POETICS

of CINEMA

DAVID BORDWELL

Routledge
Taylor & Francis Group
N e w York London
 For my colleagues at the Department of Communication Arts,
University of Wisconsin-Madison,
who make 821 University Avenue, despite leaky ceilings,
a place that fosters ideas and mirth

Routledge Routledge
Taylor & Francis Group Taylor & Francis Group
270 Madison Avenue 2 Park Square
New York, NY 10016 Milton Park, Abingdon
Oxon OX 14 4RN

© 2008 by Taylor & Francis Group, LLC

Routledge is an imprint of Taylor & Francis Group, an Informa business

Printed in the United States of America on acid-free paper

10 9 8 7 6 5 4 3 2 1

International Standard Book Number-13: 978-0-415-97779-1 (Softcover) 978-0-415-97778-4 (Hardcover)

No part of this book may be reprinted, reproduced, transmitted, or utilized in any form by any electronic, mechanical, or
other means, now known or hereafter invented, including photocopying, microfilming, and recording, or in any i n f o r m a -
tion storage or retrieval system, without written permission f r o m the publishers.

Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for
identification and explanation without intent to infringe.

Library of Congress Cataloging-in-Publication Data

Bordwell, David.
Poetics of cinema / David Bordwell.
p. cm.
ISBN 0-415-97778-9 (hardback : alk. paper) — ISBN 0-415-97779-7 (pbk.: alk. paper)
1. Motion pictures. I. Title.

PN1994.B5735 2007
791.4301—dc22 2007004036

Visit the Taylor & Francis Web site at

iittp://www.taylorandfrancis.com

ind the Routledge Web site at

hittp://www.routledge.com UNIV.
LEIDEN
Contents

Acknowledgments X I

Introduction 1

I Q 1 J L e s t ions of T h e o r y

1. Poetics of Cinema 11
The Tradition 12
Domains and Tendencies 17
One Poetics of Film 19
Poetics: A Program 23
An Excursion on Reflections and Zeitgeists 30
From Shriek to Shot 32
What Snakes, Eagles, and Rhesus Macaques Can Teach Us 43

2. Convention, Construction, and

Cinematic Vision 57
Shot/Reverse Shot: A Convention? 57
Primary Theory and a Continuum of Conventions 61
A Package Deal 66
Contingent Universals and Us 73
Afterword 75

vii
viii Contents Contents ix

II Studies in Narrative Sustaining the Network 242

Network Narrative: A Working Filmography 245
3. Three Dimensions of Film Narrative 85
Some First Moves 88 III Studies in Style
Protagonists and Their Problems 90
Narration 93 8. Cinecerity 253
Plot Structure 102
Style and Film History 255
The Narrative World 110
Style and the Critic 257
Afterword: Narrators, Implied Authors, and Other Superfluities 121
Envoi 261

4-. Cognition and Comprehension

9. Taking Things to Extremes
Viewing and, Forgetting in
Hallucinations Courtesy of
Mildred Pierce 135
Robert Reinert 263
Narrative Norms 137
Revelations at the Cinema Verdi 266
Two Methods of Murder 138
Close-Up Depth 269
The Partial Replay 143
Problems and Solutions 276
Secrets and Lies, and Narration 149
lO. CinemaScope
5. The Art Cinema as a Nlode of The Modern JS/liracle You See
Film Practice 151 Without (jlasses 281
Realism, Authorship, Ambiguity 152 The Big Picture 284
The Art Cinema in History 157 Hollywood Cadillac 287
Afterword 158 A Lack of Scope 290
Taming a New Technology 301
6. Film Futures 171
Some Virtues of Clotheslines 307
Rules of the Game 174 Reinventing the Tableau 312
Some Sources 185 The End of Screen Ratios? 320

7. Nlutual Friends and 11. Who Blinked First? 327

Ciironologies of Chance 189
The Tightrope 327
Protagonists and Projects 192 The Strength of the Stare 331
This Particular Web 194 Streamlined Behavior 334
Major Players 198
Familiars, Strangers, and Random Walks 200 12- Visual Style in Japanese Cinema,
Only Connect, or at Least Collide 204 1925-1945 337
Narration Makes Networks 207 A Classical Cinema? 338
Compare and Contrast 211 Voices in the Dark 345
Criss-Crossers Cross Over 214 Exercising the Eye 347
Four Small Worlds 218
Constructing and Reconstructing Japaneseness 352
Nashville (1975) 221 Stylistic Trends 356
Magnolia (1999) 227 Blending and Refining 363
Favoris de la lune (1984) 233
The Pacific War: Toning Down Technique 369
Les Passagers (1999) 237 Afterword 372
X Contents

13. A Cinema of Flourishes

Decorative Style in 1920s and
1930s Japanese Film 375
Some Functions of Style 377
Games With Vision 380
A Tradition of Ornamentation 388

14-. Aesthetics in Action

Kung-Fu, Cjtinplay, and
Cinematic Expression 395
Acknowledgments
Hollywood Action: The 1980s and After 396
The Mechanics of Movement 399
Toward an Ecstatic Cinema 406

15. Richness Through Imperfection

King Hit and the Cj Limps e 413
Inherited Norms 414
A Problem and Some Solutions 416
Cuts and Slashes 423
Glimpses of Marvels 429
Some of the essays in this book originally appeared in the journals and anthologies
Notes 431 listed below. The author is grateful for permission to reprint them in revised form.

Index 481 "Historical Poetics of Cinema." In Barton Palmer, ed. The Cinematic Text: Methods and
Approaches. New York: AMS Press, 1989, pp. 369-398.
"Convention, Construction, and Cinematic Vision." In David Bordwell and Noel Carroll,
eds. Post-Theory: Reconstructing Film Studies. Madison: University of Wisconsin
Press, 1996, pp. 87-107.
"Cognition and Comprehension: Viewing and Forgetting in Mildred Pierce." Journal of
Dramatic Theory and Criticism VI, 2 (Spring 1992): 183-198.
"The Art Cinema as a Mode of Film Practice." Film Criticism 4,1 (Fall 1979): 56-64.
"Film Futures." Substance no. 97 (2002): 88-104.
"Sarris and the Search for Style." In Emmanuel Levy, ed. Citizen Sarris, American Film
Critic: Essays in Honor ofAndrew Sarris. Lanham, Maryland: Scarecrow Press, 2001,
pp. 165-173.
"Taking Things to Extremes: Hallucinations Courtesy of Robert Reinert." Aura (Stockholm)
VI, 2 (2000): 4-19.
"Schema and Revison: Staging and Composition in Early CinemaScope." In Jean-Jacques
Meusy, ed. Le CinemaScope Entre art et industrie. Paris: AFRHC, 2004, pp. 217-232.
"Who Blinked First? How Film Style Streamlines Nonverbal Interaction." In Lennard
Hojbjerg and Peter Schepelern, eds. Style and Story: Essays in Honor of Torben
Grodal. Copenhagen: Museum Tusulanum Press, 2003, pp. 45-57.
"Visual Style in Japanese Cinema, 1925-1945." Film History 7,1 (Spring 1995): 5-31.
"A Cinema of Flourishes: Japanese Decorative Classicism of the Prewar Era." In David
Desser and Arthur Noletti, eds. Reframing Japanese Cinema. Bloomington: Indiana
University Press, 1992, pp. 327-345.

xi
xii Acknowledgments

"Aesthetics in Action: Kung Fu, Gunplay, a n d Cinematic Expressivity." In Law Kar,

ed. Fifty Years of Electric Shadows. H o n g Kong: U r b a n Council a n d H o n g Kong
International Film Festival, 1997, pp. 81-89.
"Richness t h r o u g h Imperfection: King H u a n d the Glimpse. In Law Kar, ed. Transcending Introduction
the Times: King Hu and Eileen Chan. H o n g Kong: U r b a n Council a n d H o n g Kong
International Film Festival, 1998, pp. 19-24.

The essays collected here, spanning 30 years, represent some of my efforts to answer
questions about cinema from the standpoint of what I call film poetics. These efforts
might be characterized as pushing a doctrine—most will call it formalist—but I think
I'm doing something else. Granted, these essays put the film as an artwork at the
center of study; they analyze form and style. But they also try to mount explanations
of how films work, and why under certain circumstances they came to look the way
they do. Those explanations invoke a wide range of factors: artistic intentions, craft
guidelines, institutional constraints, peer norms, social influences, and cross-cultural
regularities and disparities of human conduct.
Taken together, the essays are at once critical, in the sense of looking closely at
movies, and historical, in the sense of trying to explain how they got the way they
are. Are these essays also theoretical? They are, but not as theory in the academic
humanities is currently understood. Here's what theory looked like as of December
29, 2005, when I received the following e-mail from the Visual Culture program of
my university. The message announces an upcoming conference called TRANS.

This is a practical call to participate in an important transitional moment. After

all the appeals to think beyond the "post" and the "inter," after all the gestures
asking us to move beyond the divisions of history, theory, and art making,
imagined and lived communities, scholarship and activism, area studies and
the disciplines, the sciences, social sciences and humanities, deconstructive and
reparative work, knowledge-production and critique, what kinds of knowledges
are we producing, how, and to what ends? What methodologies, pedagogical
techniques, curricular structures and programming agendas do we actually put
into practice and toward what goals?

The conference takes the transsubstantiating [sic] challenge of the "trans" in

Transdisciplinarity, Transgender, Transethnic, Transart, and Transracial not
just as its theme but also as its point of departure. How might the cultural and

1
2 Poetics of Cinema
political processes of the "trans" in transplanting, transmitting, transculturat-
ing, and transferring mark not only hybridizing crossings but also the forging
of structural transformations?
The event will be held in Transylvania.
Okay, I made up the last sentence, but the rest is exactly as I received it.1 If this is
theory, the essays that follow aren't.
Most humanists' conception of theory—or as we should call it, Theory, aka Grand
Theory—is at once too broad and too narrow. It's too broad because it presumes that
all human activity can be subsumed within some master conceptual scheme (even
though some postmodernists advance the conceptual scheme that all conceptual
schemes are fatally flawed). The current conception of Theory is too narrow because it
presumes a limited conception of how one does intellectual work. The rise of Theory
crushed theories and discouraged theorizing. Grand Theory created bad habits of
mind. It encouraged argument from authority, ricochet associations, vague claims,
dismissal of empirical evidence, and the belief that preening self-presentation was a
mode of argument. Above all, it ratified what I call doctrine-driven thinking as the
principal mode of humanistic inquiry. Proponents of Theory routinely play up the
differences among theoretical positions, but they ignore what unites them—the idea
that any program propelled by doctrines can be applied, via imaginative extrapola-
tion, to one phenomenon or another. The cluster of doctrines isn't questioned skepti-
cally; the effort goes into diligent application.
Or at least some of the effort. A lot, perhaps the bulk of it, goes into rhetoric
of a peculiar kind. My TRANS instance, deadly serious and yet playful in a self-
congratulating way, illustrates what Frederick Crews has called the "ponderous
coyness" of this tradition. But there's an element of sheer obfuscation too.
Is the insistence that nature is artifactual not more evidence of the extremity of
the violation of a nature outside and other to the arrogant ravages of our techno-
philic civilization, which, after all, we were taught began with the heliotropisms
of enlightenment projects to dominate nature with blinding light focused by
optical technology? Haven't eco-feminists and other multicultural and inter-
cultural radicals begun to convince us that nature is precisely not to be seen in the
guise of the Eurocentric productionism and anthropocentrism that have threat-
ened to reproduce, literally, all the world in the deadly image of the Same?2
Catching us up in a jungle of catchphrases and vague and unsupported claims (you
have to admire the bravado of "precisely"), this passage may make us forget that it's
a pair of rhetorical questions, to which one can always answer, "No." It would take
pages to untangle this rodomontade. We have lived with this writing for 30 years.
Its limping cadences, convulsive syntax, and strategic confusions have dulled our
senses. Very likely, no one in the history of English ever published prose as incompre-
hensible as that signed by Theorists.
The masses, Nietzsche once remarked, consider something deep as long as they
cannot see to the bottom. Not just the masses, but also the Modern Language
Introduction 3
Association. Combine Hegelian ambitions for a world system of thought, patched
together from a passel of incompatible doctrines, with prose that wants to strut and
be evasive at the same time, and you have a trend that dodges the task to which we
thought academics had pledged their professional lives: producing knowledge that
is reliable and approximately true. There's a difference between getting a buzz and
getting things more or less right.
It needn't be so. The best means to produce reliable knowledge, it seems clear, is the
tradition of rational and empirical inquiry. By rational inquiry, I mean probing con-
cepts for their adequacy as descriptions and as explanations of problems. Problems are
stated as questions to be answered; the more concrete, the better. Empirical inquiry—
not "empiricism," as humanists have to be told over and over—involves checking our
ideas against evidence that exists independent of our beliefs and wishes—not evidence
delivered in pristine innocence, without conceptual commitments on the part of
the seeker, and not facts that "speak for themselves." What is evidence? It's what is
corrigible in the light of further information. And to those who believe that facts are
inevitably relative to your standpoint, I'd reply that both concepts and evidence can
cut across different research frameworks. Suppose we ask how to explain the accel-
erated cutting rates of films between 1908 and 1920. Some researchers will suggest
looking to craft norms; others will point to wider cultural factors, like modernity; and
some will suggest combining these or other causal inputs. But all researchers share
to a high degree the concept of what a shot is, what a cut is, and what would count as
a fair measure of accelerated editing pace. Film studies, like most of what is pursued
in the humanities, is an empirical discipline. It isn't ontology, mathematics, or pure
logic. A beautiful theory can be wounded by a counterexample.
So this collection isn't just critical and historical. It has one foot in film theory, but
it doesn't conceive theory as an all-purpose explanation, a Weltanschauung ready for
exploitation. The essays center on middle-level questions. How do particular film-
making traditions create normalized options for visual style, and how have creative
filmmakers worked with these? What staging strategies do we typically find in
CinemaScope films? What are the conventions of certain storytelling formats, like
forking-path plots and network narratives, and how do they engage us? What regu-
larities of film technique can we find in classic Japanese cinema or more recent Hong
Kong filmmaking? Such questions urge us not only to forge concepts (that is, mount
theories) but also to look closely (analyze films) and to study the contingencies of
time and place (investigate history). Out of midlevel inquiries can ripple bigger issues,
such as the degree to which popular culture can be artistically innovative, or the way
in which our minds engage with narrative. Noticing minor things, like actors' eye
movements, can lead us to broader conceptions of how films affect us. At the risk of
looking fussy, I try to study manageable problems, but I also try to tease out some
larger implications.
Some will say I'm actually aiming at "science." I'd say, rather, that I'm trying to
join the tradition of rational and empirical inquiry, a broader tradition than what we
usually consider to be science. This tradition includes historical research and a mix of
inductive and deductive reasoning that tries tofitthe answer to the question. My aim
4 Poetics of Cinema
is to produce reliable knowledge, both factual and conceptual, about film as an art
form, in the hope that this knowledge will deepen people's understanding of cinema.
Rational-empirical research programs have been undertaken by many other
film scholars, perhaps more by historians than by critics and theorists, but I try to
answer questions from a distinctive angle. That angle I call the poetics of cinema, and
I explain what I mean in the first essay. Poetics seems to me to provide a good tool for
probing some intriguing midrange problems about film as an art. It won't secure us
against error, but it does help make our mistakes corrigible. Big Theory comes and
goes, but imaginative inquiry of any sort, poetics-based or not, that is grounded in
argument and evidence remains our best route to understanding cinema, its makers
and viewers, and its place in our lives.
Still, science does sometimes raise its head in the pages that follow. I occasionally
invoke social-scientific studies and even evolutionary accounts as components of causal
explanations, and I fear that these efforts will be greeted with the usual resistance
from humanistic circles. There are good grounds to consider this resistance flat-out
dogmatism. One of my female graduate students became interested in evolutionary
psychology, spurred by distinguished feminists who argue that it should be part of
any comprehensive investigation of sexism. 3 The student found that her teachers in
women's studies courses resolutely refused to let her write papers on the subject. If you
don't like a one-off anecdote, consider this. Although university presses like MIT,
Harvard, California, Chicago, and the like routinely publish books on evolutionary
theory in the biological and social sciences, as of this writing none has produced a
book on evolutionary theory of art and literature. That task has been left to less salient
houses. 4 The editors of a recent collection of essays, The Literary Animal: Evolution
and the Nature of Narrative, found securing a publisher unusually difficult.
Time after time, the science editor of a given press would express great interest,
only to encounter the resistance of the literary studies e d i t o r . . . . We therefore
want to express our gratitude to Northwestern University Press for their
courage—no other word will do—in publishing our volume. 5
All the evidence indicates that poststructuralist humanists, who purportedly
revel in a Bakhtinian play of discourses, have tenaciously resisted giving the floor to
discussions of art in cognitive or evolutionary terms. When the Modern Language
Association (MLA) was launching a study group in evolutionary psychology of
literature, a well-known scholar in the area told me that it could have been started only
by a graduate student. The MLA wants to encourage its junior members, but a senior
scholar would have seemed to be leading a cabal. On a much smaller scale, 10 years
ago a group of media scholars formed the Center for Cognitive Studies of the Moving
Image, but as of this writing, none of the U.S. members teaches in a graduate program
devoted to cinema studies. I have yet to hear that any department on the lookout for
talent has decided that it needs a cognitivist to balance out its postmodernists, post-
colonialists, and cultural studies adherents.
By and large, humanist intellectuals dismiss cognitive theory and evolution-based
explanations because they distrust science. As the most visible instance of reliable
Introduction 5
knowledge in our culture, scientific research has been a target for relativists who doubt
all claims to authoritative knowledge (except those claims to authoritative knowledge
made by the relativists themselves). In addition, many progressive people believe that
egalitarianism is threatened by expertise, so science, in setting up standards of theory
and proof, seems to be "policing" discourses. But of course expertise has proven itself
more reliable as a source of knowledge than intuition, superstition, and political fiat.
In addition, there's the danger that considering science "just another discourse" plays
into ignorance and oppression. We in the United States are all too aware that religious
faith can be whipped up to support dangerous political policies. Any progressive
person ought to deplore the results of a 2005 poll finding that two-thirds of Americans
believe that humans were directly created by God, and two-fifths claim to believe
that "living things have existed in their present form since the beginning of time." 6
Whatever your politics, it's better to act from accurate information and coherent ideas
than from lies and mistakes. 7 That means acknowledging that the tradition of rational
and empirical inquiry, however subject to error, is our most reliable path to knowl-
edge, which can be used for progressive ends.
Too often, humanists recoil from science because of its social uses. Science, many
feel, is to blame for many of our current woes, from pollution to the threat of nuclear
war, and its sins include eugenics and the ghastly experiments in the Nazi camps.
Undoubtedly scientists have sometimes been recruited to immoral enterprises, and
like all knowledge, scientific knowledge doesn't automatically confer virtue. Still,
science as a communal endeavor can better the human condition. Many human-
ists justifiably believe that racial inequality, class prejudice, and global warming are
threats to civil society. W h o offers the compelling evidence that young Black men
are the United States' most at-risk population and that working-class citizens have
suffered most under Republican regimes? Not professors of literature but sociologists,
economists, anthropologists, and political scientists. Attorneys, legal researchers,
and forensic scientists have used DNA evidence to free unjustly imprisoned people.
Warnings about global climate change come from the united efforts of biologists,
geographers, geologists, and other experts. Medical professionals struggle to eradi-
cate HIV and cancer, and some risk their lives to inoculate children in the inferno of
war. It's shameful for comfortable academics to believe that these heroes labor under
a flawed epistemology.
In any event, despite what Theorists say, they don't believe it. A postmodernist
Who gets the flu hurries to the doctor as fast as anybody else. The doctor's diagnosis,
backed up by the research of thousands of specialists in the life sciences, is relied
on, not dismissed as a culturally biased interpretation or a text to be read under
suspicion. "Show me a cultural relativist at thirty thousand feet," notes Richard
Dawkins, "and I'll show you a hypocrite." Only in the seminar room (and the pulpit)
is science deeply suspect. 8
In subscribing to the antiscientific stance of Theory, film studies risks remaining
provincial. Reading the Theory pick hits of the 1970s and 1980s, you wouldn't know that
Chomskyan, not Saussurean, linguistics was revolutionizing the study of language, or
that cognitive psychology and neuropsychology were teaching us more about the mind
6 Poetics of Cinema Introduction 7

than Lacan could imagine (and he had a capacious imagination). Film studies, even Doug Gomery, Erik Gunneson, Meg Hamel, Debbie Hansen, Dave Hellenbrand,
with its professed "historical turn," continues to emphasize "methods" over questions, Linda Henzl, Boyd Hillestad, the late Chris Hoover, Lea Jacobs, the late Nietzchka
catchwords over concepts, and doctrines over free-ranging inquiry. 9 Too many film Keene, Vance Kepley, Jared Lewis. J. J. Murphy, the late Ordean Ness, Beth Onosko,
scholars promote a limited conception of interdisciplinarity, borrowing ideas only the late Tim Onosko, Sandy Rizzo, Matt Rockwell, Mary Rossa, Paddy Rourke,
from trends that fall within the Continental hermeneutic tradition (theories of literary Frank Scheide, Ben Singer, and Andrew Yonda. While conducting my education
interpretation, Lacanian psychoanalysis, postmodern anthropology, and the like). By in public, I've learned a lot from people in adjacent wings of our department: Julie
asking questions from a broader purview, we open ourselves to ideas from comparative DAcci, Michael Curtin, Michele Hilmes, Shanti Kumar, Joe Cappella, Joanne Cantor,
narratology, cognitive psychology, Darwinian theoretical programs, network theory, Mary Anne Fitzpatrick, Zhangdong Pan, the late Ed Black, Tim Haight, Jim Dillard,
and other progressive trends in the human sciences. We needn't look to them for Blake Armstrong, Ray McKerrow, Lloyd Bitzer, Michael Leff, Steve Lucas, the late
ultimate answers—we don't need another dogmatism—but we should canvass widely Michael McGee, and Sue Zaeske. Chuck Wolfe, Tom Gunning, Richard Maltby, Pete
in seeking out help in answering the research questions we pose. Parshall, and many other visitors to Vilas Hall over the years have taught me more
Most of the essays collected here have appeared in print before. All have been than they probably realize. So, too, have the friends I've made in graduate school and
revised, some a lot. Now I appreciate why so many authors prefer reprinting old at archives, universities, film festivals, conferences, and the normal networking of
articles to recasting them. It's harder to patch up an old piece than to weave a wholly academic life. And so as well have the thousands of students (no kidding) I've had the
new one. Although I haven't tried to summarize developments in any exhaustive way, pleasure of teaching here at Wisconsin. A teacher is said to be the only person who
a few essays in this book include codas that develop some themes in the light of recent talks in somebody else's sleep; happily my students have been awake, demanding, and
research. New to this volume are "Poetics of Cinema," "Three Dimensions of Film good-humored. My 30-some dissertators, from Brian Rose (1975) to Jonathan Frome
Narrative," "Mutual Friends and Chronologies of Chance," and "CinemaScope: The (2006), have been particularly patient with me.
Modern Miracle You See Without Glasses," although the first and last have their roots Finally, I must single out Bill Germano, a friend of long standing who made this
in older pieces.10 Further reprints of essays will, I hope, be appearing on my Web site, anthology possible, and Kristin Thompson, whose loyalty and love have sustained me
http://www.davidbordwell.net, along with new material from time to time. for even more years than this collection spans.
In assembling this collection, I was helped by Eric Crosby and Brad Schauer,
who took care of text matters, and Jake Black and Kristi Gehring, who prepared the
illustrations. Many institutions have helped shape the original essays, notably the
University of Wisconsin-Madison, which helped my research in a generous variety of
ways. I must also thank several people at film archives, most notably the late Jacques
Ledoux and Gabrielle Claes of the Royal Film Archive of Belgium; Pat Loughney and
the late Kathy Loughney of the Motion Picture and Recorded Sound division of the
Library of Congress; Charles Silver and Mary Corliss of the Museum of Modern Art;
the late James Card, Chris Horak, and Paolo Cherchi Usai of George Eastman House;
Bob Rosen, Charles Hopkins, and Eddie Richmond of the UCLA Film and Television
Archive; Schawn Belston of the Twentieth Century Fox Film Archive; Mike Pogorzelski
of the Academy of Motion Picture Arts and Sciences Archive; Elaine Burrows of the
National Film and Television Archive of London; lb Monty, Karen Jones, Dan Nissen,
and Thomas Christensen of the Danish Film Institute Archive; Matti Lukkarila and
Antti Alanen of the Finnish Film Archive; Okajima Hisashi of the Japan Film Center;
and Chris Horak (again), Stefan Droessler, and Klaus Volkmer of the Munich Film
Museum. My thanks extend as well to the helpful people staffing all these archives.
Among all the people who have shaped the original essays, my associates at the
University of Wisconsin-Madison have been inestimable sources of ideas, resources,
and criticism. Over the years, my work has been aided by Jeannie and David Allen,
Joe Anderson, Tino Balio, Sally Banes, Jim Benning, Joe Beres, Ben Brewster, Mary
Carbine, Noel Carroll, Kelley Conway, Beth Corbett, Jim Cortada, Don Crafton, Jim
Danky, John Davis, Susan Davis, Maxine Fleckner-Ducey, Tom Flynn, Kevin French,
 I
Questions of Theory
 1.

Poetics of Cinema

Sometimes our routines seem transparent, and we forget that they have a history. It's
commonplace for academics in the humanities to assume that every field consists of
objects of study and diverse "methods" for studying them. In literary studies, you
have texts, and you try to understand them by applying a variety of doctrines about
literature or language or life. You may be a phenomenologist or a Lacanian, a follower
of deconstruction or poststructuralism or cultural studies, but everybody, explicitly
or unawares, subscribes to some method.
Familiar as it is, this way of thinking isn't eternal. It emerged only 60 years ago,
out of the boom in college literary criticism that followed World War II. The bench-
marks are Stanley Edgar Hyman's The Armed Vision (1948) and Rene Wellek and
Austin Warren's Theory of Literature (1949). Both books set forth the novel idea that
literary studies played host to distinct "methods." 1 Intrinsic and extrinsic; textual
and contextual; sociological, Marxist, psychoanalytic, and archetypal: The categories
invoked by Hyman, Wellek, and Warren have a distinctly modern ring. Thereafter,
book-length studies explored one method or another, applied to this or that author,
and editors compiled anthologies pitting one method against another for the sake
of classroom instruction. With the emphasis on "practical criticism," the professor
could take a poem or play and train upon it the guns of competing methods. The
anthology-of-approaches genre became a going concern in the 1950s and 1960s, and
it continues to flourish.2
Film studies quickly subscribed to critical Methodism. One of the most popu-
lar anthologies, Movies and Methods, first appeared in 1976.3 You can argue that
this tactic helped give media research a path into the university. Even if somebody
thought that the object of study lacked importance—what intellectual would study

