a
wrong done to ourselves, or to those we hold near and dear to ourselves, or
to our interests, then we are prepared to "get even." This, needless to say, is
anger. Other affective responses in this neighborhood include sorrow, pity,
indignation, reverence, awe, hatred, love, shame, embarrassment, guilt,
humiliation, comic amusement, loyalty, and so forth. This is the realm of
the emotions.
Emotion is the realm of affect in which differential computational
appraisals of stimuli relative to certain interests give rise to visceral feelings
which typically prime behavioral tendencies to act. Or, to put the matter
152 AFFECT AND THE MOVING IMAGE

graphically, the perception of danger (an appraisal of the large, hulking

thing in my vicinity) leads to a chill down my spine (a visceral feeling)
which makes me freeze in place (an evasive behavior). Altogether these add
up to an instance of the emotional state of fear. Since, as we have seen,
there are other sorts of affects, like the startle reflex, which are more
primitive structurally, we will reserve the label the "emotions" for only
those affects that have the preceding, complex structure that integrates
differential computations with feelings.
The emotions are a good thing to possess from an evolutionary stand-
point. Compared to processes of conscious, rational deliberation, the
emotions are very fast, "down and dirty," decision-making routines. They
scope out situations quickly and ready the organism to react, sometimes
within the blink of an eye. This is no small advantage in a tight spot. Of
course, the emotions are sometimes mistaken. What I may at first size up
as a large, potentially dangerous creature standing in the shade may in fact
turn out to be nothing but a curiously shaped bush. However, way back
when in the environmental circumstances on the African veldt where the
emotions first developed, it was better to be safe than sorry.
The emotions were evolved by natural selection in the first instance to
respond to percipient's conception of the situation. That is, if the percipient
believed that the large shape in the shadows was a dangerous creature, then
the organism was thrown into a state of fear, even if the stimulus was really
only a bush.
Yet if this is how the emotions are supposed to work, it seems to raise a
question immediately about how our emotions can — for better or worse —
be aroused by fictions, including motion picture fictions. For, with respect
to such fictions, the audience clearly knows that the aliens in War of the
Worlds are made up and that these monsters pose no threat to either the
viewers or to anyone they care about. The saga has been concocted from
whole cloth. So how is it possible for audiences exercising their natural,
selectively given emotional powers to feel fear upon exposure to Steven
Spielberg's completely fabricated spectacle? For if they know the film is a
fiction, they patently cannot believe that Earth is being invaded from outer
space. And the audience must know it's a fiction, since that's what they
paid for; no one pays to be shot at by space monsters.
Earlier we hypothesized that experiencing an emotion like fear — where
one did not have to pay the usual entrance fee of risking potential danger —
could be pleasurable. However, that brings up a prior question, namely:
how is it even possible to fear that which one knows does not exist? We
calm the upset child by telling him, "There are no such things as ghosts."
 AFFECT AND THE MOVING IMAGE 153

So how can we be seized with fear when we know there are no such things
as alien invaders?
Indeed, the problem extends to other emotions beyond fear. How can you
feel sad about someone's loss of a loved one in a movie melodrama, when
you know that there was never a loved one to be lost? And, of course, if you
are sitting in a movie theater watching a well-known actress emote, you do
know — and therefore believe — that it is "just a story." And yet we cry. Doesn't
this defy reason?
Some philosophers have found this phenomenon quite puzzling and have
even conjectured that, if movie audiences are emotionally moved by fictions,
they must be, at least temporarily, irrational (or, in other words, insane for
the duration of the fiction). This conundrum is sometimes called "the
paradox of fiction" — the mystery of how one can be emotionally moved
(for example, frightened) by something you know does not exist. And yet,
on the basis of our understanding of the ways in which the emotions work
naturally, it does not appear that we actually are compelled to agree that
fearing fictions is in any way paradoxical or self-contradictory, nor must we
be forced to such a desperate conclusion as the conjecture that movie
viewers must be momentarily deranged.
One consideration against the temporary insanity charge is that, as we
have already observed, many emotion-like, affective states — such as
reflexes and phobias — can be elicited sans belief. We start at the loud
noise — the balloon bursting — even though we know that it is harmless. No
one thinks that is irrational. Insofar as emotions are near relations — cousins
perhaps — of these more primitive affective states, isn't it possible that they
do not require beliefs in order to be launched? And, in any event, since
various reflexes, phobias, and affect programs can be set in motion while
bypassing the belief-centers of the brain, there would not appear to be a
paradox of fiction with respect to affects across the board.
Furthermore, although it is true that an emotional state can, and often
is, ignited by a state of belief, this is not the only mental state that can
arouse an emotional reaction like fear or anger. We may also simply
imagine a state of affairs and then take note of an emotion welling up
within us. And there is nothing abnormal or irrational about this.
Consider: we are about to ask our boss for a day off in order to visit a
dying relative. We imagine how things might go down. In our mind's eye,
he denies our request and makes a characteristically cutting remark. Our
ire mounts, though we do not literally believe that our boss has either
denied our request or insulted us. We only imagine that he does so, and
this is enough to start our emotional engine churning. Or, in a moment of
154 AFFECT AND THE MOVING IMAGE

frustration, entertain the idea of saying something hurtful to a person who

is mentally disabled, confused, utterly dependent upon you, and who, as
well, worships you. A feeling of shame may overcome you, though your
lips remain sealed.
Belief—holding the content of proposition x before the mind assertively —
can bring forth emotions. But so too can something that has been merely
imagined — that is, a prepositional content entertained in thought non-
assertively. Or, in other words, believing that x and imagining that x — in the
appropriate circumstances — are both psychologically empowered to elicit
our emotions.
Moreover, it is easy to see the evolutionary advantage to be had in
possessing the capacity of the imagination to arouse emotions. For by this
means we can enhance the survival prospects of our conspecifics by instilling
fear in them with respect to counterfactual states of affairs. In order to warn
children off trusting strangers, we may, for instance, tell them tales about
what might happen to them if they wander away with just anybody.
Here, of course, we can see an immediate connection between the natural
history of our emotions and our emotional responses to fictions. For are not
many of the earliest stories that we tell (and show) children counterfactual,
cautionary tales — such as what befalls Walt Disney's Pinocchio when he
throws in with the fast-talking strangers on his way to school? That is, our
capacity to respond to fictions emotionally is rooted in the capacity —
endowed by natural selection because of its adaptive advantageousness — of
the emotion system to be aroused not only by that which we believe, but also
by that which we imagine.
Motion picture fictions are sense-bearing vehicles that mandate viewers
to imagine the states of affairs and events that they depict audiovisually.
With respect to Captain Horatio Hornblower, we see the actor Gregory Peck

The power of the imagination to engage emotions is also connected to

planning. That we can envision alternative futures and gauge what our emotional
responses would be to finding ourselves in those states of affairs is clearly
extremely advantageous. Imagining how you would feel if your child hurt herself
at play may encourage you to baby-proof your house. Human life flourishes in
large measure because of our ability to live in the future imaginatively. The
emotions, we may hypothesize, are sensitive to imaginings just because they
appraise or evaluate those future, possible, counterfactual states of affairs in
terms of our abiding interests. Moreover, it does not make much sense to discount
this capacity as irrational.
 AFFECT AND THE MOVING IMAGE 155

on screen in a three-cornered hat and then imagine that the eponymous naval
officer is steering his frigate out of harm's way because that is what, given
the context of fiction, we realize we are intended to do by the director Raoul
Walsh (and his team of fellow fictioneers). In effect, a fictional motion
picture instructs its audiences to hold certain propositions before the mind
unassertively — for example, to imagine that Hornblower is under fire, or
to suppose that Hornblower is being blasted, or to entertain the unasserted
thought (rather than the belief) that Hornblower is just barely eluding the
enemy attack.
And then, upon entertaining or imagining said state of affairs, we feel
suspense for Hornblower and his crew. This is possible precisely because,
due to evolution, our emotions are susceptible to imaginings as well as
beliefs. The cultural institution of fiction, including the precinct of fictional
motion pictures, rests upon our innate capacity to be moved emotionally
by representations of counterfactual states of affairs and events. The
phenomenon is neither paradoxical nor irrational, but natural.*
Furthermore, emotional states — like fear, pity; levity; anger, sadness,
and so forth — can be activated not only by mental occurrences such as
believings that x and imaginings that xy but also by nonpropositional states
such as patterns of perceptual awareness or attention. That is, upon
identifying the animal before me as a people-eating tiger and, thence,
believing it to be dangerous, I am reduced to fear and trembling. Likewise,
prompted by the visage on screen, I imagine that The Predator exists and
then, recognizing all his malignant properties, my flesh crawls with visceral
revulsion. But; in addition, an emotional reaction may be thrown into gear
before I have fully computed or re-cognized the identity of that which I am
in the process of encountering.
For whatever it is may command attention just because it satisfies a very
general profile for that which is potentially harmful. That is, we may not know
what x is specifically, but perception registers that x is large and advancing
toward us very quickly. And, as a result, sans conscious identification; our
fear-alarms put us on red alert, prior to any further processing. Of course,
once we recognize that it is a charging rogue elephant, that then reinforces

But, you may say, what of cases where we quell the child's fear by dispelling
his beliefs? Doesn't that establish that emotions require beliefs to take hold? No,
for there are also cases on the other side of the ledger. For example, we may
overcome our disappointment at failing in some goal today by imagining that we
will succeed tomorrow. (I owe this observation to Rianna Oelofson.)
156 AFFECT AND THE MOVING IMAGE

our fright all the more. But the relevant point at this juncture is that there a
may be enough information in these very early, very general stages of the
tracking process to send the emotion system into a state of terror, even
though there is not yet sufficient information to determine the exact nature
of the threat.
Emotions can originate near the site of perception and prime the body
for action without any further need for computation: the groom slips on a
banana peel and we burst into laughter. Or the emotion may arise after
being processed cognitively, either tacitly or consciously. Professional envy
with regard to your colleague's executive bathroom privileges takes a lot of
tliinking. Motion pictures, of course, afford opportunities for the emotions
to erupt through a variety of routes — some mediated by conscious
cognitions, some by tacit ones, and others even more immediately. The
quickly moving, dark shadows may send the icy rush of fear down our
spine without our apprehension of whatever is casting those shadows. On
the other hand, in order to admire the bravery of the hero, we must
cognize his actions under the concept of courage and also recognize that
that satisfies the pertinent criteria for admiration. Of course, whether a
given motion picture involves a greater degree of primarily perceptually
motivated emotional states versus ones calling for more cognitive processing
can only be determined on a movie-by-movie basis.
Nevertheless, what these emotional states have in common is that they
comprise appraisals or evaluations relative to certain recurring human
themes, such as personal dignity, which appraisals then engender bodily
states of feeling that dispose the organism to behave in certain ways. The
samurai, for example, appraises the mud splashed on his brother's sandals
as an affront: this causes his blood to boil, and, in consequence, he draws
his sword from its scabbard, with lethal consequences. Of course, with
respect to motion picture audiences, the behavioral portion of the emotion
scenario rarely obtains, since we are only imagining that such-and-such is
happening. For example, we are only supposing that someone's sandals have
been muddied; thus, even if we are partial to that character, we have no
reason to act, since, among other things, no one has really been insulted.
It is a remarkable fact about motion pictures that to an arresting extent
they are able to elicit — across diversified audiences — roughly the same or
converging general emotional responses to the fictions on screen. Suppose,
as might happen any day of the week, that an ordinary couple is arguing on
a street corner. The affective responses, if any, of real-world passers-by are
likely to be all over the map. But, contrariwise, a couple argues onscreen
and we all — or nearly all of us — feel indignation at the way in which the
 AFFECT AND THE MOVING IMAGE 157

heroine is being demeaned. Whether we are in Madras or Manhattan,

whether we are Lutheran, Jew, or atheist, Republican or Democrat, any
motion picture worth its salt can usually elicit roughly the emotion it
intends from its target audience, no matter how far flung, most of the
time, its members may be. How is this possible?
To begin to answer this question, first recall that emotions are appraisals —
they appraise situations in light of certain interests. The emotion of fear
appraises events, objects, and people with respect to concerns for the safety
and protection of oneself, one's interests, and the interests of one's confed-
erates foremost in mind. Moreover, these concerns, which we can summarize
as a preoccupation with harmfulness, function as the criterion used to appraise
or evaluate a predicament as fearsome. That is, emotional appraisals are
governed by criteria, as the emotion of fear can only be mobilized upon at
least a suggestion of harmfulness.
Upon detecting circumstances that are perceived to be harmful or
dangerous, fear takes over. It operates like a searchlight. It directs our
attention, organizing the details before us into significant wholes or gestalts.
In the first instance, it throws certain of the elements of the situation into
bold relief, leading us to attend to the mugger's knife, rather than to the
striped design on his tie. Fear then further guides our scansion of the array to
pick out further relevant features for notice — for example, that the mugger
has several large, equally well-armed companions.
Note here that the presiding emotional state does not batten on these
details willy-nilly. They have all been chosen because they are pertinent to
alerting us to the level of danger that confronts us. That is, these elements
all meet the evaluative criterion of harmfulness. The emotion of fear solves
the problem of what we must attend to. It weighs elements of our
circumstances in light of our interests, specifically our interests in self-
preservation and the avoidance of harm. Fear organizes situations in terms
of the evaluative criterion of harmfulness.
Like fear, our other emotions are also criterially governed. In order to
be angry, I must believe that me or mine have been wronged. To pity
someone, I must believe that they have suffered misfortune. Envy requires
that I think that someone else has something I want. And so on. The
emotions appraise situations in terms of criteria. The bodily feelings that
are then provoked by the emotional state, in turn, bias or organize our
apprehension of the eliciting state of affairs in accordance with the criteria
that govern the prevailing emotion. Our anger first fixes our attention on
the smile of the guy who has just insulted us and then directs us to each
of the other laughing faces that surround us.
158 AFFECT AND THE MOVING IMAGE

This is how the emotions work in everyday life. But what does it have to
do with viewing motion pictures? In particular, what does it have to do
with the way in which moving pictures quite frequently dispose diverse
audiences to vent extremely like-minded emotions in response to onscreen
fictions — to the degree that all or most of us feel suspense, at the same
time, for example, over the issue of whether or not the diminutive
protagonist will make the football team of his dreams or whether or not
the over-the-hill boxer stands a chance against Apollo Creed?
In life, in contrast to fiction, our emotions have to select the pertinent
objects upon which to focus from a plethora of largely unstructured
stimuli. But in fictions, including motion picture fictions, things are
different. Our emotions are not called upon to organize the situations
before us, so to say, de novo. To a much greater extent than usually
encountered in everyday life, the situations in fictional motion pictures
have already been structured for us by the director and his team. We do
not typically have to depend, from the first instant, upon our emotions
to organize fictional events for us as much as we rely upon the emotions to
perform this task for us in the ordinary course of events. For, in the main,
the states of affairs and events in motion pictures have been, in a manner of
speaking, emotionally predigested for us by the creators of the fiction.
That is, the creators of the motion picture have already done much of
the work of emotionally sculpting scenes and sequences for us through the
ways in which the salient features of the fictional situation have been
carefully designed to satisfy the criteria for drawing forth the emotional
state intended by the production team. Details that suit the conceptual
conditions of the emotional response desired by the moviemakers have
been selected, filtered, foregrounded, and emphasized in the narrative,
the dialogue, and the composition, and through the camera positioning, the
acting, the musical commentary, and so forth.
In contrast to the way that the emotions have to start from scratch when
it comes to managing our attention in daily life, when it comes to the
general run of motion pictures, the events on screen have been emotively
prefocused for us by the creators of the movie. They have selected the
elements of the scene or sequence that they think are emotively significant
and thrust them, to put it bluntly, in our face. The means to this end at the
filmmaker's disposal include: camera position and composition, editing
(including the processes of bracketing, scaling, and indexing discussed in
the previous chapter), lighting, the use of color, and, of course, musical
accompaniment, acting, dialogue, and the very structure of the script
or narrative.
 AFFECT AND THE MOVING IMAGE 159

Quite frequently in everyday life, when an acquaintance or a colleague

slights us — perhaps by a passing remark — we are not immediately angry,
even if we are hurt, because we may wonder whether the insult was an
intentional wrong rather than merely a thoughtless gaff. But as such remarks
recur, anger besets us and inclines us to begin to recognize a discernible
pattern of nastiness directed at us. However, in typical movie fictions, we
rarely have to waver so long. So often characters wear the meanness of their
actions on their sleeve, and if that were not enough, we also have access to
the disapproving judgments of the people who surround them in the fiction.
From the outset, we not only have a pretty unmistakable gestalt of wrong-
ness served up with arresting clarity, but we also have the reactions of the
characters around them to reflect and to reinforce our assessment of
the situation.
It is very difficult not to respond negatively to the snarling, arrogant,
cruel pronouncements of the dictator in Vfor Vendetta, shot as they are in
oppressive close-ups and rasped out with gravelly harshness by the actor
John Hurt. The performance is designed to elicit a gut reaction of
political defiance, aversion, and even loathing. Generally in fictional
motion pictures, the detection work that our emotions need to do for
us in daily life is somewhat minimized, since the scenes and characters in
these motion pictures have most commonly already been constructed
from the vantage of the intended emotional point of view. The aforesaid
dictator, for instance, was expressly designed from and for the perspec-
tive of loathing. Or, to say it differently, the character was emotively
predigested or prefocused for us. And the image was made in such a way
that any distracting detail was winnowed or filtered out and features
pertinent to the elicitation of the intended emotion were exaggeratedly
foregrounded.
But what does it mean for a character, or a scene, or a sequence to
be emotively prefocused? Remember: the emotions are governed by
criteria. To be happy for my cousin, I must believe that he is doing
well. If he is doing poorly, then, all things being equal, happiness would
be an odd response to his condition. Moreover, just as the emotions in
daily life are governed by certain criteria of appropriateness, so too must
the emotions mounted in response to fictions be governed. Thus, a
motion picture is emotively prefocused by being criterially prefocused —
that is, by being so structured that the descriptions and depictions of the
pertinent objects of our attention in the movie clearly and decisively
satisfy the criteria for the emotional state intended by the creators of the
moving picture.
160 AFFECT AND THE MOVING IMAGE

The makers of mummy-monster movies want to put us into the

emotional state of horror. Horror is a compound emotion, involving fear
and disgust. As we have seen, harmfulness is a criterion for fear. Impurity is
a criterion for disgust. Thus, to elicit the horror response from viewers,
the makers of a mummy picture will emphasize certain of his properties —
namely, those that count as fearsome and those that count as disgusting. His
capacity for harmfulness, of course, is satisfied by the lavish attention
paid to his killing innocent people, while his impurity is established by
elaborating on the fact that his dead body is rotting and badly in need
of rejuvenation.
Likewise, suspense is arguably an emotion with two criteria—improbability
and the desire that the good shall triumph. When, due to the nefarious plot of
some Lex Luthor type, the two trains steam toward each other on a collision
course, it is unlikely that disaster can be averted. Moreover, the death of all
the nice people on the train is clearly evil. Can even Superman save the day?
The narrative and the editing forcefully direct our attention in such a way
that it is virtually impossible to miss the unUkelihood that good can come of
this state of affairs, and, all things being equal, for the normal viewer the
result is suspense.
Once we perceive the objects of our attention under the criterially
relevant categories — such as harmfulness and disgust with regard to horror
fictions or the improbability of the morally correct outcome with respect to
suspense — the intended emotion is apt to be raised in us. That is, as a result
of entertaining the appropriate appraisals — usually the ones implicated by the
criterial prefocusing — we are likely to undergo the bodily responses, such
as laughing, crying or squirming in our seat, that the moviemakers have
planned.
Of course, not all motion pictures aim at evoking such brash emotional
responses as my examples so far might suggest. Some films may aspire to
induce more complex, ambiguous, nuanced, and even ambivalent emo-
tional states than horror or suspense movies do. In those instances,
however, it is not the case that the process of criterial prefocusing has
been retired, but rather that the fictional world may have been constructed
so as to satisfy simultaneously the criteria for more than one emotional
condition, including possibly even those that conflict. Or, in other
instances, the satisfaction of pertinent emotive criteria may only be hinted
at subtly; the audience may need to mull over and reflect closely upon
the details of the depiction before it arrives at the emotional response the
filmmaker intends. For motion pictures may trade in quiet emotions as
well as loud ones.
 AFFECT AND THE M O V I N G IMAGE 161

2 Some Affective Relations between Audiences and

Characters in Popular Movies
The structure of emotional elicitation that we have called criterial
prefocusing can operate with regard to situations either populated by
characters or not. Shooting and editing a dark and stormy night can excite
apprehension in an audience, since it meets the criteria of a place inhospit-
able and potentially dangerous for people, even if no people are yet in
evidence in the motion picture. However, though affect can be raised in
audiences where no characters are in the vicinity, clearly among the most
powerful ways for arousing affect and emotion in viewers, particularly in
mass-market movies, is through our relation to the fictional characters,
notably the protagonists and the antagonists. So in order to round off our
introductory and non-exhaustive exploration of some of the modes of
affective address employed by the movies, a section about the ties that
bind us to fictional characters is in order. But, as was noted in the previous
section, there is not just one channel of affective address here, but several. In
this part of the chapter, we will first examine a way in which movies may
promote our emotional connection with characters and then we will conclude
by discussing what we call "the mirror reflex," a mode available to audio-
visual fictions, but not literary ones.

Identification
A natural place to initiate a discussion about the emotional relationship
of viewers to fictional movie characters is the notion of identification.
There are several reasons for this. First, when asked for an account of our
emotive relation to fictional characters, especially the protagonists, most
people are likely to invoke some notion of identification. Moreover,
identification is probably the oldest account in the Western tradition of
our emotional relationship to characters, for it was first propounded by
Plato, who feared that the goodly citizens would become possessed
by undesirable emotions, such as the fear of death, when exposed to actors
shuddering onstage about the prospect of Hades. And this, of course, was
hardly a desirable state to incite in potential militiamen.
Today, similar Platonic anxieties are abroad, undoubtedly underlying
the recurring suspicion that the representation in the movies of unpalatable
sexual feelings and aggressiveness will contaminate the hearts of audiences —
most worrisomely, those of impressionable adolescents. But identification is
162 AFFECT AND THE MOVING IMAGE

not merely thought of as a source of antisocial tendencies. More often than

not, identification is characterized positively as what bonds us to those most
upstanding fictional characters, the protagonists.
Typically, it is supposed that to identify emotionally with a character is
for one to feel precisely what she feels. By inhabiting the character from
the inside, so to speak, her world becomes ours and we embrace her point
of view as our own. We grow close to characters — or, at least, to the
protagonists — because our hearts beat as one; identification is, in other
words, the secret of the bond that we forge with the protagonists. Indeed,
a common complaint — though perhaps, as we shall see, an unwarranted
one — about popular movies is: "I couldn't identify with the characters."*
Of course, "identification," as it is used in common parlance, is
somewhat ambiguous. It can signify a range of things, many of which are
completely unobjectionable. However, the version of identification that
concerns us is a very specific one. It involves the claim that the audience's
affective bond with the pertinent fictional characters is a result of the
audience's suffering putatively the self-same, identical feelings to the ones
the character is undergoing. But, this may not be what people always have
in mind when they say they identify with fictional characters.
Sometimes when talking about identification, people make no claim to
any sort of identity, affective or otherwise, between the character and
themselves. When I say that I identify with Aquaman, I may mean that
I wish I were as fearless as he is, but I can't swim and am afraid of the
ocean. I make no assumption that Aquaman and I feel the same way about
the deep blue sea. I only wish that it were so. Perhaps this shouldn't even
be called identification, but rather wishful fantasizing. And, in any event, it
is not clear how much of this daydreaming can be indulged while watching
a movie flash by without losing track of the story.
Frequently, when folks, especially young folks, say that they identify
with a character, they mean nothing more than that they think he's "cool."
But you can be a nerd — and know and feel like a nerd — and still think Rico
and Sonny are cool. At other times, all people mean when they say they
identify with a character is that they've had a similar experience — like the
character, they've been laid off, for example. But this really seems to come
down to no more than feeling some affinity for the character rather than

As will become evident after our discussion of sympathy, I think claims like
"I couldn't identify with character x" should be translated as "I couldn't or didn't
care for character x." Or, more colloquially, "I couldn't care less about soandso."
 AFFECT AND THE MOVING IMAGE 163

endorsing the conviction that one has been infected by the self-same
feelings that the onscreen protagonist is now enduring. Might this not be
more accurately termed "affiliation" rather than "identification"?
Alternatively, "identification" is often parsed in terms of putting myself
in the place of the character. This is not a matter of putting myself in the
character's shoes, as they say, but of putting the character in my shoes. But
why suppose that this entails that the character and I are in the same affective
state? Shouldn't this be called "projection" rather than "identification"?
In what follows, I am presupposing that the core concept of the leading
version of identification — and on many accounts, empathy as well — involves
the audience member being in the same type-identical emotional state in
which the viewer believes the fictional character finds himself. The protag-
onist is indignant about the treatment of minorities and so are we; the heroine
is uplifted by the sight of her Savior, as are we; the hero is terrified of the
Alien — us too. However, identity of emotion-types, even if it is necessary, is
not enough to constitute identification of the sort we are now considering.
Why not?
Consider this: the fans of a certain team at a soccer match may all be in
the same emotive state — they all hate the opposing team. Yet we wouldn't
call this identification in the relevant sense. For with whom are they
identifying? Although they may all be inflamed to the same degree with
hatred for the rival team, they are not identifying with each other, since
they may not even be aware of the presence of the others, so wrapped up
are they in the game that they are witnessing as it unfolds before them. Or
imagine each is watching the game alone on television in the solitude of his
own den; they have no idea who else is watching; thus, they could not be
identifying with all those unknown others.
Nor is it plausible to speculate that these fans are, in the pertinent sense,
identifying with any of the soccer players — the players are probably too
absorbed in the activity on the field to be emoting anything, and in any
event most of them probably don't literally hate their opponents. That kind
of sports-hatred is for the fans, not for professionals.
But if the sharing of type-identical emotional states is not sufficient for
identification, what needs to be added? That the viewers be in the
emotional state in question because that is the state they think that the
fictional characters are in. That is, putatively, in this sense of identification,
I identify with Ann Darrow when I am horrified by King Kong because she
is — or I imagine her to be — horrified by King Kong. In short, the version
of identification on the table maintains: a viewer x identifies with a fictional
character/ if and only if (1) x is in the same type-identical emotive state
164 AFFECT AND THE MOVING IMAGE

a s / is in, and (2) becausey is in — or x imaginesy to be in — that emotive state.

To speak metaphorically, the viewer has been "infected" by the ostensible
emotive state of the character. For convenience, we can refer to this as the
infection model of identification.

The infection model of identification

This version of identification, I submit, is the dominant notion that most
people today have in mind when they speak of identifying with fictional
characters. It is the relationship that many appear to think is their primary
bond with fictional characters, notably protagonists, and it is the emotional
state that plain viewers and professionals alike think explains why we cheer
the protagonists on.
Now I have no reason to believe that something like the phenomenon of
affective infection never occurs, especially in real life. There is empirical
evidence that infants feel distress when they detect distress in their caregivers.
But I am not sure that this is a full-blown emotional state, rather than merely
a certain kind of affective reflex — called the mirror reflex — which we will
discuss at the end of this chapter.
However, that notwithstanding, it is not the sheer possibility of occur-
rences of the infection model that concerns us. Rather, we want to ask
whether the infection model is up to the task for which its proponents
recommend it. That is, is it a comprehensive or nearly comprehensive
account of our emotional relationship to characters in popular fictions?
Does it account for our standard emotive relationships with fictional
characters? Is it the emotional bond that typically ties us to them?

Asymmetric emotions
Even a cursory review of cases indicates that the infection model of
identification is unlikely to provide anything even approaching a general
account of our emotional relationships to the fictional characters in motion
pictures. Imagine: the candidate is pumped up by the adulation with which
his acceptance speech has been greeted by the adoring crowd; but we know
that he is standing in the cross-hairs of the high-powered, laser-guided rifle
of a merciless hired assassin. We do not feel the thrill that the candidate
does; we feel suspense, even anxiety. Our emotional state is not type-
identical with the candidate's. Nor should it be, since the movie mandates
that we should fear for the candidate.
Moreover, this species of asymmetry of affect is rife throughout comic
fictions. Every time the would-be suitor is discovered in a compromising
 AFFECT AND THE MOVING IMAGE 165

situation, we are amused, while he is discomfited. Recall, for example, Meet

the Parents. And with respect to the BBC series Jeeves and Wooster, notice that
in virtually every situation where Bertie is flustered, we are merry.
Situations in which the emotional states of the characters diverge from
those of the audience abound in motion picture fictions of all sorts. In a
romance, for example, the wholesome and endearing daughter is head over
heels in love with the handsome stranger; but we do not share her joy,
since we know that he's a cad. The reason that such situations are so
common in motion pictures is that there is generally a significant differential
between what we know and what the characters know, and, this, of course,
can have a tangibly different impact on what is felt on both sides of the
audience/fiction divide. In many cases, we know more than the characters;
we tremble for them as they plunge ahead oblivious of peril. Curiosity
draws them into the old dark house, into the attic, the basement, the cave, or
whatever, though we know better. They haven't got a clue of what awaits
them, so they're cool, but were already shivering with terror. On the other
hand, the asymmetry can run in the opposite direction: Sherlock Holmes
always knows more than we do, so we never share his aplomb.
Obviously, circumstances like these are not rare. They may even
predominate statistically. But, be that as it may, there are more than
enough cases of asymmetric affect like these to establish that the infection
model of identification cannot be very comprehensive on readily observed
empirical grounds. Furthermore, there are also conceptual considerations
that invite us to suspect the inadequacy of the infection model. Quite often
both the cause and the object of the audience's emotional state differ from
those of the protagonist's affective condition. We are presented with a
situation in which we learn that Martha's son Henry died last year and we
learn that this event is the cause of her grief; the object of her grief is her
son, Henry. But the cause of the audience's emotional state is Martha's
bereavement and its object is the grieving mother.
Our emotion organizes or gestalts a wider state of affairs than does
Martha's, while at the same time also including the object of Martha's
grieving as a constitutory component. Moreover, our emotion is pity for
the mother, not grief— who is Henry to us? It is the mother with whom we
have become acquainted through the movie. In the case we have in mind,
Henry has been kept off screen throughout; we only know of him through
Martha.
Thus, as the previous example indicates, our emotional states often have
different causes and take different objects than the putative mental states of
the protagonists. Consequently, the conjecture that our emotions always,
166 AFFECT AND THE MOVING IMAGE

or even very frequently, perfectly match those of the pertinent characters

is highly dubious. So, again, the infection model looks like it must be
abandoned as a comprehensive picture of our emotive relations to movie
protagonists.
Indeed, there are certain cases, perhaps many, where the audience
member's emotional state can only plausibly be thought to be that of an
onlooker rather than one that could correspond to the mental state of its
object, the fictional protagonist. When we feel nail-biting suspense as the
protagonist claws his way to safety, he is highly unlikely to be feeling
suspense or, for that matter, any other emotion. He is probably numb with
concentration. He is apt to be so caught up in and focused upon his task
that his anxiety is on hold.
In my own experience of extreme situations, such as skidding at high
speed on ice in my car, I have noticed that I do not go into an emotional
state, like panic, but react very deliberately, as if affectively anesthetized,
and do what my driving instructor told me to do decades ago. If I were to
respond emotionally — if I were racked with suspense — I'd crash. My
passengers can afford suspense, but I can't. Isn't it likely that the same
asymmetrical distribution of emotion occurs quite often with respect to
motion pictures, particularly action-packed ones?
Nevertheless, the friend of infectious identification may retort that, even
if there are many cases where the presumed inner states of audience members
and those of the characters diverge, there are also a significant number of
cases where the emotive states at issue would appear to be type-identical.
The character recoils in contempt at the child molester's ploys; so do we.
Emotive symmetry obtains. Therefore, it may be maintained, even if the
infection model fails to be applicable across the board, perhaps the argument
can be made that it has compelling authority in cases like the preceding one,
in which the emotions of the viewers and the protagonists correlate. How-
ever, in order to probe this suggestion, we need to introduce a distinction
between emotions that are held in common or coincidentally and emotions
that are shared due to some intimate causal relation between them.

Coincident versus connected emotional states

American jet bombers have streaked past their fail-safe points and they are
winging their way to Beijing. They are freighted with nuclear devices.
Atomic warfare looms; millions will die. One of the protagonists, the
President of the United States, is stricken with fear; us too. Isn't this a case
where the infectious identification holds sway?
 AFFECT AND THE MOVING IMAGE 167

No. Why not? Remember that the infection model of identification

requires not only that our emotions match those of the protagonists, but
also that our emotions be a causal consequence of the protagonists being in
precisely the self-same kind of mental state. However, in a great many of
the cases in which we find ourselves in the same type of emotive condition as
the protagonist, including the preceding example, it is pretty clear that we
have gotten there by our own route, so to speak. We are not anxious because
the President is anxious. We are anxious because we have been encouraged
to imagine that a nuclear armageddon, threatening a catastrophic number of
deaths, is in the offing.
Maybe some evidence for this is that the fiction could be told without
reference to the President's anxiety and we would still feel anxiety. Were
the President made of steel, I speculate that we would still feel suspense.
We would feel the same species of consternation precisely because the
fictional situation has been structured in a way that makes certain features
that are appropriately conducive to the state of fear salient — such as
repeated assessments of how much explosive power those jet bombers are
carrying, their capacity to evade radar detection, their imperviousness
to any and all anti-aircraft defenses, as well as the putatively uneasy
diplomatic relations between the United States and China.
In other words, in the vocabulary of the first part of this chapter, the
film has been structured emotively in terms of criterial prefocusing. That
is, by means of its visual depictions, enactments, and/or verbal descrip-
tions, its circumstances have been organized or filtered in such a way that
the features the moviemakers have elected for emphasis are predominantly
exactly those suitable for or criterially apposite to the emotional states
intended to be excited by the work.
Iii the Odessa Steps sequence of his film Potemkin, the director Sergei
Eisenstein selects out of the streaming riot and highUghts the callous
massacre of old people, a mother and her adolescent son, and then another
mother and her infant child — in short, he places brackets around those
persons who are culturally figured as harmless and defenseless. His selection of
these vignettes for emphasis — rather than shots of the clouds overhead — was
designed to activate the viewer's emotions of moral outrage. The sequence
was compellingly crafted so that factors that are criterial to moral indignation
unavoidably command attention in a way that leads audience members,
unless they are Cossacks, to process the episode under the heading of evil
and, in consequence, to experience visceral distress.
Most often, I guess-estimate, when the feelings of audience members
are congruent with those of the protagonist, it is a result of criterial
168 AFFECT AND THE MOVING IMAGE

prefocusing, not infectious identification. The difference is that in the case

where criterial prefocusing leads to emotive uptake on the part of viewers,
the correspondence between what the audience feels and what the char-
acters are imagined to feel is co-incident or conjoint rather than causal.
That is, the audience has been effectively led to the emotional state it is in
by a pathway that can be causally independent of what, if anything, the
protagonist feels. Thus, cases of congruent emotions between viewers and
protagonists, though admittedly quite frequent, are typically not true-blue
instances of infectious identification, but are better regarded as cases of
coincidentally congruent emotional states engineered by means of criterial
prefocusing.
Indeed, it seems to me that postulating infectious identification in
most cases is to take on excess theoretical baggage, since typically congru-
ent emotional states, where they occur, can, all things being equal, be
adequately explained in terms of criterial prefocusing. For what it is worth,
in all the cases that I have examined, criterial prefocusing gives a perfectly
acceptable account of the rhyming emotional states. Thus, in the main,
talk about infectious identification violates the principle of explanatory
parsimony.
Of course, it may be pointed out that often the way in which situations
are criterially prefocused in movies tends to parallel the way in which the
protagonist sees things. Even if the narration is omniscient and not
channeled explicitly through the point of view of the protagonist, movie
directors often depict the fictional world from the perspective of leading
characters. The gloominess of the portrayal of the environment, for example,
may echo the apprehensiveness of the hero as he enters the hiding place of the
evil wizard Voldemort.
This is a fair point, yet it does not revive the model of infectious
identification. For, on the one hand, the viewer need not be aware that
it is the character's viewpoint that he is being invited to take on; rather, he
may suppose it to be the perspective of the narrating agency (howsoever
he understands that). And, furthermore, the criterial prefocusing here will
work in the same way whether or not it is crafted in a manner that reflects
the point of view of the protagonist.
Here it is interesting to think about cases where the fiction overtly
establishes that the way in which events are criterially prefocused is
congruent with the emotive states of the protagonist. Striking examples
of this are the use of point-of-view shots in motion picture editing. The
character looks offscreen, her face contorted with horror; then there is a
shot of what she sees — a shot of a suppurating creature, part reptile, part
 AFFECT AND THE MOVING IMAGE 169

arachnid, part lawyer, with a maw like a chainsaw — and we are horrified
too. Isn't that patently a case of infectious identification?
Again, I think that it is not, for the simple reason that we would probably
feel the same level of horror if the sequence were shown without the
character looking offscreen. An interesting experiment might be to remove
the emotive povs from a movie like The Descent in order to assess whether
our repulsion at the sight of those slimy, albino cave-dwellers dwindles.
Of course, this leaves open the question of why such point-of-view
shots — and other perspective-disclosing devices — are used by fictioneers.
The short answer, I think, is that they are a means of priming or preparing
or communicating to the audience in a very broad way the general kind of
affect (dysphoric or euphoric) that the audience should bring to bear on the
objects, persons, situations, and events they are about to encounter. In this
way, the point-of-view structures reinforce the affective information avail-
able on screen. But I will have more to say about how this communication
works in my subsequent remarks about mirror reflexes.

Vectoriallj converging emotive states

So far the second condition of the model of infectious identification has
been hammered on the grounds that, even if the audience is in an emotive
state congruent with the imagined mental state of the protagonist, that is
generally the result of the viewer having arrived at that state by a process
located in his or her own emotive-appraisal system, independently of any
necessary causal input involving the fictional character being in the type-
identical emotional state. In other words, as those jet bombers race toward
Beijing, I fear for humanity, because human life as we know it is endangered
in the fiction, and not because the fictional president fears for humanity.
My fear is co-incident or conjoint with the protagonist's, while, at the same
time, being causally independent from it. Or, at least, this seems to be what
is going on by and large in the most significant number of cases.
Nevertheless, it is important not to misinterpret this example. The
claim is not that we are never emotively influenced by the emotional states
of characters, especially the protagonists and others to whom we have a
favorable attitude. For example, at the end of Charlie Chaplin's The Gold
Rush, the Tramp accidentally meets up with Georgia, the woman of his
dreams. The two embrace, they kiss, and they fade out into the land of
happily ever after. They are in love, and their successful match gladdens us.
Scenes like this happen all the time in popular fictions. Our previous
objections to the infection model of identification are not to be taken as
170 AFFECT AND THE MOVING IMAGE

an attempt to disavow them. For cases like these should not be taken as
cases of infectious identification in the first place.
True, the emotional states of the characters do cause us to be in a
euphoric state. But our euphoric state is not precisely the same type of
euphoric state that the lovers are in. Their emotional state is infatuation.
That is not our condition; we are happy for the couple. I am not in love
with Georgia nor am I identifying with that aspect of the Tramp. Were
I in love with Georgia, I wouldn't be so happy. I'd be jealous of the Tramp.
So I am not in a state of infectious identification.
Yet I am in state of roughly the same emotive valence. They are, let us say,
euphoric and I am euphoric as well. Our emotional states converge — they
both belong on the positively charged side of the scale of the emotions. We
are not in the same emotional states, but our conditions are in broad
categorical agreement and we are in that vectorially converging state with
the state of the characters because that is the condition in which we imagine
them to be.
Contrariwise, when the monster in the concluding scenes of Bride of
Frankenstein is reviled by his reanimated betrothed, we feel sorry for him.
Our emotion does not match his. We do not feel the pain of an unrequited
lover. Indeed, I doubt that any viewers, no matter how desperate, harbor any
desires for the frizzy-haired, electrified corpse, played by Elsa Lanchester.
Yet we do respond to the monsters misery with sorrow. It is in this sense
that we share his misery. We are not miserable for being lovelorn, but we do
pity the monster.
Both misery and pity, of course, are dysphoric or negative emotions.
Both sit on the distressful, discomforting, disturbing, or painful pole of
emotional states. Again, our emotions are broadly similar in their general
valence. They converge vectorially in their negative directionality. Our
emotions are causally coordinated. But this does not count in favor of the
model of infectious identification, unless identification means nothing
more than a somewhat similarly charged feeling. However, why mobilize
the notion of identity to describe that?

Simulation theory
We appear to be emotionally tied to movie fictions predominantly through our
relations to characters, particularly those called protagonists. But what is the
nature of that relation — at least in the largest number of cases? We have just
been arguing that the concept of identification does not seem to do it justice.
Perhaps the relationship is an instance of what is nowadays called "simulation."
 AFFECT AND THE MOVING IMAGE 171

Simulation is a concept from the philosophy of mind that has recently

been imported into aesthetics. In the philosophy of mind, the idea is that
we understand and explain others, ascertain what they are feeling, identify
their intentions, emotions, and motivations, and predict their behavior by
simulating them. That is, we input their beliefs and desires into our own
off-line cognitive-conative system. Since, ex hyposthesi, our cognitive-conative
architecture is the same as theirs, if we run their beliefs-desires program on
ourselves — using their software on our hardware, so to speak — then we
should be able to derive a reasonable fix on what people just like us (folk-
psychologically speaking) are likely to think, feel, plan, want, and so forth.
That is, we put them into our shoes — or to talk in an equally metaphorical
manner, we put their beliefs and desires into the black box that outputs our
emotions, decisions, and behaviors — in order to project what they are likely
to feel, think, or do. This, moreover, is done on the presupposition that with
respect to processing beliefs and desires into emotions, intentions, and
deliberations, etc., we are pretty nearly the same structurally.
This theory — called simulation theory — is counterposed to another view in
the philosophy of mind concerning the way in which we go about understand-
ing others. This alternative view is called the theory-theory. This alternative
view is the theory that we understand what others are about by applying
something like a scientific theory to their behavior. Simulation is thought to
have several advantages over the theory-theory when it comes to reading
the minds of others. First, the theoretical framework ostensibly presupposed
by the theory-theory would be monumentally complicated — too complex
to postulate with any confidence. After all, it would have to be bigger and
better than our best existing social scientific theories, since its accuracy would
appear to be much greater. Do we really believe that most of us have in our
possession such a powerful theory? And if so, why can't our social scientists
access it explicitly? That would save them a great deal of effort in the lab.
Second, even if we ordinary folks did have such a mammoth theory at
our disposal, if only sub voce, it would surely take an immense amount of
real time to apply it to particular cases — simply figuring out which laws the
case fell under and how the influence of related laws was to be calculated
would seem to take forever. But this does not seem to match up to our
everyday experience of reading the minds of others. Quite often, we size
up people's emotional states and intentions in a blink of the mind; we don't
spend hours crunching our way through an elaborate theory. Or so the
simulation theorists contend.
But what does all this have to with motion pictures — specifically with
the question of our emotional relationship to the protagonists in popular
172 AFFECT AND THE MOVING IMAGE

fictions, like movies? In short: putatively our relationship with fictional

characters is generally like the relationships we have with other people in
the world outside the cinema. The movie fiction invites us to understand them,
to apprehend what they are feeling and intending, to speculate about what they
will do next, and so forth, just as we must mind-read our conspecifics in
everyday life. In order to do so — in order to follow the motion picture fiction
(the unfolding of the intentions and feelings of fictional characters) — we
simulate them, or at least the leading ones, primarily the protagonists.
(Probably it is implausible to suppose that viewers simulate the situations of
minor characters, like the plight of the lawyer in the outhouse in the first
installment ofJurassic Park — or to think that viewers are meant to do so).
When we simulate a character, like Spiderman, we are said to input their
beliefs and desires into our cognitive-conative network and thereby to dis-
cover what they are thinking and feeling by coming to think and feel broadly
similar things as our own system runs through its paces on their steam. On
this view; we are engaged emotively with motion picture fictions (as well as
other sorts of fiction) by being immersed in a virtually continuous process
of replicating the emotions and desires of (especially) the protagonists.
Unlike the proponent of infectious identification, the simulation theorist
does not claim that the emotions allegedly reproduced in us are identical,
either thoroughly or in part (aspectually), with those attributed to the
protagonist. At best, our emotions only approximate those of the characters.
However, this still leaves the simulation theorist open to one objection that
was leveled at the champions of infectious identification — namely, the
charge that simulation theory can be nothing close to a comprehensive
theory of our emotional relationships to fictional characters, and this is so
because of the indefinitely vast number of cases where the emotions of the
audience are and should be different from and sometimes even opposed to
those suffered by fictional characters.
This is the asymmetry problem once again. When the protagonist rushes
home joyously to tell her spouse that she's been promoted at the same time
that we, the audience, know she will open the door only to find her family
brutally massacred, our despondent state is nothing like hers. Moreover, if
at that moment we were simulating her psychology, we would not be
following the story appropriately. We are supposed to be feeling pity. But
since similar asymmetries are so often the rule in motion picture fictions,
simulation cannot be anything like a comprehensive account of our emotive
engagement with the fictional characters in movies.
Of course, as we conceded earlier, sometimes the emotions of the
audiences and the characters do converge. Mr. Smith is outraged by
 AFFECT AND THE MOVING IMAGE 173

corruption in the Senate and so are we. Perhaps simulation explains cases
of convergence. But this seems unlikely. Motion pictures proceed at a pace
that would seem uncongenial for simulation. Supposedly one of the
advantages of the simulation theory over the theory-theory is that simula-
tion is more temporally suited to sussing out the emotive states of our
conspecifics. But be that as it may, simulation takes time too and one
wonders whether one typically has sufficient breathing space in which to
simulate in reaction to a rapidly edited audiovisual array.
Furthermore, another problem with simulation is that it is a firmly
established fact that people are notoriously unreliable in identifying their
own emotions and intentions. So how likely is it that such unreliable
subjects will be able to extrapolate correctly from what they take to be
their own case to the case of another?
Moreover, where the other being is a fictional character of the order of a
movie hero, the likelihood that we are using simulation to predict their
behaviors is especially implausible. Movie heroes don't shirk their duty in
the face of overwhelming odds. When surrounded, Rambo lights into his
assailants. How many of us — given the black box of our cognitive-conative
system — would really reach the same decisions as Rambo does? Given his
beliefs that these bruisers are bad guys and his commitment to justice,
Rambo goes on the warpath. Wouldn't the rest of us just decide meekly to
be arrested? As moviegoers, we probably anticipate that Rambo will not go
quietly. But it is improbable that most of us reach this surmise by simulating
Rambo, since if we were actually simulating, we would anticipate that he
would surrender, wouldn't we?
But an even deeper question of how the simulation account can reach
the right answers in cases like these is the question of how often simulation
can be supposed to occur in our responses to motion pictures. I contend
that, if it ever occurs, it is nevertheless very unlikely that, with respect to
popular movies, it is occurring very often. Why?
Simulation theory in the philosophy of mind is advanced as a view about
the way in which we go about determining what our conspecifics are
thinking and feeling — a way of understanding and predicting their not
transparently fathomable behavior in the course of daily affairs. But popular
motion picture fictions are not like everyday life. They are expressly
designed to be understood — indeed, they are designed to be understood
quickly and clearly by untutored audiences.
Needless to say, this aspiration extends to the way in which the fictional
characters are constructed. Perhaps in our ordinary experience, our con-
specifics strike us as opaque in a way that calls for simulation. I would not
174 AFFECT AND THE MOVING IMAGE

say that this never happens, though I am not convinced that it is happening
all of the time. Nevertheless, I do contend that the kinds of situations that
call for simulation occur rarely with respect to the kind of fictional
characters who inhabit movies, because those characters are intentionally
fabricated in such a way that they wear their feelings and their thoughts on
their sleeves. Thus, there is little or no need to hypothesize the operation
of simulation in response to popular motion picture fictions, since we
usually know exactly what the fictional characters are feeling and thinking
faster than it would take to simulate said fictional beings.
Does it seem impossible that we might penetrate into the heart of
a fictional being so easily? To establish that it is not impossible, let us take
a brief digression into literary fiction before returning to the case of movies.
Most popular literary fictions employ the device of free indirect discourse.
This means that the author can narrate what the character is thinking
and feeling from both the inside and the outside. We have her context
described for us, often in emotionally suggestive terms, her physical states
delineated, and we are also made privy to her thoughts. In such circum-
stances there is no call for simulation. We are just told what the character
is feeling and tiiinking. Perhaps some readers use the text as a script for
attempting to raise similar emotions in themselves. I don't, but I wouldn't
want to claim that others are like me in this respect. Nevertheless, it
should be clear that there is no pressure to mobilize simulation theory in
cases like this in order to explain how we come to grasp the feelings and
thoughts of the protagonist. We are told them outright.
Of course, this feature of literary fictions is not as common in audio-
visual fictions (though there are exceptions, from Diary of a Country Priest to
The Twilight Zone). That is, we do not typically enter the minds of
protagonists as frequently in movies and TV as we do in literary fictions.
But, on the other hand, it is true that quite often the characters in these
artforms do tell us precisely what they are feeling and what they are
thinking, often by way of dialogue with interlocutors. And even in those
cases where the characters do not explicitly articulate their state of mind,
I would still urge that the characters in movies manifest their feelings and
thoughts so openly that simulation is effectively beside the point.
How is this possible? Perhaps one way to get at this is to ask whether
with respect to everyday experience simulation and the theory-theory
exhaust our way of gaining access to the feelings and thoughts of others.
Arguably, they do not. Often — in fact, probably most often — we impute
thoughts and feelings to others on the basis of schemas, scripts, prototypes,
contextual cues, exemplars, and other heuristics, rather than by means of a
 AFFECT AND THE MOVING IMAGE 175

theory or a simulation. These stratagems, in turn, enable us to infer the

inner states of others.
Confronted with a co-worker who has just been fired or a friend who has
lost a loved one, we infer on the basis of a culturally shared script that, all
things being equal, they are down. If their shoulders are slouched and their
eyelids heavy, we recognize these as signs of distress in virtue of our prototype
for distress. We have no need to perform a simulation, because, through
experience, we have amassed a repertoire of schemas, scripts, prototypes, and
contextual and/or recognitional cues, among other heuristics, to key us to the
inner states of conspecifics.
When we hear that a relative has secured a long-sought-after job, all
things being equal, we suppose that she is happy and we rejoice for her.
There is no need for simulation. We have access to a body of prototypes
regarding emotional responses in certain contexts as well as recognitional
cues, such as facial expressions and postures, which enable us to assess the
emotional states of others. These are not theories and they are not applied
by subsuming particular situations under nomologieal generalizations as the
theory-theory might have it. Rather, they are prototypes — schemas, scripts,
and recognitional cues — employed by analogy (analogy rather than sub-
sumption). In daily life, such strategies provide us with reasonably reliable
means for tracking the emotional states and thoughts of conspecifics, and
I, at least, am pretty confident that we depend on them in everyday life far
more than we depend on simulation. For, quite frequently, simulation would
just be too time-consuming.
Moreover, if this is right, then it has important ramifications for popular
fictions like movies. Inasmuch as popular or mass fictions, like movies, are
designed to maximize accessibility, they gravitate naturally toward the use
of the schemas, prototypes, exemplars, contextual and/or recognitional
cues, and other heuristics that abound in the cultures of their target
audiences. Indeed, it is part of the art of designing a character for the
purposes of mass consumption that one be able to streamline the details of
the character's persona so that it calls forth almost automatically the
scripts, schemas, recognitional cues, and other heuristics that are intended
to fit it.
In the situation comedy Sex and the City, the character Samantha snugly
fits the schema of the carnal woman — the sensual woman who loves sex
(e.g., the Wife of Bath). When her eyes open wide as a handsome swain
glides by, we do not need to simulate Samantha in order to determine
her internal state. We can infer it in a split second based on our prototype
for carnal women and the recognitional cues that Samantha supplies.
176 AFFECT AND THE MOVING IMAGE