11
 Poetics of Cinema
12
Hollywood?—a set of up-to-date approaches constituted intellectual bona fides. But it
seems to me that film studies accepted, too unquestioningly, the literary humanities'
conception of method itself.
In film studies, as in its literary counterpart, "method" comes down to meaning
"interpretive school." An interpretive school, I take it, asks the writer to master a
semantic field informed by particular theoretical concepts and then to note certain
features of films that fit that field. The writer then mounts an argument that relates
features of a film to the theory by citing the film, quoting from relevant theorists, and
creating associative links between the semantic field and the film. For example, to a
psychoanalytic critic, certain semantic features enjoy a particular saliency: semantic
oppositions, like male and female or sadism and masochism, along with concepts
like the deployment of power around sexual difference. The critic will then pick out
textual cues that can bear the weight of the semantic features, such as the narrative
roles assigned to men and to women, or the representation of the act of looking. The
critic will then mount an argument, perhaps using the rhetoric of demystification,
to show the significance of the semantic projections, from field to text, that the critic
generated. Every recognized "method"—phenomenological, feminist, Marxist, or
whatever—follows something like this routine. They all aim to produce interpreta-
tions, which I take to be ascriptions of implicit or symptomatic meanings to texts. 4
Poetics is a somewhat different enterprise. It doesn't constitute a distinct critical
school, so it isn't parallel to any of the doctrinally defined methods. It has no privi-
leged semantic field, no core of procedures for interpreting textual features, and no
unique rhetorical tactics. Although interpretations don't lie outside its province,
the status of interpretation isn't quite what it is in the doctrine-driven approaches.
Put another way, the domination of methods-based thinking has yielded various
hermeneutics, but poetics is something else again.
The Tradition
Aristotle's fragmentary lecture notes, the Poetics, addressed what we nowadays rec-
ognize as drama and literature. Since his day, we have had Stravinsky's Poetics of
Music, Todorov's Poetics of Prose, a study of the poetics of architecture, and of course
the Russian formalists' Poetics of the Cinema.5 Such extensions of the concept are
plausible, because it need not be restricted to any particular medium. Poetics derives
from the Greek word poiesis, or active making. The poetics of any artistic medium
studies the finished work as the result of a process of construction—a process that
includes a craft component (such as rules of thumb), the more general principles
according to which the work is composed, and its functions, effects, and uses. Any
inquiry into the fundamental principles by which artifacts in any representational
medium are constructed, and the effects that flow from those principles, can fall
within the domain of poetics.
Some further distinctions are useful. A research project in poetics may be pri-
marily analytical, studying particular devices across a range of works or in a single
work. You can, for instance, study assonance and alliteration in sonnet lines. Or the
Poetics of Cinema 13
project can be predominantly theoretical, laying out conditions for a genre or class of
work. Examples would be Aristotle's account of tragedy and Gerard Genette's map
of how temporal relations can be represented in narrative. 6 There is also historical
poetics, the effort to understand how artworks assume certain forms within a period
or across periods. Usually, any project will involve all three perspectives, but one or
another will predominate. For example, although Aristotle recognizes the changes
that tragedy underwent over several centuries, he concentrates on building a theory
of the genre as a more or less ideal type.
Along another dimension, a project in poetics m a y b e predominantly descriptive,
outlining the principles of "making" without preference for one option or another.
Alternatively, it may be more prescriptive, favoring certain options. Several ver-
sions of poetics, most often those mounted by artists themselves, are prescriptive.
The neoclassicists of eighteenth-century English literature promoted a poetics of
reason and generality, obliging the artist to propound universal truths. The poet, in
Dr. Johnson's phrase, "does not number the streaks of the tulip." 7 Romantic theorists
of a later generation argued for an alternative poetics, one that finds beauty in unique
forms. Coleridge finds himself entranced by the frost on a windowpane because it has
crystallized into traceries suggesting trees or seaweed.8
All these distinctions may seem rather abstract, so let me illustrate the diversity of
poetics, as I conceive it, with two exemplary essays, one about literature, and the other
about cinema. The poet W. H. Auden confessed that he enjoyed reading criticism
that could "throw light upon the process of artistic 'making,'" so it's not surprising
that several of his literary essays approach authors or genres from the perspective of
poetics. 9 In "The Guilty Vicarage," he provides a compact account of the classic detec-
tive story. His emphasis is predominantly theoretical, finding the crux of the genre in
an Aristotelian pattern of action: A murder is committed, many are suspected, and
the killer is revealed and punished. This permits Auden to distinguish the detective
story from an adjacent genre such as the suspense tale, in which the murderer's guilt
is known from the outset. From the core plot action, he deduces several other features
of the genre. The concealment of the murderer's identity raises problems of narration
(key information must be withheld from the reader) and the structuring of time (the
action usually needs to unfold in a short span, before the killer can escape). The crime
must be murder, because the absoluteness of the act forces society to act on behalf
of the victim. The plot demands a certain community (typically, a closed one) and
setting, along with particular roles assigned to victim and murderer, the innocent
participants, and the detective. The thematic crux of the genre is that of an Edenic
society that is ruptured. Murder creates a crisis because it "reveals that some member
has fallen and is no longer in a state of grace."10 Auden concludes by considering the
role of the reader in enjoying the tale's recovery of innocence, whereby the guilty
party is revealed to be radically different from the reader. Auden's anatomy of the
genre doesn't trace historical conventions or analyze a single story in detail, and it is
shot through with judgments about what is preferable in the genre, such as obedience
to the classical unities of time, place, and action.
14 Poetics of Cinema
Andre Bazin's classic essay "The Evolution of the Language of Cinema" illustrates
a more historical approach to poetics. 11 Bazin argues that Western film style is not
best considered as a development from silent cinema to sound cinema, but rather
as a process in which two tendencies, present at the start, collide or blend. On one
hand, there is the tendency toward recording reality; on the other, there is the urge
to abstract from reality, to create artifice. The earliest films relied on recording, but
in the mature silent cinema of the 1920s, this tendency became a minority option. In
various ways, D. W. Griffith, the German expressionists, the French avant-garde, and
the Soviet montage filmmakers exemplify the triumph of stylization, making cinema
a vehicle for abstract concepts and formal experiment. But the coming of sound, with
its unremitting tie to recording speech, banished the formal trend and led to a middle
way represented by classical studio style. The sound cinema, for Bazin, marked
the decline of antirealistic filmmaking and the emergence of a relatively realistic,
moderately manipulative style relying on analytical cutting, shot/reverse shot, and
other features that do only a little violence to the event in front of the camera. But at
the same time, directors like William Wyler and Orson Welles recovered the record-
ing capacity of the camera, with long takes and deep-space staging presenting the
event in all its completeness. Still, having learned the lessons of classical cutting,
Wyler and Welles organized their images so that they were more articulated than
the primitive frames of the Lumieres and Georges Melies. Wyler and Welles discov-
ered how to present a scene in a single shot but retain all the changes of emphasis
to be found in an analytical breakdown into several shots. Citizen Kane (1941) and
The Little Foxes (1941) constitute "a dialectical step forward in the evolution of film
language," a powerful reconciliation of opposed tendencies.
Throughout, Bazin relies upon theoretical distinctions, such as intershot effects
versus intrashot effects, types of montage, distortion versus fidelity, spatiotemporal
unity versus discontinuity, and shallow space versus depth. Note that these aren't like
the semantic fields governing doctrine-driven interpretive methods. Bazin, quite rea-
sonably, holds his concepts to be principles determining the stylistic construction of any
film whatsoever. The analyst can correlate the choice of devices with intended effects.
So Bazin proposes that the highly visible montage in Sergei Eisenstein's films yields
a step-by-step layout of meaning, whereas Welles' and Wyler's depth of field conveys
dramatic information through the simultaneous presence of various elements.
While Auden is interested in plot and theme, Bazin focuses on style. Auden is
almost wholly unconcerned with the history of the detective story, treating it as an
ideal type, but Bazin puts his theoretical categories into motion, letting them measure
the development of Western cinema from primitive filmmaking to neorealism. Like
Auden, however, he can be prescriptive. It's plain that he favors cinematic styles that
preserve the spatial and temporal integrity of reality, so the history he traces carries
a note of triumph. Welles and Wyler have ushered in a Hegelian synthesis of all
previous cinema.
One can quarrel with Auden's and Bazin's essays,12 but they illustrate some of the
options available to the poetician. They also exemplify the possibilities of a poetics
that is grounded in both rational and empirical inquiry. Each writer's categories
Poetics of Cinema 15
are explicit and open to criticism on conceptual grounds. Just as important, each
writer invokes concrete evidence that allows us to appraise his claims. This appeal to
empirical evidence, or "facts," does not make poetics into an "empiricism," at least in
any interesting sense of that term. A poetics can be rationalist or empiricist, Kantian
or phenomenological, deductivist or inductivist, idealist or positivist. Whatever its
ontology or epistemology or discovery procedures, a poetics appeals to intersubjec-
tively available data that are in principle amenable to alternative explanation.
To a great extent, an exercise in poetics typically takes as its object a body of
conventions—genre conventions in "The Guilty Vicarage," and stylistic conventions in
the "Evolution of the Language of Cinema." Conventions, in film as in other domains,
lie at the intersection of conceptual distinctions and social customs. Auden charac-
terizes the detective story as at once an adventure in reasoning and a social ritual of
casting out the sinner. Bazin's realist aesthetic leads him to range stylistic devices
along a continuum whereby some are less "conventional" than others. Nevertheless,
in studying patterns of editing and mise-en-scene, he is invoking a structured set of
options that are quite salient in Western cinema. These options constitute norms, a
central concept in poetics that is explicitly signaled by our two writers.
The two sample essays exemplify still other aspects of the line of inquiry I'm
exploring. They show that we might join observation of general tendencies with a
scrutiny of particulars, such as Auden's discussion of various detective heroes and
Bazin's account of influential films and directors. The writers consider both "texts"
and "contexts": Auden situates the detective story within Christian societies, while
Bazin looks at trends of genre and narrative that affect style. And both poeticians
presume that the artwork results from choices within a craft tradition. The fictional
detective may be propelled by purely intellectual motives, as Sherlock Holmes is, or
he may be seeking to redeem a soul, as Father Brown is. Likewise, Bazin's case rests
upon the possibility that Welles could, if he'd wished, have shot and cut Citizen Kane
in the manner of It Happened One Night (1934). Craft practices always offer a range of
options, and the choices made by the artist will be correlated with some purpose—the
design of the work or an effect on the perceiver.
My initial questions and my exposition of the Auden and Bazin essays should raise
several questions about how this approach works. What, for example, is the status of
the "principles" studied by poetics? I'd argue that the principles should be conceived
as underlying concepts, constitutive or regulative, governing the sorts of material that
can be used in a film and the possible ways in which it can be formed. At what level of
generality are these principles pitched? The degree of generality will depend upon the
questions asked and the phenomena to be studied. If you want to know what makes
Hollywood narratives cohere, "personalized causality" may suffice as one construc-
tive principle; if you want to know what distinguishes a film noir from a musical,
that principle isn't up to the job. For some poeticians, principles are held to be laws
on the model of covering laws in physical science, but we needn't push that far. You
could assert that a concept—say, Bazin's distinction between what happens within
shots and what happens between shots—is foundational, but that the ways in which
16 Poetics of Cinema
filmmakers use that as a constructive principle vary so much that we can trace only
general tendencies.
Are the principles conceived as "specific" to cinema in some sense? Although
certain poeticians have assumed a distinction between the cinematic and the non-
cinematic, this view isn't a postulate of poetics as such. You can assume that any
film could be studied by poetics, with no film laying any closer to the essence of the
medium than others. You could, though, also argue that the distinction between
cinematic and noncinematic is not a substantive but rather a functional one, to be
filled out in different periods with different content. Or you could use the distinction
in an explanation by seeking to show that in particular circumstances, this pair of
concepts entered into the norms of filmmaking practice because filmmakers believed
in some version of it.
Poetics is often assumed to aim merely at descriptions or classifications, so
I should elaborate a little on the range of explanations it offers. There's no need to
assume any one model of causation and change. Bazin argues for a broad dialectic
through which cinema evolves toward an ever more faithful capturing of phenomenal
reality. This is a teleological explanation. One could also propose an intentionalist
model that centers on filmmakers' localized acts of choice and avoidance. Then
there's the possibility of a functionalist model of explanation, whereby the institu-
tional dynamics of filmmaking set up constraints and preferred options that fulfill
13
overall systemic norms.
Nor need poetics be confined to "immanent" explanations that refuse to leave
the field of cinema, art, or representational media. Nothing in principle prevents
the poetician from arguing that economics, ideology, cultural forces, or inherent
social or psychological dispositions operate as causes of constructional devices or
effects. There is likewise no need to cast poetics as offering "scientific" explanations
(although, again, some poeticians have done so). Poetics has the explanatory value
of any empirical undertaking, which always involves a degree of tentativeness about
conclusions. On the other hand, one shouldn't dismiss historical research's affini-
ties with science too quickly, because there are many scientific disciplines, such as
geography and archaeology, which fall short of predictive accuracy but have good
records of ex post facto explanatory power.14 It's probably best to say that poetics
joins the overarching tradition of rational and empirical inquiry to which science
and kindred disciplines belong.
Finally, and to return to a difference with the doctrine-driven methods of film
studies, explanation in poetics doesn't confine itself to issues of what films mean.
Of course, meaning in one (very general) sense comprises a big part of what poetics
describes, analyzes, and explains. But meaning in the narrower sense that is the
product of film interpretation (a "reading") isn't necessarily the goal of the poetician.
Films produce many effects, ranging from perceptual ones (why certain color
schemes dominate films of a particular period) to conceptual ones (how we know that
X is the protagonist), and these matters film interpretation never seeks to elucidate.
Historical poetics, in particular, tends to offer explanations rather than explications.
Still, critics are makers too, and we could analyze their materials, principles, and
Poetics of Cinema 17
concerns with effect. That is, we can try to explain interpretations. Finally, practical
criticism, focused on particular films, can be fruitfully informed by poetics. Several
of the essays that follow try to show how.
Domains and Tendencies
Traditional poetics in any medium distinguishes among three objects of study:
thematics, large-scale form, and stylistics. Thematics considers subject matter and
theme as components of the constructive process. The researcher may study motifs,
iconography, and themes as materials, as constructive principles, or as effects of
constructive principles. Auden does this with Christian themes in "The Guilty
Vicarage." Similarly, film scholars have revealed how genres present recurring
imagery, myths, and motifs, whereas other writers, inspired by art-historical research,
have shown the importance of iconography in popular cinema. 15
Taken broadly, thematics informs many people's thinking about cinema. Scholars
and journalists commonly scan movies for social stereotypes and political attitudes.
The stock phrase "representation of race, class, and gender" invokes themes in a loose
sense. Yet although studies of these matters can be enlightening, they wouldn't usually
constitute a contribution to poetics, because they're often not concerned to link
themes to constructive principles. They don't typically show how the overall design of
the movie, including areas not obviously related to the stereotypes on display, requires
the stereotypes in order to achieve its particular purpose or effect. Nor do they treat
the themes discovered as part of historical traditions of art making.
For example, the films of Ozu Yasujiro often evoke the theme of the transience of
human life. This theme is common in the world's art, and it's especially prominent in
Japanese poetic traditions. It was reworked in Japanese popular culture early in the
20th century, when a newly modernizing Tokyo was seen to embody the ephemeral-
ity of existence: Yesterday's building would be demolished and replaced by something
up-to-date tomorrow. The theme is highlighted in Ozu's films, when characters talk
about past pleasures and look forward to a moment in the future when they will
recall what they're doing now. The theme also finds visual expression in Ozu's use of
conventional imagery of transience like clouds, smoke, and streetlights switched off
or on. Often items of setting, like laundry on a line or household utensils, disappear
or change position from scene to scene, teasing us to recall an earlier moment in the
film. Ozu carries the theme of constant change, itself a cultural cliche, from broad
social sources into the intimate drama and down into the details of filmic texture. 16
Sometimes there's a tension between thematic givens and the film's overall design.
Early in Laura (1944), a sharp opposition is set up between McPherson, the rough-
edged police detective, and Waldo Lydecker, the effete bon vivant who has been the
patron of the murdered Laura. McPherson is a real man, and Waldo a sissy. Before she
died, Laura threw over Waldo for Jacoby, who is at once a handsome, athletic man and
a sensitive artist. In any realistically motivated plot, Jacoby would be a prime suspect.
Remarkably, however, McPherson doesn't investigate him at all. It seems likely that
Jacoby, being a blend of the extremes that define the major male characters, is too
18 Poetics of Cinema
much of a threat to the final purpose of the plot, the romantic union of McPherson
and Laura. So the plot is distorted to make the themes fit.17
Thematics lies perhaps closest to method-dominated criticism, but within the
tradition of poetics there have been wide-ranging theoretical and methodological
debates of a kind not seen in interpretive approaches. What are themes, and where
might we find them? Should we consider them unique particulars (e.g., Romeo), com-
mon motifs (lovers), generalizations (love), or semantic oppositions (romantic love
versus social obligation)? Is a theme a priori, something that the artist inherits and
reiterates, or is it post facto, something that perceivers create in order to endow the
artwork with coherence and significance? For many people, the discovery of themes
is a major reason for engaging with artworks; for some poeticians, the discovery of
themes has little bearing on interpretation as usually conceived. In the 1960s, two
Soviet researchers created a "poetics of expressiveness" that treated themes as "deep
structures" undergoing various transformations before being concretized as surface
patterns of the text.18 Here, Yuri Shcheglov explains, the theme is not a message or
separable content that the reader carries away, but rather a principle provided by the
analyst in order to account for the formal features of the text.19 Somewhat similar
was the effort of Michael Riffaterre to show that a poem is generated from an under-
lying verbal formula, often a linguistic cliche. The text elaborates this donnee without
necessarily naming it. Thus a Cocteau poem elaborates a traditional motif, the inn of
death," through imagery of travel, dying swans, and other figures of speech. 20
A historical poetics of cinema is likely to consider themes as given materials that
are transformed by traditions of form and style. Soviet films of the 1920s and 1930s
were charged with representing the emergence of "revolutionary consciousness," and
filmmakers like Eisenstein and V. I. Pudovkin could take these themes for granted
and explore increasingly allusive ways to signal them. The prepackaged nature of
Leninist ideology helped directors in the montage style and Socialist Realism create
oblique, flamboyant ways of saying what everybody already "knew." 21 It isn't only
openly rhetorical cinema that can be illuminated by the idea of theme as a cultural
given. Noel Carroll has pointed out that many narrative films can be seen as illustrated
homilies. They presuppose vague commonplaces in order to be intelligible, as Back to
the Future (1985) assumes that anything can be altered by individual striving. 22 By
studying commonplaces in circulation in a given period's culture, we can often link
cinema to other media and social life.
A second domain of poetics is that of large-scale form. The poetics of literature
explores principles of progression and development governing the well-made play,
the sonnet, or the adventure novel. We students of cinema lack a term for those
transmedia architectonic principles that govern the shape and dynamics of an
entire film. The most prominent research domain here is the theory and analysis of
narrative, which is a fundamental constructive principle in films.23 In this book I've
devoted a separate essay to narrative form and several essays to particular films or
traditions, so here I'll just mention that there are other compositional principles that
poetics should investigate. A film can be organized as a rhetorical argument, or it can
collect an array of categories, as in a catalogue. The form may be associational, as in
Poetics of Cinema 19
a film lyric like Stan Brakhage's Scenes From Under Childhood (1967-1970), or based
wholly on abstract similarities and differences (Ballet Mecanique, 1924).24 In practice
these types can combine, as when films like Fahrenheit 9/11 (2004) recruit narra-
tive patterns to a larger rhetorical impulse. P. Adams Sitney's classic book Visionary
Film proposes another taxonomy of large-scale forms, as manifested in the postwar
American avant-garde. 25
Stylistics, the third leg of the poetics tripod, deals with the materials and pattern-
ing of the medium as components of the constructive process. Bazin's "Evolution"
essay is a model of stylistic history of cinema. Most of the stylistic studies in this
book concentrate on visual patterning—staging, shot scale, composition, editing, and
camera movement—but that doesn't mean that sound doesn't matter. Scholars with
better-trained ears than mine have studied how techniques of sound recording and
reproduction shape the stylistic texture of a film, or how the score contributes to the
overall stylistic dynamics. 26 On this last front, many scholars have refined our under-
standing of film music; the number of studies of film music far exceeds the number of
studies of cinematography, editing, or other techniques of the image track.
Philosophers of art have long debated how to define the concept of style. Expres-
sive theories treat style as the manifestation of artistic personality or emotional states,
rhetorical theories treat style as a matter of impact on the audience, and objective
theories consider that it consists of objective properties of the artwork's formal
design. Then there are conceptions of period style, national style, and the like. All of
these ideas have proven fruitful for researchers studying the poetics of the arts.
One Poetics of Film
Given the great variety of research programs within the broad domains of poetics, I'm
going to use the rest of this essay to characterize the threads of reasoning that wind
through the pieces in this book. I propose a version of poetics that rests upon film
analysis. For me, the most interesting questions grow out of particular films. This angle
of approach can draw inspiration from rich traditions in adjacent fields. Art historians
such as Heinrich Wolfflin, Alois Riegl, Erwin Panofsky, and E. H. Gombrich show us
how to systematically track forms and styles in the visual arts and explain their changes
causally.27 In literary theory, the Russian formalists and the Prague structuralists—
most notably, Viktor Shklovsky, Yuri Tynianov, Boris Eikhenbaum, Jan Mukarovsky,
and Roman Jakobson—proposed both concrete analyses of literary works and larger
explanations for how they functioned in historical contexts. 28 More recent literary
theorists, particularly Meir Sternberg, have also provoked me to formulate a position
on narrative in cinema. Leonard Meyer, Charles Rosen, and other musicologists have
likewise furnished models for thinking about form and style in relation to historical
change. The essays in this book are explicitly indebted to these thinkers.
Most academic books about cinema carry at least a dollop of theory, so it's best
to be clear about the role of theory in the essays that follow. I sometimes draw upon
film theorists of the pre-1970 period, such as Andre Bazin, the Soviet filmmakers
of the 1920s, and Noel Burch, but that doesn't entail that I'm committed to Bazin's
20 Poetics of Cinema
phenomenology or Eisenstein's odd version of materialism or the early Burch's serialist
theory of film. If we take a film theory to consist of a set of propositions explaining the
fundamental nature and function of all cinematic phenomena, the poetics I'm setting
out doesn't amount to a theory in that sense. It's best described as a set of assump-
tions, a heuristic perspective, and a way of asking questions. 29 It's frankly empirical
and tries to discover facts and truths about films.
When I first floated the idea of a historical poetics of cinema, I anticipated objec-
tions to an empirical program of this sort. After all, many said in the 1980s, there
aren't facts but only "facts" (that is, social constructs that vary according to time and
place), and there can be no truth (usually identified, erroneously, with absolute truth
or final truth or capital-T Truth). Since the rise of cultural studies, an area of inquiry
that wants to discover ways in which audiences appropriate films, I don't have to
be so defensive. Film scholars have come to realize that any descriptive or explana-
tory project is committed to some grounding in intersubjective data. All intellectual
disciplines seek to find out how things are.
Of course, there's no question of letting facts speak for themselves. We can't dis-
cover plausible answers to questions about films' construction without carefully
devising analytical concepts appropriate to these questions. But not all concepts are
equally precise, coherent, or pertinent, and so we may evaluate competing conceptual
schemes. Crucially, we're not complete prisoners of our conceptual schemes. We
may specify our ideas in an open-textured way, so that exceptions leap to our notice.
The poetics I'm proposing makes claims that are theoretically defined, open-ended,
corrigible, and potentially falsifiable.
This is a direct result of its not being a general theory of film. If I'm bent on substan-
tiating the belief that every film constructs an ongoing process of "subject positioning"
for the spectator, nothing I find in a film will disconfirm it. Given the roomy interpretive
procedures of film criticism, I can treat every cut or camera movement, every line of
dialogue or piece of character behavior, as a reinforcement of subject positioning. The
theory becomes vacuous, because any theory that explains every phenomenon by the
same mechanism explains nothing. On the other hand, I can ask how Hollywood films
secure unity among successive scenes, and answer with something more c o n c r e t e -
say, that one scene often ends with an unresolved causal chain that is soon resolved in
the following scene. Here I've said something that's informative. It isn't self-evident,
it isn't discoverable by deduction from a set of premises, and it's fruitful, leading to
further questions. Does this constructive principle suggest some hypotheses about
the nature of narrative norms in Hollywood? (It does.) Do films in other filmmaking
traditions utilize more self-contained episodes? (They seem to.) Most important, the
answer I supply could be disconfirmed. If it is discontinued, I need to rethink the
data and, indeed, the question itself. Shklovsky's counsel of skepticism should be our
guide: "If the facts destroy the theory—so much the better for the theory. It is created
by us, not entrusted to us for safekeeping." 30 Ideally, our hypotheses are grounded in a
theoretical activity rather than a fixed theory. This activity moves across various levels
of generality and deploys various concepts and bodies of evidence. It seeks to be driven
Poetics of Cinema 21
by data and midrange concepts rather than by abstract or absolute doctrines, and it
can be recast or rejected in the light of further investigation.
In being question centered and focused on particular phenomena, the poetics
I envision somewhat resembles the practices of inquiry in scientific endeavor. Stephen
Jay Gould writes,
Progress in science, paradoxically by the layman's criterion, often demands
that we back away from cosmic questions of greatest scope (anyone with half a
brain can formulate "big" questions in his armchair, so why heap kudos on such
a pleasant and pedestrian activity?). Great scientists have an instinct for the
fruitful and the doable, particularly for smaller questions that lead on and even-
tually transform the grand issues from speculation to action Great theories
must sink a huge anchor in details. 31
Likewise, here is Francois Jacob, discoverer of RNA's function as a "message molecule"
for transmitting genetic information:
The beginning of modern science can be dated from the time when such general
questions as "How was the Universe created? What is the essence of life?" were
replaced by more modern questions like "How does a stone fall? How does water
flow in a tube?" While asking general questions led to very limited answers, asking
limited questions turned out to provide more and more general answers. 32
Someone will object that this appetite for midsize questions may suit the hard
sciences, but studying culture and history can't be so precise. Yet C. Wright Mills, no
positivist by any description, suggests that the "sociological imagination" is charac-
terized by a middle way.
Classic social science, in brief, neither "builds up" from microscopic study nor
"deduces down" from conceptual elaboration. Its practitioners try to build and
to deduce at the same time, in the same process of study, and to do so by means
of adequate formulations and re-formulations of problems and their adequate
solutions. To practice such a p o l i c y . . . is to take up substantive problems on the
historical level of reality; to state these problems in terms appropriate to them;
and then, no matter how high the flight of theory, no matter how painstaking
the crawl among detail, in the end of each completed act of study, to state the
solution in the macroscopic terms of the problem Controversy over differ-
ent views of "methodology" and "theory" is properly carried on in close and
continuous relation with substantive problems. 33
Mills' duality echoes the two options we're usually offered in the humanities. You
may tackle very tightly focused projects, which supposedly lead to steadily accumu-
lating knowledge; you can't make bricks without straw, as they say. Alternatively, you
do Grand Theory, where you can't make a move without getting all your abstract
doctrines correct beforehand. Elsewhere I've advocated that film scholars could pitch
a project at a middle level, asking questions of some scope without deep commitments
22 Poetics of Cinema
to broad doctrines, and using the answers to those questions to build hypotheses of
34
greater generality.
I quote these worthies at length not to show that poetics is a science. My point is
that as compared with Grand Theory, it aims at satisfying general demands of ratio-
nal and empirical inquiry. Take, for example, the notion of norms. In the essays that
follow, I assume that it's often useful to ask how a film relates to sets of transtextual
norms. These operate at various levels of generality and possess various degrees of
coherence. For instance, in most studio-made narrative films, the credits sequence
characteristically occurs before the first scene, but it may also, as lesser options,
occur after a "precredits sequence" or during the first scene. Such norms, although
"codified," are not reducible to codes in the semiotic sense, because there is no fixed
meaning attached to one choice rather than the other. Jean-Luc Godard's decision, in
Detective (1985), to scatter the credits sporadically through the first 14 minutes yields
rather unusual effects on our apprehension of the story, but no definite meanings auto-
matically proceed from it.
For a long time, people training to be composers or performers studied "music
theory" and aspiring painters studied "art theory." These terms didn't refer to inqui-
ries into the nature and functions of their respective art forms, still less to the Grand
Theory that permeates the humanities today. Music theory was about how to write
counterpoint, how to orchestrate effectively, or how to build a symphony up from
phrases. Art theory was about composition, color values, and the like. Music theory
and art theory were repositories of craft knowledge, stated in more or less principled
fashion and invoking the proven success of inherited norms. To a large extent, poetics
is a systematic inquiry into the presuppositions of artistic traditions. It's a practice-
based theory of art. We want to know the filmmakers' secrets, especially those they
don't know they know.
Craft norms are historically variable; the music theory taught in Paris conserva-
tories in the 1880s was very different from that taught to musicians in India. That's
why we need a historically inflected poetics, one that recognizes that art is made
differently under different circumstances. A historical poetics, it seems to me, should
also be alert for commonalities among apparently diverse norms. Conventions shared
across distinct art traditions can be as important as those of narrower provenance, as
I try to show in the essay following this.
A poetics can reveal both change and continuity among norms by reconstructing a
historical context. How does this work fit into a tradition? How does it repeat, revise,
or reject its forerunners? This sort of thinking is commonplace in mature disciplines.
Consider, for instance, Boris Eikhenbaum's essay on the stories of O. Henry. Here
Eikhenbaum traces changes in the writer's oeuvre against the background of the
history of the American short story and its masters, Washington Irving, Poe, Harte,
and Twain. O. Henry's work, he claims, displays a series of formal experiments moving
from cyclical construction to apsychological characterization and reflexive parody.
He discusses how the writer ironizes the sentimental style that was then dominating
mass literature. He shows causal connections between O. Henry's innovations and
changes in American literary tastes and magazine publishing. He concludes with the
Poetics of Cinema 23
claim that O. Henry at the end of his life looked forward to writing something more
straightforward about manners and morals, thereby looking ahead to the loosely
plotted, slice-of-life form characteristic of Dreiser and Sherwood Anderson. "The
O. Henry story, with parody at its core, opened the way for this regeneration." 35 I'm
sure that literary historians could dispute Eikhenbaum's claims, but his essay shows
how one can fruitfully blend formal analysis with a coherent conception of historical
continuity and change. It's a model of a midlevel research project.
Finally, a few words about form, formalism, and "formalism." Given the intellec-
tual lineage I've claimed, and the fact that in the 1980s the approach I sketch here
came to be called neoformalism, I should try to block possible misunderstandings.
Sometimes "formalism" implies an art-for-art's-sake position. But if that view implies
that artworks don't have consequences for morality, behavior, and society, I don't hold
it. Some people use the term to indicate that poetics considers only "form" and not
"content," or "culture," or whatever other subjects the critic thinks more important.
More accurately, the poetics I propose looks at artistic form as an organizing principle
that works not on "content" but rather on materials: not just physical stuff like film
stock or the items set before the camera but also themes, subjects, received forms,
and styles. Out of these materials, the relevant principles create a whole that aims
to achieve effects. By studying form in the sense I mean here, we can understand
how cinema turns materials circulating in the culture into significant experiences
for viewers.
Pretty obviously, those experiences both shape and are shaped by a variety of
cultural forces. Several essays that follow indicate that by studying cinematic form
in the three domains already indicated, the poetics I'm proposing need not cut off
cinema from larger dynamics of social life. True, my questions bear more upon the
"how" of film than the "what," but both are necessary for full understanding. I hope
that any reader of "Three Dimensions of Film Narrative," "Convention, Construction,
and Cinematic Vision," or "Who Blinked First?" will see that I do engage with matters
of culture, though my conception of it may not correspond completely with that
promoted by practitioners of cultural studies.
Poetics: A Program
As I conceive it, a poetics of cinema aims to produce reliable knowledge by pursuing
questions within two principal areas of inquiry. First is what we might call analytical
poetics. What are the principles according to which films are constructed and through
which they achieve particular effects? Second, there's historical poetics, which asks,
How and why have these principles arisen and changed in particular empirical
circumstances? In my view, poetics is characterized by the phenomena it studies
(films' constructional principles and effects) and the questions it asks about those
phenomena-^their constitution, functions, purposes, and historical manifestations.
This research program doesn't put at the forefront of its activities phenomena
such as the economic patterns of film distribution, the growth of the teenage audi-
ence in the 1950s, or the ideology of private property. We may need to investigate
24 Poetics of Cinema
such matters, but they become relevant only in the light of the questions about filmic
construction that guide the inquiry. Underlying this hierarchy of significance is the
assumption that, although in our world everything is connected to everything else,
we can produce fresh and precise knowledge only by making distinctions among core
questions, peripheral questions, and irrelevant questions.
I can specify further. At the risk of seeming cute, I can characterize the research
framework I propose by six P-words: particulars, patterns, purposes, principles, prac-
tices, and processing. These are related, so that by examining any one of them, we re
likely to find connections to others.
Putting a film or set of films at the center of our concern can lead us toward the
most atomic items we can detect. Perhaps we're struck by a line of dialogue, or a
certain cut, or a moment in a performance, or an unusual sound. Details are always
worth noticing, and they're often what critics point to in justifying an appreciation or
an interpretation of a film. Some critics, from the surrealists to the present, have made
a virtue out of celebrating the isolated particulars as ends in themselves. Details can
give us a buzz. 36 Sometimes, though, the items that seize our attention seem puzzling.
Why, for instance, do characters in film stare at each other so constantly and intently,
and why do they blink so seldom? Despite theorists' interest in The Look and The
Gaze, eye behavior of this sort hasn't attracted a lot of attention, but it's an intriguing
feature of filmic storytelling. So the particulars that attract our attention can seem
either unique to the film or something, perhaps even something trivial, that it shares
with other films.
Any poetics goes beyond particulars. The items that we notice belong to patterns.
The hero's single wisecrack belongs to a stream of comments that he makes, a lot of
them wisecracks. Low-key lighting may become associated with a certain character
or locale. Most of the essays that follow take patterns of narrative or visual style as a
primary object of inquiry. But there's a problem here. Any element that we spot can be
situated within an indefinitely large number of patterns. What makes some patterns
salient, either in the act of watching the movie or in the course of our analyzing it?
Our best candidate is the purpose that we can assign to the pattern. Our hero utters
wisecracks because characterizing him this way fulfills some functions in the story.
Perhaps his insolence gets him into trouble with his boss, or makes him appealing to
us, or serves as a foil to a more phlegmatic character, or all of the above. The way char-
acters quite unrealistically stare at each other in films has particular functions, as I try
to show in a later essay. Sometimes filmmakers will acknowledge the purposes that
their strategies fulfill, but more often we have to posit some plausible ones ourselves.
And it goes without saying that anything we pick out may be serving many functions,
and several devices may be working in harmony to achieve one overall purpose.
Recall a device in what's come to be called the classical Hollywood tradition. We
might notice that in Nick of Time (1995), the climax depends upon a race against the
clock. Watch a few more thrillers, and you find the same thing. But then you watch
some romantic comedies or domestic dramas, and you find that often they too focus
their resolution on the pressure of time. We have a conventional pattern of action, one
that today's screenwriters call the "ticking clock." This pattern in turn functions to
Poetics of Cinema 25
increase dramatic tension, leading us to expect that the dramatic issues will come to
a definite climax. It's significant as well that fiction films in other traditions, such as
Antonioni's L'Avventura (1960) and Hou Hsiao-hsien's The Puppetmaster (1993), don't
rely on deadlines to bring the action to a resolution. These narratives are governed by
other purposes.
Often it's useful to conceive the artist's purpose in terms of problems and solutions.
At a mundane level, a filmmaker wants to achieve some pattern or effect. Something
blocks this, so the filmmaker contrives a way to achieve the effect he or she wants. The
result may turn out to be more complicated than what was initially planned. The essay
on CinemaScope in this volume argues that in facing technological constraints, some
filmmakers returned to a form of staging that was thought to be outmoded, with
intriguing results. Or an adventurous filmmaker may actually court problems, laying
down self-imposed constraints in order to stimulate her or his imagination. Both Ozu
and Mizoguchi Kenji did this, the one refusing dissolves and usually situating his
camera always at a lower height than what he photographed, and the other favoring
long takes and framing the action from a high angle. Each director's narrowing of the
artistic bandwidth not only yielded surprising expressive resources within the field he
staked out, but also forced him to deal with a cascade of new problems that a simple
technical choice brought in its wake. 37 We should therefore remember that functional
explanations can sometimes make things too tidy. Every decision is a trade-off, yield-
ing not only benefits but also costs.
When we find repeated items, patterns, and functions across several films, we can
ask about the principles underlying these factors. Most often those principles will be
in the nature of norms, those explicit or implicit guidelines that shape creative action.
I've already suggested that conventions are central subjects for poetics, and we can
think of norms as the principles that govern conventions. Some norms operate at
the small scale, whereas others shape the formal design of whole films. Sometimes
norms are formulated as crisp rules, but most often they are rules of thumb and oper-
ate in the background, learned and applied without explanation or even awareness.
(Filmmakers know a great deal more about their activity than they articulate.) We're
often left to infer the relevant norms by noting regularities and then seeking out
evidence that could count for or against.
Take, for instance, the staging techniques that emerge in feature filmmaking in
Europe during the 1910s. For several decades, most film historians were content to
call this tradition "theatrical" because it relied on lengthy takes of action recorded at
some distance. 38 Although the historians recognized that the editing innovations of
Americans (chiefly Griffith) were highly patterned, the European films didn't seem
organized to the same degree. The historians could support their claims by the fact
that several critics and filmmakers writing in the 1910s and 1920s praised films that
displayed close-ups, alternating editing, and accelerated cutting. But once stylistic
historians began to look closely at the European films of the period, the films no
longer seemed backward. It was evident that the films were patterned to provide clean,
clear uptake of story events within an integrated space. The patterns turned out to be
governed by deep-space staging and, more fundamentally, by the perspectival space
26 Poetics of Cinema Poetics of Cinema 27