Moreover, with respect to this particular TV series, our surmise will almost
always be confirmed when Samantha slyly, albeit redundantly, confides
her desires to her friends, Carrie, Miranda, and Charlotte. Nor does this
seem to me to be a peculiar example of the way in which characters in
popular motion picture fictions function. Instead it appears to be the norm.
Perhaps supporting evidence for this hypothesis can be drawn from our
reactions to movie villains. Movie villains, especially very, very evil ones
like Michael Meyers, are probably among the most psychologically opaque
beings around. By rights, then, we should try to simulate them. But
I conjecture that we never do. Why? Because, we render them intelligible
by means of our schema for stalker/slashers (though this is a schema that
admittedly derives more from the movies than from daily experience —
thank God).
The point of emphasizing the operation of recognitional cues and other
heuristics in negotiating our emotive relationships to fictional characters
in movies, of course, is to show that there is little motivation for hypothe-
sizing the operation of simulation in response to the protagonists in popular
motion pictures. For the problem that simulation is supposed to solve with
respect to understanding the feelings and thoughts of others in everyday life
does not generally arise with regard to popular motion picture fictions because
movies intensively exploit the schemas, scripts, prototypes, exemplars, and
contextual and recognitional cues that comprise our heuristics for discerning
the inner lives of others.
Moviemakers build characters, economically sculpting their features,
precisely to trigger quickly and effortlessly the mobilization of those
prototypes and heuristics by audiences. Thus, the need to postulate the
operation of simulation as the means by which we apprehend the emotions
and thoughts of others is largely otiose.
But what, the simulationist might ask, of motion pictures, of a more
exacting sort — such as the art cinema of the 1960s? Those characters do
not show their psychology so unabashedly. Aren't they, at least, ripe for
simulation? I suspect not — for the simple reason that most of those characters
are simply too opaque to simulate. Their beliefs and desires are often far too
murky for us to process through our own cognitive-conative architecture. We
don't know what to put into our own off-line system. This is not said to
chastise these motion pictures. Often their point is that the human heart is
ultimately too mysterious to plumb. Rather, my point is simply that even in
these cases, simulation appears irrelevant.
Therefore, once again, the need to posit simulation appears beside the
point. But if we are not simulating the emotions of the fictional characters
 AFFECT AND THE MOVING IMAGE 177

in moving pictures, then it is not the case that were are typically bound
emotionally to them by a continuous process of sharing congruent feelings.
So what is going on here emotionally?

Sympathy
We are emotionally tied to movies in large measure through our relation-
ship with characters, especially the protagonists. But neither the model of
infectious identification nor that of simulation appears to explain satisfac-
torily the general structure of this relationship in a comprehensive manner.
Whether or not infectious identification or simulation ever occur, they
occur far more rarely than is often supposed and, therefore, cannot afford
a comprehensive account of the emotive address of characters in movies.
A problem with the model of infectious identification and simulation
theory in this regard is that they postulate a more, rather than less, closely
shared emotional state between viewers and characters, whereas, as we
have seen, so often the relation is asymmetrical. Consequently, perhaps the
place to look for an account of our relation to fictional characters is a
condition where what we feel and what the protagonists are thought to feel
are categorically different.
An obvious candidate for such a bond is sympathy. Sympathy is not an
emotional state that persons bear to themselves. It is, by definition, directed
at others. For our purposes, let us construe sympathy broadly as non-
fleeting care, concern, or, more widely, a non-passing pro-attitude toward
another person (or fictional character, including anthropomorphized beings
of all sorts). Sympathy qua emotion is a supportive response. It provides an
impulse toward benevolent action with respect to those to whom it is
directed, though, of course, that impulse need not and often is not acted
upon, frequently because it conflicts with other interests that we might have.
And, needless to say, with the fictional characters in movies, the sympathetic
impulse cannot be acted upon. Perhaps, one reason why we are so free with
our sympathies toward fictional characters is that, since we need not ever act
on their behalf, their needs never threaten to fall afoul of our interests.
Indeed, it is probably because such sympathies come so cheaply that
moralists, such as Augustine and Rousseau, have perennially distrusted the
benevolent feelings elicited by fictions.
But, in any case, sympathy, conceived as an emotion, involves visceral
feelings of distress when the interests of the objects of our pro-attitude
are endangered, and feelings of elation, closure, and/or satisfaction when
their welfare is achieved. The emotion in question has as a component the
178 AFFECT AND THE MOVING IMAGE

enduring well-being of its object — a desire that things work out well for
her. In order to be the object of this pro-attitude, the person in question
must be thought to be worthy of our benevolence in light of our interests,
projects, values, loyalties, allegiances, and/or moral commitments. When
some fictional character so-and-so is appraised to be worthy of our non-
passing desire that things go well for her and this is linked to positive
feeling tones when gratified and negative ones when frustrated, we are in the
emotional state that I am calling sympathy. If "sympathy" strikes you as too
saccharin-sounding, then you are welcome to refer to it as "benevolence," or
even more aridly as "a pro-attitude."
Though sympathy may initially appear to be just another example of a
vectorially converging state, it is important to note that it need not be. For
sympathy does not always — and in any event does not necessarily — track,
even vectorially, the way in which the character feels. This is due to the fact
that sympathy concerns what we believe to be the genuine well-being of the
character. Should the heroine fall head over heels for some lounge-lizardly
Lothario, she might be in ecstasy, whereas our sympathetic response would
be anxiety-ridden, since we surmise that she is headed for trouble. Although,
on occasion, sympathy may converge vectorially upon the emotions of the
characters we care for, this is not required for the state in question to count
as sympathy.
The suggestion that sympathy plays a role in our emotional involvement
with fictional characters is fairly unexceptionable. However, what is being
claimed now is more than that. I am arguing that sympathy, along with
antipathy (which we will discuss presently), constitutes the major emotive
cement between audiences and the pertinent movie characters. But why
suppose that sympathy holds this place of privilege?
Obviously, during the course of a motion picture, the viewer undergoes
many emotional states. One is angry for awhile, then sad, then happy, then
gripped with suspense, some laughter erupts, some tears, and then we are
happy again. Sometimes sympathy for the protagonist is so strong that you
can feel it. At other times, it may appear to take the back seat for an
interlude of comic amusement. So why select out of this welter of affects
sympathy as the premier emotion?
The first reason might be called its breath. Sympathy for the protagonist
is the most pervasive emotion from the beginning to the end of the movie.
As soon as sympathy is secured, unless it is later intentionally neutralized
by the creator of the motion picture, it stays on the alert, following the
protagonist's fortunes throughout, registering distress as they waver, and
pleasure as they rise.
 AFFECT AND THE MOVING IMAGE 179

Sympathy, once enlisted, is constantly on call throughout the movie.

Generally no other emotive stance — save perhaps antipathy toward the
villain — is so long-lasting. The indignation we feel toward the surly prison
guard who cuffs the hero comes and goes. But our sympathy for the
protagonist endures. It provides the emotive optic through which we
survey the narrative from one end to the other. Typically, each event in
the story is weighed in light of our sympathy for the protagonist ■ of every
event it is pertinent to ask whether it advances or deters her fortunes, even
if, in some cases, the answer is neither.
The protagonists in motion pictures have goals and interests that are
hard to miss. The narrative trajectory usually involves the accomplishment
of these goals and the satisfaction of these interests in the face of various
obstacles. We follow this quest from the perspective of sympathy, cheering
the protagonists onwards as they advance and feeling consternation when
they falter. Of course, as already noted, sometimes our sympathy for the
character puts us out of synch with them — as when we believe the person
the protagonist trusts to be a scoundrel. But still we track the unfolding
narrative in terms of a sympathetic viewpoint, one that disposes us to care
about his best interests rather than his subjective assessment of them.
Sympathy is the most persistent emotional bond that we have with
respect to the fictional protagonists; in this sense it generally possesses
more breadth than other emotions elicited by moving pictures. Further-
more, sympathy also has what might be called special depth. For it is our
sympathy toward the protagonist that shapes our overall reception of the
action. When we are angered by the way in which the heroine is mis-
treated, that anger itself is subsidiary to the sympathy that we bear toward
her. That sympathy underlies and reinforces our anger. It is our sympathy
for the character that disposes us to regard her as inside our network of
concern, and, therefore, to assess an injustice done to her as something
perpetrated against one of "our own." The negative emotions that we
muster in response to the protagonist's setbacks are a function of our
sympathy for her. Sympathy is the real foundation here. That is why we say
it has depth, even special depth.
Two considerations then can be marshaled to endorse the hypothesis that
sympathy is the leading emotional bond between viewers and movie prot-
agonists, namely, that sympathy appears to have greater breath and greater
depth than any other competing candidate. The obvious exception here
may be the antipathy, distaste, and hatred we often direct to antagonists,
a.k.a. movie villains. However, though more needs to be said about this, since
antipathy for the bad guys is usually the reverse side of — and, indeed,
180 AFFECT AND THE MOVING IMAGE

a function of— the sympathy we feel toward their rival protagonists, the case
of antipathy is less of a counterexample to our hypothesis than it is a
corollary.
Granting that sympathy is the glue that keeps us connected emotionally
to the protagonists, the question remains about the way in which movies
are able to recruit our sympathies as effectively as they do. In life outside
the movies, our benevolent or altruistic attitudes toward others depend on
factors such as kinship, group memberships of all sorts, and group inter-
ests. Of course, for both artistic and financial reasons, the creators of
movies are aiming at larger audiences than a single extended family and
often at audiences that cross regional, ethnic, national, and religious
boundaries. And even where their targets are less than global, they must
be careful not to trigger the sectarian differences that always exist in
virtually every large group. This clearly presents the creators of popular
motion pictures with a problem to be solved, namely, how to enlist the
care and concern — the sympathetic feelings — for their fictional prota-
gonists from mass audiences of heterogeneous backgrounds and different,
often potentially clashing, interests.
That is, if sympathy is the crux of our relationship to the relevant
characters in the movies, how is this sympathy mobilized? In everyday life,
we extend our sympathies to those with whom we share interests or projects
or loyalties, or to those who exemplify values of which we approve, or to
those who fall under the protection of certain moral principles. But most
of the interests, projects, and loyalties upon which we base many of our
quotidian sympathies are highly specific to us. Needless to say, the movie-
maker cannot hope to activate on behalf of the protagonists the individualized
interests of every viewer. Rather, she must aim at engaging the audience at a
fairly generic level of interests, projects, and loyalties. That is, she must find
some common ground or touchstone amongst the diverse audience which
will encourage us to find the protagonists to be worthy of our good will.
This is a design problem for the popular moviemaker. She must find a
way in which to elicit from a disparate audience the converging desire that
the protagonists do well — that is, she must elicit our felt conviction that it
would be good for the protagonists to do well, or that they deserve to do
well. What is the solution to this problem?
As a matter of empirical generalization, I conjecture that the most
common answer to this challenge is the creation of protagonists who
command the audience's moral endorsement. In other words, morality, of
an extremely broad cast, provides the moviemaker with an interest, or
project, or loyalty upon which viewers of diverse backgrounds can converge.
 AFFECT AND THE MOVING IMAGE 181

Moreover, inasmuch as there are certain moral touchstones shared across

wide populations, the moviemaker has a lever on their allegiance.
By presenting protagonists who are morally appealing, the moviemaker
secures the criterial wherewithal to garner the sympathy that is required
for her intended audiences to be absorbed by the story. Such protagonists
will meet the criterion of being deserving of our benevolence because they
are morally deserving. It is no accident that the protagonists in movies are
good guys. Good guys are precisely what the movie doctor calls for —
characters likely to engender a pro-attitude from heterogeneous audiences
of otherwise varied and often conflicting interests and loyalties. Morality of
the fairly generic sort found in movies is just what people from different
backgrounds are apt to agree upon, at least roughly. For instance, few
would disagree that Maximus is wronged when his family is slaughtered in
Gladiator, or that Cinderella is badly treated by her stepmother and sisters,
or that Superman is a very nice person or that Story in Lady in the Water
should not be eaten by the Scrunch.
Morality, especially of a fairly widely shared and often nearly universal
gauge, supplies the moviemaker with the interests, or projects, or loyal-
ties, or touchstones of affirmation which audiences from similar cultures,
and even, frequently, from dissimilar ones, can agree. The protagonists, for
example, will typically go out of their way to protect the lame and the halt,
the helpless and the sick, the very young, the old, and the defenseless,
while simultaneously treating them with dignity and respect. They evince a
sense of fairness, justice, loyalty, honor, and honesty, and are altogether
pro-social and especially pro-family (in principle, even if they re single), at
least where the families in question are portrayed as wholesome ones (as
opposed to the Texas Chainsaw Clan). These characters tell the truth and
their keep their promises to good people, because they, themselves, are
what we call good people.
The protagonists in movies are presented as morally righteous. This is
undeniable in the greatest number of cases. But even so-called anti-heroes
usually oppose some form of compromised moral order in the name of a
deeper sense of rectitude. Once you get past their gruff exteriors, hard-
boiled detectives always seem to discover that society is even more
disreputable than they are. In the end, with respect to his girlfriend,
Sam Spade seems almost Kantian in his sense of duty. Likewise, alienated
gunslingers protect the little folk against greedy cattle barons, bullying
railroad magnates, and, in the case of The Magnificent Seven, marauding
warlords. Disaffected teenagers, in turn, really care about problems that
adult society culpably neglects or misunderstands. And so on. No matter
182 AFFECT AND THE MOVING IMAGE

how anti-social the protagonist appears at first glance, he or she is quickly

revealed to be pro-social at heart. Moreover, this is how it should be if
sympathy is to take hold across a diverse audience whose likeliest point of
convergence is apt to be morality, very generously construed.
By "generously construed," I do not simply mean that the audience
shares some rough-and-ready principles or rules of thumb, but also a sense
of what counts as virtuous. Protagonists win our approval and thence our
sympathy, because they are typically portrayed as persons of a variety of
virtues. Recognition of these virtues then precipitates in audiences the
yearning that the protagonists do well and that their rivals do badly. Our
desire is usually so strong that we generally stay to the end of the film just
to make sure that things work out this way.
Of course, it must be conceded that it is not always the case that the
protagonists with whom we align ourselves are what we would call morally
upright; and yet they still get our sympathy. Consider, for example, caper
films. However, when one examines the structure of these exercises one
notices that the fictional world has been constructed in such a way that
the aforesaid protagonists are the most virtuous characters in evidence.
The forces of everyday society are either kept off-screen so that their
countervailing claims never impinge on our sympathies or they are shown
to be venal and/or inexcusably stupid. That is, where the universe in the
moving picture is a fallen one, as it is in the TV series The Sopranos, a figure
like Tony can warrant our sympathies, because, even though he is not
moral in any absolute sense, compared to everyone else in the fictional
New Jersey where he lives he is the best of a very, very bad lot.

Solidarity

If sympathy is as central to our emotional response to movie protagonists

as has been argued, then instances of infectious identification cannot be
very common, since sympathetic feelings, conceived as emotions, are not
ones that we share with their objects. Sympathy, as mentioned earlier, is,
by definition, directed at others. The protagonist does not feel sympathy
for himself; were he to do so to any appreciable extent that would probably
turn us off. However, though essential to our emotional bonding with
protagonists, sympathy pure and simple is usually not the whole story of
our emotional connections to protagonists. It is generally supplemented by
another emotional state which we may call solidarity,
Some fictions have only protagonists. The film Just My Luck is an
example. There are two major characters — Ashley Albright and Jake
 AFFECT AND THE M O V I N G IMAGE 183

Harden. Their names are clues to their most important attributes. She is
always the beneficiary of deliriously good fortune. When she leaves her
apartment without an umbrella, it immediately stops raining. When she
hails a cab, it arrives from nowhere. And so on. Jake is just the opposite.
If he bends down, his pants will split; if he picks up a five-dollar bill, it
will be smeared with canine feces. But their luck magically changes hands
at a masquerade ball when they kiss, anonymously, while dancing. Now
everything Ashley attempts leads to disaster, while Jake becomes a very
successful record producer. The rest of the story involves Ashley tracking
Jake down in order to reclaim her good fortune with another kiss.
However, they fall in love and kiss their way to some kind of providential
equilibrium; they will live happily and unhappily ever after in the normal
proportions.
What is striking about Just My Luck — and this is also true of many other
romantic comedies — is that there is no real villain. There are some people
who present temporary obstacles to the main characters, but they are not
full-fledged antagonists. They are not on the scene long enough for our
antipathy toward them to take root, and, anyway, by the end of the film,
they all may turn out, by twists of fate, to be nice people after all. With
Just My Luck, we are encouraged to feel care and concern for Ashley and
Jake. But there are no real bad guys.
The majority of movies, however, are not like this. Most pit the
protagonists and the other nice people against some adversaries. We are
not only prompted emotionally to embrace the good people as members of
a generic "Us"; their opponents belong to "Them." Moreover, these
"Them" are not just regarded as the opposing team. They are usually
presented as people we hate, indeed, often love to hate.
If sympathy toward "Us" is characteristically elicited by portraying the
protagonists and the other nice people in the fiction as morally good* then
the antipathy generated toward "Them" is generally provoked by repre-
senting them as morally blemished. Whereas the protagonist is nice to nice
people — treating good people with good manners — the villain is at least
rude to those he perceives to be his inferiors and very often what he does
to them is much worse. The antagonists pillage, cheat, lie, rob, rape, kill,
and so on. The hero pets the old sleeping dog on the doorstep; the bad guy
kicks it out of his way.
Movies are generally political in the sense defined by the philosopher
Carl Schmitt. The fictional population in the motion picture is standardly
partitioned into friends and enemies — into Us and Them. Sympathy,
motivated by morality, disposes us to assimilate the protagonists into Us.
184 AFFECT AND THE MOVING IMAGE

But that bond is then further strengthened by introducing enemies who, in

movies, are customarily marked by being constructed as bad people. As
I am now using the concept of "solidarity," it is the name of the complex
emotive relation of sympathy-for-the-protagonists plus antipathy-for-the-
antagonists. This is a psychological condition that most moving pictures
strive to raise in audiences. They aspire to dispose the viewer to feel
emotionally allied to the protagonist and against the antagonist. The
antagonist is designed to instill anger, indignation, hatred, and sometimes
moral disgust in us.
Unlike sympathy in everyday life which we tend, all things being equal,
to extend quite readily to most of those around us by default, in movies
our sympathy must be won. The most efficient way of doing this is to
render the relevant characters morally appealing. Our bond with these
characters can then be reinforced by setting an array of nemeses against
those whom we find repulsive, customarily because their various failings —
from petty vices to outright viciousness — are emphatically foregrounded.
I suspect that, in most cases, our antipathy for the villains is engendered
by their vices, whereas sympathy for the various positive characters is
primarily elicited by their superabundantly evident virtues.
However, in some instances, characters may secure our sympathy on the
rebound, so to speak. That is, we may be so appalled by the villain that our
care and concern goes out to whomever he opposes. In other words,
sometimes solidarity in movies may result from the "enemy-of-my-enemy-
is-my-friend" phenomenon. This, furthermore, explains why it is that we
sometimes find ourselves siding with characters whose morals we do not
otherwise share, like Hannibal Lector. We side with Lector in Ridley
Scott's Hannibal, I suspect, because we find his enemies — those who
plan to feed him alive to pigs — to be so much more morally depraved
than he is. ^ ''
Solidarity involves sympathy and antipathy viscerally felt. Though our
sympathetic feelings toward the protagonists are not shared by them, our
hatred for the villains often is. But we do not hate the villain just because
the protagonists do. Rather, we have detected co-incidentally on our own
many of the same morally noxious traits of the antagonists that have
impressed the protagonists so deeply. Thus, when our hatred is congruent
with the protagonist's hatred of the bad guy, the antipathy component of
our feelings of solidarity with the hero is essentially conjoint with his. We
may air our hatred for the villain at the same time the protagonist does, but
primarily as a result of our own process of moral appraisal which has been
elicited by criterial prefocusing.
 AFFECT AND THE MOVING IMAGE 185