Figure 1.3 Jeanne Dielman, 23 Quai du

Figure 1.1 In Victor Sjostrom's Ingeborg Figure 1.2 Ingeborg Holm-. Later, a similar
Holm (1913), the clerk in the family's store camera angle shows Ingeborg's crisis. The bill
slips perfume to a pretty customer as Inge- collector presses her at the cash register, while
borg comes in from the back room. The cam- the guilty clerk pops into a patch of space
era position lets us see that the cash register behind them. Again, this shot wouldn't work
blocks the clerk's theft from Ingeborg's view, on the stage because the clerk's expression
an action that wouldn't be evident from every wouldn't be visible from most seats. Sjostrom
seat in a stage theater; indeed, from some realized that cinematic space, unlike theatri-
sightlines, the register would conceal Inge- cal space, is resolutely monocular; only the
borg's arrival from the spectator. camera's eye matters.
provided by the camera lens. These principles are in significant ways wntheatrical.
Filmmakers used both two-dimensional composition and three-dimensional block-
ing to guide the spectator's attention to the unfolding story in ways impossible to
achieve on the stage (Figures 1.1-1.2). Once the system of norms became apparent,
one could go back to documents and find records by filmmakers that pointed to their
self-conscious awareness of the pyramidal playing space of cinema. 39 Those state-
ments had been available to historians for decades, but they sprang into relief only
after close viewing revealed that there could be alternative norms.
Some norms are probably quite local, such as the deadline-driven climax charac-
teristic of Hollywood cinema. Other norms apply to a surprisingly wide variety of
films. We expect that mainstream filmmakers will tend to place the chief action at the
center of the frame, but an otherwise transgressive movie like Jeanne Dielman, 23 Quai
du Commerce, 1080 Bruxelles (1976) has recourse to the same tactic (Figure 1.3).
Across the world, filmmakers of the 1980s and 1990s tended to build scenes out of
"singles"—shots, usually quite close, of individual characters—rather than ensemble
framings. But norms are systematic and hierarchical, so any dominant principle will
mold others to its needs. In the Viennese classical style of Western music, Charles
Rosen argues, the centrality of the articulated phrase shaped rhythm, texture, and
dynamics. 40 Similarly, once a filmmaker accepts the norm of centered composition,
she or he will tend to adjust staging procedures, lighting, color choices, and editing
patterns accordingly. The reliance on singles in modern cinema forces filmmakers to
cut more frequently, in order to trace the flow of the conversation and to remind the
viewer what characters are present. 41
Commerce, 1080 Bruxelles (1976): When
Jeanne Dielman fails to button her coat
properly, traditional centering procedures
make sure we notice this as a sign of her
breakdown.
As my examples indicate, studying norms is an exercise in extrapolation. By trying
to chart the range of constructional options open to filmmakers at various historical
moments, we come up with results that are always open to revision. In practical
research terms, attention to historically changing devices, patterns, principles, func-
tions, and norms moves us beyond the single film to groups of films. By positing
alternative norms, our work becomes comparative in a rewarding way. Instead of the
couplet norm-deviation, we can posit competing systems of principles, operating at
roughly the same level of generality. We find varying norms of narration and style
in Hollywood cinema, "art cinema," Soviet montage cinema, and other modes. 42 In
Hollywood cinema, for instance, the norm of cogent storytelling favors not only a
ticking clock but also a coordination of that with other conventions, such as causal
continuity and a duplex plotline involving both work goals and romantic goals.
Although it may be momentarily helpful to characterize art cinema narration as
a "deviation" from Hollywood principles, it's more enlightening to characterize it
positively, as possessing its own fairly coherent set of storytelling principles (as I try
to do in a later essay in this book). Recognizing that we're engaged in a comparative
exercise allows us to give equal weight to one norm and another.
Moreover, we don't have to postulate every historical change as a deviation from
a norm. I've already suggested that we can often think of changes as driven by
problems, some inherited from tradition, others devised by the filmmaker. There are
as well many ways to realize norms, some obvious, some subtle. The most striking
stylistic changes in film history often don't stem from absolute innovation but rather
from a recasting of received devices. Welles' deep-focus staging in Citizen Kane is a
famous instance, 43 but we could say much the same of Godard's cutting in Breathless
(1960), which recasts orthodox continuity principles (matching on movement and
eyelines) into new patterns, to new effect. An innovation isn't necessarily a devia-
tion. I suggest in a following essay on Robert Reinert that his rather odd-looking
films are the result of taking to a limit certain staging principles that governed
 Poetics of Cinema
28
mainstream European cinema of the 1910s. Reinert, we might say, broke with the
norm by carrying it to extremes.
Some people think that studying norms necessarily celebrates them. That's not
the case. Norms can be studied objectively without marginalizing alternatives as
freakish or unacceptable. In fact, studying what's norm driven can make us sensitive
to what runs athwart the norm. And, again, what is most fruitful isn't just celebrat-
ing the demolition of standard principles but also the putting forth of alternative
systems. Kristin Thompson, for instance, has been concerned to demonstrate how
the works of Eisenstein, Jacques Tati, Godard, Jean Renoir, and others provide not
fitful deviations from norms but rather systematic innovations in thematic, stylistic,
and narrative construction. 44
We can balance a concern for revealing the tacit con-
ventions governing the ordinary film with a keen interest in the unusual film that,
subtly or flagrantly, challenges them. Accordingly, new concepts will often have to
be forged. To account for Ozu's editing, Thompson and I had to devise the concept of
the "graphic match" and to spell out how Ozu's across-the-line shot/reverse shots do
not willfully transgress rules but rather achieve particular functions within a larger,
idiosyncratic system of 360-degree space. 45
The aims and principles we detect in films are rooted in activities. Filmmakers
work with tools and materials, operating within institutions that offer both constraints
and opportunities. These factors can be summed up under the rubric of practices.
How shall we understand these practices? Two ideas can guide us. First, there is a
rational agent model of creativity. This follows from the idea that the filmmaker
selects among constructional options or creates new choices. Our task becomes that
of reconstructing, on the basis of whatever historical data one can find, the creative
situation that the filmmaker confronts. The rationality at stake is largely one of
means-end reasoning. Assuming a certain end in view, certain options are more
likely to fulfill it than others. If you want to raise tension at the end of a film, then it's
not unreasonable to add a deadline, especially if the tradition in which you're working
offers you several ways to indicate that deadline's approach (including shots of ticking
clocks). Filmmakers have reflected, to various degrees of detail, upon their creative
choices, and this literature offers a rich legacy of insights into practices. 46
This isn't to say that the filmmaker becomes the sole source of the film's construc-
tion and effects. A second, institutional dimension of practice forms the horizon of
what is permitted and encouraged at particular moments. The filmmaker works, most
proximately, within a social and economic system of production, and this involves tacit
aesthetic assumptions, some division of labor, and standard ways of using technol-
ogy. When we want to mount causal accounts of some features or forms, the mode of
filmmaking practice is a good place to look. 47
It's not just that the filmmaker's choices
are constrained; they are also actively constituted in large part by socially structured
factors ofthis sort. In the Hollywood studio system ofthe 1920s and 1930s, for instance,
the continuity script not only became a way to rationalize production. It also encour-
aged workers to think of a film as an assemblage out of discrete bits (shots, scenes), and
the individual filmmaker found choices and opportunities structured accordingly. 48
Poetics of Cinema 29
In most sorts of filmmaking, practices are crystallized in routine ways of doing
things. These form the filmmakers' craft. Filmmakers of the 1910s laid down sticks or
chalk marks to show their players the pyramidal space that the camera was taking in,
and they marked out a "front line" that signaled the point of closest focus. Screenwriters
today often plot their films around a three-act structure; if you don't have a turning
point about 25 minutes in, you're flouting standard practice. Certain choices of lenses
or film stocks come to be preferred, for aesthetic, institutional, and practical reasons.
Hollywood directors are expected to shoot a great deal of coverage, piling up alterna-
tive camera setups that offer a great deal of choice in the final editing. "The financiers
want every conceivable option," notes director John Madden. "They want you to shoot
a wide shot, they do not want you to cover it in one shot, they do not want you to say,
'I don't need that reverse.'"49 However strong tradition maybe, though, filmmakers still
have choices about how they will utilize the options available to them. Working with
shared technology and traditions, directors of early CinemaScope films differentiated
themselves. Hong Kong filmmakers of the 1980s did the same.
By invoking means-ends reasoning and the options offered by tradition, I don't
want to suggest that filmmakers brood deeply on every decision, that there's no spon-
taneity or flash of insight guiding their creative choices. When we want to mount a
causal account, starting with assumptions about rationality and craft practices can
serve as a methodological default. All other things being equal, and as a point of
departure, it's fruitful to assume that a filmmaker makes choices in order to achieve
some purposes, as those might be defined by the tradition in which the filmmaker
works. Moreover, the rational agent model doesn't rule out lucky accidents or flights
of inspiration. Where creative ideas come from is fairly mysterious, but once the artist
has the idea, she will make choices about how to integrate it into the work at hand, and
these will be inflected by means and ends, purposes and patterns. A burial scene in
Red River (1948) is enhanced by a cloud passing over the assemblage. The filmmakers
took advantage of the wayward cloud because it helped fulfill the purpose they had
in view for the scene.
Treating norms and craft practices as traditions also implies that some conti-
nuity underlies changes we might observe. In this respect, I suppose, the historical
side of poetics is conservative, often trying to remind people that things that seem
brand-new almost always proceed from longer-lived conventions. If we call Crash
(2005) and Happy Endings (2005) "hyperlink narratives," we're implying that their
formal principles are pretty new, arising from recent technological changes like the
Internet. 50 But when we look closely, those principles are revealed to be modifications
(sometimes slight ones) of norms that have been used in cinema for decades. In a later
essay, where I call such movies "network narratives," I argue that social networks,
which have always been with us, have recently become salient for reasons having to do
with culture and pressures within the film industry. The films rely on long-standing
traditions of multiple-protagonist plotting, making it possible for audiences to track
the action easily. The motto of historical poetics might be that of Shakespeare's Lear:
Nothing comes of nothing.
30 Poetics of Cinema
A n Excursion on Reflections and Zeitgeists
All my talk of conventions and practices and individuals acting within institutions runs
afoul of some long-standing intuitions. For many educated people, the most important
questions about cinema revolve around its relation to culture. The persistence of this
concern is itself puzzling to me. In no other domain of inquiry I know, from the history
of science and engineering to the history of music, literature, and visual art, is there such
unremitting insistence that every significant research project must shed light on society.
Scholars can freely study iambic pentameter, baroque perspective, and the discovery of
DNA without feeling obliged to make vast claims about culture's impact on said subjects.
Is cinema important and valuable solely as a barometer of broad-scale social changes?
In any event, many will suggest that the framework sketched here seems oblivious
to the ways in which films reflect their cultures. How can I forget that social anxieties,
economic crises, and cultural tensions govern the form and content of movies? Remark-
ably, nearly everybody believes this. The idea shapes the Sunday New York Times
think piece about how the movies of the last few months capture the current Zeitgeist.
It informs the belief that we can define periods in American popular art by presidential
eras—Leave It to Beaver as cozy Eisenhower suburban fantasy, Forrest Gump (1994) as
an expression of Clintonian post-Cold-War isolationism. 51 Reflectionism may be the
last refuge of journalists writing to deadline, but it also underlies a great deal of what
academics pursue under the rubric of cultural studies. That mass entertainment some-
how reflects its society is, I believe, the One Big Idea that every intellectual has about
popular culture. Yet there are good reasons to be skeptical of it.
It's commonly felt that cinema, being a popular art, tends to embody the attitudes or
emotions of the millions of people living in a society. Yet this argument needs shoring
up, because it easily becomes circular. (All popular films reflect social attitudes. How
do we know what the social attitudes are? Just look at the films!) We need indepen-
dent and pretty broadly based evidence to show that some deep needs of the audience
exist and are being addressed by a film. Just because Spiderman (2002) was a huge
success doesn't automatically mean that it offers us access to America's national mood
or hidden anxieties. People spend time with a piece of mass art for many reasons: to
kill an idle hour, to meet with friends, to find out what all the fuss is about. After the
encounter, consumers often dislike the artwork to some degree, or remain indifferent
to it. Because people must buy the work before they experience it, there can't be a
simple correlation between mass sales and mass mood. You and lots of others may be
suckered into going to a film you dislike, but just by going you've already been counted
as among those who support it. Doubtless many people enjoyed Spiderman. But it s
very difficult to say why, at least if we want to move beyond claims about strategies
of storytelling and cinematic presentation. And did all of the patrons enjoy it for the
same reasons? That remains to be shown, and it's hard. We know that a movie may
appeal to several audiences at once, packaging a range of appeals. Must we find reflec-
tions of cultural needs in every aspect of a movie that might appeal to someone?
A primary explanatory prop for reflectionists is politics. Talk about an American
film o f t h e 1950s, and sooner or later you'll invoke the reign of blandness that was the
Poetics of Cinema 31
Eisenhower administration. But why do we assume that America's mind-set switches
its course whenever a new president is elected? What percentage of the electorate
votes? And what percentage goes to movies regularly? And do these demographics
overlap? It's well-known that a large slice of the audience since the 1960s consists of
people too young to vote. So how are anyone's anxieties about presidential policies
being reflected in the works the kids consume?
In addition, the reflectionist typically ignores the range of incompatible material
on offer. If 1940s film noir reflects some angst in the American psyche, how to explain
the audience's embrace of sunny MGM musicals and lightweight comedies in exactly
the same years, indeed on the same double-feature program as a murky noir? The year
1956 saw the release of The Ten Commandments, Around the World in 80 Days, Giant,
The King and I, Guys and Dolls, Picnic, War and Peace, Moby Dick, and The Searchers.
Pick any one, find some thematic concerns there that resonate with contemporane-
ous social life, and you have a case for any state you wish to ascribe to the collective
psyche. But take any other film, or indeed the industry's entire output, and you have
a problem. The alternatives are to find common themes of an insipid generality or to
float the rather uncompelling claim that several hundred films reflect many different,
and contradictory, facets of the audience's inner life.
Moreover, reflectionists have always been reluctant to offer a concrete causal
account of how widely held attitudes or anxieties within an audience could find their
way into artworks. This is one reason that the usual invocation of presidential terms
is unsatisfactory. Through what specific causal processes could changing the occu-
pant of the White House affect popular culture? How exactly does a party platform
or a candidate's charisma get translated into Hollywood movies for the multitudes?
Furthermore, if there ever were a dominant mood at large in the land, it would be
very difficult for that mood to be expressed in a current movie. There's often a lag of
several years before a script finds its way to the screen; many of the films released in
1997, though read as responding to current crises, were bought as projects in 1993
and 1994. More important, movies are made by particular people, all with varying
agendas, and they are inevitably going to shape the initial project in particular ways.
Thus the preoccupations of the screenwriter, the producer, the director, and the stars
rework the given idea. And these workers, we are constantly reminded, are far from
typical, living their superficial lives in Beverly Hills. How can the fears and yearnings
of the masses be adequately "reflected" once these atypical individuals have finished
with the product? It now seems likely, for instance, that the violence in American
films "reflects" not the taste of the mass audience but the egos of the makers, who
enjoy the bravado of seeming to push the envelope.
In sum, reflectionist criticism throws out loose and intuitive connections between
film and society without offering concrete explanations that can be argued explicitly.
It relies on spurious and far-fetched correlations between films and social or political
events. It neglects damaging counterexamples. It assumes that popular culture is the
audience talking to itself, without interference or distortion from the makers and the
social institutions they inhabit. And the causal forces invoked—a spirit of the time,
32 Poetics of Cinema Poetics of Cinema 33

a national mood, and collective anxieties—may exist only as reified abstractions that
the commentator turns into historical agents. 52
It comes down, very often, to realizing that large-scale events need not have
large-scale and distant causes, and small and medium-size events can have small
and proximate causes. After 1920, the deep-space tableau style on display in films
like Ingeborg Holm (Figures 1.1-1.2) went extinct, and an American-style continuity
cinema came to the dominance it still enjoys. There's no need on the face of it to posit
worldwide social upheavals as the direct cause for the new style. Some large-scale
events, such as World War I's damaging effects on European film industries, surely
served as preconditions, but those didn't directly cause the aesthetic changes. More
Figure 1.4 Rope (1948): The first "invisible" Figure 1.5 . . . disguises a change in the
proximate causes included the renewal of national film industries, the saliency of new cut from Brandon's b a c k . . . camera reels, which will be spliced together
templates of cinematic storytelling (such as the feature-length fictional narrative), for projection on a bigger reel.
and the emergence of a younger generation of filmmakers attuned to what seemed
cutting-edge technique. These and other more proximate factors go a long way toward
explaining the worldwide absorption of continuity premises. Likewise, the style has how two gay college men, Bfandon and Phillip, garrote their friend David and hide his
remained constant in its essentials for about a century, in the face of profound social, body in a decorative chest. They then invite his friends and relations to a dinner party
political, and economic changes—largely, we suppose, because it continues to fulfill at which the buffet is arrayed, ghoulishly enough, on the chest. The killers' former prep
functions that filmmakers deem worthwhile. school teacher Rupert eventually realizes their crime and reveals it to the police.
This isn't to say that society has no impact on films. Of course it does. But that What made everyone keen to see Rope was its reputation as a technical tour de
impact isn't single or simple. 53 I'm proposing that causal explanation in poetics can force. Before production Hitchcock had announced that each shot would last nearly
best proceed in steady steps, moving from the artwork to the proximate conditions of a full camera reel. A camera reel normally held 1,000 feet of film, but the Technicolor
production (agents, institutions, and communal norms and practices). These in turn camera's maximum capacity was 952 feet, and some of that had to be wasted getting
may be influenced by both immediate social causes and longer-term preconditions; the camera up to speed, so the maximum length of a take for Rope would be around
we have to look and see exactly how. That Japanese films of the Pacific War period
10 minutes. The publicity for the film had made it seem that every shot was about
were shaped by political demands is undeniable, but the works of Ozu, Mizoguchi,
the same length. Each take averaged 925 feet," proclaimed an article in American
and their peers reworked the assigned materials in distinctive ways. Likewise, the
Cinematographer, which also mentioned that the scenes were rehearsed in sections of
vague demands of socialist realism in the USSR were fulfilled in ways that both reject
about 9 minutes each. 54 For decades thereafter, the 80-minute film was described as
and rework Soviet montage norms of the 1920s. In any instance, a social command
consisting of eight 10-minute takes.
will be mediated by the film industry, existing traditions, and the varied ingenuity of
Things are more complicated than that, because the film contains 11 shots, and
filmmakers. Similarly, long-standing social attitudes, such as racism or homophobia,
their lengths vary considerably. Just as important are the ways in which the shots
supply stereotypes, but those can be transformed by the process of production and
are connected. Rope begins with a comparatively brief high-angle shot establishing
the dynamics of the particular film. Poetics is in a good position to show how that
the street outside, seen under the credits, before the camera pans to a window and
works. It reminds us that themes will be recast, form and style will transform social
we hear a cry. The next shot is a very long take, but the shots that follow aren't all the
givens, and filmmakers will still choose among importantly different ways to tell the
same length, and most run significantly shorter than 10 minutes. More striking is the
story. Ideology doesn't switch on the camera.
fact that some cuts are disguised by stretches of blackness, typically when the camera
moves up to a character's back and away again. Yet other cuts are conventional eyeline
From Shriek to Shot matches: when a character looks offscreen, we get a cut to another character.
I've suggested that all the constructional factors are connected, that inquiring into Is there a pattern to these particulars? After the external establishing shot, there's
particulars and patterns can lead us to principles, purposes, and practices. To illus- a cut inside to the parlor, where Brandon and Phillip are strangling David. This shot
trate this, let me provide a tentative example. Before the arrival of videotape, Alfred plays out for several minutes while the two young men jam their victim into the chest,
Hitchcock's Rope (1948) was a mythical beast, unavailable in 16mm and seldom shown close it, and start to talk. At the shot's end, we get a "hidden cut" as Brandon's back
on television. On a trip to a film archive in the mid-1970s, I watched it, eager to enjoy blocks the frame and a new shot starts (Figures 1.4-1.5). But at the end of that shot,
what had become celebrated as one of Hitchcock's greatest experiments. The film tells number 3, there's a conventional eyeline-match cut. Kenneth looks off left—cut—a
34 Poetics of Cinema
Figure 1.6 Rope: The first visible cut, from Figure 1.7 . . . to the two killers leaving.
Ken...
shot shows the two killers walking to the door (Figures 1.6-1.7). The rest of the movie
follows this pattern. A long take ends with a blackout cut, then the following take ends
with a visible cut. To get schematic about it, with slashes indicating the visible cuts:
Shot 1 (the exterior)/
Shot 2—blackout cut—Shot 3/
Shot 4—blackout cut—Shot 5/
Shot 6—blackout cut—Shot 7/
Shot 8—blackout cut—Shot 9/
Shot 10—blackout cut—Shot 11
But this pattern raises a question. Why disguise some cuts and not others? A viewer
might mistake the shots linked by blackouts as all one take, but there was no effort to
hide their neighbors (3/4, 5/6, 7/8, and 9/10). Why not black out all the cuts?
The answer lies in exhibition practices. 55 Today a film is usually mounted on a
single big platter and run through one projector. But before the 1980s, theaters used
two projectors, with the projectionist switching between machines to project one
reel after another. Although camera reels held at most 1,000 feet of film, projection
reels held 2000 feet (a maximum of 22 minutes running time). Knowing that the film
would be projected on five double reels, Hitchcock shrewdly created the blackout cuts
(linking shots 2-3, 4-5, 6-7, 8-9, and 10-11) for shots that would be spliced together
on a reel. But there would be no similar way to disguise the reel change from one pro-
jector to another. Then, too, the start of each 2,000-foot projection reel would suffer
some wear and tear, so hidden cuts wouldn't survive repeated projections. Hitchcock
reconciled himself to presenting visible cuts between the shots that would be run on
different projectors. 56
Like many artists, Hitchcock submitted himself to fairly strict constraints to see
what he could make of them. He created, we might say, fresh problems in order to find
idiosyncratic solutions. But production practices and projection technology governed
his choices only in an external sense. That is, nothing about camera magazines and
projector reel changes dictated his finer-grained decisions about precisely how long
Poetics of Cinema 35
Figure 1.8 Rope: Phillip's wounded reply Figure 1.9 . . . Rupert watching thought-
that he wasn't afraid to strangle a chicken fully. It's the beginning of his suspicions,
yields a cut to . . . and the cut marks a shift in point of view,
attaching us to him during the next phase of
the film.
a shot should last or what should be shown at the end of one shot and the beginning
of another. A detailed analysis of this remarkable film would take me too far afield
here, but we might notice that Hitchcock clearly didn't use the "10-minute take" as an
invariable yardstick. Even putting aside the establishing shot under the credits (a little
over 2 minutes), the long takes vary in length quite a bit. Three shots run approx-
imately 10 minutes, five last between 7.15 and 8.11 minutes, the introductory shot
runs a little over 2 minutes, and the last two shots are comparatively brief, running
4.6 minutes and 5.6 minutes. 57 Miklos Jancso, in Sirocco (1969) and Electro. (1974),
sustained each shot until the camera reel nearly ran out, but Hitchcock timed his cuts
to articulate the unfolding drama. 58
The most evident instances are the eyeline-match cuts. Hitchcock gives this
normalized device a fresh force, not only because any cuts at all are rare but also
because these create a powerful progression. The first shot change is low-key, when
the guest Ken is left bewildered by Brandon's casual suggestion that tonight Ken might
rekindle a romantic spark in his former girlfriend Janet. Ken watches Brandon and
Phillip walk away assuredly (Figures 1.6-1.7). The next eyeline-match cut, between
shots 5 and 6, is more dramatically charged. Phillip has just blurted out that Brandon's
account of him killing chickens is a lie, and the cut takes us to their teacher Rupert,
watching appraisingly (Figures 1.8-1.9). Hitchcock puts Brandon and Phillip's ensu-
ing quarrel offscreen as we are allowed to study Rupert's reaction, a mixture of
bemusement and wariness. This cut launches the central portion of the film (and the
third projection reel), when Rupert's suspicions steadily grow. During this reel and
the next, the attached point of view shifts from the murderous couple to Rupert, who
scrutinizes them and questions the maid, Mrs. Wilson. The next eyeline cut occurs
when Rupert and his two pupils look offscreen in reply to the maid's announcement
of a phone call. Just before the cut, Rupert is telling the killers that there's something
36 Poetics of Cinema
Figure 1.10 Rope: Phillip's hand clutching Figure 1.11 . . . marks another high point,
the revolver . . . as Rupert realizes that his pupils are capable
of violence.
going on that's upsetting them a great deal. The cut marks another stage in Rupert s
growing sense that they've done something reprehensible.
The last visible cut forms a kind of climax. Rupert has just been imagining how the
murder might have been enacted, with the camera tracing the path of the action, as
if following his gaze. The framing comes to rest on Brandon's pocket, where his hand
clutches a pistol (Figure 1.10). Cut to Rupert, staring (Figure 1.11). Across the film,
the shot joins have set up an internal norm, with the eyeline match presumed to be
the prime linkage device. All these cuts show symmetrical variation too. The first and
third cuts are motivated by a glance at the end of a shot. In the second and fourth cuts,
the object of a glance—a distraught Phillip, a pistol in a pocket—ends the shot, and
the follow-up starts with a character or characters looking off. In the first cases, the
cut is perhaps somewhat more predictable, because a close view of a character's look
sets up the expectation of an eyeline match. In the other instances, the cut becomes
more unexpected and interruptive. After all, either the close-up of Phillip or that of
the pistol could easily be part of a sustained shot panning among the characters.
A parallel progression is provided by the blackout cuts. The first one (shots 2-3)
and the third one (shots 6-7) are motivated by tracking in and panning past Brandon's
back, with his jacket blotting out the cut (Figures 1.4-1.5). The second blackout exe-
cutes the same maneuver, this time using Ken's back (shots 4-5). Again, an internal
norm is set up. Even the viewer who isn't keeping strict track will probably come
to expect that these somewhat contrived blackouts will be part of the film's stylistic
unfolding. Unlike the eyeline matches, the first two cuts don't mark distinct phases
of the drama; they all continue scenic action fairly fluently. The third blackout cut
is more marked, occurring as Rupert mentions David's absence to Brandon. Over
the black frames we hear Rupert say, "As a matter of fact, I'm beginning to miss him
59
myself." The final blackout cut gathers the most force. When Rupert realizes that
David's body may be in the chest, he pulls up the lid and the camera tracks abruptly
forward. As the chest fills the screen, the image goes dark. Coming out of the transi-
tion, the camera lifts over the edge of the lid to reveal Rupert's sickened expression
Poetics of Cinema 37
Figure 1.12 Rope: Rupert lifts the lid and
discovers David's body in the chest.
(Figure 1.12). At this climax the blackout gains visceral impact, suggesting what can't
be shown and providing a purely graphic thrust from light to darkness to light, as
Rupert's stricken face fills the frame. The shots linked by this ominous blackout are
the film's shortest ones, accelerating the film's denouement.
We've identified some localized patterns and functions, but what's the broader
purpose? Why did Hitchcock go to all this trouble to sustain lengthy takes? Why
try to make a film with so few cuts? The choice is particularly odd in that Hitchcock
had long taken pride in his mastery of editing. He famously proclaimed, "If I have to
shoot a long scene continuously I always feel I am losing grip on it, from a cinematic
point of v i e w . . . . What I like to do always is to photograph just the little bits of a scene
that I really need for building up a sequence." 60 Until Rope, Hitchcock indulged in a
flashy long take now and then (Young and Innocent, Notorious), but across a whole
film his cutting rate tended to be fairly fast. Most Hollywood features of the 1940s
had an average shot length (ASL) of 8-11 seconds. Sometimes Hitchcock's work fell
in that range, but he was also inclined toward a more brisk decoupage. 61 Foreign
Correspondent (1940), Saboteur (1942), Lifeboat (1944), and Notorious (1946) all have
ASLs falling between 6 and 7 seconds. Hitchcock's producer, David O. Selznick,
thought that his films tended to be "cutty" and sometimes tried to slow the edit-
ing pace by replacing Hitchcock's single close-ups with more sedate two shots. 62 In
1947, The Paradine Case (1947), which Selznick recut, averaged 7.3 seconds per shot.
The following year, Rope's shots averaged 7.3 minutes. So again, why did Hitchcock
change his style so radically?
Probably several factors worked together. Hitchcock claimed that the long-take
procedure saved money, not an unimportant element at a moment when he was
considered a somewhat budget-straining director. On Rope, his first venture as
an independent producer, he may have wanted to show that he kept an eye on the
bottom line. Still, throughout his career he had presented himself as fiscally efficient,
largely because he claimed to plan each film out on paper to the smallest detail. Press
accounts of the tension-filled Rope set, in which any missed cue or bungled line forced
the whole production to start the shot over, suggest a precarious undertaking that a
38 Poetics of Cinema
prudent man wouldn't try. In any event, from the standpoint of historical poetics we
can propose that another factor was at work.
New norms were emerging in the 1940s, and some directors were making flam-
boyant use of them. Welles' Citizen Kane (1941) had treated long takes not just as
stylistic flourishes but also as dramaturgical building blocks. A scene might consist
of a string of long takes, or one long take alongside a few briefer shots. On rare occa-
sions the scene consisted of just one long take, what the French came to call the plan-
sequence or single-shot sequence. In Kane's follow-up, The Magnificent Ambersons
(1942), a late-night kitchen conversation consumes a single shot lasting 5 minutes.
Once the self-conscious long take was on the agenda, some filmmakers tried to test the
resources of the device. Otto Preminger, George Cukor, Vincente Minnelli, and others
made the long take a cornerstone of their filmmaking. Joseph Mankiewicz's The Late
George Apley (1947) has an ASL of 16 seconds, whereas Preminger's Fallen Angel (1945)
clocks in at 33 seconds. The most flagrant experiment along these lines was the pseudo-
subjective movie The Lady in the Lake (1947), which averages over 2 minutes per shot.
As if not to be outdone by such upstarts, Welles' Gothic Macbeth (1948) presents the
murder of King Duncan in a single camera reel running nearly 11 minutes. 63
Most of the long-take films of the period also made use of the mobile camera.
Again, Welles' flamboyant style made the option prominent, notably in the ballroom
scenes of Ambersons. A few years later, intricate tracking shots were made easier
thanks to several new cranes and dollies, notably the crab dolly; its tight turning radius
allowed the camera to spiral around a prop or an actor. Observers inside Hollywood's
technical community began noticing the "fluid camera" technique. Joseph LaShelle,
who shot Laura and Fallen Angel for Preminger, was praised for
breaking a scene down into various forceful compositions and joining these
different 'points of view' together through smooth camera movement.... On the
screen a close-up gives way to a long shot which then evolves into a follow shot. 64
This is exactly what some observers would claim that Rope does—translate ortho-
dox editing patterns into panning and tracking movements that connect distinct
camera setups.
We tend to think of Hitchcock as a self-motivated innovator, but he seems to have
been highly sensitive to what his peers were up to. He sometimes adopted the aggressive
foregrounds and deep staging popularized by Kane (Figure 1.13). Likewise, Selznick
encouraged him to shoot longer takes, and it seems likely that the emerging "fluid
camera" aesthetic aroused the competitive instincts of Hollywood's most self-conscious
experimentalist. The film Hitchcock directed just before Rope was The Paradine Case
(1947), for which Morris Rosen, the head grip, devised an early crab dolly. (Rosen would
operate the camera boom on Rope.) The Paradine Case included several lengthy shots,
and one was trumpeted in a technical journal as the "Three and a Half Minute Take"
(though it didn't survive Selznick's final cut).65 Hitchcock's rationale, as paraphrased by
one reporter, amounted to a repudiation of his cutting-based aesthetic. "The big advan-
tage gained artistically is the simulation of stage continuity.... Too often in the past,
[Hitchcock] believes, a good dramatic picture is hampered by too frequent cuts, not
Poetics of Cinema 39
Figure 1.13 In Notorious (1946), Hitchcock Figure 1.14 Striking frontality and compact
works a variant on the aggressive foregrounds arrangement in Lifeboat (1944).
of Wyler and Welles.
enough continuous action."66 Hitchcock likewise defended Rope's style on the grounds
that it preserved a stage play's fluidity, but he acknowledged that the technical chal-
lenge was another attraction for him. "Right from the beginning," recalled screenwriter
Arthur Laurents, "he'd tell me that he was going to do it as a play, and with, I think,
nine takes, or nine reels—something like that. And that interested him because that
hadn't been done."67 Hitchcock, in effect, created a new set of problems for himself.
He explained to Francois Truffaut, "I undertook Rope as a stunt."68
The stunt blends other Hitchcock preoccupations. In previous American films he
experimented with what we might call the floating close-up, a prolonged tracking shot
attached to a player's face. This satisfied Selznick's concern for sustained takes and
glamorous portraits of female stars, 69 but Hitchcock tended to use them for suspense,
often locking them within editing patterns that indicated the character's moving point
of view. He also began huddling his characters close to one another in a tight medium
shot, often turning all their faces to the camera in a quite artificial way (Figure 1.14).
This staging strategy is very salient in Rope (Figures 1.15-1.16), largely because with-
out cutting Hitchcock can't easily alternate over-the-shoulder reverse angles. Further-
more, Ropes sequences were shot in story order. Hitchcock had experimented with
this tactic in The Paradine Case. He shot the film chronologically, and several of the
courtroom exchanges were filmed in real time with four cameras running simultane-
ously.70 Similarly, Hitchcock had already explored the possibility of restricting a film's
action to one setting—in a partial way during the train scenes of The Lady Vanishes
(1938), and more systematically in Lifeboat (1944). Rope offered a chance to try the idea
again. According to a contemporary report, "The idea [of a long-take film] had been
one of Hitchcock's pet dreams for a long time. But he needed a story that had no time
lapses, and a story that took place on one set."71 Patrick Hamilton's play, whose three
acts aren't broken by time gaps, became the basis of a bravura synthesis of long takes
and camera movement within a single locale, all presented in strict continuity.
True to Hollywood's alibi for formal experiment—the story is all—Hitchcock
claimed at the time that the viewer shouldn't notice the outre technique. "The
40 Poetics of Cinema Poetics of Cinema 41