Mirror reflexes
We have been exploring several kinds of emotional relationships between
audiences and movie characters. These have included circumstances:
(1) where our emotions may be type-identical with those of the protagonists,
but due to our own independent appraisal of the relevant situations
(in response to criterial prefocusing); (2) where our emotions vectorially
converge on the emotional states of characters; (3) where we sympathize
with the characters (which, of course, may, but need not, be an instance of
a vectorially converging emotion); and (4) where we emote in solidarity
with the protagonist (where the antipathy component may be a response
congruent with the protagonist's though one we find ourselves in as a result
of our own appraisal of the situation). None of these cases corresponds to
the popular model of infectious identification which requires that the
audience member be in an emotional state that is type^identical with that
of the protagonist just because the pertinent character is in that state. But
let us conclude by briefly examining a fifth affective relation — one which
comes closer to the model of infectious identification, and whose very
existence has probably lent some credence to the ubiquitous talk about
identification and simulation with movie characters.
What I have in mind here are what may be called mirror reflexes. By
calling them "reflexes," I mean to signal that they are not full-fledged
emotional states. Consequently, the existence of mirror reflexes does not
corroborate the model of infectious identification, though it is understand-
able that some might think it does. But, at the same time, no account of
our affective relationship to motion pictures would be adequate without a
discussion of mirror reflexes. So, what are mirror reflexes?
Occasionally when speaking to another person, we suddenly realize that
we have adopted their facial expression. They are frowning; we start to
frown. Or, they are smiling and we find ourselves smiling. We have an
unmistakable tendency to ape our interlocutors — to mirror them. This is the
case not only with facial expressions, but also with gestures and postures.
We tend to fall into step with our companions; when they're walking tall, so
are we. If our informant bends toward us conspiratorially, we bend toward
them in response. When we watch the outfielder stretch to intercept a fly
ball, the muscles in our arm tug slightly in that direction. And so forth.
We have an involuntary tendency to mirror automatically the behavior,
especially the expressive behavior, of our conspecifics.
Putatively, we do this in order to gain some sense of what they are
feeling. By configuring our own facial expression after the fashion of a
186 AFFECT AND THE MOVING IMAGE

fellow conversationalist — grimacing when she grimaces — we secure some

inkling of her inner state. The feedback from the disposition of our facial
muscles buzzes our autonomic nervous system in a way that is presumably
isomorphic to what is going on in her system, so long as she is not
dissembling. This does not give us full access to her emotional state, but
it provides an important clue, since it yields something like a facsimile of
the bodily-feeling component of her overall state. It does not tell us her
precise appraisal of the object of her state, but it does relay the visceral
sensation that goes with that appraisal.
Similarly, gestures and postures are also mimicked — though usually only
in a highly truncated manner — in order to gather information about what
is percolating inside our conspecifics. If everyone around us starts looking
upward, we do too. This mimicry is predominantly automatic, not inten-
tional. It is, in all probability, part of our biological heritage.
Children on their caregiver's knee evince mirror reflexes in abundance.
Clearly this behavior is a boon for learning the emotional repertoire — and
much else — of one's culture, for it enables the child to discern the
situations the caregiver associates with feelings of distress or elation.
Among adults, mirror reflexes are also highly adaptive, since they facilitate
social communication. Albeit subliminally, one can intuit — at least very
broadly — the temperament of a room or the disposition of one's spouse by
using one's body as a detector of the kind of internal sensations that are apt
to be associated with the manifest expressions of others. These communi-
cative advantages, moreover, are so adaptive that they are probably bred
in the bone.
When by mimicry I "catch" the negative vibes or feelings of distress of
another, I immediately survey the environment in order to locate the
source of his discomfort. His negative affect alerts me in my own muscu-
lature to the likelihood that I may soon need to mobilize some vectorially
converging emotion, such as fear or possibly anger. Mirror reflexes, which
may be linked to what neuroscientists call mirror neurons, are not only
relevant for gleaning information about what surrounds other people. By
relaying to us something of what they are feeling, mirror reflexes help us
cope with others. Detecting that one's boss is in a foul mood via mirror
reflexes is useful in deciding when to ask him — or not ask him — for a raise.
And, in addition, mirror reflexes are immensely functional for coordinating
group activities — for getting the troops reeved up as they march off into
battle in lock-step camaraderie. In sum, the human capacity for mirror
reflexes is, in all likelihood, a stunning asset from the perspective of natural
selection, since it is a means for gathering affective information about
 AFFECT AND THE MOVING IMAGE 187

conspecifics and for synchronizing joint ventures, from face-to-face

conversations to cheerleading to military maneuvers.
Aestheticians have been aware of the phenomenon of mirror reflexes at
least since the writings of Theodor Lipps. And the filmmaker Sergei
Eisenstein explicitly attempted to exploit Lipps' insights not only in his
theory of filmmaking, but in his practice as well. Though neither he nor
Lipps employed the terminology of mirror reflexes, Eisenstein included
close-ups of stereotypical facial expressions and body parts, such as
clenched fists, in the hope of inspiring the sort of mimicry in viewers
that would prepare them viscerally for the kinds of emotional states that he
wished to arouse in them. Less theoretically inclined motion picture makers
also discovered the importance of such affective modeling intuitively and it
has been a staple of popular cinema since the 1920s. This is a legacy,
furthermore, that film has bequeathed to TV. Undoubtedly, a significant
amount of the affect stirred up by audiovisual entertainments is connected
to the way in which they educe muscular mimicry in their audiences.
Watching a videotape of Riverdance, the audience taps its feet, accessing a
simulacrum of the spirited pulse of the dancers. As Bruno in Hitchcock's
Strangers on a Train reaches for the lighter that he has dropped into the
sewer, our arm muscles flex, within the circumscribed ambit of our theater
seat, in a manner echoing his in order to help us feel his intention within
our body. It is not that we identify with Bruno morally or emotionally, but
we ape his gesture in order to help us determine a glimmer of what he is
feeling. At the conclusion of Alien Resurrection, we get a feeling, strangely
enough, for the monster's pain that results from his mother's betrayal, not
because we identify with Ripley's progeny but because we involuntarily
mimic his facial expression.*
And even before we see the monster displayed on screen, the screaming
visage of its victim, inscribed with horror and framed in close-up, prompts
us to tense up our faces analogously in a way that signals through our
muscles that things are about to get unpleasant. The activity on screen
primes mimicry of a partial or limited variety which can deliver informa-
tion about the internal states of characters which we sample in terms of
similar sensations in ourselves. Though not full-scale emotions — but only
feelings sans objects and, thus, without appraisals thereof — these sensa-
tions may nevertheless be a serviceable source of the affective grip such
motion pictures have on us in at least two ways.

I owe this example to Amy Coplan.

188 AFFECT AND THE M O V I N G IMAGE

First, these mirror reflexes contribute to keeping the excitement level

in our bodies elevated, thereby realizing one of the abiding promises of
popular movies. And second, they may make available information that we
can integrate into our more encompassing emotional responses to the
characters. The bodily feelings of distress that are imprinted on the
contorted features of the protagonist are relayed to us by our selective
imitation of his expression so that we can use the dysphoric taste of that
sensation as a recognitional cue for the kind of emotion that is appropriate
on his behalf — as, for example, sorrow would be appropriate were we to
detect by motor mimicry that he is feeling some sort of pain. That is,
mirror reflexes may function as sub-routines in the formation of our
emotional responses to fictional characters, not only alerting us to the
general valence — whether positive or negative — of their mental states, but
also calibrating the kind of passion we need to send back their way.
Mirror reflexes are not examples of emotional identification since
they do not involve the replication of complete emotional states by only
the feeling component and, in fact, they need not duplicate precisely the
same feelings, but only vectorially similar ones. Nor are they what are
sometimes nowadays called simulations of the emotional states of charac-
ters, since they do not employ beliefs and desires. On the other hand, they
are important elements of affective address in motion pictures. They can
make our bodies vibrate with feeling in the presence of certain characters,
and, even more importantly, they can facilitate our recognition of charac-
ter emotion and modulate our own emotional response to it. In this
regard, it would appear that motion pictures, including movies and TV,
have resources for exciting the affective reactions of audiences that literary
fictions, popular and otherwise, lack.
This capacity is perhaps particularly evident with respect to a struc-
ture alluded to several times already in this chapter, namely, point-of-
view editing. This figure, at bare minimum, comprises two shots — what
have been called the point/glance shot and the point/object shot. The
point/glance shot involves a character looking (glancing) offscreen; then
the point/object shot shows us what (the object) the character is looking
at. The hero in Rear Window looks toward the camera — presumably across
the courtyard in the world of the fiction; then there is a cut and we see
what he sees — a man in an opposite apartment mauling the onlooker's
fiancee.
This cinematic figure can be put through its paces either progressively
(starting with the point/glance shot) or retrogressively (starting with the
point/object shot and backing up to the point/glance shot). The figure can
 AFFECT AND THE MOVING IMAGE 189

also be iterated; the editor can repeat the figure several times progres-
sively, retrogressively, or in some combination of the two. In early cinema,
when filmmakers were somewhat anxious about whether or not people
would understand this cinematic figure, it was often reiterated through
several cycles. However, today, repetitions are less pronounced, except for
dramatic effect.
The comprehensibility of point-of-view editing rests, first and foremost,
on certain biological tendencies of the human organism (as well as other
mammals) — specifically, the tendency to track the glance of our conspe-
cifics (and, indeed, of other animals) to its target. This is a way in which to
derive some insight into what is going on in the minds of others — we
follow the gaze of the hostess of the party as she keeps stealing glances at
the door and then we speculate that she is expecting someone. This
tropism is inborn; a child on his mother's knee naturally follows her eyes
wherever they point. When a caregiver introduces a word, the child looks
at what mother is looking at, thereby, in the process, being inducted into
the conceptual scheme of the culture.
But the point-of-view editing structure does not only exploit our innate
tendency to trace the gaze of our conspecifics for the purpose of mind-
reading. It also mobilizes the mirror-reflex system for the sake of cinematic
communication. For, when the character's face appears in close-up on
screen, the viewer has an automatic tendency to imitate her expression,
thereby getting an intimation of the range or tenor of her inner state.
If mimicking the character's demeanor causes unpleasant turmoil within
us, we ready ourselves — we calibrate our expectations — to muster some
dysphoric emotion, such as anger, fear, horror, or the like. The character's
face in the point/glance shot, in other words, functions as an emotional
range-finder, demarcating the valence of the emotions appropriate to
the objects we are about to see in the point/object shot and suggesting
the approximate scope of the emotions we need to enlist in response. In this
way, point-of-view editing capitalizes upon an evolutionary adaptation that
took place countless millennia before the birth of cinema.
Although the moving image is a cultural invention, it is useful never to
lose sight of the fact that in many ways it succeeds as well as it does because
it freely avails itself of our biological heritage. The motion picture is an
artform, but, as with art in general, we must not suppose that it is solely
an affair of the mind. Motion pictures address our bodies as well. Though
we have seen this phenomenon in evidence in other chapters, the way
in which moving images interact with our bodies is perhaps never
more blatant than with regard to the hold motion pictures can exert
 190 AFFECT AND THE MOVING IMAGE

over our affective life, including, notably, our emotions, that nexus of
mind and body.
Movies can, so to speak, reach under our skin and stir up our feelings.
Moviemakers, in this regard as in others, are amateur psychologists,
experimenting intuitively with the human sensory apparatus for the
purpose of art, fame, and money, but often with results that sometimes
reveal how we, as incarnated beings, work. This is not said in order to
attempt to reduce the moving image to a repertoire of biomechanical
triggers. The moving image is undeniably a cultural creation. But it is
important not to lose sight of the fact that culture, including our affective
and emotional life, is constructed out of the biological possibilities deliv-
ered up by natural selection, often for purposes never dreamt of on the
sprawling savannahs where those possibilities took root.

Suggested Reading
The leading volume with respect to motion pictures and the emotions is Passionate
Views, edited by Carl Plantinga and Gregory Smith (Baltimore, MD: Johns
Hopkins University Press, 1999). Though not devoted to affect in motion pictures
in particular, but to art and the emotions in general, another very important
volume is Emotion and the Arts, edited by Mette Hjort and Sue Laver (Oxford:
Oxford University Press, 1997). On the range of affect in motion pictures, see
Jinhee Choi, "Fits and Startles: Cognitivism Revisited," Journal of Aesthetics and Art
Criticism 61:2 (Spring 2003), 149-57.
The contemporary discussion of the paradox of fiction was introduced by Colin
Radford in "How Can We Be Moved by the Fate of Anna Karenina?," Proceedings of
the Aristotelian Society, supplement 49 (1975), 67—81. One of the most important
attempts to solve the paradox of fiction is Kendall Walton's "Fearing Fictions,"
Journal of Philosophy 75:1 (Jan. 1978), 5—27. Interesting comments on Walton's
article include Alex Neill, "Fear, Fiction and Make Believe," Journal of Aesthetics
and Art Criticism 49:1 (Winter 1991), 47—56, and Richard Moran, "The Expression
of Feeling in Imagination," The Philosophical Review 103:1 (Jan. 1994), 75—106.
Kendall Walton responds to some of his critics in his article in Emotion and the Arts.
Concerning the relation of the emotions to movie genres, sources include Flo
Leibowitz, "Apt Feelings, or Why Women's Films' Aren't Trivial," in Post-Theory:
Reconstructing Film Studies, edited by David Bordwell and Noel Carroll (Madison:
University of Wisconsin Press, 1996), and The Philosophy of Horror, by Noel Carroll
(New York: Routledge, 1990). Carroll's view is criticized in many places,
including Berys Gaut's "The Enjoyment of Horror: A Response to Carroll," British
Journal of Aesthetics 35:3 (July 1995), 284—9. On suspense, see Noel Carroll, "The
 AFFECT AND THE MOVING IMAGE 191

Paradox of Suspense," in his Beyond Aesthetics: Philosophical Essays (Cambridge:

Cambridge University Press, 2001).
On the issue of identification, see Amy Coplan, "Empathic Engagement in
Narrative Fictions," Journal ofAesthetics and Art Criticism 62:2 (Spring 2004), 141—52.
See also Berys Gaut, "Identification and Emotion in Narrative Film," in Passionate
Views, ed. Plantinga and Smith. Criticism of Gaut and Coplan can be found in Noel
Carroll, "Sympathy for the Devil," in Philosophy and the Sopranos, edited by Richard
Greene (LaSalle, ILL: Open Court, 2004), and id., "On the Ties that Bind:
Characters, the Emotions, and Popular Fictions," in Philosophy and the Interpretation
of Popular Culture, edited by William Irwin and Jorge Garcia (Lanham, MD: Rowman
& Littlefield, 2006).
The most seminal articles regarding simulation theory with respect to aesthetics
are Gregory Gurrie's "The Moral Psychology of Fiction," Australasian Journal of
Philosophy 73:2 (1995), 250—9, and his "Imagination and Simulation: Aesthetics
Meets Cognitive Science," in Mental Simulation, edited by Martin Dayies and Tony
Stone (Oxford: Blackwell Pubhshing, 1995). Criticism of this view may be found
in Aaron Meskin and Jonathan Weinberg, "Emotion, Fiction and Cognitive
Architecture," British Journal of Aesthetics 43:1 (Jan. 2003), 18—34. See also Carroll,
"On the Ties that Bind," and also his "Simulation, Emotions, and Morality" in his
Beyond Aesthetics. The major book-length study of the emotional relation of audiences
and movie characters is Murray Smith, Engaging Characters (Oxford: Clarendon
Press, 1995).
On mirror reflexes, see Carl Plantinga, "The Scene of Empathy and the Human
Face," in Passionate Views, ed. Plantinga and Smith. For psychological background
on this phenomenon, read Elaine Hatfield, John T. Cacioppo, arid Richard L.
Rapson, Emotional Contagion (Cambridge: Cambridge University Press, 1994),
especially chapter 2.
For an up-to-date philosophical discussion of the emotions, see Jesse Prinz, Gut
Reactions (Oxford: Oxford University Press, 2004). On art and the emotions,
see Jenefer Robinson, Deeper Than Reason (Oxford: Oxford University Press, 2005).
One aspect of affect not discussed in this chapter is mood. For an introduction
to this topic, see Noel Carroll, "Art and Mood," The Monist 84:4 (Oct. 2004),
521—55. See also Gregory Smith, Film Structure and the Emotion System (Cambridge:
Cambridge University Press, 2003).
Chapter Seven
Evaluation

In the preceding chapter, and throughout much of this book, we have been
concerned with the relationship between various motion picture structures
and the corresponding experiences said structures engender in audiences.
Thus, in a manner of speaking, we have stayed inside the movie house. But
in this chapter, let us walk outside onto the street and listen to how we talk
about moving pictures after we've seen them.
You meet a friend. She says that she has just seen a motion picture — say,
Cache. What do you say? What do you ask? Supposing that you know that it
is a current motion picture, the odds are that you respond with something
like: "How was it?" or "Was it any good?" That is, even before you inquire
about the specifics of the story, you are likely first to request an evaluation.
Moreover, your friend will be happy to gratify you; for evaluating motion
pictures is an activity that most people enjoy. We not only engage in
moviegoing in anticipation of the experience of it; most of us, in addition,
delight in assessing that experience afterwards and then comparing our
evaluation of it with those of other viewers.
Perhaps when you first consider the topic of evaluating motion pictures,
you are prone initially to think of movie critics. These are people who are
in the business of pronouncing upon the value of motion pictures. There
are so many movies to see and so little time. Consequently, almost all of us
have to fall back on the advice of movie critics from time to time in order
to inform our choice of viewing fare.
There are several different ways in which a movie critic may carry out
her role. Two of the most common ones are those of the consumer
reporter, on the one hand, and the taste-maker, on the other. Consumer
reporters try to predict which movies most of their readership will love
and which they will hate. For such readers, these critics are consumer
guides. Such critics treat themselves as live-detectors, presuming that what
they like, their readers will like too. Critics who aspire to the function of
consumer reporters speak in the vox populi. It is probably the case that
 EVALUATION 193

most, though not all, journalistic movie critics intend to be of this sort. Or,
at least, a great many of their readers expect this of them.
Other critics aim at being taste-makers. They do not attempt to reflect
the inclinations of others; instead they hope to shape the taste of their
readers. They are always on the lookout for motion pictures that are special,
even if— and perhaps sometimes especially if— they are not obviously so;
and, in the best of cases, these critics also suggest how the rest of us might
learn to go on to appreciate these outstanding motion pictures as well.
The taste-makers and the consumer guides, of course, are often at odds
with each other, given their respective aims. What the consumer guide
dismisses as boring, the taste-maker may recognize as a deliberate strategy
that effectively articulates the theme of the pointlessness of modern life, as
in the case of La Notte by Michelangelo Antonioni.
Readers who prefer their tastes confirmed will gravitate toward the
consumer guides. Those who want their taste expanded are apt to seek
and to admire those critics who are able to pick out singular cinematic
achievements and to contextualize and explain them.
However, though the role of the movie critic commands an important
position in cultures like ours that are awash with movies 24/7 (in movie
theaters, on television, on videocassettes, DVDs, personal computers, and
even cellphones), as hinted earlier, it would be a grave mistake to think of
moving picture evaluation as exclusively a professional affair. Evaluating
movies is something that we all do all of the time.
Nor do I mean by this merely that we automatically form preferences
for some of the movies we see over others and rank some of them as better
than the rest. As humans, we naturally tend to do this with most of our
experiences. But with regard to moviegoing, this is not something that
simply happens to us automatically. It is, as already remarked, an activity
that we avidly pursue. Think about it: when we talk about motion pictures
with others, most people spend most of their time airing their evaluation
of the movie rather than summarizing, let alone analyzing, it.
And when we converse back and forth with acquaintances about moving
pictures we've all seen, most of our energy is devoted to trading evalu-
ations, comparing them, sometimes sharing them, and often arguing about
them. Mary says that so-and-so's performance was great; Marty responds
"If that's an Irish accent, I'll eat my shillelagh." Indeed, we frequently read
critics after we have seen a motion picture in order to enter into imaginary
conversations with them, sometimes agreeing with them (and congratu-
lating them on their extraordinary intelligence and sensitivity), though also
often wondering aloud what galaxy they come from.
194 EVALUATION

Moviegoing is frequently portrayed as a solitary transaction in which each

individual spectator communes with the brightened screen in a darkened
room. But moviegoing, as a folkway of contemporary life, is more sociable
than this sketch allows. Moviegoing is part of our social life. After we have
seen a motion picture, we want to talk about it with others. We want to tell
them what we have seen, to exchange ideas with others who have seen the
same movies, and to discuss our reactions and theirs. Moviegoing is a fixture
of everyday conversation; it is a virtually natural pretext for sociability. On
occasion, we may even swap opinions about movies with strangers. And, as a
moment's reflection should reveal, evaluation is at the crux of this very
elemental form of social exchange and consolidation. It is through such
encounters that many of us express and develop ourselves as cultural beings.
This is all highly abstract. But what I have in mind can be clarified by
recalling a very common experience. You attend a moving picture with
some friends. Afterwards, as often happens, you go out for tea and coffee,
or sweets, or a drink. You talk about the motion picture that you have just
seen, remarking upon the parts you liked or disliked. But, of course, quite
often the conversation does not just end with a summary of personal
preferences. The discussion may often turn to whether or not the motion
picture is good or whether it worked. That is, there is an almost ineluct-
able tendency for the conversation to drift from reports of subjective
enjoyment or lethargy to questions of objective evaluation.
Experiences like these are part of the typical life of moviegoing — which
is to say that evaluation is near the heart of this practice for most of us.
It adds zest to the activity. It is something that ordinary moving picture
audiences care about deeply. It is something that viewers want to do and
are interested in doing well. That too is one of the reasons that audiences
read motion picture critics — not simply for advice about what to see, but
also for models of how to talk about and to conceptualize what they see.
And yet this is an aspect of the practice of motion pictures about which
contemporary philosophers and theorists of cinema are virtually silent.
So, in this last chapter, let us join the moviegoing audience where they live
and speak of evaluating moving pictures.

Movie Evaluation as a Philosophical Problem

Thus far, I have alleged that evaluating motion pictures is something that
we come to almost naturally. So, you might ask, what business does
philosophy have here? "Bug off philosophy!" you might be tempted to say.
 EVALUATION 195

In order to motivate philosophy's intrusion into the discussion, let

us return to the conversation between you and your friends that we
inaugurated two paragraphs ago.
Imagine that we have just seen a movie together — for example, Spike
Lee's Inside Man. You like it; I don't. As often as not, the colloquy will not
stop there, as it probably would if we were comparing our preferences in
condiments — with me voting loudly for mustard and you plumping
enthusiastically for ketchup. Rather, you are liable to press further and to
defend your liking of Inside Man on the grounds that it is good — asserting,
in other words, that your liking of it is justified.
I reply that I do not deny that you liked the motion picture — I concede
that this is a psychological fact about you. I take your word for it. I grant
your authority in the matter of your likes and dislikes. But I add that your
liking of it does not prove that Inside Man is a good movie. For> it is
eminently consistent for someone to like a fiction and for the fiction in
question to be bad, even abominable, at the same time. For example, I've
always had a soft spot in my heart for the novels of Edgar Rice Burroughs,
yet, entre nous, most of them are god-awful.
There is no logically reliable connection between a person's liking of
x and x's prospective goodness. The correlation between what is liked by
people and what is objectively worthwhile is radically contingent. Many
love fast food that is nutritionally compromised. So, I accept that your
liking of Inside Man is sincere, but then retort that this is perfectly
compatible with Inside Man being a bad movie. Indeed, I up the ante;
I contend that Inside Man is actually flawed, in fact, irreparably so.
Yet, at this point in our dialogue, take notice that we have left off talking
simply about our likes and our dislikes. We are speaking in the idiom of
good and bad. Consequently, we have entered the realm of objective
evaluation. In effect, you are claiming that any reasonable, informed, and
suitably backgrounded viewer ought to regard Inside Man as a good film,
whereas I am denying precisely that.
Of utmost importance, observe: we are arguing. There would no point
in our arguing about what we like. You like ketchup; I like mustard. That's
it. Period. Full stop. We accept each other's preferences as facts; I don't
suppose that it makes much sense to attempt to argue you out of your
relish for ketchup. Indeed, what sorts of reasons might I bring to bear?
However, in stark contrast, we are arguing about Inside Man. And that
entails, furthermore, that our discussion revolves around coming up with
reasons that we believe should sway or even compel others to accept our
viewpoint. It is just this — our implicit appeal to third parties and their
196 EVALUATION

standards of evidence and inference — that marks our interchange as

objective. For we intend to be playing by the inter-subjective rules, or,
at least, various ascertainable protocols of rationality.
As has already been stressed amply, a subjective report of a like or a
dislike is a psychological fact about us. There is no disputing your likes,
even if they are unusual, such as a taste for extremely charred toast. For
the expression of a preference of this kind does not, on the face of it, aspire
to command the agreement of others. If you dislike broccoli that is not
a reason for me to dislike broccoli, since reasons — as opposed to sheer
likings — are objective.
Reasons are offered to others with an implicit claim upon their assent.
Thus, if you provide a reason for saying that Inside Man is a good film and if
that reason is relevant and acceptable, then you expect, with sound
warrant, that others should concur with your assessment. In our discussion
of Inside Man, our behavior — our exchange of putative reasons — gives
every appearance of being committed to objectivity. You do not, red-faced,
pound the table and shout "But I like it!" Instead, you try to find objective
reasons that will convince others, notably me, that your assessment
is correct.
Therefore, you have exited the domain of psychology and embraced
normativity. In the face of disagreement, you are now conversationally
implicated in establishing that Inside Man is a good movie, for anyone with
functioning eyes to see, ears to hear, and a mind, suitably prepared, to
appreciate it. However, as all of us have experienced, our reasons do not
always dissolve disagreements about movies as easily as a dehumidifier
dispels dampness. So, how are such disputes to be resolved?
Here in a nutshell, then, is one particularly nettlesome problem. The
very possibility of the objectivity of motion picture evaluation (as a sub-
species of artistic evaluation in general) is a conceptual challenge which
invites, indeed beckons, philosophy onto the scene. For it falls to philoso-
phy to engage with and to relieve the seemingly intractably stubborn fact
of such disagreement. That is, we not only diverge in our preferences
for movies. We think that these divergences, at least in important cases, can
be negotiated, at least some of the time, in terms of objective evaluations.
But how is this feasible? It the task of the philosophy of motion pictures,
since it is a precinct, as established in the first chapter of this book, of the
philosophy of art, to address questions like this.
We believe that our contrasting opinions can ultimately be backed up
with reasons that others — in fact, nearly all of the relevant others — ought
to find probative. We do not think that we are merely airing personal
 EVALUATION 197

whims or confessing guilty pleasures when we say that Jean Renoir's Rules
of the Game is a better film than Stanley Kramer's Guess Who's Coming to
Dinner. We think that a disagreement like this one can be resolved
rationally — that is, with reasons. Reasons, properly so called, portend
objectivity. Nevertheless, where do these reasons come from and what
grounds them? Inquiring minds would like to know; and they are surely not
perverse in this desire. Clearly, it is the responsibility of philosophy to
speak to this urgent issue — the question of the adjudication of apparently
relentless disagreement.
In order to get a handle on how such seemingly immovable disagree-
ments might be negotiated rationally, let us follow our earlier, imaginary
dispute about Inside Man just a little bit further. You say Inside Man is good;
I say it is bad. What happens next? Very often we start mentioning other
movies. You refer to comparable suspense films (with hostage-taking), such
as, for instance, William Wyler's The Desperate Hours and/or Sidney
Lumet's Dog Day Afternoon or F. Gary Gray's The Negotiator. Your purpose
here might be to get me to agree that these are good films; since, if I agree
that they are gopd films, then you will argue that consistency (a funda-
mental test for rational objectivity) requires that I admit that Inside Man is
also good. For, once granting the premise — that Inside Man shares essential
features with these other suspense thrillers which I supposedly have already
accepted to be good — I am bound by logical analogy to judge Inside Man
to be good as well.
At this point, the debate could go in several directions. I might deny that
Inside Man is really analogous to the three exemplars of suspense thrillers
that you have adduced, or I might reject the attribution of goodness to any
of the suspense thrillers that you recommend as paradigmatic ones. If I take
the first option, I must show that the analogies which you are attempting
to enlist are either insubstantial or implausible; this will, perforce, be a
matter of close analysis.
But let us imagine that I accept your analogies and, instead, that I go
for the second option: I deny that any of the analogues that you have
dragooned on behalf of Inside Man are themselves good films. Therefore,
even if Inside Man is strictly analogous in essential respects to your
hypothetical paradigms, that premise affords no support in our debate
for your contention that Inside Man is good, since, by my book, none of the
other films cited have anything going for them either. So the fact that Inside
Man is analogous to them counts for nothing.
However, if I reject all of your paradigms at this juncture, then you
certainly have the right to demand that I produce one of my own. What
198 EVALUATION

exactly, you ask, is an example by my lights of a good motion picture?