Figure 1.15 Rope: In the absence of over-
the-shoulder staging, the characters huddle
frontally.
Figure 1.16 Rope: Characters strung out
horizontally. The depth relations couldn't be
as baroque as in Figure 1.13 because lighting
for color film didn't permit the extreme focal
distances available for black and white.
Figure 1.19 . . . and settles on the two killers,
in a long-take version of the eyeline match-
ing that has linked shots through cuts earlier
(Figures 1.6-1.11).

Figure 1.17 Rope: At the climax, Rupert Figure 1.18 Rope: The searching camera
unfolds the rope while commenting that he scans the room as we hear Rupert speak of
won't ride with his pupils. suspense...
audience must never be conscious of i t . . . . Hie result I'm after is to excite the audience
by making the picture flow smoother and faster." 72 Yet the blackout cuts call atten-
tion to themselves, and several critics of 1948 pointed out that what Hitchcock called
the "roving camera" could be quite obtrusive. 73 Some moments flaunt the prolonged
take quite explicitly. The camera's anxious probing of the apartment during Rupert's
voice-over replay of the murder is one instance (another device Hitchcock had tried
out earlier, when Maxim recounts his wife's death in Rebecca). An even more overt
passage comes when Rupert fishes the rope out of his pocket, turns to his pupils, and
extends it (Figure 1.17). Over a close-up of his hands, we hear his voice say, "Driving
with you and Phillip now might have an additional element o f . . . suspense." Before
the last word has been spoken, the camera starts to glide rightward away from him,
passing a corner and a window (Figure 1.18) before settling on the faces of Brandon
and Phillip (Figure 1.19). The refusal to cut obliges the camera to traverse the space
completely and make us wait for the men's response.
In an ordinary film, using a tracking shot to postpone their reactions would seem
ham-fisted, but such self-initiated camera movements, independent of character
movement, have become prominent in the later phases of this film, so this tactic seems
a logical culmination. Naming the shot's overt purpose—to provide "an additional
element of suspense"—on the dialogue track marks the device even more explicitly.
It might seem merely a critic's fancy that the lethal rope is a kind of emblem for the
stretched-out take and the unwinding camera movements, but a poster advertising
the film reinforces the link. The taglines play up the opening, the closing, and the
sinuous continuity in between: "It begins with a s h r i e k ! . . . It ends with a shot! From
beginning to end, nothing ever held you like Hitchcock's Rope."
Having made the longest-take movie in studio history, Hitchcock could afford
to let up a bit. His follow-up, Under Capricorn (1949), averaged 44 seconds per shot
(placing it second, I believe, in Hollywood's long-take sweepstakes). The most sus-
tained shot, when the distraught Henrietta tells the protagonist of the secrets in her
past, runs about 8.5 minutes. Although that sequence is relatively stationary, relying
on subtle reframing and refocusing, for other shots Hitchcock posed himself new
problems, tracking characters through several rooms and up and down staircases. 74
Under Capricorn often uses editing to ratchet up the dramatic tension, as we saw
with the eyeline matches in Rope, but it also revives the concealed-cut device, giving
a subliminal sense that a shot runs even longer than it did on the set. Across film
history, directors tend to love long takes, but producers hate them because scenes
can't be tightened up in postproduction. With The Paradine Case, Hitchcock was at
Selznick's mercy, but he produced both Rope and Under Capricorn himself and was
able to indulge his experimental ambitions. Naturally, he made sure that the films'
bravura techniques were fully covered in the press, both popular and professional.
It would be worth studying how the problem of filming theater was being rethought
by several other directors at the time. Bazin wrote a brilliant essay about this devel-
opment, suggesting that Laurence Olivier's Henry V (1944) and Jean Cocteau's
Les Parents terribles (1948) had devised various cinematic equivalents for the expe-
rience of staged drama. Bazin noted that Olivier starts his film with a performance
before an audience, whereas Cocteau confines his camera to an apartment in an
42 Poetics of Cinema
effort to suggest the suffocation of the play's single set.75 One might add Jean-Pierre
Melville's Les Enfants terribles (1950), which flaunts theatrical interpolations like a
descending curtain. Wyler's The Heiress (1949) and Detective Story (1951) offer less
flamboyant but no less intriguing instances. Like Bazin, Wyler speaks of trying to
steer a middle course between simply photographing a stage play and opening it up
so broadly that it loses its theatrical flavor. In Detective Story, "I did not change the
construction of the play; I changed the set. Instead of having two little rooms, as in
the play, I have five or six rooms." 76 Hitchcock made an effort along similar lines with
Dial Mfor Murder (1954). Because "the film will have to follow the play very closely,"
he explained, "I am treating it in a modified Rope style."77 Dial M boasts an ASL of 9.1
seconds (the same as Rebecca and Spellbound) and trim, efficient staging. Hitchcock
obeys the Bazin-Wyler dictum, opening up the play only a little by showing the apart-
ment's terrace, bedroom, and outside hallway.78 The absolute confinement of Rope
had defined one pole of theatrical cinema, and most directors, including Hitchcock
himself, retreated to a middle way.
As for the long take, Hitchcock and most of his peers abandoned it as a structural
unit in the 1950s. Yet the dream had great staying power. The neorealist screenwriter
Cesare Zavattini imagined 90 minutes of real life presented in a single shot. 79 Many
people believe, mistakenly, that Andy Warhol films like Sleep (1963) and Empire (1964)
consist of static takes many hours long. Aleksandr Sokurov's Russian Ark (2002) uses
digital video (and an orchestration of action at least as intricate as that in Rope) to
capture uninterrupted time. Josh Becker's low-budget feature Running Time (1997)
resorts to hiding its cuts in the Hitchcock manner. In the 1940s, filmmakers began
to acknowledge shot duration as a formal parameter, and we might conjecture that
Hitchcock, like many of his peers, took the long take as a challenge, an occasion to
reshape contemporary norms of cinematic storytelling.
I could have introduced this brief analysis of Rope's decoupage by identifying the
general problem Hitchcock set himself, that of a feature film presented in something
approximating a continuous shot. I could then have discussed the functional con-
sequences of this goal, including the new problems it posed (the absence of reverse
angles, the need to allow reel changes) and the rational solutions that Hitchcock
found (frontal staging, hidden cuts). Instead, just to give the flavor of an inquiry into
poetics, I tried to show how the analyst can frame and revise questions that move
among several pertinent aspects of a film, from details to patterns, from functions to
principles, and from internal dynamics to historical context.
I wanted as well to highlight how the rubric of practices includes the institutional
forces at work, like production and projection routines, and the technology employed,
such as standardized reel lengths. A further lesson here is that practices include
informal relations among personnel, such as the urge to show one's skill. Filmmakers
may address their work to their peers as well as to their public. No historian of painting
would be surprised to learn that artists compete in displaying their virtuosity. We can
explain important aspects of how movies work by considering filmmakers as creative
agents working with craft practices within a community. Members of that commu-
nity may be sharply aware of traditions and trends. They may replicate well-tried
Poetics of Cinema 43
norms or explore emerging ones. They may solve problems in routine ways or pose
new difficulties in order to triumph over them. And some of the most ambitious and
gifted creators are likely to treat constraints as opportunities.
W h a t Snakes, Eagles, and Rhesus Macaques
Can Teach Us
I once projected a kung fu film, Snake in the Eagle's Shadow (Yuen Woo-ping, 1978),
in a Cantonese-language print lacking both English dubbing and subtitles. The
question was, How much of the film could the audience grasp without knowing its
native language?
Actually, quite a lot. It might seem too obvious to mention, but we in the audience
perceived the film. In Snake in the Eagle's Shadow, we saw patches of color, patterns of
light and dark, trajectories of movement, and changing shot displays. But our percep-
tion wasn't really of abstract configurations. Humans evolved to detect objects and
actions in a three-dimensional world, and in watching Snake we definitely recognized
things. We saw young and old men and women, all going about activities in a volu-
minous space. We saw a youth involved in social interactions, in locales—a village,
a clearing—that we could recognize, at least generically. We heard speech, and if it
had been in English, we could have grasped it as quickly and involuntarily as we
grasped the sight of a human face. We heard noises, such as the blow of fists on flesh,
or labored breathing, or the sound of a cobra hissing. We heard music, mostly in a
tradition we recognized, and it registered as such.
But also, and more interestingly, we viewers of Snake in the Eagle's Shadow under-
stood a lot of the story. We understood that the protagonist was a young servant in
a kung fu school. He wants to learn martial arts and meets an old man who, despite
his shabby appearance, is a master fighter. The youth undergoes arduous training and
eventually comes to defeat a villainous master. These features aren't simply given in per-
ception; we had to bring in large domains of knowledge to arrive at this story. Viewers
who were familiar with the kung fu genre could structure the film along familiar
lines, but even those who weren't martial arts fans understood a good deal of the
action because they had skills in understanding any type of story. At one point, when
the protagonist sees a cat fight a cobra, all of us realized that he was inspired to model
his kung fu technique on the cat's attack. Call our activity of this sort comprehension,
a grasp of the concrete significance of the perceptual material as patterns of social
action. In this case, the patterns are presented in the form of a story.
Finally, spectators used the film in various ways. Snake in the Eagle's Shadow
wasn't intended to be shown in a college classroom, but I drew it into my own agenda.
Some of the students took the film as an occasion to celebrate the prowess of Jackie
Chan. Others took it as proof of the artistic bankruptcy of Hong Kong cinema. Some
students from Hong Kong read it as a statement of local pride in the face of adversity.
Those who practiced martial arts themselves spotted techniques that they could try
themselves. Let's say that all of us appropriated the film, in however disparate ways.
44 Poetics of Cinema
These types of activity suggest how poetics can address what I'm calling the
processing of films by viewers. If poetics is concerned with how filmmakers use the
film medium to achieve effects on spectators, we ought to have some idea of how those
effects might be registered. Film researchers aren't psychologists or sociologists, but
we can draw upon the best scientific findings we have to mount a plausible framework
for considering effects. The poetics I propose is thus mentalistic: It assumes that we
can characterize the spectator's embodied mind as engaging with the film. It's also
naturalistic, presuming that scientific investigation of mental life is likely to deliver
the most reliable knowledge. I'd also propose that the best mentalistic and natural-
istic framework we have available is that provided by what we can broadly call the
cognitive approach to mental life.
Adopting this perspective makes some researchers worried. Some object that it
neglects the influence of society, ideology, or culture on viewers. But this is to assume
that a mentalistic and naturalistic framework focuses wholly on individuals. It
doesn't. Cultural activities are mental in an important sense: They're learned, recalled,
rethought, and so on by the embodied minds of social agents. The framework pre-
sumes some inter subjective regularities of mental activity across individuals, but
cultural theorists do the same thing when they discuss how members of a subculture
come up with a resisting reading. Other critics have argued that conceiving of the
spectator in the way I propose neglects the differences of race, gender, ethnicity, and
other markers of identity. Yet clearly there are common effects across such groups;
people of all sorts feel suspense in a thriller and sadness in a melodrama. Studying
such commonalities isn't on the face of it unreasonable or uninteresting. Moreover,
there's always a degree of idealization in discussing spectatorship. Just as linguists
create the idealization of "the native speaker" in order to understand grammatical
principles, virtually all researchers are obliged to idealize the spectator, even the
female or African American spectator, to some degree. Finally, although not every
conceptual framework fits well with every cluster of research questions we might want
to float, I think that some identity differences can be understood from the standpoint
of poetics, as I'll try to show shortly.
We can start to understand the effects of films by borrowing a distinction from
classic cognitive psychology, that of top-down and bottom-up mental processing.
Top-down processing is concept driven; bottom-up processing is data driven. A clas-
sic instance of top-down processing is problem solving. Given a crossword puzzle,
you draw upon your stored knowledge about language and the world (including the
stratagems of crossword puzzle designers) to fill in the blanks properly. By contrast,
bottom-up processing arises from a moment-by-moment encounter with the world.
As you enter an unfamiliar room, for instance, your visual system picks up informa-
tion about edges, brightness differences, and a host of other features that coalesce into
a spatial whole.
Our brains can process information in both "directions" at the same time, so any
particular experience will be a mixture. While searching your memory for the right
word, you get bottom-up information about the crossword puzzle from the written
clues and the array of empty spaces and black ziggurats on the page. Upon entering an
Poetics of Cinema 45
unfamiliar room you quickly rework the perceptual input in the light of knowledge.
Identifying a chair in the shape of a beanbag and a lamp bubbling like lava leads you
to make a higher-level inference about the tastes of the people living there. On the
whole, bottom-up processes are fast, involuntary, cheap in cognitive resources, and
fairly consistent across observers. 80 In an important sense, all TV viewers watching
the horrendous crash of the airliners into the World Trade Center saw and heard the
event in the same way. Top-down processes are slower, more voluntary, more expen-
sive in cognitive resources, and more variable across observers. Having seen the Trade
Towers assault, viewers interpreted its significance in different ways—as an act of war,
as a response to globalization, and/or as a counterthrust to U.S. imperial ambitions.
Perceptual uptake occurs in milliseconds, and for good reasons. Our brains evolved
in situations in which survival demanded reasonably accurate information about
spatial layout and other agents. Consequently, the activity of our perceptual mecha-
nisms is hidden from us; we can't watch our retinal image or our neuronal firings.
And although experimental films like James Benning's Ten Skies (2004) create notice-
able visual effects, like illusory movement, we can't really probe the mental hardware
yielding the experience. Nearly as fast are intuitive judgments, as when we sense that
a person is arrogant or kindly, or when we just know we'll like a class after hearing just
a little of the teacher's opening lecture. Although we can make these judgments in a
few seconds, they draw on stored knowledge and are thus to some degree top-down. 81
Yet even these remain fairly impervious to introspection.
The top-down-bottom-up distinction drastically simplifies a complex process
that would probably be best modeled along several dimensions rather than a single
vertical one. Doubtless neurological research will eventually show that any experi-
ential process involves complicated feedback and input-output among many mental
systems. Take mirror neurons, which can be found in various areas of the brain.
Watching someone lift a heavy weight, either in front of you or on a movie screen,
stimulates some of the neurons in your brain that would fire if you lifted a weight
yourself. Many of these mirror neurons are linked to intentional action on your
part, so that when they fire, you can spontaneously understand the actions of others
as products of their intentions. It seems that we have a powerful, dedicated system
moving swiftly from the perception of action to empathetic mind-reading. 82
This is only one instance of how contemporary research asks us to consider
that many of what we take to be learned or culturally guided mental activities will
turn out to be packed into our biological equipment. Psychological research in the
cognitive paradigm has steadily diminished claims for a blank-slate conception of
the human mind and belief in the unlimited plasticity of human capacities. More
and more activities (e.g., language, recognition of emotional signals, and attribution
of intentions) seem traceable to humans' supersensitive natural endowment. Many
specialized faculties need only triggers from the regularities of our world to lock in
and function at high levels quite quickly. As research goes on, many "higher-order"
activities will probably be revealed as grounded in a rich perceptual system present at
birth but awaiting activation and tuning from the environment. 83
 Poetics of Cinema Poetics of Cinema 47
46
The perception of film as a representation of the world emerges very early in human
A Model of Viewer Activities
development. Many ofthe experiments testing babies' reactions to facial expressions use
Concept-Driven
processes televised images of the mother, which indicates that children spontaneously identify an

II H audiovisual representation of the caregiver.85 Paul Messaris has shown convincingly

that people in cultures without images recognize films and photographs as present-
ing persons, places, and things. 86 The perceptual mechanisms that film engages seem
Appropriation
to be shared with other primates. Experimenters routinely use videos to test chimps,
t * t ' * monkeys, and their cousins, and the results indicate that these creatures identify their

Comprehension I'iU counterparts on the screen. Other experiments suggest that perception is also attuned
to displays of emotion. When rhesus macaques who are unafraid of snakes watch a film
t |t |t of other monkeys shrinking from a snake, they begin to show fear themselves. 87
I A ;
i > A This isn't to say that the processes of filmic perception are innate, as if a newborn
Perception could enjoy Snake in the Eagle's Shadow. Perceptual development unfolds in response
to the environment. Humans and other primates have evolved to be ready for first

t t t t
Data-Driven
processes
y V A A / W V W V
Phenomenal film
Figure 1.20 A schematic m o d e l of t h e spectator's activities. C o n t i n u o u s arrows indicate
the p r i m a r y direction of processing, w i t h d o t t e d arrows indicating a degree of feedback
a m o n g processes.
So I grant the schematic quality of my distinctions. As a first approximation,
however, they can clarify how our minds interact with movies. I suggest that we
can characterize viewers' interactions with films along a continuum of activities:
perception, comprehension, and appropriation. 84 (See Figure 1.20.) Sensory input
drives perceptual processing; perceptual processing feeds into comprehension and
appropriation, in the "bottom-up" direction. Appropriation drives comprehension to
some degree and perception to a lesser degree. There are secondary feedback effects,
too (indicated by the dotted arrows in Figure 1.20), as when the manner of appropria-
tion can recast perception or comprehension. For example, a decision to interpret the
film a certain way can lead us to look more closely at the film and notice or compre-
hend aspects that might otherwise be missed. I'd argue that such feedback systems
can't go all the way down or all the way up, because perception can't in every respect
determine appropriation, and appropriation can't completely reshape perception or
comprehension. Wishing that Thelma and Louise don't die won't make it so.
Perception evolved in large part to give us reliable information about the three-
dimensional world in which we live. Representational films solicit this activity straight-
forwardly: We involuntarily see the world depicted on the screen. We recognize our
conspecifics and their surroundings. We hear noises, music, and language. We also
see movement where there is only a stream of rapidly projected still pictures.
encounters with the world's regularities. By the time people watch movies with any
degree of perceptual understanding, they have developed the capacities to negotiate
the three-dimensional world as well. And perception isn't terribly plastic. Given a
stable input, perceptual capacities will develop along well-marked pathways. No
one learns to see in infrared, or to move as if the world were two-dimensional, or to
differentiate the separate frames of film flickering past. We are not so made.
The activities of filmic perception tend to be neglected by scholars today, but
there's a long tradition of film aesthetics that places importance on the moment-
by-moment effects of composition, lighting, cutting, and the like. From the Russian
montagists through Rudolf Arnheim, Andre Bazin, and Noel Burch, theorists have
paid attention to fine-grained creative choices that structure the viewer's perceptual
uptake. Although strict experiments on filmic perception are welcome, there's a lot
to be said by those of us not wearing lab coats. We can be sensitive to how patterns
and practices of the medium shape such apparently simple strategies as directing the
viewer's attention. Some of the essays that follow try to show ways in which style
shapes our moment-by-moment perception of the flow of images.
In comprehending a film, we construe the outputs of filmic perception as repre-
senting a hierarchical pattern of actions, a conception, or simply a train of sensuous
elements (as in an abstract film). The viewer applies a wide range of knowledge to
make sense of film, segment by segment or as a whole, and to give it some literal
meaning. Narrative comprehension is the clearest instance. In my Snake experiment,
spectators were able to build out of a perceived world a story about an ambitious young
man who wants to master kung fu. Comprehension also comes into play when we're
asked to grasp a cinematic argument or lyric. Comprehension is evidently a matter of
degree; some viewers get more, some get less. In Snake, the viewers who knew Jackie
Chan recognized him as the star and probably hypothesized that he would triumph
through vigorous abuse of his body. Those who didn't recognize Jackie were probably
surprised by the punishment he inflicted on himself. But the fact that comprehension
varies in degree only indicates the extent to which it's a top-down process. Not every-
one has the same set of conceptual schemes.
48 Poetics of Cinema
Again, poetics has a lot to contribute to understanding comprehension. The tech-
nical choices made by filmmakers organize perception in ways designed to enhance
comprehension. Filmmakers design their shots and scenes so that spectators can
follow the movie's large-scale form. Focusing on certain traditions or particular
films, we can study how principles of style, narrative, and the like aim to provide
a distinct experience for the viewer. This will be a central concern of most of the
essays that follow. For instance, since the 1920s most films in most countries have
organized their perceptual surface according to some basic principles. Commercial
storytelling cinema has long followed the conventions of analytical editing: master
shot, followed by a two shot or over-the-shoulder shots, followed by singles highlight-
ing each participant in shot/reverse-shot fashion. In fact, seldom do we find in any
art a style with such pervasive presence and 100-year longevity. These norms have
provided easy, comprehensible ways for narrative action to be understood.
Beyond stylistic patterning, it seems clear that comprehension also rides upon
action schemas that the spectator can activate. When one man slaps another, and
the second responds with a punch, there's not much doubt about what is going on in
the story world at this point: insult and physical conflict. Likewise with a theft, an
abduction, or glances that suggest attraction between man and woman. The large-
scale form of the film is designed to create a flow of cues that ask viewers to apply
schemas for typical situations and h u m a n actions and reactions, locking them into
place quickly. Indeed, there's good reason to believe that these action schemas enable
us to learn the stylistic schemas that present them. We know that people tend to
face one another when they converse, so this regularity of social interaction makes
comprehensible the stylistic option of shot/reverse-shot editing. As Gombrich puts
it, "It is the meaning which leads us to the convention and not the convention which
leads us to the meaning." 88
The perceptual surface can also be so roughened that holistic action patterns become
difficult to grasp. In The Two Minutes to Zero Trilogy (2003-4), Lewis Klahr presents
shaky fragments drawn from comic books, showing only bits of words and imagery of
bank holdups and police chases. The images are so broken up and they shudder past us
so quickly that we never get time to figure out character relations or construct a com-
plete story (Figure 1.21). The slender cues summon up action-based schemas but also
frustrate our efforts to absorb them into scenes and larger narrative patterns.
Comprehension occupies an intermediary place in my framework, balancing
between data-driven and concept-driven features. Appropriation is much more top-
down. Here the viewer uses the film in a more or less deliberate way, drawing it into
her personal projects, and she may stray far from the phenomenal film. I appropriated
Snake in the Eagles Shadow as a classroom example, but fanboys have appropriated
it as a cult object. Films are appropriated by individuals and communities for all
manner of purposes. People employ favorite films for mood management, watching
Die Hard to pump themselves up or Sleepless in Seattle to have a good cry. Bloggers
may use films to flaunt their tastes or strike a posture, whereas academics interpret
films to validate a theory. Social groups appropriate films to a multitude of ends,
treating some as praiseworthy representations of political positions and castigating
Poetics of Cinema 49
Figure 1.21 A bank robbery is evoked in
jittery comic book imagery in The Two Minutes
to Zero Trilogy (Lewis Klahr, 2004).
others as harmful. Filmmakers are always surprised by the range of ways in which
people take their movies.
Accordingly, much of what interests cultural critics are acts of appropriation. Some
Asian Americans have attacked Charlie Chan movies as exemplifying Hollywood
racism, not only in their plots and characterizations but also because the Chinese
Hawaiian detective is played by a Westerner in "yellowface." They have perceived
the films and comprehended them; but they have appropriated them in a way very
different than the makers intended, or could probably have imagined. Much of what
Janet Staiger attributes to "perverse spectators" consists of unusual forms of appro-
priation. She writes,
Knowing that fictional narratives are produced permits many viewers to con-
centrate on narrational issues related to the production of the text. A study of
some 1950s gay male viewers of A Star Is Born (1954) revealed that they were
much more interested in constructing the story of the production of the film
(when did Judy Garland shoot which scene) than in the film's plot—which at
any rate was already "known." 89
It's an interesting fact about films that groups (and individuals) can build unforeseen
inferences out of particular aspects of a film that interest them. Nonetheless, what
Staiger calls "uncooperative spectators" tend to perceive and comprehend the film
in quite convergent ways, as she indicates by saying that the gay audience already
grasped the film's plot.
Staiger's example typifies the tendency of cultural critics to stress divergence
of response among groups. We know as well that there's likely to be considerable
differences among individuals in any group we pick out. Both sorts of divergence are
explicable in the light of top-down appropriation, for in this domain an indefinitely
large range of conceptual schemes can be brought to bear on any phenomenon.
A balanced account will also note the high amount of convergence at all levels.
Academic interpretations display great agreement at the levels of perception and
50 Poetics of Cinema
comprehension, as well as a surprising degree of overlap in interpretations too, as
when one critic revises the reading of her or his predecessors. 90 More importantly,
there are many convergences among spectators at the level of comprehension. All
storytelling traditions evidently deploy such concepts as protagonist, goals, personal
agency, conflict, and causal change—all concepts relevant to comprehension. Patrick
Hogan has shown that some prototypical narrative patterns, such as romantic
91
tragicomedy, are to be found throughout the world's literatures.
This three-stage framework helps us understand the range of control available
to filmmakers and viewers. Critics often ask, How much does "the text" control its
"readings"? This framework lets us give some focused answers. As we move up the
ladder, from bottom-up to top-down processing, the filmmaker's control diminishes
and the spectator's power increases.
By constructing the phenomenal film, the filmmakers control very strongly,
though not absolutely, the viewer's perception of it. It's impossible for a viewer to
perceive the hero of Snake in the Eagle's Shadow as, say, tall and blonde. In addition,
all the options of film style and structure can be mobilized to guide the viewer's notice
to certain material. Framing can center key information, whereas cutting can high-
light a detail. Style operates at all levels, but among its basic tasks is organizing the
stimulus for uptake—even if that uptake is made difficult by an oblique technical
choice or a problematic narrative.
At the level of comprehension, the filmmaker still has a lot of control, because
features of theme and subject, style, and large-scale form are mobilized to guide the
spectator's overarching understanding of the material. Still, no formal pattern can
anticipate every question that can be asked about it. In grasping narrative form, for
instance, the spectator contributes a lot—picking up the cues planted by the film-
makers, as well as inferring, extrapolating, filling in gaps, and the like. Most of this
inferential elaboration is foreseen and governed by the filmmaker, but not all of it
is. Shakespeare famously leaves us with the question of whether Macbeth and Lady
Macbeth ever had children. She claims she has "given suck" to a babe, but Macduff
says, apparently of Macbeth, "He has no children." Did the couple have a child who
died? Did Lady Macbeth suckle another woman's baby? Spectators will differ in
the ways that they deal with such zones of indeterminacy. Most will ignore them,
but others will use them as occasions for appropriation, as when fans write fiction
filling in gaps in Star Trek or The Lord of the Rings. No narrative can avoid leaving
some openings for inferential elaboration of this sort. Louisa May Alcott couldn't
have anticipated Geraldine Brooks' novel March, a fictional biography of the father
depicted in Little Women. Of course, some artworks deliberately introduce gaps that
they decline to fill, as with ambiguous endings.
Clearly, filmmakers have the least control over the activities I've gathered under
the rubric of appropriation. Having perceived and comprehended (to a greater or
lesser degree) Snake in the Eagle's Shadow, you're free to do with it as you will. Viewers
sympathetic to gay rights can take Brokeback Mountain as a plea for tolerance,
whereas those opposing gay rights can treat it as Hollywood propaganda for alter-
native lifestyles. Director Ang Lee and his colleagues can seek to shape the film's
Poetics of Cinema 51
reception through publicity campaigns, but they can't anticipate every way the movie
can be appropriated. The wave of mashup trailers that swept the Web (Brokeback to
the Future, The Empire Brokeback) weren't foreseen by the filmmakers, however much
they may have welcomed them. In sum, as we move up my chart, filmmakers' freedom
wanes and spectators' power increases.
I haven't mentioned a prime component of film's effects: emotions. In my past work,
the historical poetics I've proposed has slighted emotions, leading some people to
think that a cognitive perspective can't tackle such matters. In part, leaving emotion
out of the picture is a simple piecemeal idealization of the phenomenon; studying the
grammar of a joke may not yield insights about what makes it funny. In addition, in
the early 1980s, when the cognitive perspective was hitting critical mass in several
disciplines and when I imported some of its observations into film studies, emotional
matters were set to one side. But they weren't legislated out of existence, and the 1990s
saw vigorous efforts to incorporate emotional life into the cognitive framework.
This was also reflected in film studies, in the work of Murray Smith, Ed Tan, Torben
Grodal, Greg Smith, Carl Plantinga, and many other researchers since.92 It's not an
area of specialization for me, but I think that the processing framework I've proposed
can accommodate emotion as an integral part of a film's effects.
Emotion is part of our evolutionary heritage, and it has largely served in tandem
with cognition. That is, rather than being the foe of emotion, reason has used emotion
and emotions have exploited reason. Certain sorts of reasoning would be maladaptive
without some emotional upsurge that halts thinking and forces action. The hominids
who lingered to investigate whether the stripes glimpsed in the underbrush belonged
to a predator didn't leave as many offspring as those who, driven by fear, simply fled
at first glimpse. Emotions offer quick and dirty solutions to problems that make
thinking risky. Alternatively, so-called commitment emotions may have evolved to
strengthen group bonds, even if they work against self-centered rationality. Fathers
have no rational reason to hang around after a woman is impregnated, but it seems
likely that men who had romantic attachments to the mother had more children
who survived, and so love helps unite father and mother across the lengthy period in
which children grow to self-sufficiency. 93 Within specific cultural contexts, of course,
people learn to judge the proper moments to express feelings, to mask them with
other feelings, or to send emotional signals.
In cinema, I'd suggest that emotion operates at all three of the levels I've sketched
out. Most obviously, acts of appropriation are shot through with emotion. Fans
cherish their favorite movies, critics get worked up in attacking a film they loathe, and
unhappy viewers can wax indignant about a film's moral shortcomings. Less apparent
are the ways in which emotions function in perception. A controversial case would be
our startle response, which can be triggered quite automatically, as when you jump at
a sudden burst of sound in a horror film. Startle isn't a prime candidate for being an
emotion—it seems to prepare the way for the emotion of surprise—but it does lead to
physiological arousal of a sort that primes affect.
More common and central is our sensitivity to emotional signals sent by other
humans. Just as the rhesus macaques recognize signs of distress in their mates in a
 Poetics of Cinema Poetics of Cinema 53
52

movie, we are prepared to grasp many facial expressions. Newborn babies can reliably
read their mothers' smiles, eye movements, and eyebrow play. A film's soundtrack can
arouse us quite directly by cries, bellows, and other signals, just as infants respond to
the mother's coos and baby talk. 94 The weight of the evidence shows that evolution has
primed us to engage in encounters with others by making us sensitive to the slight-
est signs of their emotional states. And these "affect programs" seem cross-cultural
in large part, although we should expect them to vary extensively from person to
person within any given culture. 95 There is some evidence that in-group familiarity
leads to faster recognition of facial expressions; Asians, whether living in Asia or the
United States, recognize emotions on Chinese faces somewhat more readily than Figure 1.22 The first shot of Ohayo (Ozu Figure 1.23 Ohayo: The last shot presents
non-Asians do.96 Yasujiro, 1959) shows a neighborhood and its the same configuration, but from the oppo-
More obvious are the emotions that fund comprehension. As we come to under- clotheslines squatting under electrical towers. site angle. Ozu's typical reshuffling of shot
stand a narrative, we begin to run scenarios that require "emotional intelligence"— elements is here applied to landscapes, empha-
sizing the washed underpants of a boy who
good guesses about how characters will react to the story situations. At the same time, can't control his bowels
we gauge a character's personality or current attitude on the basis of their emotional
responses. Our inferential elaboration of the cues we're given is guided by the emo-
tions that characters register. At the same time, the emotions we feel shape our sense and that modeling in turn rests upon sensitivity to emotional signals. Here all I want
of the film's macro-action. If we feel that a character has been wronged, we may mimic, to indicate is that my last p, processing, can support systematic studies of the range
in weakened form, her anger and self-righteousness. Screenwriters provide strong of emotional effects that a film can have. More broadly, the cognitive perspective I've
prompts for sympathy, such as making sure that the protagonist is treated unfairly, sketched here has informed several, quite different research programs, and many prac-
and many screenplay manuals argue that the skilful filmmaker evokes both hope titioners wouldn't endorse all the claims I've made.101 My efforts are simply to show
(for the character's success) and fear (of the character's failure).97 Again, there may
how a historical poetics of cinema of the sort exemplified in the essays that follow can
be considerable cross-cultural regularities in these emotions, most of which depend
usefully adopt some cognitive principles as a way to chart a film's manifold effects.
upon recurrent social situations that people in most cultures encounter—sympathy
Sometimes, however, we can point to patterns and principles and purposes with-
for children, anger at being wronged, and a sense of fairness or justice.
out being able to specify effects very well. This is largely because we don't have a tight
In comprehension, emotion and thought mesh. Greg Smith argues that narra-
theory allowing us to trace the consequences of artistic choices. Occasionally we may
tive films tend to sustain moods and then punctuate them with bursts of emotion
find constructional principles that don't correlate with any plausible effect. The begin-
proper. 98 These in turn can focus our attention on story developments. In Rope, our
ning shot and final shot of Ozu's Ohayo (1959) are reverse-angle views of one another
knowing where the corpse is hidden generates suspense, a mood that in turn makes
(Figures 1.22-1.23). Given Ozu's well-known interest in rejuggling objects and view-
us hyperattentive to every movement toward the chest. When Rupert lifts the lid, the
points and his use of 180-degree cutting patterns, it's likely that these two shots were
mood prepares us to concentrate on his face, which betrays his shock and distress.
planned to have the sort of regularity we find here.102 But when I point this out, some-
Emotion also affects memory; in real life, a traumatic event becomes sharply etched
one will say quite reasonably that no one could perceive this pattern while viewing.
into our minds. Films exploit this tendency by making the most vividly emotional
Only by geeky concentration on Ohayo's overall architecture can we detect it. My
scenes crucial for the plot—a death, a separation, a reunion. Ben Singer has proposed
response is that a few filmmakers build their films as objects as much as experiences,
a catalogue of prototype scenarios, drawn from intense emotional experiences in
as patterned constructs that may or may not fit snugly in a viewing unfolding across
ordinary human life, that melodramas draw upon. For example, the pathos we feel
when seeing people degraded by misfortune forms the basis of a scene in Mother time. Like a poet who plants hidden acrostics or numerical codes, Ozu expects some
India (1957), when the mother, in the aftermath of a flood, digs desperately through viewers to look at his film as if it were existing in a virtual space in which every shot
can be compared with every other.
the mud for something that will feed her dying children. 99 Patrick Hogan's survey
of transcultural story patterns traces their constant features to the way they make More generally, I'd suggest that we adopt the circuit particulars-patterns-purposes-
salient certain emotion-based prototypes of happiness. 100 principles-practices-effects as a default to guide our inquiry. Most of the time it will
There's a great deal to be studied about how emotion works within our cinematic serve us well. In a few cases, we may have to register the possibility that a film's orga-
experience, including the bonding effects of watching a film with others. In some of nization can outrun its effects on any viewer. Yet by pointing out this state of affairs,
the essays that follow, I propose that narrative films often model social intelligence, the poetician may help the artist realize his or her design. Now that I've highlighted
54 Poetics of Cinema Poetics of Cinema 55
the reverse-angle shots separated by an entire film, viewers are free to be alert for empirical research, and theoretical explicitness, has worn fairly well. Despite errors
them when they see Ohayo, or other Ozu films. At this point, a connoisseur's appre- of fact, thought, and judgment—not least my own—the conversation has advanced.
ciation for tiny felicities becomes the most relevant effect, and of course this too can By concentrating on particular questions and then comparing our reasoning and
be fruitfully investigated as part of an art tradition to which Ozu belongs. research with that of others asking congruent questions, we have begun to produce
Let me sum up. I've argued that broadly speaking, the central question of film reliable knowledge about film.103 A historical poetics of cinema isn't the only vehicle
poetics, posed as a methodological point of departure, can be understood in this way: for this enterprise, but the essays that follow try to show that it can be a sturdy one.
How are films made in order to elicit certain effects?
The first part of this formulation, dealing with the making, invites us to explore
two domains. Analytical poetics studies the materials and forms of films to bring
out the principles shaping them. Here we study theme and subject matter, large-
scale form (such as narrative), and audiovisual style. Analytical poetics promotes
functional explanations.
A second domain is historical poetics, the study of principles of filmmaking as they
inform films in particular historical circumstances. This requires not only analysis of
the films but also research into norms and craft practices impinging on the principles
informing the films. It investigates how film artists, as historical agents, work within
the zones of choice and control offered by their circumstances. Historical poetics thus
traffics in both functional and causal explanations.
The last part of the initial formulation—the role of films in eliciting reactions—
invites us to postulate that spectators play a role too. Call this a poetics of effect. Here
we ask what activities are elicited by the thematic-formal-stylistic dynamics of the
film and the principles undergirding same. I've suggested that studying viewing effects
can fruitfully adopt a cognitive perspective, understood as involving perception,
comprehension, and appropriation, all invested with emotion.
Overall, I've proposed a framework within which a variety of questions can be
plausibly asked and answered. I hope it's clear that not every question that we pursue
has to take all the particulars, patterns, and so forth into account. Sometimes just
working on one or two components is quite fruitful. Nor is my anatomy an account
of stages of inquiry. In studying any of these components, we are constantly moving
back and forth among them. Sensitive critics have always shuttled between part and
whole, material and form, form and function. I'm not urging that we create a rigid six-
step procedure. I suggest only that we systematize our intuitions to a greater degree
than usual, while allowing that historical analysis and recognition of the viewer's
activity can enrich our sense of the film's constructive principles.
In stressing my own views, I've probably not done justice to the fact that poetics
can host a variety of disparate research programs that can usefully debate, alterna-
tive positions. For example, my solutions to certain problems will not be exactly
compatible with those of the Russian formalists. Still, I owe to their great 1927
anthology, Poetika Kino, much more than the title of this collection. I share their broad
theoretical ambitions and their methodological commitment to conducting rational
and empirical inquiry into principles of art making within and across cultures. When
I began writing from this perspective in the early 1980s, film studies was dominated
by ideological critique and feminist psychoanalysis. From today's vantage point,
I think that historical poetics, with its commitment to dialectical argumentation,
 I.
Convention, Construction,

and. Cinematic Vision

Cinema is partly pictorial representation, and we have come to expect, especially

after the dissemination of structuralist and poststructuralist theories, that the
most enlightening accounts of pictorial representation will involve a theoretical
account of conventions. Yet the humanities have not yet solved the problem of how
to understand conventions; indeed, I am not convinced that we know very well
what a convention is.1
This essay aims to clarify the operations of visual conventions in cinema, but it has
broader goals as well. I shall suggest that we can make progress toward understand-
ing artistic convention by rejecting some tenets of structuralist and poststructuralist
doctrine—notably, the equation of "convention" with "arbitrariness." I go on to sketch
a criticism of the radical "constructivist" position that is often associated with such
doctrines. The essay also points toward the relevance of cross-cultural regularities for
understanding even the most local and idiosyncratic conventions.

Shot/Reverse Shot: A. Convention?

The problem of convention in filmic representation can be strikingly posed by consid-
ering one film technique. What is called "shot/reverse-shot" editing typically involves
displaying two figures in face-to-face interaction. The camera shows each one alter-
nately, with either the other character absent or only partly visible. The filmmaker
cuts from one shot to another, following the flow of the conversation and the facial

57
58 Poetics of Cinema
<9
w /-
Figure 2.1 Metropolis (Fritz Lang, 1927). Figure 2.2 Metropolis.
reactions. Sometimes the view is taken from slightly behind each character, putting
the other character's shoulder in the foreground in what is called an over-the-shoulder
(OTS) shot (Figures 2.1-2.2).
The shot/reverse-shot device deserves to be called a stylistic invention. It wasn't
determined by the technology of the cinema, and I can find no plausible parallels in
other nineteenth-century media, such as comic strips, paintings, or lantern slides.
It wasn't utilized as a stylistic device in the first 15 years or so of filmmaking; that
period was dominated by the so-called tableau style, which showed the entire scene in
a single shot. In the early 1910s, some fiction films used the shot/reverse-shot device
occasionally, whereas by the end of the teens it was common in American features. 2
Fairly soon after this, shot/reverse-shot cutting was adopted around the world. It con-
tinues to be one of the most commonly used techniques in film and television.
What makes the shot/reverse shot comprehensible? Theorists have offered two fairly
distinct answers to this question. The first, and older, view is that the device offers a
kind of equivalent for ordinary vision. In an early discussion, Soviet filmmaker V.
I. Pudovkin says that editing aims to guide the spectator's attention to important
elements of a scene. "The lens of the camera replaces the eye of the observer, and the
changes of angle of the camera—directed now on one person, now on another . . .
—must be subject to the same conditions as those of the eyes of the observer."3 This
is rather vaguely put, but the idea that editing simulates the change of glance of an
observer makes shot/reverse shot a kind of heightening of our ordinary perception
of an event involving participants. More recently, Barry Salt has compared such
editing to "what a spectator before the scene would see, standing there and casting his
glance from this point to that point within it."4 For these theorists, then, filmmakers
discovered in the shot/reverse shot a correlate to spontaneous perceptual activity. Call
this the "naturalist" position.
The naturalist position answers several questions. What enabled shot/reverse shot
to be discovered? Presumably, filmmakers seeking to engage audiences hit upon it
by trial and error, perhaps guided by their own perceptual intuitions. Why was it so
rapidly taken up? Because it achieved the requisite purposes of presenting an intel-
ligible structure of information to the spectator. Why has shot/reverse shot been so
Convention, Construction, and Cinematic Vision 59
Figure 2.3 Class Relations (Jean-Marie Straub Figure 2.4 Class Relations.
and Daniele Huillet, 1983).
enduring and pervasive? Because, as an obvious correlate to perceptual experience
outside the movie house, it does not require viewers to have special training in order
to understand it.
The chief problem with this account is that shot/reverse shot is in several respects
quite unfaithful to perceptual experience. The best equivalent to a viewer moving her or
his glance from one character to another would seem to be obtained by simply swiveling
("panning") the camera from speaker to speaker. But this is a very rare stylistic option
in mainstream cinema. The instantaneous transfer of attention given by the cut would
seem to be a conventional substitute for this swiveling of the imaginary spectator's
attention—a substitute that has no exact correlate in ordinary perceptual experience.
The shot/reverse-shot device is also unfaithful to ordinary vision because it
changes the camera position so as to favor 3/4 views. When you're a third party to a
conversation, you don't typically watch each speaker from an oblique angle, let alone
from the changing angles provided by reverse shots. When we watch a face-to-face
interaction, we are not perceptually capable of shifting our angle of view as drasti-
cally as is normal in shot/reverse-shot cutting. And you certainly don't watch from
over each character's shoulder. In the absence of panning from face to face, a profiled
shot/reverse shot (such as that in Figures 2.3-2.4) would provide a closer equivalent
to "what a spectator before the scene would see" than does the angled OTS views
presented by the majority practice.
Such difficulties were noticed in Pudovkin's day. He therefore added a proviso: The
camera allowed the director to create not an actual observer but an "ideal," omni-
present one. Similarly, as Karel Reisz and Gavin Millar point out, the change of angle
within shot/reverse-shot cutting has "no analogous experience in real life."5 The
director aims at creating "a ubiquitous observer, giving the audience at each moment
of the action the best possible viewpoint. He selects the images which he considers
most telling, irrespective of the fact that no single individual could view a scene in
this way in real life."6 This is justified as artistic selection.
But this deviation from the natural-equivalent premise opens the door to quite a
different theoretical position. Once shot/reverse-shot cuts depend at least partly upon
60 Poetics of Cinema Convention, Construction, and Cinematic Vision 61

purely artistic considerations, we can ask if they are not simply conventions. Any Primary Theory and a Continuum of Conventions
artistic device as widely used as shot/reverse shot, if not significantly motivated by
In contrasting the two views of shot/reverse shot, I followed precedent in distin-
perceptual equivalences, is likely to be seen as a stylistic convention.
guishing something called nature from something called convention. The first step in
This presumption, I think, dominates film studies today. From this perspective,
forging a more comprehensive theory, I believe, is to discard these notions and offer
shot/reverse-shot cutting is an arbitrary device, having no privileged affinities with
some more flexible concepts in their place.
natural perception. But what is a convention, on this view?
The term nature comes to us fraught with connotations. To most film theorists,
Minimally, I suppose, most contemporary scholars would say that shot/reverse
it suggests either biologically innate capacities or universal laws operating in the
shot is a convention because it is a piece of artifice, and because it must be learned.
physical world generally. It also suggests the realm of necessity, that which cannot
Most theorists are content to leave the matter there, but neither point really blocks the
be changed by human will or skill. Such conceptions of the "natural" have been
naturalist position. The naturalist position does not have to claim that shot/reverse-
frequently attacked by structuralist and poststructuralist theorists, who insist that all
shot editing is not artificial in some sense. After all, it is an invention; it was not
signification is constructed, conventional, and culture bound.
present at the birth of cinema, and people decided to use it. Nor does the naturalist
Still, only dogmatists would deny that representation, especially visual representation,
position have to deny the role of learning. Once we have learned to perceive the world,
relies at least partly on the perceiver's psychophysical capacities. It seems very unlikely
the naturalist might argue, we can learn to grasp artistic devices that provide equiva-
that our ability to perceive humans and objects in images owes nothing to our biologi-
lents to the world. Accordingly, our ability to grasp those devices ought to ride upon
cal heritage. Our understanding of images could hardly be unconnected to our capaci-
the appropriate sorts of perceptual skills.
ties to move through a three-dimensional environment and to recognize conspecifics.
To this, a contemporary theorist might reply that the typical convention is arbi-
The individual's development of language, according to the most powerful theories now
trary. Here, arbitrariness must mean something like this: In principle, an indefinitely
available, is as much a biological capacity as the inclination to grow arms rather than
large number of other representations would serve as well; the one chosen is simply
wings.8 Certain relevant abilities may not even be species specific: Pigeons and monkeys
assigned that task by the rules of the langue or code in force. A dog might as easily
respond to photographs as if recognizing the sorts of things represented.9
have been called a chien or a hund.
Nonetheless, I propose that we can make some progress if we bypass the nature-
There are some problems with extending to nonlinguistic phenomena a conception
culture couplet for the moment and concentrate upon some "contingent universals" of
of arbitrariness derived from the lexical items in a language. Consider the turn signals
human life. They are contingent because they did not, for any metaphysical reasons,
on a car. To an observer on Sirius, the fact that I flash the right signal when I intend
have to be the way they are; and they are universal insofar as we can find them to
to turn right might appear arbitrary. Other options are logically possible: People
be widely present in human societies. They consist of practices and propensities that
could signal a right turn by activating the left signal. But in fact such mechanisms
arise in and through human activities. The core assumption here is that given certain
are designed to fit our propensities to signal rightward movement by something that
uniformities in the environment across cultures, humans have in their social activities
stands in a rightward relation to our body. Our nonverbal symbol systems, like our
faced comparable tasks in surviving and creating their ways of life. Neither wholly
technical gadgets, are engineered to our fixed dispositions, including innate ones, and
"natural" nor wholly "cultural," these sorts of contingent universals are good candi-
the choice among all possible options is not indifferent. 7
dates for being at least partly responsible for the "naturalness" of artistic conventions.
A similar case can be made about the shot/reverse-shot technique. If the director
Paradigm cases of contingent universals would seem to be practical skills such as
seeks to represent two people looking at each other, it is less arbitrary to show them
the ability to use language for communication, to divide labor tasks, to distinguish
looking at each other than to show them, say, looking away from each other, or at the
between living and nonliving things, and so on. I have stated these rather generally;
moon. A visual "code" that showed figures looking at each other in order to signify
it is an empirical question as to whether there are not much more specific contingent
that they are not looking at each other would be bizarre in the extreme. We would,
universals, such as recognizing focal colors or taking turns during conversation.10
I think, be inclined to call that alternative code "arbitrary," but not the normal case,
I have stressed contingent universals as involving behavior, but it seems likely that
which reflects naturalistic assumptions about the image's representation of the state
they constitute a conceptual frame of reference as well. The anthropologist Robin
of affairs. For creatures like us, the two options are not equiprobable.
Horton calls such a framework "primary theory" and characterizes it as follows:
Nevertheless, the naturalist's position on shot/reverse shot remains problematic
because of the undeniably "unrealistic" qualities present in orthodox uses of the device. Primary theory gives the world a foreground filled with middle-sized (say
And something theoretically stronger is probably required to allay the conventional- between a hundred times as large and a hundred times as small as human
ist's worries. At this point, I want to suggest a middle way between the two positions, beings), enduring, solid objects. These objects are interrelated, indeed, inter-
one that captures the intuition that such visual devices are constructed and signifi- defined, in terms of a 'push-pull' conception of causality, in which spatial and
cantly artificial while also preserving the idea that they are not utterly arbitrary. temporary contiguity are seen as crucial to the transmission of change. They
62 Poetics of Cinema
are related spatially in terms of five dichotomies: 'left' / 'right'; 'above' / 'below';
'in-front-of / 'behind'; 'inside' / 'outside'; 'contiguous' / 'separate'. And tempo-
rally in terms of one trichotomy: 'before' / 'at the same time' / 'after'. Finally,
primary theory makes two major distinctions amongst its objects: first, that
between human beings and other objects; and second, among human beings,
that between self and others.11
Horton suggests that although different communities may emphasize some aspects of
primary theory and leave others comparatively undeveloped, as a conceptual frame-
work it does not vary significantly from culture to culture.
Note that no decisive claim need be made that contingent universals, whether
practices or "primary theory," are either biologically prewired or culturally acquired.
In a trivial sense, the capacity to undertake any action must precede that action, so
there must be some "natural" capacities. More strongly, those capacities result from
evolution. Like other species, humans have evolved in tandem with their environ-
ment, and so we're equipped to detect the sort of primary-theory regularities that
Horton points out. We should also remember that our environment includes other
humans. As social animals, we're attuned to interact not only with sticks and stones
but also with our conspecifics. (I'll argue in some of the following essays that cinema,
like other narrative arts, relies on displays of social intelligence, some aspects of which
are plainly cross-cultural.) In sum, many cross-cultural convergences can be traced
to our evolutionary heritage. At the same time, a great many aspects of artworks rely
on particular cultural traditions, so our perceptual capacities, our primary theory,
our attention to other humans, and so on will still be shaped and fine-tuned by
circumstance and culture. We can be quite agnostic about the sources of this or that
feature of artworks. As students of visual art, we can assume that, say, the ability to
discriminate colors or the skill at working material with tools is a contingent universal
of human activity, and leave the detailed story behind that activity to research within
the appropriate disciplines.
This perspective casts the concept of "convention" in a fresh light. "Arbitrariness'
as a measure of conventionality stems, I think, from a misapplication of Ferdinand
Saussure's claims about the arbitrariness of the linguistic sign.12 There is another way
to conceive of conventions: as norm-bound practices that coordinate social activities
and direct action in order to achieve goals.13
If we think of convention in terms of practical action, "arbitrariness" is not a very
fruitful way of characterizing it. In one important sense, an action counts as arbitrary
if the same goal could have been achieved by an alternative means, with no additional
costs or difficulties. If I want a bag of potato chips and I am equidistant between
two stores selling the snack, all other things being equal the choice is arbitrary. But
most artistic conventions are not arbitrary in this sense. First, for reasons already
mentioned, some choices are weighted because human proclivities favor them. It is
nonarbitrary that the right rear turn signal on an automobile announces that the
driver intends to turn right, not left. Moreover, many artistic conventions are more
Convention, Construction, and Cinematic Vision 63
appropriate to certain ends than others. If I am a film director and I want spectators
to study an actor's expression, my choice of a close-up isn't arbitrary, because that's
an option more favorable to achieving my purpose than, say, selecting an extreme
long shot.14 I conclude that we want an account of convention that accommodates
two demands: The "engineering" ought to fit human predispositions, and the means
ought to be weighted in relation to ends.
A final piece of brickwork needs to be laid in place. One of the attractions of the
concept of culturally conventional "codes" is the premise that works transmit or pro-
duce meanings. Meanings are cultural; where there is meaning, so goes the reasoning,
there must be codes. Instead, though, we may think of works as producing effects, of
which meanings are certain types. If we take the artist's goal to be that of eliciting
discriminable effects, we can consider a wider range of theoretical possibilities. Now
we can conceive of conventions as part of the artist's means for producing effects of
many sorts. And these effects take their place in a fabric of human action; they are
consequences of practical action on the part of artists, and grasping the conventions
is bound up with larger activities pursued by perceivers.
Our middle way between sheer naturalism and radical conventionalism, then, is
signposted by the notions of contingent universals, conventions as norm-governed
patterns of behavior, and artistic goals conceived as effects. The map I propose involves
a scale of visual effects, with distinct regions but loose boundaries between them. Here
I am picking up on E. H. Gombrich's hint that we could consider "representational
method" as ranked on "a continuum between skills which come naturally to us and
skills which may be next to impossible for anyone to acquire."15
At one end of the scale lie visual effects that are dependent upon cross-cultural,
even universal factors. Roughly, these effects would seem to be of two types.
First there are what we can call sensory triggers. These are cues that automatically
stimulate spectators. In the pictorial arts, contrasts of tonality and texture would
seem good candidates. Gombrich's interest in visual illusions has led him to insist
particularly on the importance of such triggers. He has frequently drawn analogies
to the behavior studied by ethologists, such as the ability of a rigged scrap to draw
attack from the stickleback fish. But Gombrich also suggests that such triggering
mechanisms need not be in the service of illusion; they can also stimulate a search
for meaning, creating perceptual anticipations that run ahead of the evidence. One of
Gombrich's great accomplishments is to have discovered that sensory triggers play a
much larger role in the visual arts than most theorists had recognized. 16
All nonlinguistic arts exploit such sensory triggers: scale and volume in sculpture,
rhythm and loudness in music, and so on. They are among the best candidates we
have for wired-in responses. In cinema, we do not have to look far for such triggers.
Apparent motion, the basis of cinematic movement, is an obvious one. We still do not
know exactly how apparent motion works; it may involve a cluster of specific mech-
anisms, possibly including motion-detecting cells in the visual system.17 Apparent
motion is a prime instance of a contingent universal: We did not evolve in order to be
able to watch movies, but the inventors of cinema were able to exploit a feature of the
64 Poetics of Cinema
design of the human optical system to create a pictorial display that is immediately
accessible to all sighted humans. Other sensory triggers available in cinema are the
use of extreme contrasts of visual tonality; the startle response evoked by sudden
intrusions into the frame; and, if Gombrich is right, the use of lighting to create
texture and volume within the shot.
Apart from sensory triggers, there are visual effects that draw upon contingently
universal factors. These rely on regularities of experience that are reasonable candi-
dates for being cross-cultural. Recognizing and reacting to these activities almost
certainly require some learning, but their ease of recognition among adult members
of all cultures makes them function as contingent universals, instances of Horton s
"primary theory."
As in our case of face-to-face interaction in the shot/reverse shot, these contingent
universals are so firmly fixed that we can scarcely imagine what arbitrary alternatives
would be. For example, we are so used to thinking about the variability of the rep-
resentation of pictorial space across different periods and places that we often forget
that these variations stand out against the background of a remarkable constancy in
the portrayal of human beings. If visual representation were truly arbitrary, then we
ought to find humans portrayed with four eyes or five legs as frequently as with two
of each. Yet in art across the world, the human body is represented in broadly com-
parable terms: the right number of limbs, the anatomically correct placement of head
and feet and hands, approximately similar canons of proportion, and so on. Indeed,
deities and monsters are marked as such partly by violations of such norms. Just as we
can recognize other members of our species in ordinary life, so too can we recognize
the human being in art of very distant or ancient societies. Surely cinema draws upon
this cross-cultural ability to recognize our conspecifics without any special training.
Returning to our example of shot/reverse-shot cutting, I suggest that face-to-face
personal interaction is a solid candidate for a cross-cultural universal. This is probably
why a visual code is unlikely to represent shared glances by divergent glances, as
noted above. It is also perhaps why the situation portrayed in shot/reverse shot is
instantly recognizable across cultures and time periods.
Moving along the continuum, we can turn our attention to visual effects that
depend on culturally localized skills but can be learned easily. Easily here translates
into "quickly, on the basis of comparatively limited exposure, and/or without special
training or expert guidance." These are norm-bound practices that can be picked up
largely through participating in a culture's life as a whole.
For instance, line drawings seem not to be culturally universal, and so they rely
on skills specific to certain cultures. Yet we have no reason to doubt that someone
can learn the conventions of understanding line drawing very easily—certainly much
more easily than learning a first or second language. Drawings of objects or persons
taken singly are highly recognizable across cultures. As Gombrich has pointed out, a
lot of the case for the relativity of pictorial representation has rested on examples of
overall systems of spatial representation that depict the relative locations of objects.18
These systems do vary more across cultures than do techniques of portraying
Convention, Construction, and Cinematic Vision 65
individual items. Although the ancient Egyptians developed a spatial system that
differs in many ways from post-Renaissance systems in the West, we can recognize
people, animals, and other items within the array. In fact, Egyptian images of fish and
birds are so faithful to appearances that specialists can identify the different species
portrayed. Perhaps even the broader spatial systems, such as the Egyptian one or
various forms of optical perspective, select out of real environments certain recog-
nizable features to preserve. Margaret Hagen has argued that the history of picture
making contains only a few families of spatial systems, and all of them operate under
projective geometries that capture real-world relationships.19 Although these systems
are artificial, her survey suggests, they are quickly and easily learned.
Sometimes such conventions are acquired in the course of normal human devel-
opment within the culture. There is evidence that in our culture at least, children
learn to understand line drawings of objects in tandem with learning to distinguish
(name, indicate, use) such objects in the world. 20 For an adult learner, such norms
require only the most minimal exposure and the most nontechnical, ostensive
training. Often, these norms can be guessed from context, as the speed lines in
comic strips are.
Cinema is full of such easily learned visual effects. Arguably, most transitions,
such as dissolves or fades; most acting styles; and most stylistic innovations, such as
crosscutting or complex camera movements, call upon such skills. Moreover, once
the viewer has mastered narrative structure to a useful degree, she or he has a suffi-
ciently strong sense of context in which to situate particular cinematic devices. Once
the viewer has the working concept of a scene, for instance, she or he can hazard a
guess that the darkening and lightening of the screen serve to mark one scene off from
another. If cinema does have codes, they are mostly codes of this very easily acquired
sort—which makes them significantly different from the codes governing other sign
systems, like semaphore or calculus.
At the other end of the continuum are those visual effects that depend on cultur-
ally specific skills requiring more learning. Acquiring them is time-consuming and
requires wide exposure to exemplars and/or special training and/or expert guidance.
In these respects, there is perhaps a genuine analogy to language—not to speech com-
prehension and production, but to reading and writing. (It's much easier to learn to
talk than to learn to spell and to punctuate.)
In the arts generally, modernism is marked by such comparatively recondite
conventions. Cubist painting, the novels of Joyce, the poetry of Pound or the Acmeists,
serial or minimalist music—all demand that the perceiver cultivate highly special-
ized skills. Such skills need not be wholly of form or style, because understanding
the depicted material itself could require training. Sacred iconography would be a
key instance. In Quentin Metsys' Mocking of Christ (1466), one might recognize a
man standing on a balcony, hands bound and thorns on his brow, without know-
ing the tradition whereby the situation and objects identify him as Christ before
his persecutors.
66 Poetics of Cinema Convention, Construction, and Cinematic Vision 67