Imagine I say Fast, CheapticOutof Control by Errol Morris.
Undoubtedly, you will be taken aback. For, though you agree that Fast,
Cheap ScOut of Control is a good film, you cannot help feeling that somehow
I have missed the point. Why? Because that is not the kind of motion
picture that Inside Man is. Fast, Cheap ScOut of Control is an essayistic film —
an open-textured documentary that encourages the viewer to contemplate
the thin line between genius and loony obsession as well as that between
the appearance of intelligence and massive parallel processing. What
possible relevance can a meditative nonfiction film like Fast, Cheap <SL Out
of Control have to gauging the quality of a fictional exercise in thriller
suspense such as Inside Man? Inside Man and Fast, Cheap Sc Out of Control
belong to very different categories — they have different purposes, and,
consequently, they mandate different standards of evaluation. Isn't it as they
say — apples and oranges? Surely, to devalue Inside Man because it is not
like Fast, Cheap ScOut of Control strikes most of us as a gross non sequitur.
One feature of evaluation that the preceding debate illustrates is the
degree to which our assessment of the merits of a motion picture depends
upon placing the movie in question in a certain category. For example, we
situate Inside Man in the category of the suspense thriller in order to assay
its value. It is a psychological fact about human beings that we think this
way — to evaluate, we categorize. The rationale for this is that categories
are generally connected to acknowledged purposes which, in turn, provide
standards of evaluation. For instance, when we classify a piece of cutlery as
a steak knife, we identify that its purpose is to slice cleanly through thick
pieces of beef, and, consequently, that a good-making feature of such
an implement is its sharpness. Sharpness, in other words, is a standard
of goodness in steak knives because it is in virtue of its sharpness that a
steak knife can discharge its purpose. In this way, categories, purposes, and
standards of evaluation form a matrix.
When we view a moving picture, we also — as a matter of our
psychological make-up — put it in a category, and this shapes our expect-
ations and, thereby, recruits a certain set of standards. Our disagreement
about Inside Man at the present juncture turns upon the fact that you have
situated Inside Man in one category or comparison class and it seems very
likely that I have located it in another.
That evaluation is tied to categorization is, as already indicated, a matter
of psychology, not logic. Thus, we do not resolve the question of whether
or not Inside Man is good by merely citing the different categories into
which we are disposed to slot it. That may clarify our differences by
 EVALUATION 199

pinpointing precisely why we hold such different views, but it does not
settle our dispute. For our debate will not be adjudicated rationally until
we are able to determine which of our competing categories is the correct
one (or, at least, the more correct one). To decide who is right on the issue
of the quality of Inside Man, we need to establish who has accessed the right
(or the predominantly right) category.
Note that our disagreement about the goodness or badness of Inside Man
has escalated, in a manner of speaking, into a debate about categories.
Which category or categories is/are appropriate or correct in this instance?
So, our dispute, thenj looks like it can be brought to a reasonable
conclusion just in case we can prove which category is the apposite one
to bring to bear on Inside Man.
Indeed, a great many problems with respect to movie evaluation —
though not every one of them — could be solved, if only we had a way
to fix the correct category or categories for weighing the moving picture in
question. But how does one find the correct category? That is the million-
dollar question. For without an answer to that question, it appears that our
dispute over the right way to classify Inside Man is as rationally intransigent
as the issue of whether or not Inside Man is good.

The Classical Solution

As we have just seen, one of the central philosophical problems of motion
picture evaluation is what we can call "the problem of the correct category."
Though this is not a problem that has preoccupied many contemporary
theorists of the moving image, it was a matter of overriding concern for
what might be called "classical film theory" — that is, roughly speaking,
film theory before the advent of semiotics and post-structuralism. This is
the kind of theory we encountered previously in the chapter on medium
specificity. Figures such as Rudolf Arnheim and the Soviet montagists are
representatives of what might be thought of as the first generation of film
theorists, whilst realists, such as Andre Bazin and Siegfried Kracauer,
belong to a second generation.
One way of characterizing classical film theory is to identify it as
centrally preoccupied with isolating the essence of cinema, which, as
discussed in chapter 2, is what many conveniently label as the cinematic.
The cinematic, as we saw earlier, was believed to be that which differenti-
ated film from other artforms, such as theater and painting. Moreover,
classical film theorists were not simply interested in finding which property
200 EVALUATION

or properties in fact distinguished film from other artforms. The differ-

entiae they sought were also often assumed to have normative implications.
That is, classical film theorists were hunting for the way in which film
should distinguish itself from the other arts.
Speaking summarily, the classical film theorists maintained that cinema,
which, for them, meant film, should be cinematic. Of course, the various
classical film theorists defined what they took to be cinematic in different,
even conflicting ways. However, in the main, each had a conception of
what counted as cinematic, and they generally also agreed that films ought
to aspire to be as cinematic as possible. For the classical theorist, the failure
of a film to be cinematic, however that was to be construed, is a defect.
But what does this talk of the cinematic have to do with the problem of
the correct category? Simply this: in effect, for the classical film theorist,
all films — or, at least, all the films we would want to talk about from the
artistic point of view — belong in a single category, which we might tag
with the label oifilm as film. Therefore, debates about the goodness or
badness of a given movie can be referred back to the question of whether
or not the specimen under examination is cinematic — that is, whether or
not it is a proper instance of the category of film as film. Thus, in order to
determine whether Inside Man is a good film, we would have to demon-
strate that it is cinematic.
One obvious advantage of the classical approach to the problem of the
correct category is that it makes the evaluation of moving pictures an
extremely unified practice. The notion of the cinematic, in the hands of a
given classical theorist, provides a single scale upon which all motion
pictures can be weighed and compared. Whether or not a movie is good
(relative to the category of film as film) can be ascertained by showing that
it is cinematic, while exactly how good it is can be gauged in virtue of how
cinematic it is. Likewise, one can say that a motion picture is bad if it fails
to be cinematic, or that it is blemished to the degree that uncinematic
elements intrude into its style. Moreover, moving pictures can be com-
pared and even ranked as good, bad, better, and worse in accordance with
a single measure — their degree of cinematicity.
Undoubtedly, this is a very tidy program. It would provide a way of
rationally brokering all disagreements about evaluating moving pictures
just in case we are able satisfactorily to characterize what constitutes
the cinematic. But there is a fly in the ointment. For, as is well known,
different classical film theorists tended to argue for different and often
conflicting (or non-converging) conceptions of the cinematic, no version of
which, truth be told, has survived unscathed.
 EVALUATION 201

Early film theorists, particularly those associated with silent film, for
example, tended to locate the cinematic in terms of two contrasts — that
which differentiated motion pictures from theater, on the one hand, and
that which differentiated film from the slavish recording of reahty, on the
other hand. As we saw in our opening chapter, they stressed the difference
between film, properly so called, and mere recording, in order to establish
the credentials of the motion picture as an artform — that is, as something that
did not simply duplicate reality but that could reconstitute it expressively,
formally, and/or creatively.
But the early film theorists also desired to differentiate the moving
picture from theater. The point here was to demonstrate that cinema was a
unique artform, an artistic category unto itself — film as film. Thus, these
film theorists, of whom Arnheim is a leading example, argued that film,
properly so called, was neither an imitation of nature nor an imitation of
any other artform, notably theater.
Whereas theater narration typically relies upon words, cinematic nar-
ration, it was asserted, could and should emphasize movement, image,
action, and rapidly changing points of view. If theater is primarily verbal in
its typical mode of address, cinema, ideally, is primarily visual. Similarly,
cinema has resources, particularly editing, that enable moviemakers to
manipulate spatial and temporal transitions with more fluidity than is
customary in theater. Thus, editing, or montage, was generally celebrated
as the most important, essential characteristic of cinema. On this view,
the moving image is essentially visual, its natural subject of representation
is highly animated action, and its primary means of expression is editing
or montage.
Other techniques were also regarded as cinematic — including close-ups,
assertive camera angulation, trick photography, visual devices such as
fades, wipes, and superimposition, camera movement, and the like. As
with editing, these techniques were prized because they both declared the
difference between film and theater (inasmuch as the effects available
through these procedures were not easily achievable in theater) and also
departed from the "straight" shooting of reality. In other words, the use of
these devices standardly indicated that film differed from theater and from
what was thought of as the merely mechanical recording of reality (which
was sometimes equated, oddly enough, with something called "normal
perception"). These techniques, in other words, displayed film as a unique
artistic category — film as film.
A feature of a film was cinematic, then, so long as it deployed techniques
that underscored the putatively unique capacities of cinema. Stylistic features
202 EVALUATION

that failed to do this — such as the use of extended dialogue or stolid tableaux
shots — were uncinematic. Excessive reliance on words at the expense of
animated action, or of a single camera position to record the declamation of
speech-ridden dialogue (rather than exploiting the powers of editing)' was
not only uncinematic, but downright theatrical. And to be theatrical, a.k.a.
uncinematic, flew in the face of the canonical standards of film as film.
Theatrical or uncinematic films were bad films, inappropriate or defective
examples of the category of film as film.
A movie such as Inside Man does nicely on this conception of cinema.
Dialogue here is primarily in the service of action — of which there is a
goodly amount — which, in turn, is nicely articulated through editing. The
camera shifts its perspective often and there are no visual dead spots. On
the other hand, Fast, Cheap ScOut of Control would not fare so well because
much of its screen time is simply a matter of straightforward interviewing.
Thus, this approach to motion picture evaluation could settle our earlier
dispute in short order, if only its account of the essence of cinema is found
to be compelling.
But this conception of cinema is hardly incontestable. In factj it was
challenged by subsequent theorists in the classical tradition, of whom Bazin
and Kracauer are noteworthy examples. These theorists were often called
realists because they thought that the essential feature of cinema is pho-
tography (cinematography) and that this feature committed cinema to
meeting certain standards that emphasized the recording and disclosure
of reality.
So, where earlier classical theorists, like Arnheim, thought that the
capacity of film to diverge from the recording of reality implied that
cinema should employ assertive devices such as editing to reconstitute
reality, realist theorists, like Kracauer, looked more favorably on cinema's
provenance in photography and inferred that this argued for the realistic
usage of film. Indeed, the realist Bazin even argued that the putatively
realist origin of cinema in photography privileged certain techniques like
the multiplanar, depth-of-field, sequence shot later embraced by directors
like Hou Hsiao-Hsien over the rapid montage championed by Soviet film-
makers of the 1920s like Alexander Dovzhenko.
As so often occurs in debates like this, the "repressed" of the first
generation of theorists, not only returns but is valorized by the next
wave of thinkers. Nevertheless, both waves — as represented by Arnheim
and the Soviets, on the one hand, and Kracauer and Bazin, on the other — were
classical film theorists, since both groups believed in the cinematic. However,
they disagreed about what constituted that category and, consequently, they
 EVALUATION 203

disagreed about what standards were suitable to bring to bear on evaluating

films as films.
In fact, their theories led them in opposite directions. Highly styliz_ed
films, like The Last Laugh, Variety, and The Docks of New York, of the sort
favored by Arnheim's theory were Hkely to be castigated by the lights of
second-generation film theorists because of their disregard for the canons
of realism.
And in any event, from Kracauer's perspective, a film like Inside Man
would be criticized as uncinematic because of the various ways in which it
violates Kracauer's conception of the (normative) essence of cinema. For
instance, Kracauer thinks that cinematic realism mandates a disposition
toward the use of open-ended narrative structures, whereas the storyworld
of Inside Man, with its tricky plot turns, is what he would disparage
as "closed" and, therefore, as unrealistically contrived; Inside Man, that
is, is the kind of suspense thriller that comes with a high quotient of
mystery — n o t of the whodunit variety, but of the "How did they do it and
why?" brand — whose neat, "last-minute" answers are just too convenient.
Indeed, the movies that employ closure, as discussed in chapter 5, are
generically uncinematic, according to Siegfried Kracauer. O n the other
hand, if only because it is a documentary, Fast, Cheap &,0ut of Control would
probably get far better marks from Siegfried.
In the debate over the cinematic, proponents of the earlier view — that
film should reconstitute reality — and champions of the later view — the
realists — could both score points against each other. Each side could argue
that the other team failed as a comprehensive theory of the nature of
the moving image, because each side believed that the other was blind to
certain kinds of cinematic achievement. The defender of realism could
claim that theorists such as Arnheim ignored the accomplishments of
films whose style gravitated m o r e toward recording, such as the work of
the Italian neo-realists (whose aspiration, of course, could be elegantly
explained from a realist point of view).
O n the other hand, those who favored assertive stylization and editing
could account for many avant-garde experiments which were anything but
realistic. Each side pretended to be a comprehensive theory of the nature
of film, b u t opponents of either side could point out that the other side
neglected or attempted to sublate, in an extremely high-handed and ad hoc
manner, massive amounts of the data of film history.
Furthermore, it did not appear possible to patch up this problem by
merely combining the two schools of thought of classical film theory,
because, in important respects, they contradicted each other. Adding the
204 EVALUATION

two lines of theory together would not lead to a more comprehensive

theory, but only to an inconsistent one.
An explanation of why these two strands of classical film theory failed
to deliver the goods is that, though each side presented itself as a com-
prehensive theory of all motion pictures, this was not, in actuality, what
either of these theoretical approaches was about. Both were really critical
characterizations — couched as theory — of different stylistic tendencies that
had crystallized at different points in the history of the moving picture.
Theorists such as Arnheim and the Soviet montagists, like V. I. Pudovkin
and Lev Kuleshov, were particularly sensitive to stylistic developments
in the period of silent filmmaking, whereas realist theorists — notably Bazin
and Kracauer — were especially attuned to stylistic developments in the
sound film of the 1940s and the early 1950s.
Both sides, in other words, mistook certain period-specific develop-
ments in motion picture history to reveal the essence of cinema. This is
why their theories were suited to certain bodies of work, while, at the
same time, being doggedly undiscerning with regard to the accomplish-
ments of other stylistic traditions. Whereas a theory of the nature of the
movies should apply to motion pictures of all styles, these theories
inadvertently privileged certain styles over others — to the extent that
certain avenues of cinematic achievement were deprecated by them as
uncinematic. Thus, their theories failed to be comprehensive because they
were biased or stylistically partisan.
Undoubtedly, another — logically less flattering — way to put the matter
is to observe that these theorists proposed to demonstrate that certain
options of film stylization were uncinematic by deducing the essence of
cinema; but, in fact, what they did was to allow their stylistic preferences
to shape — indeed, to infect — their conception of the essence of cinema.
Kracauer asserts that photography is the essential feature of cinema. But
how does he know that photography rather than, say, editing is the essential
property? The actual answer is because he already presumes that realist
filmmaking demarcates the recognized body of achievement in the medium
which a theory, like his, needs to track. That is, his "data" have been
gerrymandered to elicit his conclusion. But this, of course, begs the
question, since Kracauer has already "selected" the "evidence" that induces
his "conclusion."
Of course, the same kind of interrogation could be aimed at Kuleshov.
How does he know editing is the essential feature of cinema? Indeed,
Kuleshov's position is almost blatantly the mirror image of certain realist
positions. Kuleshov begs the question no less than Kracauer. Moreover, this
 EVALUATION 205

is the sort of specious reasoning that runs rampant throughout the corpus
of classical film theorizing.
One very ingenious attempt to reconcile the differences between the
different schools of classical film theory can be found in Victor Perkins's
brilliant book Film as Film. This is an especially interesting text for us, given
the concerns of this chapter, because it is one of the only explicit efforts to
provide a rational foundation for film evaluation. Although Perkins would
probably bridle at this suggestion, his book is an example of classical film
theory, as his title — Film as Film — indicates. Like other classical theorists,
Perkins tries to develop a unified canon of movie evaluation for all motion
pictures, irrespective of genre and period.
Perkins evolves his canon of the evaluation of motion pictures by
combining in a logically scrupulous way some of the insights — such as
those of the montagist Arnheim and the realist Kracauer — into a single,
non-contradictory formula. On Perkins's account, in order to be good, a
moving picture must abide by certain realist standards of verisimilitude. In
this way, Perkins pays tribute to the theoretical tradition of people like
Kracauer and Bazin. But Perkins also pays his respects to the tradition of
assertive stylization in classical film theory. For he maintains that the extra
credit that a movie accrues (over and above the minimal accreditation as
good that it receives for being realistic) is to be added to the movie's
account in terms of the extent to which it is stylized — via editing, set
design, camera angulation, camera movement, costume, and so on — just
so long as this stylization is articulated within the bounds of realism.
That is, a film as film is good simpliciter, if it plays by the rules of realism,
as set forth by Perkins. How good — or how much better — it is, then, is a
matter of how artfully stylized it is, so long as that stylization is constrained
reahstically. For example, a film such as Elmer Gantry can employ hyper-
active montage metaphors and still win Perkins's aj/plause for doing so,
because those metaphors are motivated realistically Within the world of
the fiction. \
In Perkins's theory, realism and stylization do not contradict each other,
but can coexist peacefully, where the two tendencies'cooperate in accord-
ance with the principle that realism constrains the legitimate compass of
stylization. That is, realism and stylization are coordinated by a rule that
says a good film contains both, but is only good to the degree that the
stylization conforms to the discipline of realism. In Elmer Gantry, figura-
tively expressive cutting between a swelling fire and the uncontrolled
delivery of a "hell and brimstone" sermon is commendable cinematically
because the fire that comments upon the frenzied preaching is of a piece
206 EVALUATION

with the naturalist settings and narration of the fictional world. It is, so to
say, motivated from within the storyworld of Elmer Gantry.
In Elmer Gantry, the story establishes that the fire started literally
backstage, in the same locale as the pulpit; a workman threw a butt on a
flammable heap which smolders as the rhetoric mounts. So the fire is given
as something that happened in the fiction world; it is not sheerly a
comment from an authorial elsewhere. This use of editing putatively
contrasts favorably with those Eisensteinian montage-sequences which
resort to similes that intrude upon the decoupage from outside the setting
of the fiction. That is, when in October, for example, Eisenstein compares
Kerensky with a peacock, he does not establish that the peacock, as an
actual artifact, dwells in the same palace or even the same storyworld that
Kerensky does.
By requiring that stylization be constrained by realism, Perkins proposes
a principled way for films to exploit both of the tendencies advocated by the
conflicting strands of classical film theory. Perkins's offers an immensely
intellectually seductive and thoughtful compromise solution to settle the
differences between the t w o opposing camps of classical film theory, camps
which, it should be acknowledged, have advanced the very best thinking
about the moving image produced to date. In this way, Perkins's approach
appears to realize the dream of classical film theory — to solve the problem
of the correct category by discovering a standard of evaluation that is
applicable to all motion pictures.
Nevertheless, if Perkins's approach represents one of the highest points
in classical film theory, it is also unfortunately (I mean that, I really wish it
were otherwise) vulnerable to some of the same criticisms that we have
seen plagued earlier forms of this kind of theory. We have already observed
that a recurrent failing of classical film theorists involved their hypostatiza-
tion of certain period-specific film styles — their tendency to mistake these
styles for the very essence of cinema itself. Ironically, a similar problem
besets Perkins's project.
Perkins takes the exploration of stylization within the bounds of realism
to be the quiddity of film as film. By why suppose that this tendency is any
more representative of the essence of film than experiments in avant-garde
irrealism? That is, why think that Otto Premingers Carmen Jones is m o r e
cinematic than Jean-Luc Godard's One Plus One or his One or Two Things
1 Know About Her?
Yet, when we scrutinize the database of Perkins's theory, an explanation
of his assignment of Preminger's position in the cinematic pantheon
readily suggests itself. For the examples that bolster Perkins's theory
 EVALUATION 207

predominantly come from Hollywood movies of the 1950s and 1960s.

During that period, constraining stylization within the bounds of realism
was, so to say, a rule of the practice or the ongoing game, which we might
even call expressive realism. But in this respect, what Perkins has done,
like Arnheim, the Soviets, Kracauer, and Bazin before him, has been to
misconstrue an essential (or fundamental) feature of a period-specific
stylistic practice for the essence of cinema as a whole. That is, Perkins's
preference for a certain kind of filmmaking has biased his account of the
nature of, as he himself calls it, film as film.
One of the greatest promises of classical film theory was that it would
solve the problem of the correct category. For, if we could make classical
film theory work, we would have the one and only category we need for
evaluating all films. By identifying an essence of film — one with normative
impact — we would have a standard that could be brought to bear in order
to resolve rationally most disagreements about the quality of given movies.
But, as we have seen, the search for this kind of essence — call it the
philosopher's stone of film theory — has proven as chimerical as the original
philosopher's stone.
In most cases, classical theorists have proceeded by begging the question.
They have built in their stylistic preferences as premises in their alleged
deductions of stylistic imperatives and cinematic standards. This is rather
like the magician who puts the rabbit in the hat and then draws it out.
But what is not really magic is not logic either.
Sadly, this sort of error recurs throughout the body of classical film
theory, even in the work of its most sophisticated practitioners, such as
Perkins. Thus, if and until this failing is repaired, the prospects for solving
the problem of the correct category by means of the strategies of classical
film theory appear foreclosed. This calls for another alternative to be
explored.