In film, there may be relatively few visual effects that depend upon such specialized
skills. The avant-garde cinema is a plausible place to look for such effects, and it seems
likely that the films of Stan Brakhage or Yvonne Rainer require an audience to be con-
versant with abstract expressionism or contemporary poststructuralist theory. More
mainstream art cinema of the 1960s may have cultivated certain comparable devices,
such as plays with narrative time or shifts between black and white and color.
The spectrum I've outlined, from sensory triggers to comparatively recondite
"expert system" effects, is intended as no more than an initial shot at how to conceive
Figure 2.5 Yaaba (Idrissa Ouedraogo, 1989). Figure 2.6 Yaaba.
conventions. Many of my examples are speculative and are open to empirical
disconfirmation. But the general aim is to produce a frame of reference for theoretical
reflection and concrete analysis. Such a continuum lets us avoid the difficulties of the more contingently universal cues lead us to make sense of more esoteric cues in partic-
naturalist-conventionalist couplet. In order to achieve certain effects, artists may tap
ular ways. This would obviously facilitate learning: Not only do we need little exposure
biological propensities and contingent universals; in order to achieve other effects,
to certain effects, but also, in each image, the universal factors reinforce our hypoth-
artists may invoke more localized and recondite skills.
eses about the proper reaction we should have to the more culturally specific ones.
Does all this ascribe too little a role to culture? I think not. For one thing, as
I suggest that shot/reverse shot is best considered along these lines—as a compos-
Gombrich has often pointed out, it is culture that generates the tasks and interests that
ite phenomenon, drawing on features from various regions of the continuum. I can't
shape the ways in which visual effects are manipulated. If conventions are relations of
itemize all the relevant cues here, but let me make a start.
means and ends, the social purposes of a representation necessarily govern how the
In its prototypical form, shot/re verse shot is predicated on a two-person, face-to-
activity is conducted, how the first region of effects is formed into more complex ones.
face encounter. This phenomenon would seem to be a good candidate for a contingent
Moreover, the centrality of artistic schemas—those inherited patterns and formulas
through which the artist achieves effects in the medium—assures that culture plays universal of social intercourse—something that would be intelligible across cultures
a central role. "Only where there is a way is there also a will," Gombrich notes. The and periods. This consideration is so rudimentary that neither the naturalist nor the
individual can enrich the ways and means that his culture offers him; he can hardly conventionalist position on shot/reverse shot deems it necessary to weigh it, but in my
wish for something that he has never known is possible." 21 argument it forms a kind of cross-cultural bridgehead. For instance, Figures 2.5-2.6
from Yaaba may present facial or gestural cues specific to rural life in Burkina Faso.
Nonetheless, the cutting and camera positions present a face-to-face encounter
A Package Deal
between the young protagonist and his elder, and they do so through a prototypical
One advantage of tracing out this gamut of factors is that it allows us to see that the shot/reverse-shot construction.
most intuitively obvious phenomena of visual representation seldom fall neatly on The pattern may capture other contingent universals at work as well. Conversa-
one point of the continuum. It is unlikely that a single image, or a design system tional turn taking, with its interchanging role of speaker and listener, might furnish
like linear perspective, or even a technical device like shot/reverse-shot editing will an approximate structure for the alternation of images we get in shot/reverse shot.
constitute a pure instance of any sort of effect along this spectrum. 22 As critics and Indeed, it would seem likely that historically this alternating editing grew out of an
historians we inherit a language of practical craft, and this set of terms has other ends effort to capture the turn-taking phenomenon in cinematic form. Another important
in view than the distinctions that produce the continuum I've sketched. We ought to
cue, at least in the prototypical instance, is the glance of the persons represented on
expect that a typical representational device will present a bundle of effects of different
the screen. Noel Carroll has suggested that the informational saliency of eye move-
sorts. Any given technique may call upon sensory triggers (such as color contrast to
ments in primates gives filmmakers a powerful opportunity to engage audiences
indicate contours), contingent universals such as identification of the human figure,
cross-culturally. 23 We don't lack testimony from filmmakers that eyes matter. Here is
easily learned effects such as outline drawing, and more complex skills, such as that
J. J. Abrams, director of Mission: Impossible III (2006): "No matter how many trucks
of identifying allegorical figures.
and trailers are at base camp, it's ultimately about those few actors—those eyes, what's
Because most of the technical devices we encounter package many sorts of appeals
being conveyed emotionally." 24 In the terms I have proposed, the direction of the
together, it seems plausible to hypothesize that the cues lying closer to the "sensory
trigger" end of the scale will specify and constrain those cues that are more culturally glance would function as a sensory trigger, informing us of the object of the person's
specific and more difficult to pick up. That's part of what we mean by understanding attention. It stands as another cross-cultural regularity of human activity that can
something "because of its context." In the representational package we're offered, the elicit effects in beholders.
68 Poetics of Cinema Convention, Construction, and Cinematic Vision 69

Out of such basic materials—the face-to-face encounter, the marked look, the turn-
taking structure of conversation—the cinematic shot/reverse shot elaborates a more
complex construction. The immediately intelligible aspects of shot/reverse shot anchor
what we might consider to be more culturally specific sorts of effects. Because it is
universally intelligible to people from a very young age, the dyadic face-to-face encounter
offers a constant that can contextually guide inferences of more specialized sorts.
Consider the propensity we already noted for shot/reverse-shot images to be
3/4 views. There is some experimental evidence that for h u m a n faces in pictures,
the 3/4 view may be more easily recognized than other orientations. 25 Whereas the
straight-on and profile views of police mug shots are aimed at recording measurable
Figure2.7 Suvorov(Vsevelod I'udovkin, 1941). Figure 2.8 Suvorov.
facial data and easing recognition of the real face, the 3/4 view has generally been
found to be strongest when pictures are to be compared with other pictures. Because a
movie viewer doesn't have to pick the actor out of a lineup, the 3/4 view in shot/reverse
shot serves the purpose of maximizing rapid recognition (at least in cultures that have
pictures). The fact that profiled shot/reverse shot seems to be rare in the world's film-
making practice suggests that filmmakers have exploited a widespread, easily learned
norm of representation.
The saliency of 3/4 views has intriguing implications for over-the-shoulder shots
too. An OTS provides relevant and redundant information. We see the face of the
favored conversant while also being reminded that her partner is present, and at a
certain distance f r o m her. We shouldn't think of this camera position as providing
Figure 2.9 The Land (Youssef Chahine, 1969). Figure 2.10 The Land.
the view of an observer, either realistic or ideal. Rather, the image constitutes a display
that makes salient key information about the encounter in a way that permits quick
pickup (as, say, a view from steeply above or below wouldn't).
be impossible to grasp the purpose of the camera positions and editing. In a meta-
As for the instantaneous change of view that is said to create the "ubiquitous" or
phorical sense, the prototype of shot/reverse shot is constructed out of such contin-
"ideal" observer, this would seem to be a special case of the immediate leap in time
gent universals: It is a refined elaboration of them, a piece of artifice serving cultural
or space caused by any cut, of any sort. And once spectators, presumably f r o m a very
and aesthetic purposes.
young age, have acquired the skill of taking a cut to signal such a shift in orienta-
The multiplicity of those purposes can best be grasped, I think, if we t u r n to a
tion, the other cues present in shot/reverse shot may suffice to motivate the distinct
final issue. Gombrich has argued that the history of style in the visual arts is usefully
changes of angle. 26
There are doubtless other cues that are ingredient to the shot/reverse-shot device, understood as a process of schema and revision. The artist takes an available pattern

such as the more localized n o r m that the figures will be observed f r o m the same side and recasts it in the light of the capacities of the medium, the purposes she or he has

of an imaginary "center line." 27 Nevertheless, these remarks indicate the directions in
which my account would move. Against the naturalists, I suggest that we don't have
to take the shot/reverse-shot technique as straightforwardly conforming to ordinary
perception. It's not necessary to posit the device as creating an invisible observer; it's
at least as likely that the shot/reverse shot presents a patterned display organized to
highlight certain information. Hence its avoidance of a panning movement to simu-
late the glance and the physical implausibility of its canonical angles. The shot/reverse
shot can best be considered as a bundle of norms, some less stylized than others.
Against the conventionalists, I suggest that this bundle of norms draws upon con-
tingent universals of h u m a n culture as well as pervasive, easily learned practices of
filmmaking. And it seems likely that the former constrain and specify the latter: If the
viewer knew nothing of face-to-face conversations, eyelines, or turn taking, it would
in view, and the available means of achieving particular effects on the beholder. Art
has a history, Gombrich suggests, because all these factors can change over time. 28
From this standpoint, the shot/reverse-shot device can be seen as a schema circu-
lating across the history of film style. Once invented and found to achieve the desired
effects, it became a formula for rendering the dramatic scenes that comprise most
narrative films. It proved enormously flexible. The shot/reverse shot could be adjusted
to include several characters (Figures 2.7-2.10). It could show characters sitting side
by side or perched at different heights (Figures 2.11-2.12). It could present characters
with their backs to each other. It could display only portions of characters' bodies
(Figures 2.13-2.14). The camera could be placed at various distances and angles as
well. For example, we may have shot/reverse shot with individuals or groups looking at
the camera—that is, at the other participant(s) in the exchange (Figures 2.15-2.18).
70 Poetics of Cinema C /On v p n t~ i C i y— i~ 1 .J f ^ , • "XT- .

Figure 2.11 Prunella (Maurice Tourneur, 1918). Figure 2.12 Prunella.

Figure 2.17 School Daze (Spike Lee, 1988). Figure 2.18 School Daze.

Figure 2.13 Gossette (Germaine Dulac, 1923). Figure 2.14 Gossette.

Figure 2.19 The Cloak (Grigori Kozintsev Figure 2.20 The Cloak.
and Leonid Trauberg, 1926).

Figure 2.2] ///,• Killci (loiin \ \ o o , 1989). Figure 2.22 The Killer.

Figure 2.15 Toni (Jean Renoir, 1935). Figure 2.16 Toni.

The shot/reverse-shot schema has proved capable of fulfilling more self-consciously

expressive purposes. In The Cloak, Grigori Kozintsev and Leonid Trauberg alter angle
and distance so drastically as to make the Important Personage loom very large and
to make the supplicating Akaky Akakevich seem tiny (Figures 2.19-2.20). }ohn Woo
stresses the affinities of cop and killer by making the graphic design of the shots very
similar (Figures 2.21-2.22). In order to convey the idea that two separated lovers are
thinking of each other, Rene Clair joins them in a shot/reverse shot (Figures 2.23-2.24).
More disturbing is the famous shot/reverse-shot series in Nosferatu, when Dracula's
attack on Jonathan Harker is halted by Nina's beseeching gesture in distant Bremen
 Poetics of Cinema Convention, Construction, and Cinematic Vision 73
72
time, to say that shot/reverse shot is somehow mimicking ordinary perception. But
neither is it to say that the schema that is here revised is wholly artificial, completely
arbitrary, for it requires some contingent universals as its bridgehead. At the same
time, it presupposes some norms, however easily learned they may be.

Contingent Universals and Us

Figure 2.26 Nosferatu.
Figure 2.27 Muriel (Alain Resnais, 1963): Figure 2.28 Muriel: The next shot, show-
The medium shot of Helene's client is 17 frames ing the client's coat button, is 12 frames long
long (0.7 seconds). (0.5 seconds).
Figure 2.29 Muriel: The next shot, of her Figure 2.30 Muriel: 7he "reverse shot" of
hat, lasts only 11 frames (0.45 seconds). Helene's hands is also 11 frames long.
(Figures 2.25-2.26). Once the shot/reverse-shot formula has been absorbed,
"modernist" approaches can push it to elliptical limits. The opening of Muriel (1963)
revises the schema by making the shots abnormally fragmentary and brief, many less
than half a second long (Figures 2.27-2.30). One can take this passage as recasting the
sort of intercutting of body parts we find in Gossette (1923; Figures 2.13-2.14). Like all
norms, shot/reverse shot can be revised for specific purposes.
Such sequences require that audiences know standard shot/reverse shot well
enough to recognize the play between continuity cues and idiosyncratic factors
working against the canonical effect. Even in the highly experimental cases, though,
certain constants remain. Without the default assumption that characters are near
one another, and without the cues of angle, distance, orientation, and cutting pro-
vided by the schema, such passages would be unintelligible. This is not, one more
If any slogan wins immediate acceptance in contemporary theory in the humanities,
it is that a given phenomenon is "culturally constructed." The term gathers its force
partly from implicit contrast to alternative positions. The phenomenon is constructed,
and thus in some sense artificial; it is the result of human praxis, not natural process.
The phenomenon is cultural, and so neither natural nor "individual"; it is broadly
social, not narrowly psychic. So far, what I've been sketching out here is consistent
with these general implications of the phrase.
What I have been trying to say is not, however, compatible with another implica-
tion. Sometimes the phrase culturally constructed is used to suggest that the phenom-
enon is not universal or even widespread; it is assumed to be specific to a particular
culture. Yet even if cultural models exercise a local validity, it doesn't follow that
all of them are unique to a single society or period. It is perfectly possible for a
phenomenon to be culturally constructed and at the same time be very widespread,
or even universal, among h u m a n societies.
Too often, advocates of radical cultural constructivism have supposed that humans
in groups dispersed across time and space never face recurring conditions or problems
and that they never develop similar or even identical solutions to these conditions.
It is a cardinal error to assume that cross-cultural convergences indicate only a shared
"biological" or natural propensity, and that all else must be a matter of divergence and
variability, somehow traceable to the vagaries of cultural differences.
Not only perceptual equipment but also the disposition to see the world as a three-
dimensional space in which free-standing objects exist independent of the observer;
not only language "in general" but also pronouns and proper names, lies and narra-
tives, grammatical redundancy and the greater frequency of short words for familiar
objects; not only tool making but also the fashioning of pounders and containers; not
only spontaneous smiling but also expressions of skepticism and anger, as well as a
fear of snakes and loud noises—all these and many more activities are current candi-
dates for being true cultural universals. Apparently all cultures distinguish between
natural and nonnatural objects, between living and nonliving things, and between
plants and animals. 29 All societies have created fibers for tying, lacing, and weaving. 30
To recall Marjorie Garber's "No culture without the hermaphrodite," we can add,
"No culture without string."
The value of recalling such anthropological data is, I hope, to help us get beyond
the knee-jerk equation of cross-cultural (or even cross-subcultural) with natural or
biologically determined. Not even the most hubristic sociobiologist would postulate
a genetic basis for proper names, containers, and twine. It seems likely that regulari-
ties of the human body, along with regularities of the physical environment and of
74 Poetics of Cinema
interpersonal relations, to which humans are attuned by species-specific propensities,
have called forth from social collectivities many similar and even universal practices.
If social life requires that humans share information, tacit norms guiding face-to-face
interactions and conversational turn taking will assist the process in any circum-
stance in which humans meet.
At the same time, we ought not to quail at the prospect that these universals
frequently have a component rooted in biological predispositions. Academic human-
ists resist the idea of a human nature, convinced that it leads to reductionist and
determinist explanations. But it doesn't, because human capacities and propensities
are always reshaped by culture—and those capacities and propensities do as much
to create culture as to respond to it. It's clear by now that the nature-nurture split
is uninformative, that genes are designed to respond to the environment, and that
nature has shaped us to be resourceful enough to adjust behavior in relation to our
surroundings. Rather than being the robotic servant of a gene for executing this or
that piece of behavior, we are flexible and resourceful. "Nature," writes Matt Ridley,
"can act only via n u r t u r e . . . . The environment acts as a multiplier of small genetic
differences, pushing athletic children toward the sports that reward them and push-
ing bright children toward the books that reward them." 31
We ought not, therefore, to balk when the metaphor of construction leads us
to recognize that social practices may be "built out" of contingent universals. I've
argued elsewhere that a constructivist theory of social convention and mental activity
requires some conception of materials out of which a representation is fashioned. 32
These materials need not be raw, nor even material in the strict sense (because
constructivism is a metaphor to start with). As to the source of these materials, we
can be quite agnostic; it's not up to film scholars to do the work of anthropologists,
population geneticists, and the like. All we need do is note that some features of the
films we study, for whatever reason, are manifested across cultures and may thereby
create convergent effects.
Theoretically, the most comprehensive and powerful explanations of conventions
in any art would seem to be those that show them to be functional transformations
of other representations or practices, some of which may be sensory triggers or con-
tingent universals. Methodologically, the best strategy would seem to be constantly
on the alert for the cross-cultural factors that would be part of any representational
process. Sometimes, these may go without saying; at other times, examining these
may shed light on how familiar formulas achieve their distinctive power.
Something like this position, I think, has the best of both naturalism and conven-
tionalism. This view also points toward ways of understanding how conventions may
develop in specific social circumstances. Perhaps most tellingly, a moderate construc-
tivism along these lines points toward an understanding of the cross-cultural powers
of visual art. 33
Convention, Construction, and Cinematic Vision 75
Afterword
This coda tries to block one line of objection to the preceding essay's argument,
correct a published misrepresentation of my position, and clarify my position on
evolutionary explanations.
Many film scholars would object to my position on the grounds that it inhibits
progressive political thinking, but this doesn't follow. Since this essay was first
published, a compatible argument has been set forth by Peter Singer in A Darwinian
Left. He proposes a continuum fairly congruent with mine: "While some areas of
human life show great diversity, in others, human behavior stays fairly constant
across the whole range of h u m a n cultures, and some aspects of our behavior are also
shared with our closest nonhuman relatives." We find, he points out, great variation
in the ways people produce food, arrange their economies, practice religion, and
govern themselves. We find less variation in certain sexual arrangements (e.g., sexual
intercourse before marriage) and in ethnic identification (e.g., racism). We also find
universals, such as concern for kin, willingness to cooperate, concern with status and
hierarchy, and assignment of gender roles. To recognize the biological component
in contingent universals isn't to reduce them to it; it's just being clear-sighted. Social
experiments like that undertaken in the USSR that treat humans as utterly malleable
are likely to come to grief. Singer warns that for progressive people, "to be blind to the
facts about human evolution is to risk disaster." 34
The position outlined in this essay received criticism from two writers. Slavoj
Zizek raised objections in The Fright of Real Tears: Krzystof Kieslowski Between
Theory and Post-Theory. I've replied to those in Figures Traced in Light (Berkeley:
University of California Press, 2005, 260-265), and to related ones on my website
(http://www.davidbordwell.net/essays/zizek.php). Miriam Bratu Hansen proposed
criticisms of this essay and of related works of mine in her essay, "The Mass Produc-
tion of the Senses: Classical Cinema as Vernacular Modernism." I replied to those as
well in Figures (pp. 275-276), but because Hansen's comments reflect some common
misapprehensions about my claims in this piece, I take this opportunity to expand
my rejoinder. 35
Hansen has suggested that the cross-cultural appeal of mainstream or "classical
Hollywood cinema" (which she calls "an international modernist idiom") does not
rely on regularities of perceptual and cognitive pickup. She characterizes the position
that she rejects in this way:
Once "the system" is in place (from about 1917 on), its ingenuity and stability
are attributed [by D. B.] to the optimal engagement of mental structures and
perceptual capacities that are, in Bordwell's words, "biologically hard-wired,"
and have been so for tens of thousands of years. 36
First, it's worth noting that Hansen takes my phrase entirely out of context. The
passage she quotes, from a 1996 essay on Louis Feuillade, is not talking about cross-
cultural appeals of classical Hollywood style, or indeed about style at all. My original
essay is questioning Walter Benjamin's thesis that human perceptual mechanisms
76 Poetics of Cinema
are sufficiently plastic to be remade by the shocks of modernity. His general claim is
that "the mode of human sense perception changes with humanity's entire mode of
existence." Here is the relevant passage from my piece:
If true, Benjamin's claim would have stunning implications, but there seems
no good reason to accept it. . . . Since " h u m a n sense perception" has evolved
over millions of years, we would need a remarkably full story to explain
how social factors arising in fifty years or less could alter biologically hard-
wired mechanisms. 3 7
There is nothing here to indicate that I consider biological hard wiring to be the
source of the "ingenuity and stability" of the classical style. Hansen simply transfers
my phrase into a wholly different topic area. And I can't explain why she turns my
millions of evolutionary years into "tens of thousands."
As for the essay you have in hand, Hansen cites this in her "hard-wiring" reference
too. Yet you will not find the term in this piece. If she intends her objection regarding
classical style to apply to this essay, the effort fails again. Readers of my essay will note
that I claim that many aspects of film style are transformations of a variety of cross-
cultural skills and practices, both biological and cultural. My central example in the
essay is that of face-to-face interaction in conversation. It isn't for me to say whether
the cross-cultural constants in such interactions are strictly a biological adaptation, a
by-product of biological adaptation, a pure product of culture, or a mix of all. That's
why I call such regularities "contingent universals." As indicated in the essay, I'm
inclined to think that a representational practice like continuity editing packages
together many sorts of features, including ones specific to certain cultures.
In all, though, the sources of such universals don't matter for my argument. My
point is that such face-to-face interaction is a cross-cultural universal, and cinema
has found stylistic figures (e.g., shot/reverse-shot framing and cutting) that amplify
and streamline this for easy uptake. As I also indicate, though, it does seem plausible
that many aspects of perception activated by style, such as the detection of edges and
surfaces, the recognition of other humans on the screen, or the ability to discrimi-
nate among facial expressions, are biologically hardwired. Even such mechanisms,
however, require exposure to environmental regularities in order to mature normally.
As two theorists of language point out, "Innate ideas need not be inborn." This is
because "the relevant cognitive resources may develop according to a maturational
timetable, or they may be triggered by experiences that come later in life."38
I mention another aspect of Hansen's essay because it neatly illustrates the distinc-
tions I tried to urge in this book's first essay (pp. 11-15). She doesn't explicitly argue
against the existence of contingent universals (perhaps because she considers any
appeal to "biological hard wiring" as sufficient to induce doubt in academic humanist
readers). Instead, she asserts that a better explanation for the cross-cultural success of
mainstream cinema is that it offers a profusion of possible interpretations.
If classical Hollywood cinema succeeded as an international modernist idiom
on a mass basis, it did so not because of its presumably universal narrative form
Convention, Construction, and Cinematic Vision 77
[not the same as style, which is what she was just talking about—D. B.] but
because it meant different things to different people and publics, both at home
and abroad.
The claim in the last clause is asserted, not demonstrated. Moreover, "meant different
things" is not explained, but what Hansen has in mind is suggested from the questions
she goes on to ask:
How were the films programmed in the context of local film cultures, in par-
ticular conventions of exhibition and reception? Which genres were preferred
in which places (for instance, slapstick in European and African countries,
musical and historical costume dramas in India), and how were American
genres dissolved and assimilated into different generic traditions, different
concepts of genre? And how did American imports figure within the public
horizon of reception which might have included both indigenous products and
films from other foreign countries? 39
Whatever the value of these questions, a reading of my essay shows that they aren't
congruent with the ones I ask. The matters that intrigue Hansen move beyond the level
of face recognition, social interchange, and the construction of space. The varieties
of interpretive reception she mentions are of a different order than the regularities
I sought to explain.
To put it another way, and given the framework I set out in the first essay in this
book, Hansen is seeking to explain the pervasiveness of continuity devices because
they generate diverse appropriations. I am arguing that their pervasiveness stems at
least partly from widely appealing features of perception and comprehension, though
as I indicate in the first essay, there can be convergent appropriations too. (Indeed, if a
group of people in other countries appropriates Hollywood movies in some divergent
way, there would have to be convergence . . . among members of that group.) But even
if appropriations differ, perception and comprehension can be convergent. So these
need not be conflicting explanations, because they operate at different levels.
But it would seem that my framework has theoretical priority. Before distinct audi-
ences can appropriate a story's meanings in discrete culture-driven ways, they would
have to understand at least some aspects of the story, along with the stylistic presen-
tation of those aspects. So not only are the stylistic transformations of contingent
universals on a different logical level than the plurality of abstract meanings issuing
from reception; the former are also preconditions of the latter. There seems to be no
getting around recognizing the ubiquity of the sort of perceptual-cognitive skills that
stylistic features prompt.
Hansen doesn't offer any evidence that a massive diversity of interpretations was
actually taking place. She does indicate that sometimes Russian distributors changed
the happy endings of Hollywood films, but this doesn't show incompatibility of
uptake. Russian audiences saw Hollywood movies and recognized the endings as
happy; they preferred, at least so the distributors thought, unhappy ones. Once the
endings became unhappy, presumably the Russian audience saw them that way. So,
 Poetics of Cinema
78
presumably, would audiences in many other cultures. So how does the example show
that the movies "meant different things to different people and publics, both at home
and abroad"? It just shows that different audiences have different tastes.
Remarkably, Hansen's essay mentions not a single film title. Nor does it suggest
how the stylistic and narrative regularities that historians of classical cinema pick
out have facilitated, blocked, or simply proven irrelevant to creating the profusion of
appropriations that she postulates. Consider the counterfactual: Had American cin-
ema remained in its "tableau" phase from the 1910s on, would a plethora of overseas
interpretations not have been forthcoming? If diverse appropriations would still have
emerged, then perhaps the classical style contributed nothing to them. Hansen's case
would also be more compelling if she could show that the output of other national
cinemas didn't evoke the profusion of interpretations she discerns with respect to
Hollywood. For all we know, French, German, and Danish films activated diverse
interpretations among nondomestic viewers. Apart from misrepresenting my posi-
tion, Hansen has advanced a vague hypothesis that needs deeper analysis and more
supporting evidence.
Finally, it might be worth clarifying one last time the role of evolutionary
explanations in what I'm proposing. So I append a codicil from my contribution to a
collection of papers exploring film in relation to ecological psychology.
Moving Image Theory: Ecological Considerations, edited by Joseph D. Anderson
and Barbara Fisher Anderson (Carbondale: Southern Illinois University Press, 2005),
is a collection of research articles that shed new light on the psychological dimensions
of cinema and television. My foreword makes reference to several of those essays, as
well as to the founder of ecological psychology, James J. Gibson. 40 From 1950 until his
death in 1979, Gibson argued that our perceptual systems evolved in order to detect
regularities, or "invariants," in our environment. Our senses supply an enormous
amount of information to us, he maintained, but the artificiality of laboratory experi-
ments often disguises that fact. In real environments (hence the ecological label of
his theory), as mobile and exploratory creatures, we don't need elaborate cognitive
machinery to solve many of the problems our species has faced. 41
What processes enable us to perceive, comprehend, and respond emotionally to
moving pictures? Here, in gross outline, is one answer. As humans we have evolved
certain capacities and predispositions, ranging from perceptual ones (biological
mechanisms for delivering information about the world we live in) to social ones
(e.g., affinities with and curiosity about other humans). Out of these capacities and
predispositions, and by bonding with our conspecifics, we have built a staggeringly
sophisticated array of cultural practices—skills, technologies, arts, and institutions.
Moving pictures are such a practice. We designed them to mesh with our percep-
tual and cognitive capacities. What hammers are to hands, movies are to minds: a
tool exquisitely shaped to the powers and purposes of human activity.
A great deal of movies' effects—more than many contemporary film theories
allow—stem from their impact on our sensory systems. We are prompted to detect
movement, shape, color, and sounds, and this is surely one of the transcultural capac-
ities that movies tap. Similarly, films from all nations and times draw upon more
Convention, Construction, and Cinematic Vision 79
"cognitive" skills, such as categorizing an object as living or nonliving, or seeing a
face as furious—abilities that, it's reasonable to think, are part of our evolutionary
heritage. And because affective states and counterfactual speculation are of adaptive
advantage, it is likely that an artistic medium that permits emotional and imaginative
expression would have appeal across cultural boundaries.
If we consider culture to be an elaboration of evolutionary processes, there's no
inherent gulf between "biology" and "society" in this explanatory framework. True,
these elaborations vary historically, yielding (among other things) what we usually
call conventions—local practices that seem "artificial" and that differ from one society
to another. Yet some conventions are less artificial than others. 42 Writing a verbal
language, mastering chess, and solving differential equations take years to learn,
largely because all rest upon hard-core conventionality. Other conventions can be
picked up fast because they are functionally similar across cultures. Some countries
require you to drive on the right side of the road, and others on the left, but the idea of
ordering the traffic flow is congruent across both. Still other conventions require only
the slightest adjustments of our natural proclivities. In a picture, if the most impor-
tant element occupies the center of the format, viewers from any culture will probably
not be surprised. Centering (manifesting the principle of symmetry) is in some sense
a convention of pictorial composition, but it seems to run with the grain of our visual
predispositions, taking the line of least resistance. Strategic decentering, on the other
hand, may be a convention that requires a little more tutoring.
Films use conventions. In most movies, characters face each other in an odd way:
Their bodies and faces are conveniently tilted in 3/4 view for the camera. Scenes are
cut according to the tactics of continuity editing. We may hear music that does not
issue from the locale of the scene, and a dissolve or fade may convey a passage of
time. Still, such conventions are mostly of the quickly learned variety. Many of them
piggyback on our natural predispositions; others require only slight adjustments.
Several amplify and streamline regularities of human interaction, as when movie
characters talking to one another stare more fixedly and blink far less than they
would in real life. We understand movies fairly easily because in many respects their
conventions are easy to learn: They are simplifications of things we already know.
Of course, a particular filmmaker may wish to block that easy understanding—to
be, as we say, unconventional—but very often, she will have to tap into other capaci-
ties and proclivities we have. If the story is told out of order, then we will need some
redundant cues to that design as well, such as Pulp Fiction's replay of the opening
dialogue when the action returns to the diner for its climactic scene. Nevertheless,
a great deal of what is conveyed in a movie is conveyed naturally—through those
perceptual-cognitive-affective universals that are part of our biological inheritance.
This story, I believe, is likely to be true. Yet it would be stoutly rejected by most
film scholars. The reasons are partly due to certain strongly held opinions within the
humanities, and partly due to the history of film studies as an academic discipline.
My fuller version of the story can be read elsewhere, 43 but in brief it goes like this.
The framework I just sketched presumes contingent universals of h u m a n makeup
and experience, but most scholars in the humanities tend to doubt the existence (or the
80 Poetics of Cinema
importance) of empirical universals. Further, the framework hypothesizes causal and
functional explanations for social practices. Most humanists, though, prefer interpre-
tation to explanation. When they do seek explanations, they rule out biological causes
or functions as too deterministic and prefer some form of social-learning theory. The
framework I traced also takes rational-empirical inquiry, of which science is our
most successful exemplar, as the most promising way to explain cultural practices.
But academic humanists on the whole mistrust science and, sometimes, rational-
empirical inquiry more generally.
Film academics are on the whole even more suspicious of this framework than
their peers in other disciplines, I believe. This is largely because film studies, entering
university humanities departments in the late 1960s, became rather quickly attached
to certain doctrines. Most of these, such as semiotics and psychoanalytic theory, were
deeply antinaturalistic (at least in the versions that became influential). Although these
particular doctrines have lost their grip, an extreme version of cultural constructivism
is at the base of most film studies. Consider just a few premises.
All personal experiences—identity, concepts, feelings, and even perceptions—
are socially constructed.
(Constructed out of what? That matter is not addressed.)
Because everything is socially constructed, there is no such thing as a more or
less realistic representation; every sign is equally arbitrary.
(Can the concept of the arbitrary sign be intelligible without a concept of
the nonarbitrary sign? Shouldn't one then consider the possibility that
there might be nonarbitrary signs? And why should all signs be equally
arbitrary? These questions are not asked.)
Realism is a myth because no representational system provides total access to
some "reality out there" (if, indeed, such a thing exists).
(Doesn't this set the bar unreasonably high? A realistic representation need
not preserve all aspects of its referent in order to be reliable, as we see in
architectural design and forensic photography. But these objections are
caricatured as "naive realism.")
Every culture creates its own web of meaning. There may be "hybridity" when
cultures come in contact, but there is no universal human culture.
(If every culture is sui generis, how could theorists have grasped enough
features of alien cultures to arrive at this generalization? This self-
refutation isn't considered.)
Despite some claims that the discipline has become more pluralistic since the 1980s,
premises like these, invoked ritualistically in the literature and taught by rote and exem-
plar in courses, have become operational assumptions of most academic film writing.
Film studies also got off on the wrong foot methodologically. Instead of framing
questions, to which competing theories might have responded in a common concern
for enlightenment, film academics embraced a doctrine-driven conception of research.
Academics embraced a scholastic conception of their work, holding that certain
theorists had revealed core truths and that their gospel could be applied, in a more
Convention, Construction, and Cinematic Vision 81
or less mechanical fashion, to particular movies. First came Laura Mulvey's "gaze"
theory, then postmodernism, then versions of identity politics, multiculturalism, and
"modernity theory"—none weighed as candidate answers to a puzzle or problem but
were accepted unskeptically, then used to churn out interpretations of film after film.
Film studies remains, in a word, dogmatic.
In these circumstances, the appearance of Moving Image Theory: Ecological
Considerations can only be welcomed. The editors have assembled a distinguished
cast of empirical researchers and film theorists to explore, within a naturalistic
framework, the ways moving images mesh with our minds. Every essay bristles with
insights and fruitful suggestions for further reflection and experiment, and all point
toward ways of reconsidering some of the tenets I've already outlined.
Take, for example, the very issue of ecological psychology. Most generally, this
means treating evolutionary considerations as one constraint on theorizing about
the psychology of film. It is one of J. J. Gibson's longest-lasting contributions to have
brought evolutionary issues into the study of perception. At this level, any examina-
tion of moving-image media that reckons evolutionary constraints or tendencies into
account—as Torben Grodal, Dolf Zillmann, and Dirk Eitzen do in their accounts in
Moving Image Theory—deserves the name ecological. From the same adaptive per-
spective, certain candidates for contingent universals can be illuminated by robust,
nuanced overviews like that provided by Ed Tan's discussion of facial expression, or
by John Kennedy and Don L. Chiappe, who (fittingly enough in a paper on meta-
phor) offer us the image of human cultures as islands linked into an archipelago by an
unseen common ground—what used to be called human nature.44
We also encounter ecological theorizing in a narrower sense, that is, as proceeding
from the "direct-perception" theories Gibson developed. Here we find essays ranging
from selective treatment of some ideas (e.g., William Evans' contribution) to explo-
rations of Gibson's system as a whole (e.g., the essays by Claudia Carello et al. and
Sheena Rogers).45 I have no competence to assess the latter, but they promise to be
as much of value to psychologists as to those of us interested in the psychology of
art. Perhaps they will link up in time to the emerging ecological strain in cognitive
theory, such as Gerd Gigerenzer's concept of "ecological rationality." 46
Encouragingly, all these essays allay any concerns that a Gibsonian view commits
one to preferences for art that preserves natural appearances. Nearly every study
shows how a realist psychology gives special meaning to artists' efforts to violate
ecological validity (Shaw and Mace), to defeat our normal responses as well as to
build upon them (Zillmann and Grodal again), and to create filmmaking traditions
that preserve certain invariants and stylize others (for example, Cutting on—what
else?—cutting). 47 This is a subtle and supple realism, one that takes veridicality as
a bridgehead—biologically, perceptually, and cross-culturally—and then shows how
conventions might arise out of systematic revisions or rejections of it.
The difference between these contributions and most current film theory might
boil down to this: Contemporary theory assumes that cinematic representation is
almost wholly conventional (and the conventions come from culture); what is not con-
ventional is very little (often called "physiology"!) and not very important. According
82 Poetics of Cinema

to the ecological view, cinematic representation relies on a great many nonconven-

tional capacities and processes, and the conventions are correspondingly small in
number and easy to learn—riding as many of them do upon just those ecologically
constrained processes.
It's appropriate that Gibson developed his perceptual theories out of his work with
cinema. As a lieutenant-colonel in the U.S. Air Force during World War II, he was in
II
charge of testing films that could train pilots. In trying to simulate the problems of
identifying other aircraft and landing on airstrips, he was led to treat human vision
not as a snapshot but as a flowing optic array—oriented to the horizon, displaying Studies in Narrative
texture gradients, and illuminated by light from above.48 Movies were very effec-
tive in teaching young pilots what flying looks like, and so Gibson had pretty hard
empirical reasons for adopting a realist perceptual psychology. Film, he understood,
could faithfully capture essential features of a life-or-death situation. If we invited
today's postmodern academics to come up with reliable ways to represent airplane
maneuvers, I shudder to think what casualties would result. But maybe not, at least
once the researchers got off the ground. If there are no atheists in foxholes, then
perhaps there are no culturalists in cockpits.

83
r

3.

Three Dimensions

of Film Narrative

A man sitting in a bar suddenly shouted, "All lawyers are assholes!"

Another man jumped off his stool. "Those are fighting words!"

"Oh, so you're a lawyer?"

"No, I'm an asshole."

The study of narrative has a long history, but as a self-conscious body of inquiry, this
enterprise is principally a creature of the twentieth century. It was then that it came to
be called narratology, an ugly term but one that apparently we can't easily do without.
Whatever we call it, the study of narrative is very important. Storytelling is a per-
vasive phenomenon. It seems that no culture or society is without its myths, folktales,
and sacred legends. Narrative saturates everyday life too. Our conversations, our
work, and our pastimes are steeped in stories. Go to the doctor and try to tell your
symptoms without reciting a little tale about how they emerged. The same thing
happens when you go to court or take your car to a mechanic or write a blog. Perhaps
storytelling is part of human maturation, because it emerges quite early in human
development. Children only two years old can grasp certain features of narrative,
and there's evidence from "crib monologues" that the narrative ordering process is
emerging even earlier. We share stories with each other, assuring others that we have
experiences congruent with theirs. Sometimes we tell a joke, like my curtain raiser, to

85
86 Poet ics of Cinema
Figure 3.1 A Trobriand Island string figure: Figure 3.2 When one fighter wins, he departs
The headhunters face one another. with the enemy's head.
create a bond—though after some experience, I'd advise you that this one won't create
deep ties in certain situations.
We can apparently turn anything into a story. String figures akin to Cat's Cradle
may tell tales. Figure 3.1, from the Torres Straits, represents one stage in a fight
between headhunters: The two warriors are squaring off. The player then tugs on the
left-hand loops, and the headhunters clash. The outcome can't be predicted. Both
fighters may die and fall apart, or one kills the other and "travels home," bearing
the enemy's head (Figure 3.2).1 In Australian Aboriginal sand paintings, what might
seem to outsiders to be abstract squiggles and whorls represent mythical events or
2
incidents from daily life.
Narrative appears to be a contingent universal of human experience. It cuts across
distinctions of art and science, fiction and nonfiction, literature and the other arts. So
it's not surprising that studying narratives brings together students of not only literary
studies, drama, and film, but also anthropology, psychology, even law and sociology
and political science. Narratology is a paradigm case of interdisciplinary inquiry.
Widespread as narrative is, though, it retains a distinct identity. Considered as
a thing, a certain sort of representation, a story seems intuitively different from a
syllogism, a database, and an fMRI scan. My opening joke isn't exactly like other
forms of humor, such as a bumper sticker ("Today is the day for decisive action! Or is
it?"). How should we try to capture narrative's uniqueness? Perhaps narrative is like
grammar in a natural language, or perhaps it's a sign system, like traffic signals, as
semiotic theories suggest.
Narrative is more than a kind of thing; it seems to involve distinct activities as well.
One activity we call storytelling, and the other . . . well, what do we call it? Story con-
sumption? Story receiving? Story pickup? In any event, we have capacities that enable
us to grasp and present stories. This talent too opens up many questions. From one
angle, our stories come from our psyches, involving mental contents and processes.
The very act of remembering something is coming to be seen as less a retrieval of fixed
data than an ongoing construction according to principles of narrative logic. 3 Yet
narrative is as well preeminently social, a way of organizing experience so that it can
be shared. Narrative conventions invoke lots of particular knowledge, and my opening
joke wouldn't be understood in a culture that lacked bars, lawyers, and lawyer jokes.
Narratives activate social skills, and although some people become expert storytellers
(some can tell 'em, some can't), nearly all of us recognize well-formed stories when we
encounter them. Our narrative competence relies on social intelligence.
Three Dimensions of Film Narrative 87
Distinct as narrative seems, it's also polymorphous. It blurs and blends into a lot
of other forms and activities. In a novel, it's often hard to carve out the descriptive
passages cleanly from the plot, because accounts of people crowding a train station or
skiing easily pass into little suites of action. The rhetorical tradition, theorizing about
what persuades audiences, recognizes that stories can carry weight in an argument;
the summary of the facts of a law case were known to the ancient Romans as the
narratio. I could use my joke to illustrate an argument about why lawyers get no
respect or a tirade about what conservatives call the coarsening of our culture. Peter
Greenaway's film The Falls (1980) provides a purely categorical macrostructure—a
directory of people whose last names begin with the letters Fall—but soon we find
that every Fall- has a life course full of incident.
In their turn, stories are omnivorous, consuming other forms. Japanese literature
includes the genre of travel journal, which is in prose but often splices in descriptive
verse passages. Frank Capra's film The Battle of Russia (1942) spends a fair amount of
time cataloguing all the types of people living in the USSR. Mikhail Bakhtin argued
that the novel was impelled to interweave contrasting voices, but it may be that all
sorts of narrative have an appetite for assimilation.
One reason that narrative emerges as a distinct area of study rather late is that for
centuries it was identified largely with spoken language. According to ancient tradi-
tion, a narrative was a story told, whereas a story that was enacted was considered
drama. The rise of film, comic books, and the like encouraged theorists to rethink
things. Now narrative is usually considered a transmedium activity. A story can be
presented not only in language but also in pantomime, dance, images, and even music.
My lawyer joke could manifest itself in a comic strip, a radio skit, or a TV sketch. In
certain respects, we can think of narrative as a preverbal phenomenon.
Still, language remains our most important way of communicating with one
another, and language-based narrative is our default. (We do call it story telling.) So
what are the connections between verbal narrative and other sorts? Perhaps the other
sorts derive from verbal storytelling. We might be able to follow the string figure
battle and the Aboriginal stories in sand only thanks to verbal cueing. Perhaps a child
learns to understand TV shows and movies based on the fairy tales she has heard at
bedtime. Alternatively, perhaps both verbal and nonverbal narratives tap into some
more basic conceptual skills—ideas of agency, causality, time, and the like—which we
deploy to make sense of anything we encounter. Once you have the idea of a person,
you can understand characters' identity, motives, and the like, whether you meet
them in the pages of a book or on the screen.
Such questions aren't just splitting hairs. How we answer them can shape how we
analyze particular stories in different media. A great many narratologists seem to
believe that language-based narrative is the Ur-form, to which other media approxi-
mate. If language sets the agenda for all narrative, then we ought to expect all media
to follow along. So in a film the analyst will look for equivalents of first-person point
of view, or something analogous to the voice of a literary narrator. But if we think that
language is on the same footing as other media, a vehicle for some but not all more fun-
damental narrative capacities, then we might not expect to find exact parallels between
88 Poetics of Cinema
literary devices and filmic ones. Different media might activate distinct domains of
storytelling. Perhaps, that is, filmic point of view might be quite different from literary
point of view, and there may be no cinematic equivalent of a verbal narrator.
For all these reasons, it seems fair to say that in studying narrative we ought not
to forget that narrative can engage people quickly and deeply. A simple joke like the
one I started with, only 40 words long, can trigger a laugh. We reflect on narrative
because it's powerful on many dimensions. It rivets our attention; it focuses our
perception; it arouses our emotions; it teaches and pleases. But how? By what means?
What enables us to grasp and follow a story? What gives stories their enormous power
over mind and emotions?
I'd argue that our most fruitful line of investigation starts with our ordinary
understanding. Narratives exploit proclivities, habits, and skills we take for granted—
sharpening them, twisting them, and subjecting them to confirmation or question-
ing. Narratives use folk psychology, which is notoriously unreliable in certain matters
but nevertheless remains our court of first resort. In real life, it may not be fair to
judge someone on our first impressions, but we do, and narratives capitalize on this
tendency by introducing characters so that their essential traits pop out clearly.
Likewise, when I say that narratives rely on causality, I don't mean that it yields strict
deductive entailments. Because people devise narratives outside the lab, it's likely that
the kind of causality at stake won't meet the standards of scientific inquiry. Something
like commonsense reasoning or folk causality is likely to be the plausible candidate.
In studying narrative, poetics has to be more psychological than ontological. The
principles, practices, and processes we detect are unlikely to be models of rigorous
reasoning. But, then, neither are most of the ideas we entertain.
Some First Moves
For a poetics of the cinema, then, narrative begs for examination. We can start by
offering a first approximation—a toy model of the phenomenon we're trying to under-
stand. Rather than asking, "What is Narrative?" let's try for something a little more
tractable: "What is a narrative?" N a r c o l o g i s t s share a fair amount of agreement on
what a narrative looks like, though there are two principal ways of understanding it.
One tendency I'll call action-centered, the other agent-centered. From the action-
centered perspective, a narrative consists of certain elements arranged in time. The
elements are events and states of affairs. My bar joke gives the state of affairs at the
start—two men in a bar—and the events consist of what they say and do. Those
elements, arranged in time, constitute the narrative presented in the joke.
Some action-based theorists think that this doesn't go far enough. If the events
are merely connected by succession in time, we could come up with some fairly
strange stories.
On July 6, 1947, a flying saucer crashed in Roswell, New Mexico.
On July 23,1947, Marjorie Bordwell gave birth to a son, David.
On July 23,1948, D. W. Griffith died.
Three Dimensions of Film Narrative 89
Confronted with this bald string of events, we might call it a chronology or a
chronicle, but we're disinclined to call it a story. For one thing, we'd probably require
that some agents reappear; an individual ought to be undergoing at least some of the
events presented. For another, this doesn't feel like a story unless we can posit some
causal connections among the events. We'd need a sense that the alien arrival had
an effect on my birth, or that my appearance on earth is connected to the death of
Griffith. For such reasons many theorists, including me, think that both some conti-
nuity of agent and some causal connection are conditions of a minimal narrative.
In addition, an action-based theorist of narrative might remind us that a narra-
tive requires not just events in time but also change. Travel narratives change place,
psychological narratives change characters' attitudes or temperaments, and mystery
stories change the state of characters' knowledge. One thing we expect of stories is
what Aristotle called peripeteiae—changes of fortune from bad to good or good to
bad. Even our barroom joke presents changes in behavior and in our knowledge
(concerning the lengths a person will go to avoid being considered a lawyer).
This action-centered notion of minimal narrative can be traced back to Aristotle's
Poetics. Against it we can set a conception that's often identified with Romantic and
post-Romantic literary criticism. Someone might argue that all this talk of "events,"
"states of affairs," and "causality" turns narrative into a bloodless abstraction. When
we think of narrative, we think first of characters. For Aristotle, a narrative is a whole,
and agents take up a place in a larger rhythm of event-driven activity. But we can treat
the agents and their capacities as the basis of narrative, with events seen as products
of those qualities. Historically, this perspective was influenced by medieval and
Renaissance theories in which character was conceived as a mix of vital humors or
dispositions. In a reaction to neoclassical norms of proper writing, theorists pointed
to Shakespeare. His plays seemed to be weak on abstract plot geometry but unsur-
passed in their portrayal of human behavior. A. W. Schlegel wrote that Shakespeare
created unique individuals who act spontaneously but plausibly. Shakespeare endows
"the creatures of his imagination with such self-existent energy that they afterwards
act in each conjuncture according to general laws of nature." Shakespeare doesn't
laboriously tot up all of a character's motives, for that could suggest that each one's
identity is simply the sum of larger forces. "After all, a man acts so because he is so."4
It's not that this view disregards plot as such. Whereas Aristotle sees human agency
as a part of a total action, Schlegel believes that the abstract structure of events flows
from the display of human personality in the process of change. Maybe most people
would agree. They think of narratives, or at least the most valuable ones, as portraits
of human minds and hearts. True, the page-turner, the book we read with unquench-
able interest, might seem to cater to our action-based appetites. Yet even then, many
will say, we read on because we're held by characters who arouse our passions. Still, it
seems to me that the drastic split between plot and character, derived from Romantic
theory, has led to a kind of caste system, whereby character-driven stories are felt to
be inherently superior to ones that showcase suspense, excitement, and unexpected
twists. For one thing, supposedly character-driven narratives often turn out, on exam-
ination, to have a rich action-based architecture too. Shakespeare's plays are marvels
90 Poetics of Cinema
of construction, and the indie films supposedly putting character on display often
obey many conventional plot mechanics. Moreover, narrative offers many pleasures,
from psychological probing and nuanced social observation to imaginary adventure,
thunderous surprises, and Grand Guignol shocks. Flaubert and Dumas, Trollope and
Conan Doyle tap into different sources of narrative pleasure, and it's not clear that a
Merchant-Ivory adaptation is more satisfying or accomplished than Die Hard.
In any case, what follows tries to outline what I take to be a promising poetics of
filmic narrative. It suggests that we can look for constructive principles and normal-
ized practices along three dimensions. None of those dimensions is rigidly biased in
favor of action-based or agent-based models of a story, but in my application of them,
probably my predilections will shine forth.
Protagonists and Their Problems
Taken singly, the three dimensions I'll be considering seem to me uncontroversial.
All have been considered before in the vast literature on narratology. But in spreading
them out side by side, I think we gain a sense of the rich array lying open to analysis
from the standpoint of poetics.
One dimension involves what I'll call the story world: its agents, circumstances, and
surroundings. In my opening joke, that world consists of a bar (and all of the presumed
furnishings of a prototypical bar). A second dimension is that of plot structure, the
arrangement of the parts of the narrative as we have it. My joke is structured as a
series of actions and reactions, statements and replies. It has a neat symmetry (two
lines from each of the two participants), and it builds to a payoff, the punchline.
The third dimension I propose is that of narration, the moment-by-moment flow of
information about the story world. The narration of the joke is laconic, never describ-
ing the bar or the men or even how they're arrayed in the bar (except that one is
apparently on a stool). We are outside the men's minds, Hemingway fashion, whereas
other jokes are resolutely subjective. All three dimensions contribute to the point of
the joke. Though this analogy shouldn't be pressed too far, the story world is similar
to the semantic dimension of language, plot structure is comparable to grammatical
or syntactic structure, and narration is comparable to verbal style, as governed by
pragmatic context.
Before I consider each dimension separately, let me provide an example of how
making these distinctions can help us with problems in poetics. We commonly believe
that a narrative film is likely to have a protagonist. But how do we determine who or
what a protagonist is? I suggest that several dimensions of judgment are involved,
most ingredient to all narratives in any medium but one specific to cinema.
In the story world that the narrative presents, the protagonist is the agent whom the
story is about. There are many heuristic cues that help us pick out a hero or heroine.
The protagonist may be the character with the greatest power, as King David is in
certain chapters of the Old Testament. The protagonist may also be the character
with whom we tend to sympathize most keenly, as in the biblical story of Daniel. The
protagonist may be the character with whose value system we are assumed to agree.
Three Dimensions of Film Narrative 91
Or the protagonist may be the one who is most affected or changed by events, as in
James' Portrait of a Lady.
No one of these cues is decisive on its own. After watching The Godfather, many
viewers would say that Michael Corleone's wife, Kay, arouses more sympathy than
either Don Vito or Michael, and the dons' value system is unlikely to be wholly
endorsed by us. Michael especially seems a cold protagonist, like Tamburlaine. More
important, though, Vito and Michael are the most elevated characters, with the power
to decide life and death, and Michael is evidently the character who changes the most
in the course of the action. These criteria seem to weigh heavily in this story world.
Don Vito and Michael are spotlighted by narrative structure as well. The major
portions of the films pivot around them, from Don Vito's attempted assassination
to Michael's escape to Sicily. Were we to divide the film into large-scale parts, or
long chapters, the breaks would reflect major changes in their fortunes. Moreover,
the actions of these two men, both proactive and reactive, dictate the overall shape of
the plot. Don Vito's decision not to join the drug-selling business set up by Sollozzo
triggers the gang war that follows, and Michael's decision to assume his father's place
in the family business guides events along the course they take in the second half of
the film. Structurally, the character whose actions give the drama its distinctive arc
is likely to be the protagonist, as the etymology of the term suggests. Agon refers to a
contest or competition, and so the protagonist is "the first combatant," whereas the
antagonist is the warrior who opposes the protagonist.
But wait, somebody might say. In The Godfather the plot developments are really
triggered by Sollozzo's decision to start a drug business, and Don Vito merely responds
to that initiative. Why isn't Sollozzo the protagonist? Similarly, later plot develop-
ments are responses to Sollozzo's decision to wipe out Don Vito. Our intuition, of
course, is that Sollozzo is not a protagonist but an antagonist, but how do we justify
that impression?
Here we can usefully invoke our third dimension of narrative craft, that of narration.
The Godfather is designed to concentrate our attention on the doings of the Corleones,
not of the Sollozzo gang. Significantly, we don't spend much time with Sollozzo when a
Corleone isn't present. One quick measure of how narration can suggest who is a pro-
tagonist involves registering how long a character is onstage. Scenes including either
Don Vito or Michael Corleone consume nearly 75% of the duration of The Godfather,
and Michael appears in nearly half of it.5 No other characters receive nearly this much
screen time. It seems likely that the more pages or minutes devoted to a character, the
more likely we are to take him or her as a protagonist.
Just as important as sheer quantity of coverage is the way narrational restriction
attaches us to the family. We know, by and large, what Don Vito, Sonny, Tom Hagen,
and Michael know, and in Michael's case we often know it in depth. Many scenes
access his moment-by-moment psychological reactions, as when he sets up the fake
hospital protection for his wounded father or when he assassinates McCluskey and
Sollozzo. True, his final revenge scheme isn't spelled out in advance. But our earlier
access to his mind makes our realization that he's coldly ordered a massacre all the
more shocking. To put it loosely, the action of The Godfather is presentedfrom the point
92 Poetics of Cinema
of view of the Corleones, and most often that of Michael. In the spirit of Rosencrantz
and Guildenstern Are Dead, we could imagine recasting the film's narration to create
a story told from the side of Sollozzo and his allies, in which the Corleones are distant
figures. But that's not the movie we have.
Considering all three dimensions, I don't think we can come up with a single or
simple definition of how we know a protagonist. In grasping any narrative, we weigh
the dimensions comparatively. We tacitly assay a character's prominence in the story
world, her structural role, and her narrational salience. Often these factors will dove-
tail neatly. In The Untouchables (1987), Elliott Ness is clearly the protagonist. He is
powerful and sympathetic in the story world, and his character undergoes the great-
est change, moving from ineffectual rectitude to a hardheaded willingness to fight
fire with fire. His value system gives the film its moral compass. Structurally, Ness is
a prime mover; his all-out campaign against A1 Capone breaks neatly into large-scale
patterns of thrust and parry. And as is often the case, narration provides our point
of entry. Ness is the figure to whom we're restricted most closely throughout. We see
nearly all the action "from his side" and sometimes through his eyes.
Cinema, like theater and dance, has one other means of reinforcing our inferences.
Although I'm reluctant to treat it as a dimension on the same level as the others, it's
worth pointing out because I don't see that it has a parallel in literature. Often we take
the film's most famous star to be the protagonist, and usually we're correct. In many
films, the star factor reinforces the others, as when Kevin Costner is top-billed in
The Untouchables. Ancient Greek theater defined the protagonist not only as the
prime character but also as the play's "first actor."
True, filmmakers have sometimes relegated big-name actors to secondary
roles. But that just means that the star criterion has been outweighed by the oth-
ers. Going to The Untouchables on its opening weekend, we might expect that the
presence of Sean Connery's name in the credits would make his character Malone
equal to that portrayed by Costner. As we watch the film, though, we understand
that the actions of Malone in the story world (serving as guide and mentor, not
making the ultimate decisions) and his place in the unfolding structure (entering
fairly late, murdered just before the climax) work against our considering him the
protagonist. For all his rugged authority, Malone is a helper, not a hero. Being less
central in the fictional world, in the overall structure, or in the narration is what
makes a star play second fiddle.
In a later essay, I'll be proposing that the three dimensions, plus the ancillary input
of the star system, can firm up our intuitive sense that some films have two, three, or
more protagonists. For now, it's enough to see how poetics can clarify the principles
governing what we take for granted. At this point, though, those critics who find
taxonomies to be hairsplitting might protest. Isn't it artificially tidy to distinguish the
factors that govern our sense of who the protagonist is? Lots of stories play fast and
loose with such functions. Psycho starts by attaching us to Marion Crane before she
is killed, obliging us to follow Norman Bates' trajectory for a while before picking up
Marion's sister and boyfriend as the next vehicles for our knowledge and sympathy.
Don't such instances make hash of neat categories?
Three Dimensions of Film Narrative 93
Of course artworks constantly cross theoretical categories. Nevertheless, here as
elsewhere, by drawing distinctions we can illuminate how the aberrant cases work. We
already intuit that Psycho shifts the protagonist function from one agent to another,
and more radically than in The Godfather. The news is that it does so by exploiting
all the dimensions I've traced out. The narration first attaches us to Marion, in both
range and depth of knowledge. When she dies, she's fulfilled her structural role—but
the movie has lots of time yet to run. So she ceases to be the protagonist, even though
she's the movie's biggest star. Norman becomes a protagonist because he's the new
focus of narration, and he launches story action by trying to cover up his mother's
crime. By spelling out the conditions governing clear cases, we can understand what
makes fuzzy cases fuzzy.
Narration
In line with the introductory essay in this book, I propose that we conceive the poetics
of film narrative within a framework that's mentalistic. That is, we ought to assume
that a film cues spectators to execute operations, and one central goal of these opera-
tions is to comprehend the story.
So I propose an inferential model of narration. Instead of treating the narrative as
a message to be decoded, I take it to be a representation that offers the occasion for
inferential elaboration. As per the model of spectatorship I offered earlier, I suggest
that given a representation, the spectator processes it perceptually and elaborates it on
the bases of schemas she or he has to hand. These schemas aren't necessarily codes in
the strict sense, because many are loosely structured, semantically vague, and open-
ended. Still, the elaboration isn't wholly a matter of individual taste either. If you and I
see a driver swigging out of a bottle and swerving his car along the road, we'll probably
both suspect that he's under the influence. The conclusion isn't guaranteed: The bottle
might contain iced tea, and he might be avoiding roadkill we can't see. But our infer-
ence about DUI is more plausible. Films rely centrally on just such garden-variety
inferences; it's one of the ways in which narratives trade on real-world knowledge.
By focusing on comprehension as an inferential elaboration, I might seem to be
ignoring the role of emotions in responding to narrative. Isn't this a cold, cold theory?
But this objection would misunderstand how inquiry works. Consider an analogy.
People are often emotional when they speak, but it's legitimate and useful to have a
theory of language that focuses on how language is structured for understanding,
regardless of what emotions are summoned up by certain sentences. If a wife says to
her husband, "Pack up and get out," Chomsky's linguistics has little to say about the
anger she may be expressing. Rather, Chomsky's theory concentrates on how syntax
makes the sentence intelligible. Different theories pick out different features of the
phenomena they try to explain. It would be as unfair to say that "my spectator" feels
no emotions as to say that Chomsky's "native speaker" feels none. There's a degree of
idealization involved in focusing only on comprehension, but it isn't harmful if we
grant that it's only one aspect of our experience of narrative.
94 Poetics of Cinema Three Dimensions of Film Narrative 95