An Alternative to the Classical Evaluation of

the Moving Image: The Pluralistic-Category
Approach
What we had hoped to inherit from classical film theory was a solution to
the problem of the correct category. This problem is a pressing one for
motion picture evaluation because if we are able to find a way to establish
the correct category for evaluating a given film, then we are at least on our
208 EVALUATION

way to resolving rationally many — though not all — of the disagreements

that arise when we attempt to assess motion pictures. Unfortunately,
however, we have seen that so far classical film theory has failed in this
regard. So, until classical film theory delivers on its promise, we must
explore alternatives.
Undeniably, the failure of classical film theory is instructive. It shows us
something, but we need to be careful about what we think it reveals. It
does not entail that there cannot be correct categories for evaluating films.
Rather, it only suggests that it is unlikely that we can evolve a unitary
theory of evaluation based upon a single category, such as film as film. That
is, at its strongest, the failure of classical film theory may only indicate that
there is not a universal litmus test — such as the cinematic — for goodness in
movies. But the failure of classical film theory does not imply that
categories as such are irrelevant to motion picture evaluation.
The defects of classical film theory may incline us to be skeptical about
whether there is a single evaluative category that subsumes all movies. But
this should not make us skeptical about whether moving pictures fall into
various different categories, many of which come with perfectly respect-
able standards of evaluation. That motion picture practices are too diverse
to be all assimilated productively under the sort of essentialist category of
the generality that classical theorists hoped to isolate hardly indicates that
movies cannot be categorized.
In fact, isn't it overwhelmingly obvious that movies can be categorized?
We have lots of motion picture categories, indeed, many that are perfectly
unobjectionable, such as suspense, horror, structural films, melodramas, mys-
teries, thrillers, trance films, action films, war films, platoon films, German
expressionist films, science fiction films, neo-realist films, bio-pics, comedies,
mythopoetic films, art films, westerns, socialist realist films, ninja films, fantasy
films, musicals, surrealistic films, Bollywood films, costume films, new talkies,
fluxus films, anime, and various New Waves (add the country of your choice
here — France, Germany, Taiwan, Hong Kong, South Korea, etc.). Of course,
I've barely started; I could go on for pages.
Some of these categories are less widely known than others, though
many of them can be found in the movie and TV listings in your daily
newspapers as well as on the racks in video stores. Some of these categories
may be genres, such as the soap opera, while others can be movements,
like the New American Cinema, or styles, as in the case of film noir or
montage films.
Of course, these categories are not necessarily exclusive and can, in
many cases, be combined, as in Outland (which restaged High Noon on
 EVALUATION 209

Jupiter's volcanic moon of lo). Nevertheless, there is no reason to think

that regarding certain motion pictures under some of these categories
rather than others is not often correct. The Gunfighter is a western, not a
musical; An hasard Balthazar is an art film, not a romance picture; L'Age d'or
is surrealism, not socialist realism. And so on.
Furthermore, many of the aforesaid categories (and scores and scores
of others unmentioned) come with subtending purposes and standards of
excellence attached to them. For example, some movies are correctly
categorized as melodramas. Part of the point of a melodrama is to incite
the audience to feel — at least for some portion of the duration of the movie —
sadness or pity for one or more of the protagonists. A correctly categorized
melodrama that failed in this respect would be prima facie defective.
A correctly classified melodrama, as opposed to a parody thereof, that set
us laughing at the mishaps of the protagonists, as arguably John Cromwell's
Madefor Each Other, with its excessive agenda of relentless misfortunes, does,
would be a disaster. But a melodrama, like Back Street, that reduces us to
tears at the plight of its central characters is prima facie a good melodrama.
Likewise, a correctly identified classical detective movie that fails to
prompt speculation about whodunit — perhaps because the butler answer is
too blatantly obvious from the get-go — or a comedy whose gags are so
badly timed that every laugh line falls flat are both presumptively failures.
For they do not measure up to the standards of the categories to which
they are known to belong. If nothing can be said to justify the ways in which
these movies deviate from the criteria of worthiness for the categories by
which they are justly described, then we can rationally ground our negative
assessments of them.
To return to the question of Inside Man: you said that it is good; I said
that is bad. However, in the course of our discussion, it became apparent
that the likely basis for my disparaging assessment of Inside Man was that
I placed it in the wrong category. I situated it in a comparison class along
with Fast, Cheap &. Out of Control, an essayistic documentary in the lineage
of Werner Herzog with aspirations to be meditative — that is, aspirations to
invite an active exploration of certain philosophical issues on the part
of the viewer. But clearly the aims and standards relevant to this type of
documentary differ wildly from those appropriate to thriller-suspense films
(of the heist and hostage variety). Among other things, meditative docu-
mentary essays are designed to invite abstract rumination, and, though
some suspense fictions may also do this (think Vertigo here), explicitly
provoking philosophical rumination is not a standard expectation that we
require every (or even any) suspense thriller qua thriller to fulfill.
210 EVALUATION

A meditative documentary ought to be philosophically provocative. If a

suspense movie is not philosophically intriguing — as Inside Man is not — that
should not count against it (given its category); if a suspense thriller
is philosophically interesting, then that's icing on the cake, as they say
(though "icing" that needs to be philosophically interrogated, as frighten-
ing as that sounds). On the other hand, were an intended thriller-suspense
movie philosophically ambitious, yet not suspenseful in the least, we would
find it wanting and probably discount it as deadly dull, would we not?
However, in any event, the failure to engender philosophy is not a basis
for charging that a motion picture correctly classified as a suspense thriller
is bad, since philosophizing is not wherein excellence in the suspense genre
primarily resides. Thus, if you can adduce objectively creditable reasons
for categorizing Inside Man as an exercise in thriller-suspense, while also
challenging my attempt to compare it with meditative documentaries, then
you have a reasonable basis for dismissing my dissing of Inside Man. And if
I cannot undermine your categorization with reasons, nor defend my
categorization, I should (rationally speaking) concede your point.
In this example, a great deal of weight has been placed on the role of
reasons in supporting or rejecting categorizations. But this may strike the
reader as somewhat obscure. What are these reasons and in what sense are
they objective? Where do they come from? In this regard, three kinds of
reasons come into play most frequently in classifying motion pictures.
The first kind of reason pertains to the structure of the work: if the
work possesses a large number of features that are typical of the motion
pictures inhabiting a certain category, then that is a strong reason to
subsume it as, for example, under the category of a suspense thriller.
Conversely, the fewer features a movie contains of the sort that are typical
for a certain category, the less likely it is that the movie belongs to that
category. Inside Man, for example, possesses a great many of the features
common to suspense thrillers — like a daring escape against the odds — but
it has few of the salient features of meditative documentaries (since, for
example, it is neither a documentary nor meditative). So it is far more
reasonable to regard Inside Man as a suspense thriller and to award it high
marks for being good of its kind.
The number of relevant structural features a motion picture possesses or
lacks, as well as the degree of salience those properties project, may not
always provide conclusive reasons in favor of one categorization over
another, but statistics such as these generally provide evidence in the
direction of one classification versus another. Moreover, these reasons
are objective inasmuch as it can be a matter of inter-subjectively verifiable
 EVALUATION 211

fact that a movie possesses or lacks a certain number of clearly defined

features and that said features are typical of a certain category of moving
picture. That Inside Man contains a daring escape is not a figment of
personal opinion; it is incontrovertible. That daring escapes are a mark
of suspense thrillers is likewise a matter of historical record.
Another kind of reason that is pertinent to the question of the correct
categorization of a movie involves authorial intention. How a moviemaker
or a group of moviemakers intend a work to be classified is quite often
an inter-subjectively determinable matter of fact, and, in a great many of
the remaining cases, a highly plausible hypothesis about the intended
categorization of a given movie is available. We know as well as we know
anything, for example, that Martin Scorsese intended The Aviator as, at least,
a bio-pic.
Many scholars in the humanities today are chary of attributing intentions to
artists, including movie directors. But their anxieties are generally misplaced.
There is nothing more chancy about recognizing the intentions — especially
the categorical intentions — of moviemakers than there is about divining the
intentions of others in everyday life. Just as one cannot typically doubt
that the conductor on the train is reaching out for your ticket, can anyone
really doubt that Howard Hawks intended Bringing Up Baby to be a comedy
or that Henri-Georges Clouzot meant Diabolique to be a thriller? In a great
many cases, there is nothing arcane about determining a moviemaker's
categorical intentions. It is only academic fashion that seems to make
it seem so.
Of course, authorial intentions may not always provide conclusive
reasons for a specific categorization. Sometimes we may suspect the authors
of dissembling (this is often said of so-called "snuff" films) or possibly of
being confused. Or the evidence may be indeterminate. Nevertheless, very
often our information is adequate enough to support a presumption in
favor of one categorization over another. And when evidence of authorial
intention can be wedded to structural evidence, of the sort alluded to
previously, the grounds for rationally preferring one categorization over
another mount appreciably.
Further reasons on behalf of the categorization of a motion picture can
be gathered from the historical and/or cultural context in which the
moving picture emerged. That is, if a certain category of moviemaking is
alive and abroad in the historical and/or cultural context in which the
film was produced (especially when competing categorizations are not),
then that supplies rational grounds for classifying a movie one way rather
than another.
212 EVALUATION

Notice that this too is generally a question of fact. Whether a certain

kind of motion picture is common or can be found in a given moviemaking
community at a certain time period is not up for grabs. Were spaghetti
westerns being produced in Hollywood in the 1920s? No. Is this just my
opinion? No. It is a verifiable, historical datum.
That thriller-suspense films are a staple in the cultural enclave of the
Hollywood-type, mass-market moviemaking whence Inside Man originated
gives us some warrant to think that thriller-suspense is the category or
comparison class to which it belongs, especially in contrast to the suggestion
that it be assessed as a meditative documentary of the ilk of Fast, Cheap ScOut
of ControL For meditative documentaries are not native to the kinds of
popular venues for which Inside Man was crafted and distributed; their
venues are far more specialized than even Michael Moore's screeds.
Of course, historical and/or cultural reasons of this sort are not absolutely
conclusive, but like structural and intentional reasons, they supply us
with some rational motivation to move in the direction of one classification
rather than another, particularly where the rival classifications are less
convincing historically and/or contextually.
Structural, intentional, and contextual considerations, then, provide us
with reasons, often strong reasons, for categorizing motion pictures a
certain way. Where we have reasons of all three kinds, their combined
force may sometimes be conclusive, or at least as conclusive as things get in
this vale of tears. Often, these kinds of reasons dovetail. For example,
contextual considerations frequently count as evidence for — and, thereby,
reinforce attributions of — authorial intention, since moviemakers usually
have an interest in addressing audiences in terms of categories with which
they are familiar. Most moviemakers (save certain avant-gardists) are not
invested in confusing their target audiences. Moreover, moviemakers
typically signal the category they have in mind by exhibiting in their motion
pictures well-entrenched and readily recognizable structural features of the
kind associated with the type of moving picture they are presenting. That is
why so many horror pictures begin on dark and stormy nights and westerns
open with panoramas and expansive musical scores.
Of course, in some cases, our categorizations may not depend on all
three kinds of reasons in concert. Sometimes structural (or authorial or
contextual) reasons alone will be sufficient. This is especially likely to be
the case where there are no persuasive, countervailing reasons available for
competing categorizations.
Where does the current discussion of motion picture categorization
leave us? First, it lends support to our earlier conviction that there are
 EVALUATION 213

objective grounds for categorizing films one way rather than another. Thus,
if the objective evaluation of motion pictures rests upon our ability to
categorize movies correctly, then we have shown that this requirement can
be met sometimes, if not often. Furthermore, if we possess the means
to categorize moving pictures correctly — and to defend certain categori-
zations over others — then we can rationally settle some — indeed, I suspect
many — disagreements with respect to evaluating movies.
In order to defend our own evaluations of motion pictures against
competing ones, we will often proceed by demonstrating that our evalu-
ation is grounded on a correct categorization, while also arguing that rival
evaluations depend on incorrect, unlikely, or, at least, less plausible ones.
Moreover, insofar as this procedure of invoking objective categorizations
enables us to adjudicate a significant number of evaluative disputes, it
establishes that motion picture evaluation is, at least to this extent, a
rationally governed activity. Let us call this procedure for evaluating
motion pictures the pluralistic-category approach, since it depends upon
countenancing many diverse categories of moving pictures.
Admittedly, the pluralistic-category approach will not dissolve every
disagreement about motion picture evaluation. Even if the correct category
is identified, there may be debates about how to apply it — disputes, for
example, about how a given movie fits the category or categories in
question; disagreements about how to weigh the different standards that
may be pertinent to the category or categories at issue, and debates about
their underlying purposes. Perhaps most often, these controversies need to
be worked out on a case-by-case basis (which is not to say that they cannot
frequently be resolved objectively). As well, some further matters, involv-
ing cross-categorical rankings — which we will address in the next section
of this chapter — require an approach different from the pluralistic-category
approach. Nevertheless, it remains the case that a great many of our
disputes over the merits of moving pictures can be settled by establishing
the correct category, as when the husband disparages a movie because it
lacks gunfights, but the wife points out that it is a light and frothy,
adolescent romantic comedy (rather than, say, a dark one like the recent
Mr. and Mrs. Smith).
One objection to the pluralistic-category approach is that it is inherently
formalist, since it primarily involves assessing a movie in virtue of the kind
of motion picture it is, referring its evaluation to the formulaic canons of
certain categories, whether they be genres, movements, styles (period and
otherwise), and so forth. One idea behind this criticism is that if movies
are evaluated in terms of categories, they will be insulated from cognitive,
214 EVALUATION

moral, or political evaluation in favor of whatever comprises the internal

standards of the category.
But this anxiety is ill founded for at least two reasons. First, many
motion picture categories quite clearly presuppose criteria of cognitive,
moral, and political excellence. For example, many nonfiction films, as
a matter of the kind of documentaries they are, are committed to educating
their audiences cognitdvely as well as often enhghtening them morally and
politically. Furthermore, a social problem picture that failed to portray
injustice effectively would be a defective specimen of its category
Moreover, even in cases where the categories in question are not as
straightforwardly connected to other-than-formalist criteria of excellence,
cognitive, moral, social, political, spiritual, and other "extra-aesthetic"
considerations are usually not irrelevant to the evaluation of the movies
that fall under their rubric. Suspense pictures, for example, given the
argumentation of the previous chapter, will have to abide by a recognizably
pro-social ethic, if they aspire to enlist the care and concern of typical
audiences on behalf of those protagonists hanging from the cliffs. Melo-
dramas will have to reassure us that those whom we are intended to pity
are morally deserving; most normal viewers will not cry for the remorse-
less Nazi commandant who dies in a car accident while trying to run over
an ailing grandmother and her infant grandson; many may even applaud.
In cases like these, the morally flawed perspectives of the movie will
patently count against them.
Likewise, cognitive gaffes can also be demerit-making features in a
movie that is not overtly committed to enlightenment, as in the case of
the adventure flick in which the hero announces that he is running out of
bullets but continues blasting away for the next thirty-five minutes or so.
Being an action movie does not automatically get you a pass for outright
dumbness. In short, things like cognitive and moral defects can enter our
deliberations not only with respect to motion picture categories that are
designed to address such issues directly, as part of the nature of the kind
of category they are, but also with respect to the evaluation of movies in
many other categories as well.
Another objection to the category approach is that it makes a funda-
mental ontological mistake. It seems to suppose that there is one and only
one category into which each motion picture neatly fits with no remainder.
And this is just false. But, of course, we may readily concede that movies
may inhabit more than one category. However, that does not show that the
approach is flawed. It only entails that evaluation may be more complicated
than examples so far have indicated. Motion pictures may combine
 EVALUATION 215

categories and, therefore, may need to be evaluated in light of several

moving picture kinds. Bride of Frankenstein, for instance, blends horror and
dark comedy. This is a case where the mesh is nearly perfect. However, this
kind of category splicing can also lead to mixed results. A movie may turn
out to be good with respect to one category to which it belongs, but bad
with respect to another. The humor component of a comic mystery of the
Cat and Canary variety may be superlative, while the mystery element is
utterly and disappointingly transparent.
This sort of situation, however, should come as no surprise, Mixed
results often figure in our assessments of movies. Even with respect to a
single category, more than one criterion of excellence may be relevant in
evaluating a film. Several Batman spectacles have been especially commend-
able for their production designs, while being leaden narratively. Of course,
if one comes to think about it, probably most of our assessments of
movies — indeed, of most artworks, save the most magnificent master-
pieces — are mixed. Mixed results are just something — like gravity — which
those of us who evaluate movies have to learn to live with. Thus, the fact
. that the pluralistic-category approach to motion picture evaluation leads
to mixed results should not be taken as a shortcoming. In fact, that it
can usually clarify the sources of the mixed results should probably be
scored in its behalf.
Some may fear that relying so heavily on motion picture categories is too
conservative. It appears to presume that the pertinent categories are fixed.
But moving picture categories are also mutating. Some are disappearing,
others are changing (often by combining with other categories), and
sometimes, in addition, new categories are coming into existence. Can
the category approach handle motion pictures that are in the process of
evolution, not to mention movies that herald the very onset of new genres?
For, certainly any approach is of dubious value, if it ignores evolutionary
phenomena like these.
But the pluralistic-category approach need not deny or otherwise
repress these phenomena. In defense of the approach, however, it is
important to remember that the pertinent categories, like the movies
that fall under their umbrella concepts, do not hail from nowhere. New
categories arise from earlier ones by well-precedented processes of devel-
opment — such as hybridization, amplification, repudiation, and so forth.
Consequently, it is possible to track emerging categories along with their
purposes and their subtending standards in medias res.
Perhaps this seems doubtful when it comes to the avant-garde. But
with respect to avant-garde cinema, the identification of new categories,
216 EVALUATION

purposes, and their correlative criteria of accomplishment is typically

facilitated by the existence of manifestos (signaling authorial intent) and
the lively motion picture culture that surrounds such film and video
production (which supplies pertinent historical/contextual information)
as well as, frequently, information that comes through the existence of
related aspirations (and similar structural features) articulated in adjacent
artforms, like painting and music, which limn the rationale behind new
developments in cinema (that is, which afford even more informative
historical/contextual background).
For example, the avant-garde category of the New Talkie — which
appeared in the late 1970s and early 1980s — was comprehended virtually
with the arrival of the first examples of this tendency due to the manifestos
and discourse that enveloped it. Thus, there is no reason to fear that talk
of categories precludes receptivity to novel forms of motion picture
making, even in the face of the persistent challenges of avant-garde
visionaries. There is always a social context that prepares for "the
unexpected." When something is called "unprecedented" or "unexpected"
in the world of cinema, that is, more often than not, promotional rather than
descriptive.
The defense of the pluralistic-category approach in the preceding para-
graph is likely to ignite the ire of those who think that there are motion
pictures that stand outside of any category — works of such exceeding
uniqueness as to be reckoned one of a kind. But there are grounds to be
extremely skeptical of this view. A motion picture that was truly, onto-
logically incomparable with any other would be, honestly, an equally
incomprehensible artifact. Faith in the consummately singular motion
picture derives from a romantic-modernist fantasy of the genius. But the
creature dreamt up in these philosophies is a psychological impossibility.
No artist comes from nowhere. And, in any case, the artist as a young
wolf-child or an apprentice to Kasper Hauser would be indisputably
incapable of making a movie.
A recent motion picture that has been proffered as one that defies the
category approach is the Cremaster series by Matthew Barney. This is a
brilliant, sprawling work with often elusive imagery of gay import and
radical juxtapositions. But it does fit into a tradition and it does belong
to a category — namely, that of the mythopoetic cinema, as christened
by P. Adams Sitney, which includes Inauguration of the Pleasure Dome by
Kenneth Anger, Heaven and Earth Magic by Harry Smith, Dog Star Man
by Stan Brakhage, and others. There is no doubt that this category is not
on the tip of every viewer's lips — as is the label musical mostly is; however,
 EVALUATION 217

the obscurity of the category does not militate against either its reality or
its relevance.
In this regard, it pays to note that one function that motion picture
critics can perform for the general public is to inform them of the
existence of categories that are not widely known and to explain their
points, purposes, and virtues. It is unfortunate that journalistic film
criticism is often assigned to writers who possess nothing other than the
ambition to write — thus presuming that anyone can write about movies —
whereas a really accomplished motion picture critic should have a
capacious enough knowledge and understanding of motion picture history
that she ranges widely over its stunning diversity of categories, subgenres,
visionary movements, stylistic tendencies, and so forth. One of the most
significant services that a motion picture critic can perform is to enable
plain viewers to place puzzling work in unfamiliar categories where the
goodness to be had from the movie in question becomes available to
viewers upon the elucidation of the subtending purposes of the relevant
categories.
Clearly, the pluralistic-category approach to motion picture evaluation,
as its very label advertises, is not as unified as the one proposed by classical
film theory. Since classical film theory acknowledged only one category, it
supplied a unitary metric according to which every motion picture, genre
or style notwithstanding, could be ranked on a single scale. That is, every
movie could be compared for its degree of cinematicity. The pluralistic-
category approach is far more fragmentary, since it maintains, quite credi-
tably I would argue, that there are many different categories that we can and
indeed should call upon to evaluate different types of moving pictures.
Some may find this disheartening. They might complain that the plur-
alistic-category approach makes the qualitative comparison of motion
pictures from different categories impossible. This may be a big disap-
pointment, especially when contrasted to the capacity of the classical
tradition to rate all movies along a single grid. But, these lamentations
may be exaggerated, on the one hand, and Utopian on the other.
The objection that the pluralistic-category approach makes all comparisons
between motion pictures of different categories impossible is hyperbolic
for at least two reasons. First, as noted earlier, many movies inhabit more
than one category. Consequently, a high-spirited, comic spy film can be
compared qualitatively to a comedy simpliciter in terms of humor. Moreover,
there is no reason to suppose that there are never shared, cross-categorical
dimensions of evaluation, such as narrative coherence or the deft and not-so-
deft manipulation of point-of-view structures.
218 EVALUATION

At the same time, the wish to be able to compare every motion picture
with all other motion pictures is frankly unrealistic. Why imagine that a
near-perfect cine-dance like Hilary Harris's Nine Variations can be weighed
against an estimable social drama like Once Were Warriors? And even supposing
one could find some points of tangency between these two moving pictures,
note that in virtually every respect that we care about with reference to these
two achievements, the two candidates are effectively incommensurable.
Indeed, sometimes the attempt to compare certain films with others
cross-categorically seems almost silly. When I was youngs my friends and
I would get into shouting matches over whether Sherlock Holmes was
smarter than Captain Nemo, or whether Mighty Mouse was tougher than
Superman, or, my favorite: could Frankenstein's Monster take Godzilla?
This was a pleasant enough way to blow a summer afternoon, but, in point
of fact, these arguments were ridiculous, since the comparisons were being
made between non-overlapping fictions whose make-believe worlds were
utterly unconnected. Similarly, but for different reasons, it often is simply
nonsense to attempt to appraise every movie on a single axis.
This is especially obvious with regard to masterpieces. Which is
greater — A Midsummer Night's Dream (the play) or the Goldberg Variations?
Who knows? How would you go about deciding the issue? And, basically,
who really cares? At a certain level of achievement, trying to parse the
precise difference in quality between masterpieces, even if possible, seems
feckless. Both the aforesaid works belong to the highest strata of artistic
excellence; calibrating a differential between them seems akin to deter-
mining whether Captain America is stronger than Batman. Both artworks
are just "top of the line."
Just as Captain America and Batman are the non plus ultra human
warriors of their respective, ontologically disjoint fictional worlds and,
consequently, not suitable for direct comparison, so A Midsummer Night's
Dream (the play) and the Goldberg Variations dwell at the summits of at best
weakly commensurable artistic categories. Turning to motion pictures,
judging whether Some Like It Hot is better than John Huston's Maltese Falcon
seems likewise beside the point. It is questionable whether it can be done,
how it could be done, and, most of all, why it needs to be or should be
attempted at all.
Because it may not make sense to attempt to weigh every motion picture
against every other motion picture, it cannot be a failing of the pluralistic-
category approach that it is unable to do so. Perhaps this alleged short-
coming of the pluralistic- category approach is actually a reflection of its
underlying good sense.
 EVALUATION 219

But . . .
Maybe sometimes, even often, it is silly, or imponderable, or both to try to
evaluate comparatively certain works of radically disjoint and virtually
incommensurable categories. Nevertheless, it is not always silly to do so.
Aren't there cases where we are willing to indulge in this sort of com-
parison with great confidence? Does anyone feel squeamish in asserting
that Tokyo Story is a better film than The Big Broadcast of 1938 or that Father
Panchali is superior to Glen or Glenda? However, these are not the sort
of evaluations that the pluralistic-categorical approach is well suited to
advance, since it seems fruitless to attempt to situate either of the
preceding pairs of motion pictures informatively in the same category or
categories. Consequently, the pluralistic-category approach to motion
picture evaluation has to be supplemented to manage cases like these (and
probably other sorts of cases as well).
One task for the philosophy of a practice, including the practice of
moving picture evaluation, is to strive to establish the premises upon which
reasoning in that practice proceeds. So far we have tried to show that a
great deal of movie evaluation rests on trying to place the moving picture
on the docket in the correct category (or categories). But that cannot be
the whole story about movie evaluation, since there appear to be perfectly
acceptable motion picture evaluations where the category approach
appears inapplicable. Tokyo Story and The Big Broadcast of 1938, on the
one hand, and Pather Panchali and Glen or Glenda, on the other, are so
categorically remote from each other that it seems pointless to suppose
that they share any aims concrete enough to motivate the differential
assessments we are willing to make here.
Nevertheless, with these particular examples, it is not difficult to
hypothesize the principle that makes the evaluations at issue compelling.
Tokyo Story and Pather Panchali number among the most excellent achieve-
ments in their class, while The Big Broadcast of 1938 is at the bottom of its
particular heap and Glen or Glenda is the worst of its lot. The pairs in
question, though not profitably situated in shared categories, can be
compared evaluatively vis-a-vis their relative status in the categories to
which they do belong. And where one motion picture is an excellent
example of its type and another makes a very poor showing in its class, we
are likely, all things being equal, to rate the former above the latter.
Yet the preceding possibilities do not exhaust the field. Isn't it fair to
suppose that most informed motion picture goers would be prepared
220 EVALUATION

to agree that The Bicycle Thief is & better moving picture than Beetlejuice? In
this case, we have two motion pictures, each of which is at the top of its
category — humanist realism (The Bicycle Thief), on the one hand, and
horror-comedy (Beetlejuice), on the other. The narration in both films is
flawless and the performances perfect. It is tempting to think that this
might be one of those cases where we think it is fatuous to ask whether
one of these really splendid motion pictures is better than the other. And
yet it does not sound silly or strained to maintain that The Bicycle Thief
is better than Beetlejuice, even to someone like m e who really adores
Beetlejuice. Why not?
The value or disvalue of a motion picture is connected to its purposes.
As we have seen, moving pictures belong to various categories, which
categories, in turn, are devoted to the realization of certain points and
purposes. A movie that accomplishes its aims succeeds as the kind of
motion picture it is; a mystery, for example, that sustains our curiosity
throughout achieves its objective and is objectively valuable for doing so.
However, assessing a motion picture from the perspective of its category
(or categories) does n o t address the question of the value of the purpose
that the category is intended to attain. Nevertheless, it would appear to
stand to reason that an overall evaluation — a.k.a. an "all-things-considered
judgment" — of a motion picture should take into account not only the
movie's success or failure by its own (category-relative) lights, but the value
or disvalue of the purpose or purposes to which the category is committed.
Moreover, the achievement of some purposes is clearly more valuable
than the achievement of others. For example, the discovery of the polio
vaccine was m o r e valuable than my managing to get out of bed this
morning (or any other morning, for that matter). Likewise, the dramatist
who succeeds in disclosing the deep secrets of ressentiment produces some-
thing of greater value than the New Yorker cartoonist who captures perfectly
the quaint foibles of a Greenwich Villager complaining about out-of-town
bagels. In both cases, the works in question are good of their kind, b u t we
rank the purposes of the one kind mor e highly than the purposes connected
to the other kind.
Returning to the case of The Bicycle Thief versus Beetlejuice, it seems likely
that we are inclined to rank The Bicycle Thief'more highly because, all things
being equal, we think that cinematic works of humanist realism are m o r e
valuable than works of horror-comedy in cases where each of the candi-
dates in question are superb with reference to their own respective kind.
O n the other hand, we may be less willing to rank The Bicycle Thief and
La Grande Illusion, since both realize their purposes magnificently and
 EVALUATION 221

neither purpose seems obviously more valuable than the other. Indeed,
their aims seem of a piece.
Movies are evaluated in terms of whether they succeed or fail in
realizing their purposes. Generally, these purposes are rooted in the
categories the motion pictures under evaluation inhabit — which categories
include genres, subgenres, cycles, styles, movements, and so forth. Some
of these categories may belong exclusively to the history of motion
pictures, such as the cine-dance, but many belong to more than one
artform, such as the melodrama.
Yet, in either case, it is primarily from its presiding category (or
categories) that the movie, so to speak, gets its marching orders. If the
motion picture reaches its destination, then it is good as the kind of moving
picture it is; if not, not. This kind of evaluation might be thought of as
movie evaluation, narrowly construed, since it is tethered to the specific
kinds of movies there actually are and have been. Evaluation of this sort is
the type of evaluation to which movie critics and connoisseurs may lay
special claim, because it relies upon a sophisticated knowledge of motion
.picture history, its modes, and its varieties.
Yet sometimes we also incorporate into our overall evaluation of movies
an assessment of the value of the purpose which the motion picture at hand
successfully realizes. Where does this evaluation come from? The short
answer is that it comes from the culture at large. In this respect, this
species of judgment might be thought of in terms of the evaluation of
motion pictures, broadly construed. Moreover, this is the kind of evalu-
ation in which we are involved when we pronounced The Bicycle Thief to
be better than Beetlejuice. We are expressing the conviction, ceteris paribus,
that The Bicycle Thief and the kind of purpose this category of motion
picture making implements serves more important cultural interests than
Beetlejuice and the genre of horror-comedy.
Indeed, even within a single category, we appear willing to rate one very
excellent example of a kind over another very excellent instance, where
one of the pair, in addition to being splendid of its type, also serves some
higher purpose. Thus, we may feel that Vertigo is better — all things
considered — than The Lady Vanishes, since, although both films do an
especially fine job engendering suspense, Vertigo offers insights into the
philosophy of romantic love as well, thereby contributing more than simply
a thrilling adrenaline surge.
Of course, once we start to issue estimates of this variety, we are no
longer simply speaking as movie critics and/or connoisseurs with narrowly
specialized knowledge about the history of movies. Movie evaluators can
222 EVALUATION

say what makes for a good splatter film, but they have no special authority
when it comes to attesting to the value of a good splatter film, like
Halloween n + 2 , versus a satiric exposure of hypocritical posturing, like
Team America. To enter the discussion about the value of the cultural
interests served by different cinematic categories, or even by different
individual movies, is undoubtedly to broach questions of general axiology.
To assert, for example, that even a very remarkable splatter film is, ceteris
paribus, not as valuable as an equally well executed satire of contemporary
affairs obviously requires taking a stance upon the status, aims, and needs
of the civilization in which this particular discussion makes sense. That is,
this discussion involves a shift from movie evaluation, narrowly construed,
to cultural criticism — a transition that requires that the best prepared
evaluators of movies not only know film history, but be generally informed
intellectually as well.
Although the very best movie evaluation requires an informed and
refined understanding of the actual history of the moving image, it does
not entail that, in our evaluations of motion pictures, we should stick to
the narrow gauge — in fact, far from it. Instead, there would be something
peculiar about trying to sequester movie evaluation solely inside the history
of the moving image. For motion pictures penetrate the life of the culture
in every way, and therefore the discourse about them, including the
evaluative discourse, should also be multi-layered. Unquestionably, movie
evaluation requires specialist knowledge, but it also calls for an engagement
with the life of a breathing civilization.
In some eases, evaluating some motion pictures may require summing
up the value of the movie in terms of the kind of moving picture it is plus
some estimate of the social value of the purposes that category of moving
image usually serves. With reference to our previous discussion of The
Bicycle Thief and Beetlejuice, talking crudely and only heuristically, we might
try to represent this by saying that on a scale of —10 to + 1 0 , both films
get a + 9 . 9 for category excellence. But on a similarly calibrated scale for
cultural importance, let us say that The Bicycle Thief gets a +9.9 again, but
Beetlejuice only garners a + 6 . Overall, then Bicycle Thief would merit a
score of +19.8, while Beetlejuice racks up a +15.9, where the hypothetical
point spread is explicable in virtue of the variable cultural weight that we
assign to the purposes we attribute to these different kinds of movies.
Although we may often feel uncomfortable in admitting it, when we are
confident in claiming that, of two excellent motion pictures of barely
commensurable categories, the latter is superior to the former, it is
reasonable to conjecture that that is probably because we are convinced
 EVALUATION 223

that the latter (or the category to which it belongs) is more important
because the civilizational interests it serves are more important than those
served by the former. That is what really grounds our evaluations. Indeed,
it is difficult to imagine what other grounds there might be. The only
reason that we feel nervous about confessing this is that many of us are still
in the grip of the ideology of the autonomy of art, including motion picture
art, which dictates that considerations of social importance are somehow
beyond the pale. And yet as members of society we cannot but help
to harken to its claims when it comes to assessing almost everything,
including art and even movies.
Does this entail that it does make sense to compare every film in this
way? Probably not. When we have excellent motion pictures, such as
masterpieces, from disparate categories where the there is no assignable
difference in cultural value between the purposes of the categories in
question or where the difference is imponderable or indeterminate, it is
likely that an attempt at comparison will either be beside the point or silly.
Of course, as the preceding sentence concedes, there may be cases where
the cultural weighting to be assigned to the different categories will be
disputed by reasonable people. In those cases, our evaluative debates
may sometimes go unresolved. Nevertheless, there are no grounds for
thinking that none of our disagreements will ever be resolved. The cartoon
"Godzilla Meets Bambi" is perfectly delightful enough on its own terms,
but no one can seriously maintain that it is of the same order of cultural
heft as Lanzmann's Shoah and its like.
Some, even many, evaluative disputes will be settled. Some will be
beside the point. And some may remain intractable. However, the fact that
some evaluative disagreements concerning motion pictures remain unset-
tled is not a reflection of some special weakness of moving picture
evaluation in contrast to other cognitive enterprises. For we should not
expect to resolve every dispute concerning the evaluation of moving
pictures. After all, we have not done that in any other field of inquiry.

The Value of Cinema?

So far, we have been stalking the question of value in relation to cinema in
terms of the evaluation of particular motion pictures. However, a number
of previous theorists and philosophers have also been invested in claiming a
value for cinema as such — albeit cinema generally understood as film. For
example, Bazin thought that we value film as such as a kind of celluloid
224 EVALUATION

mortuary, a means to immortality (since on photographic film, Greta

Garbo will be eternal); Kracauer maintained that film redeemed physical
reality for a civilization that had become overly abstract; and Stanley Gavell
suggests that film might serve as a symbol or talisman against the anxieties
of solipsism, inasmuch as film shows us a world past, a world that, despite
our absence from it, persists.
It would be wonderful to conclude this book with a comparable flourish.
However, as the argument of this chapter and others suggests, there is no
single, unique value that film, or, more broadly, the moving image as such
achieves. Unlike Bazin, Kracauer, and Cavell, we have defined our object
of inquiry functionally rather than in terms of a specific medium. The
function of the artifacts in question is their capacity to impart the impres-
sion of movement, a function that can be multiply realized in a number of
media, including film, obviously, but also video, broadcast TV, CGI, and
technologies yet to come.
Moving imagery, in turn, can obviously serve a multitude of agendas,
from the documentation of scientific experiments, the analysis of move-
ment, security surveillance, animal studies, the diagnosis of football strategies,
ballistics testing, and weather forecasting to producing video and computer
games, advertising, political propaganda, the delivery of the evening news,
and, of course, fictional movies.
Moving imagery can implement many purposes and gets its value (and
its disvalue) from the uses to which it is put, and its successes and failures
with respect to discharging them. It has no end in itself; the moving image
is a means to an end, and each use of moving imagery ultimately derives its
value through factoring into account the worth of its goals and its actual
facility in accomplishing them.
The moving image as such has no particular value, unique or otherwise.
No one watches a moving image for the intrinsic value of seeing movement
in the image. There has to be a point to be made by the movement. And
from thence — the point and its making — comes the value.
Historically, the capacity of the moving image to project motion pictures
that can be combined into visual narratives, generally of a fictional sort, has
been its most prominent function in Western culture, and in others.
Throughout the twentieth century, visual narration — as art and /or enter-
tainment, as cinema and/or movies — has been the purpose that has driven
the evolution of the moving image and the refinement of its technical
capabilities. That is why so much of this book has been devoted to
questions pertaining to art and narrative. For as a philosopher of a practice,
one must go where the practice does.
 EVALUATION 225

However, things are changing. Motion pictures are no longer creatures

that inhabit just movie houses. They have taken over our homes (TV), our
offices (computers), and even our pockets (cellphones). They dominate
not only our lives, but our children's lives (video and computer games).
In some big cities, moving picture billboards are cropping up on subway
entrances; huge ones vault certain major intersections. O n large airplanes,
the stewardess no longer recites the safety instructions; the monitor on the
back of the seat in front of us illustrates them. A videocassette or a DVD
tells us how to assemble our computer. And so on.
Moving images are not only for Friday nights or rainy Saturday after-
noons. They are everywhere, all of the time. We are becoming a moving
image culture, n o t merely in the sense that motion pictures are our leading
form of art a n d / o r entertainment, but in the sense that they are beginning
to play a role in a remarkably diverse and comprehensive range of our
communicative activities. Perhaps someday they will be a part of almost
every one of our channels of exchange.
Thus, one predicts that in the future the philosophy of the moving image
will have a far larger waterfront to cover than the one that has been
surveyed in this text.

Suggested Reading

As indicated in this chapter, the question of motion picture evaluation has not
been the subject of much recent philosophical and/or theoretical discussion.
Nevertheless, sometimes explicitly, but often implicitly, the question of film
evaluation — as it relates to the notion of the cinematic — was a preoccupation
of much of the film theory produced through the 1960s by what might be called
the classical tradition. See the suggested reading at the end of chapter 2 for
citation of several of the most relevant texts.
One participant in that conversation who is specifically interested in laying the
foundations of rational film evaluation is Victor F. Perkins. His book Film as Film:
Understanding and Judging Movies (Baltimore, MD: Penguin, 1972) is probably the
most ambitious and thoughtful treatise on film evaluation to date.
The discussion of the pertinence of categories for motion picture evaluation
was inspired by Kendall Walton's very important article, "Categories of Art," The
Philosophical Review 79 (1970), 334—67. An earlier version of the pluralistic-
category approach can be found in Noel Carroll, "Introducing Film Evaluation,"
in Engaging the Moving Image (New Haven, CT: Yale University Press, 2003).
A useful response to that article is Cynthia Freeland's "Evaluating Film," Film
Studies 8 (Summer 2006), 154-60.
226 EVALUATION

Daniel Kaufman has recently published pioneering work on the philosophy of

evaluation. See Daniel Kaufman, "Normative Criticism and the Objective Value
of Artworks," Journal of Aesthetics and Art Criticism 60:2 (Spring 2002), 150—66,
and Daniel Kaufman, "Critical Justification and Critical Laws," British Journal of
Aesthetics 43:4 (Oct. 2004), 393-400. Both of these pieces are highly recom-
mended to any readers interested in pursuing further the topics broached in this
chapter.
A very influential approach to movie criticism that has not been discussed in this
chapter is what is called auteur theory. Originated in France by members of the
New Wave, it was imported to the United States by several critics, the most
notable of whom is Andrew Sarris.
A founding document of the auteurist persuasion was the 1948 article by
Alexandre Astruc entitled "The Birth of a New Avant-Garde: Le Camera-Stylo,"
which is reprinted in The New Wave, edited by Peter Graham (New York: Doubleday,
1968). This anthology also contains other statements of the theory. A variant of the
approach inflected by structuralism is available in Peter Wollen, Signs and Meaning in
the Cinema (Bloomington, IN: Indiana University Press, 1972).
For the version of auteurism that probably had the most impact on critics in the
United States, see Andrew Sarris, "Notes on the Auteur Theory in 1962," in Film
Theory and Criticism, 3rd edn., edited by Marshall Cohen and Gerald Mast (New
York: Oxford University Press, 1985). Pauline Kael criticizes this approach in her
"Circles and Squares," which appears in the same edition of Film Theory and
Criticism. The auteur theory is also criticized in Noel Carroll's introduction to
Interpreting the Moving Image (New York: Cambridge University Press, 1998).
The topic of evaluative disagreement which motivates a great deal of this
chapter finds its locus classicus in David Hume, "Of the Standard of Taste," in
David Hume: Selected Essays, edited by Stephen Copley and Andrew Edgar (Oxford:
Oxford University Press, 1993). For a useful overview of the twentieth-century
debate among analytic philosophers about the evaluation of art and, in addition,
for a provocative theory of art evaluation, see George Dickie, Evaluating Art
(Philadelphia, PA: Temple University Press, 1988).
 Select Bibliography

Richard Allen, Projecting Illusion (Cambridge University Press, 1996).

Richard Allen and Murray Smith (eds.), Film Theory and Philosophy (Oxford University
Press, 1997).
Joseph Anderson, The Reality of Illusion (Southern Illinois University Press, 1996).
Rudolf Arnheim, Film as Art (University of California Press, 1956).
Alexandre Astruc, "The Birth of the Avant-Garde: Le Camera Stylo," in Peter
Graham (ed.), The New Wave (Doubleday, 1968).
Bela Balasz, Theory of Film (Dover Publishers, 1952).
Andre Bazin, What is Cinema?, 2 vols. (University of California Press, 1967).
Walter Benjamin, "The Work of Art in the Age of Mechanical Reproduction," in
Illuminations, ed. Hannah Arendt, trans. Harry Zorn (Schocken Books, 1968).
David Bordwell, Narration in the Fiction Film (University of Wisconsin Press, 1985).
Making Meaning (Harvard University Press, 1989).
David Bordwell and Noel Carroll (eds.), Post-Theory: Reconstructing Film Studies
(University of Wisconsin Press, 1996).
John M. Carroll, Towards a Structural Psychology of Cinema (Mouton Publishers, 1980).
Noel Carroll, "Address to the Heathen," October 23 (1982/3), 89-163.
- "The Power of Movies," Daedalus 114:1 (1985), 79-103.
Mystifying Movies (Columbia University Press, 1988).
Philosophical Problems of Classical Film Theory (Princeton University Press, 1988).
The Philosophy of Horror (Routledge, 1990).
Theorizing the Moving Image (Cambridge University Press, 1996).
Philosophy of Mass Art (Oxford University Press, 1998).
Interpreting the Moving Image (Cambridge University Press, 1998).
Philosophy of Art: A Contemporary Introduction (Routledge, 1999).
Beyond Aesthetics: Philosophical Essays (Cambridge University Press, 2001).
"The Wheel of Virtue," Journal of Aesthetics and Art Criticism 60:1 (2002), 3-26.
Engaging the Moving Image (Yale University Press, 2003).
Comedy Incarnate (Blackwell Publishing, 2007).
"On the Ties That Bind: Characters, the Emotions and Popular Fictions," in
William Irwin and Jorge Garcia (eds.), Philosophy and the Interpretation of Popular
Culture (Rowman & Littlefield, 2006).
228 SELECT BIBLIOGRAPHY

"The Problem with Movie Stars," in Scott Walden (ed.), Photography and
Philosophy (Blackwell Publishing, forthcoming).
Noel Carroll and Jinhee Choi (eds.), Philosophy of Film and Motion Pictures (Blackwell
Publishing, 2006).
Allan Casebier, Film and Phenomenology (Cambridge University Press, 1991).
Stanley Cavell, The World Viewed, enlarged edn. (Harvard University Press, 1979).
Pursuits of Happiness (Harvard University Press, 1981).
Contesting Tears (University of Chicago Press, 1996).
Cities of Words (Harvard University Press, 2004).
Jinhee Choi, "All the Right Responses," British Journal ofAesthetics 43:1 (2003), 308-21.
"Fits and Startles: Cognitivism Revisited," Journal of Aesthetics and Art
Criticism 61:2 (2003), 149-57.
"Leaving it up to the Imagination," Journal of Aesthetics and Art Criticism 63:1
(2005), 17-25.
Amy Coplan, "Empathic Engagement with Narrative Fictions," Journal of Aesthetics
and Art Criticism 62:2 (2004), 447-69.
Donald Crawford, "The Uniqueness of the Medium," The Personalist 51 (1970),
447-69.
Gregory Currie, Image and Mind: Film, Philosophy, and Cognitive Science (Cambridge
University Press, 1995).
"The Moral Psychology of Fiction," Australasian Journal of Philosophy 73:2
(1995), 250-9.
"Visible Traces," Journal of Aesthetics and Art Criticism 57:3 (1999), 285-97.
Arthur Danto, "Moving Pictures," Quarterly Review ofFilm Studies A\\ (1979), 1-21.
Gilles Deleuze, Cinema, 2 vols. (University of Minnesota Press, 1989).
Jan B. Deregowski, Illusions, Patterns and Pictures: A Cross-Cultural Perspective
(Academic Press, 1980).
Mary Devereaux, "Beauty and Evil," in Jerrold Levinson (ed.), Aesthetics and Ethics
(Cambridge University Press, 1998).
George Dickie, Evaluating Art (Temple University Press, 1988).
Umberto Eco, "On the Contribution of Film to Semiotics," in Gerald Mast and
Marshall Cohen (eds.), Film Theory and Film Criticism, 2nd edn. (Oxford University
Press, 1979).
Sergei Eisenstein, Selected Works (BFI Publishing/Indiana University Press, 1988).
Cynthia Freeland, "The Naked and the Undead" (Westview Press, 2002).
"Evaluating Film," Film Studies 8 (2006), 154-60.
Cynthia Freeland and Thomas Wartenberg (eds.), Philosophy and Film (Routledge,
1995).
Berys Gaut, "On Cinema and Perversion," Film and Philosophy 1 (1994), 3-17.
"The Enjoyment of Horror," British Journal ofAesthetics 35:3 (1995), 284—9.
"Cinematic Art," Journal of Aesthetics and Art Criticism 60:4 (2002), 299-312.
"Identification and Emotion in Narrative Film," in Carl Plantinga and
Gregory Smith (eds.), Passionate Views (Johns Hopkins University Press, 1999).
 SELECT BIBLIOGRAPHY 229

"The Movies: Cinematic Narration," in Peter Kivy (ed.), The Blackwell Guide
to Aesthetics (Blackwell Publishing, 2004).
Alan Goldman, "Specificity, Popularity, and Engagement in the Moving Image,"
Film and Philosophy 5:6 (2002), 93-9.
Karen Hanson, "Minerva in the Movies," Persistence of Vision 5 (1987), 5—11.
Gilbert Harman, "Semiotics and the Cinema," in Gerald Mast and Marshall Cohen
(eds.), Film Theory and Film Criticism, 2nd edn. (Oxford University Press, 1979).
"Eco-location," in Gerald Mast and Marshall Cohen (eds.), Film Theory and
Film Criticism, 2nd edn. (Oxford University Press, 1979).
Elaine Hatfield, John T Cacioppo, and Richard L. Rapson, Emotional Contagion
(Cambridge University Press, 1994).
Brian Henderson, Critique of Film Theory (Dutton, 1980).
Mette Hjort and Sue Laver (eds,), Emotion and the Arts (Oxford University Press, 1997).
Julian Hochberg and Virginia Brooks, "Pictorial Recognition as an Unlearned
Ability," American Journal of Psychology 75:4 (1962), 624-8.
David Hume, Selected Essays (Oxford University Press, 1993).
Ian Jar vie, Movies and Society (Basic Books, 1970).
Philosophy of Film (Routledge, 1987).
Pauline Kael, "Circles and Squares," in Gerald Mast and Marshall Cohen (eds.),
Film Theory and Film Criticism, 3rd edn. (Oxford University Press, 1985).
Andrew Kania, "The Illusion of Realism in Film," British Journal of Aesthetics 42:3
(2002), 243-58.
"Against the Ubiquity of Fictional Narrators," Journal of Aesthetics and Art
Criticism 63:1 (2005), 47-54.
Daniel Kaufman, "Normative Criticism and the Objective Value of Artworks,"
Journal of Aesthetics and Art Criticism 60:2 (2002), 150—66.
"Critical Justification and Critical Laws," British Journal of Aesthetics 43:3
(2004), 393-100.
Haig Khatchadourian, "Film as Art," Journal of Aesthetics and Art Criticism 33:3
(1975), 271-84.
"Movement and Action in Film," British Journal of Aesthetics 20 (1980), 349—55.
Soren Kjorup, "George Inness and the Battle of Hastings, or Doing Things with
Pictures," TheMonist 58 (1974), 216-35.
Pictorial Speech Acts," Erkenntis 12:1 (1978), 55-71.
Peter Kivy, "Music in the Movies," in Richard Allen and Murray Smith (eds.), Film
Theory and Philosophy (Oxford University Press, 1997).
Deborah Knight, "Aristotelians on Speed," in Richard Allen and Murray Smith
(eds.), Film Theory and Philosophy (Oxford University Press, 1997).
Siegfried Kracauer, Theory of Film (Oxford University Press, 1960).
Joseph Kupfer, Visions and Virtue in Popular Film (Westview Press, 1999).
Flo Lebowitz, "Apt Feelings, or Why 'Women's Films' Aren't Trivial," in David
Bordwell and Noel Carroll (eds.), Post-Theory: Reconstructing Film Studies
(University of Wisconsin Press, 1996).
230 SELECT BIBLIOGRAPHY

Jerrold Levinson, "Film Music and Narrative Agency," in David Bordwell and Noel
Carroll (eds.), Post-Theory: Reconstructing Film Studies (University of Wisconsin
Press, 1996).
Andrew Light, Reel Arguments (Westview Press, 2003).
Paisley Livingston, "Cinematic Authorship," in Richard Allen and Murray Smith
(eds.), Film Theory and Philosophy (Oxford University Press, 1997).
Dominic Mclver Lopes, "The Aesthetics of Photographic Transparency," Mind
1 1 2 : 4 4 7 ( 2 0 0 3 ) , 433-^-8.
Joseph Margolis, "Film as Art," Millennium Film Journal 1 4 - 1 5 ( 1 9 8 4 - 5 ) , 8 9 - 1 0 4 .
"Le Significant Imaginaire, malgre lui," Persistence of Vision 5 (1987), 28—36.
"Mechanical Reproduction and Cinematic Humanism," Film and Philosophy
5 - 6 (2002), 1 1 4 - 3 0 .
Patrick Maynard, "Drawing and Shooting: Causality in Depiction," Journal of
Aesthetics and Art Criticism 44:2 ( 1 9 8 5 / 6 ) , 1 1 5 - 2 9 .
Aaron Meskin and Jonathan Cohen, "On the Epistemic Value of Photographs,"
Journal of Aesthetics and Art Criticism 62:2 (2004), 1 9 7 - 2 1 0 .
"Photographs as Evidence," in Scott Walden (ed.), Philosophy and
Photography (Blackwell Publishing, forthcoming).
Aaron Meskin and Jonathan Weinberg, "Emotion, Fiction, and Cognitive Archi-
tecture," British Journal of Aesthetics 43:1 (2003), 1 8 - 3 4 .
Paul Messaris, Visual Literacy: Image, Mind and Reality (Westview Press, 1994).
Richard Moran, "The Expression of Feeling in Imagination," The Philosophical Review
103:1 (1994), 7 5 - 1 0 6 .
Hugo Munsterberg, The Photoplay: A Psychological Study (Dover Publishers, 1916).
Alex Neill, "Fear, Fiction and Make-Believe," Journal of Aesthetics and Art Criticism
49:1 (1991), 4 7 - 5 6 .
"Empathy and (Fiction) Film," in David Bordwell and Noel Carroll (eds.),
PosuTheory: Reconstructing Film Studies (University of Wisconsin Press, 1996).
Victor Perkins, Film as Film: Understanding and Judging Movies (Penguin Books, 1972).
Carl Plantinga, Rhetoric and Representation in Nonfiction Film (Cambridge University
Press, 1997).
Carl Plantinga and Gregory Smith (eds.), Passionate Views (Johns Hopkins University
Press, 1999).
Trevor Ponech, What Is Non-Fiction Cinema? (Westview Press, 1999).
Jesse Prinz, Gut Reactions (Oxford University Press, 2004).
V. I. Pudovkin, Film Acting and Film Technique (Vision Press, 1958).
Colin Radford, "How Can We Be Moved by the Fate of Anna Karenina?,"
Proceedings of the Aristotelian Society, supplement 4 9 (1975), 67—81.
Jenefer Robinson, Deeper Than Reason (Oxford University Press, 2005).
William Rothman and Marion Keane, Reading CavelTs The World Viewed (Wayne
State University Press, 2000).
Bruce Russell, "The Philosophical Limits of Film," in Noel Carroll and Jinhee
Choi (eds.), Philosophy of Film and Motion Pictures (Blackwell PubHshing, 2006).
 SELECT BIBLIOGRAPHY 231

Edward Sankowski, "Uniqueness Arguments and Artist's Actions," Journal of

Aesthetics and Art Criticism 38:1 (1979), 61-74.
Andrew Sarris, "Notes on the Auteur Theory in 1962," in Gerald Mast and Marshall
Cohen (eds.), Film Theory and Film Criticism, 3rd edn. (Oxford University Press,
1985).
Roger Scruton, "Photography and Representation," in Noel Carroll and Jinhee
Choi (eds.), Philosophy of Film and Motion Pictures (Blackwell Publishing, 2006).
Alex Sesonske, "Cinema Space," in David Carr and Ed Casey (eds.), Explorations in
Phenomenology (Martinus Nijhoff, 1973).
"Aesthetics of Film," Journal ofAesthetics and Art Criticism 33:1 (1974), 51—7.
"Time and Tense in Cinema," Journal of Aesthetics and Art Criticism 38:4
(1980), 418-26.
Gregory Smith, Film Structure and the Emotion System (Cambridge University Press,
2003).
Murray Smith, Engaging Characters (Oxford University Press, 1995).
Susan Sontag, "Film and Theater," in her Styles of Radical Will (Farrar, Straus &
Giroux, 1966).
Francis Sparshott, "Vision and Dream in the Cinema," Philosophic Exchange 1
(1971), 111-22.
"Medium and Convention in Film and Video," Millennium Film Journal 14—15
(1984-5), 72-88.
Kevin Sweeney, "Constructivism in Cognitive Film Theory," Film and Philosophy 2
(1995), 33-44.
Malcolm Turvey, "Seeing Theory," in Richard Allen and Murray Smith (eds.), Film
Theory and Philosophy (Oxford University Press, 1997).
Kendall Walton, Categories of Art," The Philosophical Review 79 (1970), 334-67.
"Fearing Fictions," Journal ojPhilosophy 75:1 (1978), 5-27.
"Transparent Pictures: On the Nature of Photographic Realism," Critical
Inquiry 11:2 (1984), 246-77.
Mimesis as Make-Believe (Harvard University Press, 1990).
"On Pictures and Photographs: Objections Answered," in Richard Allen
and Murray Smith (eds.), Film Theory and Philosophy (Oxford University Press, 1997).
Nigel Warburton, "Seeing Through 'Seeing-Through' Photographs," Ratio (1988),
64-77.
Thomas Wartenberg, Unlikely Couples (Westview Press, 1999).
Thomas Wartenberg and Angela Curran (eds.), The Philosophy of Film: Introductory
Text and Readings (Blackwell Publishing, 2005).
Thomas Wartenberg and Murray Smith (eds.), Thinking Through Cinema: Film as
Philosophy (Blackwell Publishing, 2006).
Trevor Whittock, Metaphor and Film (Cambridge University Press, 1990).
George Wilson, Narration in Light (Johns Hopkins University Press, 1986).
"On Film Narrative and Narrative Meaning," in Richard Allen and Murray
Smith (eds.), Film Theory and Philosophy (Oxford University Press, 1997).
232 SELECT BIBLIOGRAPHY

"le Grand Imagier Steps Out: The Primitive Basis of Film Narration," in
Noel Carroll and Jinhee Choi (eds.), Philosophy of Film and Motion Pictures
(Blackwell Publishing, 2006).
Peter Wollen, Signs and Meaning in the Cinema (Indiana University Press, 1972).
Richard Wollheim, Art and its Objects (Cambridge University Press, 1980).
Sol Worth, "Pictures Can't Say Ain't," Vegas 12 (1975), 85-108.
 Index

actor—character relation, 69—71 Arnheim, R., 11, 199, 201, 202-3, 204
actualizes, 9 art, film as, 2—3, 7—33
aesthetic interest, 12, 18-20, 21, medium specificity, 40—6
26-30 "photography is not art" arguments,
affect, 4 - 5 , 147-90 8—12; aesthetic interest, 12,
audience—character relations, 18-20, 21, 26-30; control
161—90; asymmetric emotions, argument, 12, 16-18, 21, 23-6;
164—6, 172; coincident sheer physical causation, 12—15,
emotions, 166—9; connected 21-3
emotions, 166—9; identification, or as a recording of an artwork,
161—70, 185; mirror reflexes, 9 - 1 1 , 30-3
164, 185—90; simulation theory, Asphalt, 42
170-7, 185; solidarity, 182-4, asymmetric emotions, 164—6, 172
185; sympathy, 177-82, 183-4, Atget, E., 22
185; vectorially converging attention management, 122-4, 132
emotions, 169-70, 178, 185 audience, affect and, 149—60
compass of, 149—51 emotions, 151-60, 161-84
emotions, 151-60, 161-84 gamut of affect, 149—51
Age d'or, 11, 133 see also audience—character affective
alethic modal logic, 22—3 relations
Alien Resurrection, 187 audience—character affective relations,
Alternating Pink and Gold (sculptural 161-90
installation), 74 asymmetric emotions, 164—6, 172
Altaian, R., 142 coincident emotions, 166—9
Amarcord, 143 connected emotions, 166—9
Amis, M., 135 identification, 161—70, 185
anachronisms, 16—17 mirror reflexes, 164, 185—90
analyses, real definitions, 54—5 simulation theory, 170—7, 185
see also cinema, the essence of solidarity, 182-4, 185
anti-heroes, 181—2 sympathy, 177-^-82, 183—4, 185
antipathy, 179-80, 183--4, 185 vectorially converging emotions,
Antonioni, M., 142 169-70, 178, 185
234 INDEX

audio design Cabinet of Dr. Caligari, The, 10

film as art, 31 camera movement, 116—17, 123,
reflex responses to, 149 125-8
auteur theory, 226 Captain Horatio Hornblower, 154—5
authorial intention, 211, 212, 216 Casablanca, 57, 126
avant-garde films categorization, for evaluation, see
evaluation, 215—16 cinematic category approach
narrative, 143 cats, 88-9
variable framing, 129 causation argument, 12—15, 21—3
Avventura, 11, 142 Cavell, S., 224
CGI, 31-2, 97
Back Street, 209 Chaplin, C , 11, 44-6, 48, 50, 169
BandofNinjas, 59,60 character—actor relation, 69—71
Barney, M., 216 character-audience relations, 161—90
Batman films, 215 Chien Andalou, Le, 124
Battleship Potemkin, 31, 167 Chinese Connection, The, 61—2
Bazin, A., 3, 199, 202-3, 204, cinema, the essence of, 3-4, 53—78,
223-4 199-200
Beetlejuice, 220, 221, 222 detached displays, 56—8, 64, 72
beliefs moving images, 58—63
affect and, 153-4, 155 screen projection, 76—8
cinematic sequencing and, 131—2 seeing vs. seeing a motion picture,
cognitive illusions, 81—5 55-6
Benjamin, W , 22, 71 simplified theory, 75—8
Bergman, L, 43 templates, 65, 66-7, 70-2
Bicycle Thief The, 220-1, 222 tokens, 65-7, 69-72, 73
Big Broadcast of 1938, The, 219 two-dimensionality, 72—5
biomechanical reflexes, see bodily types, 64-72
reflexes see also cinematic category approach
Black Scorpion, The, 150 cinema, value of, 223—5
Blood of a Poet, 133 cinema studies, legacy, 2
bodily reflexes, 149-50, 153, see also film theory
185-90 cinematic category approach, movie
"Boom-Boom" theory, 150 evaluation, 5, 198-9
Bordwell, D., 120 classical film theory, 199-208, 217
bracketing, 125, 126-9, 142 pluralistic categories, 207—18, 219
Bradbury, R., 65 value of the movie's purpose,
Brakhage, S., 216 219-23
Breathless, 120 cinematic narration, 4, 128—32,
Bride of Frankenstein, 170, 215 133-44
Brooks, V., 107-8, 109 cinematic sequencing, 4, 116—33
Bullitt, 149 attention management, 122-4,
Buiiuel, L., 133 132
 INDEX

film language hypothesis, 117, Dali, S., 133

118-21, 140 definitions, real, 54—5
variable framing, 124-33, 142 see also cinema, the essence of
cinematic, the, see cinema, the demographics, 1—2
essence of; cinematic category Descent, The, 31
approach detached displays, 56—8, 64, 72
Citizen Kane, 141—2 transparency thesis of shots,
classical film theory, 199-208, 217 99-102
close-ups, 123—4 dialogue
closure, 134-44, 203 classical film theory, 202
Cocteau, J., 133 medium specificity, 37—8, 39, 4 1 ,
code thesis, 81, 102-9, 112, 113, 43,49
114, 115 disgust response, 160
cognitive considerations distortion
affect, 153-4, 155, 156 illusion thesis, 86
cinematic sequencing, 131—2 medium specificity, 41
movie evaluation, 214 Dog Star Man, 216
cognitive illusions, 81—5 Dogville, 49-50n
coincident emotions, 166—9 Don't Look Now, 139
color, 89
communication, shots for, 112, Easy Rider, 139
113-14, 118-33 Eco, U , 106
computer-generated imagery Edison films, 76
film as art, 31—2 editing
transparency thesis of shots, 97 cinematic sequencing, 116, 117,
connected emotions, 166—9 123, 125, 126
consumer reporters, 192—3 classical film theory, 201—2, 203,
contexts, movie categorization, 204-6
211-12, 216 code thesis of shots, 103
control argument, 12, 16—18, 21, film as art and, 11—12, 30—1
23-6 medium specificity, 41—2
Cotten, J., 69 point-of-view shots, 168—9,
Cremaster series, 216 188-9
criterial prefocusing, 159—60, 161, see also montagists
167-9, 184, 185 egocentric information, 99^102, 113
critics, movie evaluation, 192—3, 8 Mile, 120
194, 217 Eisenstein, S.
Croce, A., 75 classical film theory, 206
Croce, B., 8 criterial prefocusing, 167
Cromwell, J., 209 film as art, 11,31
cultural context, movie categorization, medium specificity, 42
211-12, 216 mirror reflexes, 187
cultural criticism, 222—3 Elmer Gantry, 205—6
236 INDEX ■—-

emotions, 151—60 value of, 223—5

audience—character relations, see also cmema, the essence of
1 6 1 - 8 4 ; asymmetry, 1 6 4 - 6 , 172; film editing, see editing
coincident, 166—9; connected, film language hypothesis, 117, 118—21,
1 6 6 - 9 ; identification, 1 6 1 - 7 0 , 140
185; pomt-of- n e w shots, 168—9, film projection, see projection
189; simulation theory, 170—7, film stylization, 2 0 1 - 2 , 2 0 3 - 7
185; solidarity, 1 8 2 - 4 , 185; film theory
sympathy, 1 7 7 - 8 2 , 1 8 3 - 4 , 185; classical, 1 9 9 - 2 0 7 , 217
vectorially converging, 169—70, legacy of, 2—3
178, 185 flashbacks, 139, 142
epistemic (cognitive) illusions, 81—5 flashforwards, 139
erotetic narration, 4 , 134—44 Flavin, D , 7 4
essential photographs, 14—15, 2 1 , flip books, 75—6
22, 23 F o r d , ] . , 120
evaluating movies, 5, 19 2—2 2 3 Frampton, H . , 5 9 , 60
as a philosophical problem, 194—9; Fred Astaire and Ginger Rogers Book,
categorization, 198—9; classical The, 15
fflm theory, 1 9 9 - 2 0 8 , 217; freeze frames, 62
pluralistic-category approach,
207—18, 219; value of cinema, gaze following, 189
223—5; value of the movie's General, The, 125
purpose, 219—23 General Line, The, 11
expression of thought, 13—14, 15, Glen or Glenda, 219
16, 1 7 - 1 8 , 2 1 - 2 , 2 3 , 2 5 - 6 , Godard, J.-L., 49n, 59, 120
28-9 Gold Rush, The, 1 1 , 4 4 - 6 , 4 8 , 50,
expressionists, 41—2 169-70
Goldman, A., 76
Fahrenheit 451 (novel), 65 Gorm, J.-R, 59
Fast, Cheap S^Out of Control, 198, 202, grammar
203, 209, 212 cinematic sequencing, 119—21,
fear, 150, 151, 1 5 2 - 3 , 154, 1 5 5 - 6 , 140
157, 160 code theory of shots, 104—6
feeling, see affect Grande Illusion, La, 220—1
Fellini, R, 143 Greenaway, R, 49n
fiction, paradox of, 153
fictional narrators, 145—6 Half Nelson, 120
film Iiannibal, 184
as art, 2—3, 7—33; m e d i u m Haskin, B., 150
specificity, 40—6 Heaven and Earth Magic, 216
as film, 2 0 0 - 7 Histoire du cinema, 11, 49n
moving image in contrast t o , 3—4, historical context, movie
55-6 categorization, 211—12, 216
 INDEX

History Boys, The, 50n Lady Vanishes, The, 221

Hitchcock, A., 38, 4 3 , 69, 127, Lang, R, 106, 127, 130
150, 187 language, shots as equivalents of,
Hochberg, J., 1 0 7 - 8 , 109 1 0 2 - 7 , 115
holography, 7 3 - 4 cinematic sequencing, 117, 118—21,
Homicide: Life on the Street, 120 140
House of Meetings (novel), 135 Laocoon (treatise), 3 9 - 4 0
Hume, D., 135 Last Command, 42
Last Year at Marienbad, 143, 144
identification, 1 6 1 - 4 Lee, B., 6 1 - 2
infectious, 1 6 4 - 7 0 , 185 Lee, S., 195
Iimura, T., 59 see also Inside Man
illusion thesis, 10, 80, 8 1 - 9 3 , lens movements, 117, 123, 125—6
113, 114 Lessing, G. E., 39-^-1
Illusionist, The, 139, 142 Letter to Jane, 59
imagination, emotions and, 153—5 Lipps, X , 187
Inauguration of the Pleasure Dome, 216 Lizzies of the Field, 44—6
indexing, 1 2 5 - 6 , 1 2 7 - 9 , 142 Lo Wei, 6 1 - 2
India Song, 143 Lumiere Brothers, 9, 81
infectious identification, 164—70, 185
insect phobias, 150 M, 106, 127
Inside Man, 1 9 5 - 9 , 202, 203, 2 0 9 - 1 1 , macro-questions, cinematic narrative,
212 137, 138, 141
intention, authorial, 2 1 1 , 212, 216 Made for Each Other, 209
intentionality, 13—15, 21—3 Malle, L., 43
interpretation, 66—7, 7 0 , 7 1 , 72 Man With a Movie Camera, 133
Ivanhoe (novel), 65 March of the Penguins, 141
Marker, C , 59, 61
Jetee, La, 59, 61 Matchstick Man, 120
Journeys to Berlin, 49n May, J., 42
Ju-on, 142 mechanical reproduction, film as, 8—9,
jump cuts, 119, 120 10-11,40-5
Just My Luck, 1 8 2 - 3 medium /media
definition, 35—6
Keaton, B., 125 film as art, 1 6 - 1 8 , 2 3 - 6
Keystone Kops, 44—6 purity of, see medium/media:
Kiarostami, A., 32n, 51 specificity
kinetoscopes, 76 specificity, 3, 35, 36—51
King Kong, 1 3 6 - 7 , 150 meta-cinema, 143
Kitty Safari, 8 8 - 9 micro-questions, cinematic narration,
Kracauer, S., 199, 2 0 2 - 3 , 204, 224 137, 138, 141
Kuleshov, L., 2 0 4 - 5 Mifune, T., 71
Kurosawa, A., 7 1 , 142 Mildred Peirce, 142
238 INDEX

Miller-Lyre Illusion, 82 shots, 80-114

mind, philosophy of, 171 use of term, 63
see also simulation theory; multiple-instance artforms, 64—8, 72
theory-theory Munch, E., 9
minimal (essential) photographs, Murnau, F. W , 42, 66
1 4 - 1 5 , 21, 22, 23 My Dinner with Andre, 43
mirror reflexes, 164, 185—90 mythopoetic cinema, 216—17
mirrors
film as art and, 18-19, 26, 32 Nakata, H., 142
transparency thesis of shots, 100—1 Naked Jungle, The, 150
Mizoguchi, K., 40 narration, 4, 133—44
Moby-Dick, 137 cinematic sequencing and, 128—32,
modal fallacy, 22—3 142
montagists narrators, fictional, 145—6
cinematic sequencing, 117 Nashville, 142
classical film theory, 199, 201, natural recognitional prompt thesis,
202-3, 204 108-14, 115
code thesis of shots, 102—3 North by Northwest, 150
medium specificity, 42 Nosferatu, 66
morality Notorious, 127
audience emotions, 180—2, 183—4 novels, as multiple-instance artform, 65
movie evaluation, 214
Morris, E., 198 October, 42, 206
Morris, R., 74 one hundred and eighty degree rule,
movement 119, 120
bodily responses to, 150 One Second in Montreal, 59
of cameras, see camera movement 1 in 10, 59
defining cinema, 58—64, 72—3, ontology of cinema, 53
74-5 see also cinema, the essence of
illusion thesis of shots, 87—93, 113 Oshima, N., 59
movie critics, 192-3, 194, 217 see also Band of Ninjas
movie evaluation, 5, 192—223 Our Hitler, 49n
as a philosophical problem, 194—9;
categorization, 198—9; classical painting
film theory, 199-208, 217; defining cinema and, 58—9, 61, 62
pluralistic-category approach, film as art and, 14-15, 16-19
207—18, 219; value of cinema, medium specificity, 35, 37, 39—40
223—5; value of the movie *s Paradine Case, The, 38
purpose, 219—23 paradox of fiction, 153
moving image, philosophy of, see Pather Panchali, 219
philosophy of the moving image perceptual illusions, 81, 82, 85—93
moving pictures performance, defining cinema, 64—72,
philosophy in, 6 73
 INDEX

Perkins, V, 205-7 Poetic Justice, 59, 61

philosophy of mind, 171 poetry, medium specificity, 39-4-0
see also simulation theory; point-of-view shots, 168-9, 188-9
theory-theory point/glance shots, 188, 189
philosophy in motion pictures, 6 point/object shots, 188, 189
philosophy of the moving image problem/solutions model, cinematic
affect, 4 - 5 , 147-90 narration, 138—9
cinematic sequencing, 4, 116—33 projection
defining cinema, 3-4, 53—78, defining cinema, 76—8
199^200 illusion thesis of shots, 91—3
demographic factors in emergence Prosperous Books, 49n
of, 1-2 Psycho, 38, 137
film as art, 2 - 3 , 7-33, 40-6 Pudovkin, V. I., 103, 117, 145, 204
future of, 225
intellectual factors in emergence question/answer model, cinematic
of, 2 narration, 135—44
medium specificity, 3, 35, 36—51
movie evaluations, 5, 192—223 Rainer, Y., 49n
narration, 4, 128-32, 133-44 Rashomon, 142, 144
" the single shot, 4, 80-114 real definitions, 54—5
value of cinema, 223—5 see also cinema, the essence of
phobias, 150, 153 realism, 202—7
photographic transparency, 19—20, reality
26-9, 32, 34 film as reproduction of, 8—9, 10—11
photography illusion thesis of shots, 81—93
classical film theory, 202, 204 Rear Window, 188
film as art and, 7—33; aesthetic recognition thesis, 81, 108—14, 115
interest argument, 12, 18—20, reflexes, 149-50, 153
21, 26—30; control argument, mirror, 164, 185-90
12, 16-18, 21, 23-6; or merely a reproduction, film as, 8—9, 10—11,
recording of, 9—11, 30—3; sheer 40-5
physical causation argument, Republic, The (book), 114
12-15, 21-3 response dependence, 88—93
flip books, 75 Ringu, 142
pictorial comprehension Riverdance, 187
code thesis of shots, 107—9
recognition thesis of shots, 109—13, Sansho the Bailiff, 4 0
115 Satjricon, 143
Plato, 114, 161 scaling, 122, 123-4, 125, 127-9, 142
plays, see theater Schmitt, C , 183^1-
plot construction, 136 Scott, R., 120, 184
pluralistic categorization, movie Scott, Sir W., 65
evaluation, 5, 207-18, 219 Scream, The, 9
240 INDEX

screen projection, 76—8 Sitney, P. A., 216

scripts, 68—9 slides, 59, 60, 61, 62
sculpture Smith, H., 216
defining cinema, 72—3, 74 snake phobia, 150
medium specificity, 37, 40 Snakes on a Plane, 150
seeing Snow, M., 59
defining cinema, 55—6 So Is This, 59
film as art, 19, 28 soap operas, 134, 142
transparency thesis of shots, solidarity, 182-4, 185
93-102, 113 Sopranos, The (TV series), 182
Sennett, M., 46 sound
see also Keystone Kops film as art, 31
sentences, shots as equivalents of, reflex responses to, 149
104-6 Soviet montagists, see montagists
cinematic sequencing, 118—19 space, detached displays, 57—8, 64, 72
sequence shots, 116—17 transparency thesis of shots,
sequencing, cinematic, 4, 116—33 99-102
attention management, 122-4, 132 speech
film language hypothesis, 117, classical film theory, 202
118-21, 140 medium specificity, 37—8, 39, 4 1 ,
variable framing, 124-33, 142 43,49
Seven Samurai, 71 spider phobia, 150
Sex and the City (TV series), 175—6 Spielberg, S., 152
Shadow of a Doubt, 69 Stagecoach, 120
shadow movement, 93 startle response, 149
sheer physical causation argument, stasis, defining cinema, 59—62
12-15, 21-3 Steam (sculpture), 74
Shimizu, T., 142 Steichen, E., 21
shots, 4, 80-114 still pictures, 58—62
cinematic sequencing, 4, 116—33 see also painting; photography
code thesis, 81, 102-9, 112, 113, storytelling, see narration
114, 115 Strangers on a Train, 187
film as art and, 29 structural features, movie
illusion thesis, 80, 81-93, 110, categorization, 210—11,
113, 114 212, 216
medium specificity, 42 stylization, classical film theory, 201—2,
recognition thesis, 81, 108—14, 115 203-7
transparency thesis, 80, 93—102, Sunrise, 42
110, 113, 114—15 Sunset Blvd., 139
silent films Syberberg, H.-J., 49n
classical film theory, 201, 204 sympathy, 177-82, 183-4, 185
medium specificity, 41 syntax, cinematic sequencing, 119—21,
simulation theory, 170—7, 185 140
 INDEX

Tarzan films, 43 two-dimensionality, 72—5

Taste of Cherry, A, 32n, 51 two-tiered artforms, 67—8
taste-makers, 192, 193 types, defining cinema, 64—72
television, 77—8
templates, 65, 66—7, 70—2 Vfor Vendetta, 159
temporal order, narration, 139, 142 variable framing, 124—33, 142
theater vectorially converging emotions,
cinematic sequencing, 125, 169-70, 178, 185
129-30 Vertigo, 150, 221
classical film theory, 201, 202 Vertov, D., 20, 133
defining cinema, 64-72, 73 Volver, 138n
film as art, 10, 11 von Sternberg, J., 42
medium specificity, 37—9, 49—50 von Trier, L., 49-50n
theories, real definitions, 54-5
see also cinema, the essence of Walsh, R., 155
theory-theory, 171, 173, 174-5 War of the Worlds, 152
This Is Cinerama, 149 waste avoidance, 50
To Catch a Thief, 43 Wei, Lo, 61-2
tokens, 65-7, 69-72, 73 Welles, O., 143
Tokyo Story, 219 Wiene, R., 10
Totosi, 136 Winter Light, 43
transparency thesis Wittgenstein, L., 1
film as art, 19-20, 26-9, 32, 34 words, shots as equivalents of, 103—8
moving picture shots, 80, 93—102, cinematic sequencing, 118—19
110, 113, 114—15
traveling (sequence) shots, 116—17 You Only Live Once, 130
Trial The, 143
Truffaut, E, 150 zoetropes, 75—6