I would claim, however, that with respect to most narrative cinema, comprehen-
sion must play a role in emotional uptake. It would be odd to say, "That film moved
me deeply but I found the story incomprehensible." However we explain the emotions
generated by narrative, a large part of those emotions relies upon making basic sense
of the story. We can't feel poignancy at the end of Late Spring or satisfaction at the end
of Stagecoach without at least partly understanding the events that have led up to these
climaxes and the impact those events have upon the characters. If you're interested in
how people respond emotionally to narratives, an account of comprehension would Figure 3.3 The opening of The Silence of the Figure 3.4 The Silence of the Lambs: She
presumably contribute a lot to your inquiry. Indeed, this is just what's happening. After Lambs (1991): Agent Starling comes out of the turns and runs into the forest; the camera
story world to meet us. follows and carries us into the story action.
I floated this comprehension-centered account of narrative in the mid-1980s, several
scholars who wanted to pursue questions about emotional response built upon narra-
tional concepts. 6 This is a natural and salutary way scholarly inquiry proceeds.
Someone might go on to say that my belief in convergences of comprehension is
naive. Women don't comprehend stories as men do, and people in Japan don't under-
stand their stories as Europeans do. Note that this objection does presume some
convergence, if not between social groups then within them. Why believe that only
certain groups share understanding and others can't share it? Why can't comprehen-
sion strategies crisscross groups in that hybrid fashion beloved of postmodernists? Figure 3.5 At the end of The Silence of the
Moreover, because comprehension involves such features as tracking psychological Lambs, Hannibal Lecter turns from us and
follows his prey into the story world, but we stay
states, causality, time shifts, and the like, the onus is on the critic to show that women
behind. This is a conventional mark of closure.
or cultural insiders possess different senses of cause and effect or time relations than
other perceivers do. One of the most commonly cited examples is that in watching
a Western, Native American audiences might cheer on the Navajos attacking the that follow a character from behind in exploratory fashion (Figures 3.3-3.4). On the
settlers. Even this apparently apocryphal anecdote, however, doesn't damage my
soundtrack, music sets a mood, and dialogue rises to audibility. Clearly all these tac-
case. I assume that the audience understood the story—that the settlers were crossing
tics are blended to engage the spectator's interest, parceling out information needed
Indian land, that the Indians wanted to wipe out the settlers—and that the viewers
to understand the action. The cut-ins or forward camera movements also suggest
took sides in a way not anticipated by the film's makers. To say that there's conver-
that we are being drawn gradually into the story world. Strikingly, the ending of an
gence in understanding is not to say that all spectators act upon their understanding
ordinary movie often reverses these devices. The camera pulls back, characters turn
in the same ways.
away and we don't follow them, doors and gates may shut, the music rises again, and
By focusing on comprehension from a mentalistic perspective, I hope to adhere
titles may appear (Figure 3.5). The opening literally opens up the movie and lets us
to other conditions I set out earlier. In accord with my layout of spectatorial activi-
in; the closing shuts it down and expels us. The best explanation for these regularities
ties, I assume that there's a fair amount of convergence in viewers' understanding of
onscreen is that they're manifesting principles that filmmakers share, perhaps tacitly,
the narrative. There may be some disagreement among spectators' grasp of character
motivations or consequences, and we should expect this, given the variety of schemas and they function to shape our experience of the story.
that viewers bring to films. But divergences in comprehension aren't anything like as The symmetries between openings and closings suggest that narration is a system
wide and varied as we'd find in interpretations, for reasons I've already suggested. that's put into motion across the whole film. All the factors we normally associate
Again, in accord with the sort of poetics I'm proposing, this study of narrative with narration—play with the order of events, shifts in point of view, and voice-over
treats films holistically. My conceptions of narration, plot structure, and the story commentary—shape our overall experience. They're not just one-off tactics; they
world try to take into account the overall form of a film. The assumption here is that play roles in larger patterns running across the entire movie. So once we've identi-
regularities we find across the whole artifact allow us to make inferences about the fied a passage of omniscient narration or optical point of view, we should go on to
purposes of its makers and the activities coaxed from its viewers. Take the openings look at how that functions in the broader organization of the narration. Why shift
we find in ordinary movies. Very often we get an expository title giving time and to optical point of view here? How does it shape the experiential logic of the overall
place, along with one or more long shots of an area of action. Cuts or camera move- film? Narration is more than an armory of devices; it becomes our access, moment by
ments may carry us into a scene, with characters moving toward us, or tracking shots moment, to the unfolding story. A narrative is like a building, which we can't grasp
96 Poetics of Cinema
all at once but must experience in time. We move from static spaces to dynamic ones,
enclosed spaces to open ones, peripheral areas to central ones—often by circuitous
routes. That journey has been arranged, and sometimes wholly determined, by archi-
tectural design. Narration in any medium can usefully be thought of as governing our
trajectory through the narrative.
This analogy helps us see that we don't gain by treating narration as something
like an envelope enclosing the story action. As a process, narration burrows all the
way down into the material, shaping it for our uptake. It governs how we grasp overall
structural dynamics and the immediate scene before us. It controls how we build an
inferential elaboration of any event.
Consider this sentence:
A boy saw a woman kissing a man.
By narrating the event this way, I've shaped your inferences, identifying certain
features of the action and eliding others. (We don't know the relationships among the
three characters.) Now try this rendition of the same action:
Tim saw Dorothy kissing Wally.
I've not only named the agents but also encouraged you to posit a relation among
them; Tim, Dorothy, and Wally are unlikely to be strangers to one another. By provid-
ing their first names, I've also encouraged you to assume a certain familiarity with
them. In Rex Stout's detective stories, we know Archie Goodwin by his first name and
as I (because he's the narrator), but we know Nero Wolfe chiefly by his last name. Who
would dare call him Nero? By such simple means does literary narration conjure up
intimacy or distance. This makes it rather off-putting when Dashiell Hammett calls
his protagonist "Ned Beaumont" throughout The Glass Key. We don't really know our
relation to the enigmatic figure. But like most narrative devices, this piggybacks on
our normal social interchange, with first names as marks of intimacy.
Let's return to our example, with another change:
Tim saw Mommy kissing Daddy.
Now Dorothy and Wally are presented in terms that coax us to infer a specific relation
to Tim. The sentence doesn't say he's their son, and it's possible he's not (as in the case
where a daughter tells us about her boyfriend, "I was so embarrassed that Tim saw
Mommy kissing Daddy").
Still, the narration has opened up a new range of inferences. It's only a short step to
I saw Mommy kissing Santa Claus.
Employing a traditional cue in literary narration, the I that replaces the Tim anchors
us in Tim's consciousness. The Mommy gives us better grounds to infer a kinship
with the speaker than the earlier example. The big trick comes with Wally-Daddy's
new guise. By renaming Wally, the sentence invites us to think that Daddy is dressed
up as Santa and the I doesn't know it. Here we have to go far beyond the data given,
elaborating every proper noun according to what's most likely.
Three Dimensions of Film Narrative 97
Of course, every construal is defeasible. It's easy to imagine scenarios that would
demand different inferential moves. Perhaps the I is not a child but a naive space alien
who has inferred that the woman is named Mommy and the man is named Daddy.
Perhaps Saint Nick himself is really in the house and Mommy is having a torrid affair
with him. No narration can roadblock every detour. It can only try to shape the most
likely construal by specifying the context and conjuring up the appropriate schemas.
We speak of "following a story," and that suggests that we take up leads offered to us
without seriously losing our way.
For the storyteller, choosing between narrational vehicles always has both costs
and benefits. Any one of the Mommy-Daddy accounts gives us certain information
but denies us other bits. In the first case, the neutral nouns boy/woman/man don't
give us the agents' names or personal relationships; in other versions, we know one
but not the other. This is what I mean by saying that narration goes all the way down,
into the very texture of the event. It's one reason that the theory of narration has to
include matters of film style. It's not that a piece of story action is a single kernel event
to be rendered in a variety of ways (though it's helpful to imagine alternatives). As
we watch, in real time, online so to speak, we take the event as the narration presents
it. Visual and auditory techniques are rendering the event for us, already organizing
and slanting it in a certain way. Consider a simple case, somewhat parallel to our
Mommy-Daddy-Santa instance.
Two characters are talking to one another on the telephone. The filmmaker faces a
number of choices for rendering this event. First, we can see both characters exchang-
ing dialogue, perhaps via crosscutting, split screen, or some other technique. As a
result, following the turn taking of the dialogue, we hear the entire conversation.
Alternatively the filmmaker can, throughout the conversation, show us just one of the
pair. But that offers a further choice: Shall we hear what the offscreen speaker says, or
not? If we hear the speaker but see only the listener, we can observe the reaction to the
lines. Instead, the filmmaker might eliminate the sound of the speaker's dialogue, so
that we don't get access to what's coming through the earpiece. In this case we see the
speaker's reaction, but we have to imagine what's being said that provokes it. In sum,
each choice narrates the phone call in a different way, doling out different informa-
tion for different purposes. In a comedy, we might want to see both characters speak
their lines and react to each other. In a mystery, it might serve the scene's purpose to
omit one side of the conversation, so we don't know who the speaker is, or whether the
speaker is sincere, or why the listener reacts as she or he does. All of the presentational
tactics I've mentioned—crosscutting, split screen, eliminating a sound stream, pre-
senting the sound coming into the receiver—are stylistic choices, but they're inevitably
narrational choices as well. They shape what information we get and how we get it.
I've said that through narration, the film encourages us to indulge in inferential
elaboration. What is the product of that process? Basically, what we call the story.
Most of our inferences are merely enforced perception. Our eyes and ears turn a con-
figuration of images and sounds into the simple output "The hero is running down
the street." But even this apparently brute uptake will go beyond the data given. We'll
presume that "The hero is rushing to a wedding," or "The hero is fleeing his pursuers,"
98 Poetics of Cinema
or we'll ask, "Why is the hero in such a rush?" As we encounter a stream of such con-
figurations, we build up a story world of characters, relationships, motives, decisions,
reactions, and all the rest. The configuration itself, the arrangement of information
for the sake of pattern and point, will have its own structure, as we'll see. But for now,
we should concentrate on the way that organizing the givens coaxes us to build that
story world in a particular way.
Narratologists have long distinguished between the organization of the action in
the narrative text and the action as it's presumed (inferred, extrapolated) to occur
in the story. Aristotle referred to praxis, all the events constituting the action, and
muthos, those events as structured into a plot. 7 Theorists influenced by French struc-
turalism of the 1960s distinguish between histoire (story) and discours (discourse). I've
found it most useful to follow the Russian formalists in using the concepts of fabula,
the story's state of affairs and events, and syuzhet, the arrangement of them in the
narrative as we have it. In addition, recall my claim that the fine grain of the medium
shapes our construal of events, as in the Mommy-Santa sentences or the phone call
menu. So I would add that narration must include the patterning of the film's surface
texture, its audiovisual style.
Tying all these ruminations together, and utilizing the inferential model I've
proposed, here's my claim. I take narration to be the process by which the film prompts
the viewer to construct the ongoing fabula on the basis of syuzhet organization and
stylistic patterning. This is, we might say, the experiential logic of understanding a
film's narrative, the equivalent of the tourist's guided path through a building.
Now it should be clearer why I haven't employed the story-discourse couplet. The
term discourse harbors a certain ambiguity because it suggests patterning at several
levels. Take one of the most common examples indicating the principled independence
of the two: the flashback. Here story events that occur early in the chain of events
are displaced and shown or told about later; the discourse rearranges the story. But
discourse also implies something more fine-grained, the texture of a spoken or written
language, or perhaps, in film, a shot's composition or the nature of a cut.
Discourse, in effect, bundles my concepts of syuzhet and style together. You might
ask, Why keep them apart? Theoretically, it allows for a bit more discrimination.
Practically, we'll sometimes encounter films in which syuzhet patterning and stylistic
patterning are out of sync. In films displaying what I call parametric narration, style
comes forward as a distinct organizing principle.
By treating narration as the process of guiding our comprehension of the story,
I don't mean to suggest that stories aim at full disclosure. Filmmakers want us to
construe the story, moment by moment, in a certain way, and that way can involve a
lot of diversions and blind alleys. Narration can mislead us. Yet in order to mislead
us, it has to rely on our making certain inferences about causality, ordering in time,
and the like. A common strategy is the unmarked ellipsis, whereby we're encouraged
to ignore a time gap that the narration doesn't flag—only to later come to understand
that something important took place in that gap. This ploy is at work in Fritz Lang's
You Only Live Once (1937) and The Blue Gardenia (1953) and Otto Preminger's Fallen
Angel (1945). Alternatively, by restricting our knowledge to what only one character
Three Dimensions of Film Narrative 99
knows, the narration can mislead us about events, only to surprise us later when we
get a fuller account of what happened. This is common in detective stories and film
noir tales. Cinematic narration has a great many resources. The syuzhet can juggle the
order of fabula events, providing a flashback or flashforward. It can manipulate fabula
duration, stretching out or compressing the time that story events consume. It can
present simultaneous fabula events successively (via crosscutting), and successive
events simultaneously (through split screen or other devices). Cinematic narration
can be more or less knowledgeable, claiming greater or lesser access to information,
and more or less self-conscious, flaunting the act of presenting this story to various
degrees. The syuzhet can provide an omniscient range of knowledge, as when a film
intercuts characters' trajectories, or it can restrict the flow of information to what one
character knows, as some detective films do. Stylistic devices like optical point-of-
view shots, voice-over commentary, and sound perspective can funnel information
through a character's literal standpoint. A common pattern of cinematic narration
is to attach us to one character for a scene or two, then move to another character's
range of knowledge, creating a sort of shifting restrictiveness. Cinematic narration
can also be more or less objective, remaining resolutely on the "outside" or pulling us
into characters' minds via memories, dreams, or imaginings.
Cinematic narration overlaps with literary narration, but the two aren't perfectly
congruent. For instance, filmic "point of view" is rarely as stringent and sustained as
the literary variety. A first-person narrator in a novel restricts us to a single conscious-
ness, but a film's voice-over narrator can initiate the revelation of events that she
didn't witness, or even know about, as in Ten North Frederick (1958). A long-standing
convention holds that literary storytelling mimics storytelling in life, whereby every
tale has a teller and receiver (reader, listener). This communication schema works
well for many novels, though perhaps not all. (Who "tells" a montage-based novel
like Dos Passos' USA trilogy?) In any event, a film's syuzhet and style aren't bound by
the constraints of verbal communication. Cinematic narration, being an audiovisual
display rather than a written text, appropriates bits and pieces of the communica-
tion model opportunistically. So we can have voice-over commentary from the pro-
tagonist without there being any indication that they are speaking to anyone in the
fictional world. The commentary may be taken as stream-of-consciousness musings or
as simply another conduit for story information, without any need for the real-world
baggage of speaker-listener relationships. I expand on this idea in my discussion of
the problem of narrators in cinema, which serves as an appendix to this essay.
"I don't like voiceover as exposition," Steve Martin remarks of his film Shopgirl,
"because I don't think anyone is listening." 8 No one, except the only one who matters:
the viewer. At the start of Jerry Maguire, the hero's voice introduces us to his lifestyle
and his personal crisis, and then his voice vanishes, never to return. To whom was he
speaking? The question is as irrelevant as the physics of light sabers. The film doesn't
need to anchor his discourse in a full-fledged communication situation because it
recruits part of the communication template to get information out to us. Communi-
cative logic can go hang; all that the narration cares about is cueing us to make the
right inferences.
100 Poetics of Cinema
Structuralist thinkers have brought many of these processes to light, creating use-
ful taxonomies of temporal manipulation and point of view. From the standpoint I m
indicating here, I suggest that we need to put taxonomies into motion, so to speak,
by considering the characteristic sorts of activities that distinct categories tend to
encourage. For instance, many narratologists have rightly celebrated Gerard Genette's
layout of temporal possibilities, but few have recognized that they elicit rather different
activities when situated in certain contexts or different media. Theoretically, straight
chronology is on a par with juggled time sequence, but psychologically chronology
operates as a default. It's our presupposition in following events in the world, let alone
events in narratives. Another abstract option is this: If two fabula events are occurring
simultaneously, you can present them successively or simultaneously in the syuzhet.
But literature is ineluctably successive (words follow one another), and on the page you
can't strictly show two things happening at the same time. In reading we have to infer
simultaneity from the bits of action presented moment by moment. Film, however,
presents simultaneous action very easily, both within the shot (one character in the
foreground, say, and another in the distance) and in split-screen imagery.
Meir Sternberg has been the most eloquent and persistent advocate for treating
taxonomic categories functionally. He has argued that what matters is that all the
strategies charted by the taxonomists must be gauged in relation to their capacities
to create distinctive effects on the perceiver. For example, a flashback isn't just an
abstract rearrangement of story incidents. Its function is to trigger interest in finding
out what led up to what we see. Sternberg suggests that by considering three aspects
of our narrative appetites, we can offer good functional explanations for particular
devices. Curiosity stems from past events: What led up to what we're seeing now?
Suspense points us forward: What will happen next? Surprise foils our expectations
and demands that we find alternative explanations for what has happened. Syuzhet
arrangements of events arouse and fulfill these cognition-based emotions. Sternberg's
account of the experiential logic of narration fits well with my concern for a poetics
of effect. 9
In this sequence of words, which one doesn't belong?
Skyscraper Temple Cathedral Prayer
Most people would say Prayer, because the first three terms refer to types of buildings.
But if the words are presented in this sequence:
Prayer Cathedral Temple Skyscraper
people usually say that Skyscraper is the outlier, because the first three items refer
to religion. This is what psychologists call the primacy effect. The order of events
governs how we understand them, and the first item has greater saliency. Likewise,
a film's opening will set a benchmark against which we measure what happens later.
The characters we first encounter, the point at which we enter the story action, and
other elements will shape our inferences. Sternberg speaks of the "rise and fall of first
impressions," pointing out how the narration can create distinctive effects by letting
us trust too much in what we see at the outset. 10 This syuzhet strategy has been put to
Three Dimensions of Film Narrative 101
good use in films like The Usual Suspects (1995), which makes us revisit initial action
and rethink what we thought we knew.
My account of narrational uptake may seem cerebral and juiceless. Surely, our
inferential elaborations are bound up with emotions? They are, and just as modern
cognitive science presupposes that emotions operate in tandem with perception and
thought, so I'd readily grant that our time-bound process of building the story is
shot through with emotion. Murray Smith, for example, has traced how the com-
plexities of narration can tie us to or separate us from the emotions the characters
are undergoing, creating "structures of sympathy" or dissonances between what he
calls alignment and allegiance. Thus the narration may signal us that even though
we're tethered to what a character knows, other cues indicate that we are not to ally
ourselves to that character's moral frame of reference, so that our response may blend
sympathy, empathy, and emotional distance. 11
I've had to be peremptory in surveying filmic narration, but I've discussed the
subject at length elsewhere, and it will recur in some of the essays that follow.12 An
additional advantage of treating narration from the standpoint of poetics is that
it lets us track different storytelling traditions. Classical Hollywood construction
may distract us along its path to the end, but eventually we arrive at fairly definite
and reliable inferences. By contrast, other traditions, such as that of "art cinema,"
open gaps that aren't closed, trigger inferences that don't have clear-cut conclu-
sions, and use fluctuating patterns of time and space to create a more unreliable
presentation of events. Films such as Toto le heros, Blind Chance, and Les Passagers
set into motion narrational systems that don't resolve at either the level of the story
action or that of syuzhet organization. Such films give the spectator an experience
of patterned ambiguity about events or states of mind, a play among competing
schemas, and an invitation to interpret the film more abstractly. By thinking of
narration along the lines I've sketched, we're in a good position to make our poetics
of storytelling comparative.
Finally, some people have objected that by emphasizing the flow of information
about story states and actions, I make films too dependent on revelations and plot
twists. Every movie becomes a mystery story, my critics suggest. But in an important
sense every narrative does depend on uncertainties, the most basic concerning what
will happen next. Beyond that, nearly all narratives rely upon unevenly distributed
information. Very simple stories, such as counting narratives like "The Twelve Days
of Christmas," don't display disparities in characters' knowledge, but in most cases
narration obliges us to reckon, Who knows what? Aristotle pointed out that the tragic
plot carries its protagonist from ignorance to knowledge, but most plots carry at least
some characters in this direction. Sternberg points out that any story action relies
upon gaps and miscalculations: "No ignorance, no conflict; and no conflict, no plot."
He quotes Henry James: "If we were never bewildered, there would never be a story to
tell about us."13 The interplay among agents' range of knowledge and ours shapes the
curiosity, suspense, and surprise we feel in engaging with the story, whether it has a
corpse in the library or not.
102 Poetics of Cinema
Plot Structure
The term plot structure can mean many things. I'm using it to refer to the way in
which the syuzhet is patterned in itself, without regard to the strategies by which the
narration presents t h e f a b u l a information. A prototypical example of plot structure
thinking would be Jane Smiley's claim that a novel falls naturally into four parts:
exposition, rising action, climax, and denouement. 14 These divisions bear wholly on
the syuzhet. The rising action may be a flashback, the denouement may shift point
of view, but none of these narrational techniques alters the abstract action-based
geometry she proposes. (For reasons adduced above, however, I think that the term
exposition isn't a good one for naming a portion of the plot. Exposition is best thought
of as a function-driven process of narration, because it can occur at any point in the
plot.) If the narration is like our trip through a building, the plot structure is like the
building as we might reconstruct it in a blueprint—an abstract, quasi-geometrical
layout of parts according to principles of size, proportion, and contiguity.
Understood this way, syuzhet structure in effect organizes the actions and states
of affairs in the story world according to a certain pattern of development. Usually,
there is some sort of change, and often some conflict, within the story world, and
the syuzhet structures it according to widely recognized principles. As usually stated,
though, ideas of rising action, climax, and denouement are quite vague. The same
goes for "desire encountering obstacles and finding fulfillment"—a fair summary of
many, many stories but still pretty vaporous. Can we make conceptions of plot struc-
ture more precise without losing some general applicability? Seymour Chatman has
pointed out that it's very difficult to provide a paradigm of narrative macrostructures
as precise as any we can provide for narration. He voices a justified skepticism about
structuralists' efforts to find a grammar of action that would govern every story we
might encounter. He reminds us that the action units into which we break a body
of tales are governed by our tacit understanding of what audiences in various tradi-
tions are supposed to make of them. Would a certain piece of action be considered a
"betrayal" or a "sacrifice"? What makes something a rising action? Until we can find
a generally agreed-upon basis for marking out the units, he recommends that we start
by focusing on single works and genres. 15
Looking at the grab bag that writers come up with in conjuring up the 7 or 10 or
36 "basic" plots, I can only agree with Chatman's hesitations. 16 From the standpoint
of theoretical poetics, it does seem unlikely that we'll generate a precise taxonomy
of structure applicable to all narratives. Historical poetics, however, can usefully
trace how particular traditions have built up fairly broad principles of plot structure.
Again, Aristotle leaves us some pointers. He evidently thinks that a tragic plot can
be described in a hierarchical fashion, with each level identifying different organiza-
tional strategies. In its widest compass, the plot has a beginning, a middle, and an end,
according to what triggers and concludes the chain of actions. More specifically, that
chain would also consist of a complication and a denouement. More specifically still,
tragic action consists of episodes leading from pathos to reversal to recognition. Even
Three Dimensions of Film Narrative 103
if this layered conception of structure would not apply to comedy and epic, Aristotle's
distinctions are valuable tools for showing principles of construction in tragedy.
Perhaps we can find more local principles guiding other sorts of plot structure.
As a first approximation, let's distinguish between internal and external concep-
tions. Internal models treat the syuzhet's pattern of actions according to some macro-
structural principle of design. The best examples are those that invoke geometrical
figures. A rising pattern of action can be visualized as a curve or vector. Gustav
Freytag's "dramatic pyramid" conceives the plot action as leading to a central climax
or principal turning point, the apex of a triangle, followed by a decrease in tension
(the anticlimax). When we speak of frame stories and inset stories, we're evoking
brackets or bookends. Similarly, when we encounter stories embedded in stories that
nest inside still other stories, it's hard to avoid thinking of rectangles enclosing other
rectangles. The Locket (1946) displays this "Chinese box" structure, with one flash-
back inside another, and both inside a third. E. M. Forster spoke of Henry James'
novel The Ambassadors as having the shape of an hourglass, with two lines of action
meeting at a central juncture. 17 Alternatively, distinct lines of action can be conceived
as forming parallel lines, or as entwining into a braid, with the trajectories splitting
and converging at crucial points. 18
These schemes of plot structure don't have universal validity, but they can be
heuristic guides to analyzing particular narratives or bodies of work. Thus it may be
helpful to think of the pair of stories in Chungking Express as giving the plot a dumb-
bell shape: two tales linked by one character passing between them. A later essay
in this book considers how some narratives rely on a model of network affiliations
connecting a wide range of characters.
External structures—principles for segmenting the plot by some metric not
derived from the action patterns—have a bit more historical solidity because they're
acknowledged by filmmakers fairly explicitly. One example is reel-by-reel plotting.
Here the film is chunked into groups of scenes that correspond to the length of a film
reel in projection. In the years before 1912, fiction films usually consisted of only one
reel. Projection speeds weren't standardized, but the maximum running time per reel
was about 15 minutes. So a technological constraint served as a simple boundary
for the entire story to be told. As films became longer, they were broken into several
reels. In theaters with only one projector, the end of one reel would be followed by
an interval while the projectionist threaded up the next. Even at theaters equipped
with two projectors, there might be a distinct pause between reels. Recognizing that
the presentation would be segmented, filmmakers began to build their dramatic arcs
around the reel break. Urban Gad, a Danish director who immigrated to Germany
in 1912, noted that the "mechanically necessary interruptions" demanded that the
film be divided into "acts," each one leading up to a gripping scene just before the reel
change. 19 By labeling these acts with expository titles, filmmakers invoked theatrical
precedent and perhaps also hoped to borrow some of the stage's prestige. The break-
down could be labeled in less standard ways, too; in Lang's epic Siegfried, each reel is
entitled a lay, as in a bard's song.
104 Poetics of Cinema
By the mid~1920s, most European theaters had two projectors, so there was less
need for a reel to end on a strong note. But in the USSR, single-projector venues were
the standard. Directors were accordingly advised to break the films into well-defined
20
parts. Some filmmakers, wanting to make the audience aware of large-scale form
in their films, exploited the reel structure to articulate the action of their plots quite
vividly. Sergei Eisenstein is the most famous instance. He broke Strike into six episodes,
each marking a phase in the prototypical strike. He gave Potemkin five parts, then split
each reel about halfway through, creating symmetrical actions around a caesura.
The arrival of synchronized sound standardized running speed at 24 frames
per second, making reel length 1,000 feet, or 11 minutes maximum. 2 1 Films were
shipped on 1,000-foot reels, but the biggest venues had projectors that could handle
bigger reels, so many projectionists doubled up to reduce changeovers. Hitchcock was
counting on this practice when he alternated visible cuts with camouflaged ones in
Rope (p. 33). Reel structure in world cinema still needs to be fully researched, but one
recent instance is intriguing. In Hong Kong cinema of the 1970s, script construction
became fairly loose. Filmmakers preferred to build their plots additively, stringing
together comedy, fights, and chases. One company, Cinema City, began planning its
films reel by reel, demanding that each reel contain at least one comic scene, one
chase, and one fight. Color-coded charts revealed immediately which reel lacked the
necessary ingredients. The practice influenced most directors who emerged in the
1980s, even the elusive Wong Kar-wai. His wispy plots look more firmly structured
when you realize that they're built up reel by reel in postproduction. The fragmen-
tary martial arts drama Ashes of Time (1994) devotes reel 1 to the primary protago-
nists, the swordsmen Evil East and Poison West. The plot spends its next two reels
on the story of the Murong brother-sister couple, then devotes reels 4 and 5 to the
Blind Swordsman. The film finishes with a three-reel denouement involving the
protagonists and the woman they both love. Shooting without a finished script and
welcoming spontaneous digressions, Wong used the Hong Kong tradition of reel-by-
reel construction to shape his masses of footage. 22
Reel construction is a fairly loose metric for plot structure. Provided with merely
a proportional segmentation, the filmmaker must still create more specific patterns
of action that will fill it out. Perhaps the closest analogy is the word count assigned to
serial publication of a novel's chapters, or the standard number of lines per verse in
epic recitation or popular songs.
Internal and external criteria blend in one of the paramount conceptions of struc-
ture at work in mass-market cinema today—the notion that a film narrative divides
into distinct acts. Across the history of drama, act structure is a vexed question.
Horace proposed that five acts were the correct number, and this precept was adhered
to for centuries by playwrights and publishers in England, France, and Germany.
Spanish dramatists of the sixteenth and seventeenth centuries promoted a three-act
structure, which Hegel praised as the most theoretically correct design. (It neatly
echoed his thesis-antithesis-synthesis triad.) But the five-act conception persisted
through the 19th century, encouraged by Gustav Freytag's influential argument that
Three Dimensions of Film Narrative 105
plot structure pivoted around a climax coming midway through the play. By the early
twentieth century, operas and plays seem to have favored three acts.
What of cinema? There's no doubt that the analogy between dramatic acts and film
is fairly forced, especially once there were no longer breaks between reels. Perhaps
screenwriters adopted the three-act model simply because people took at face value
Aristotle's remark that every plot has a beginning, a middle, and an end. Although
there's some evidence that the three-act structure held sway during the classic studio
years, it was widely disseminated in screenwriting manuals after the 1970s, chiefly
thanks to Syd Field's influential book Screenplay.23 Field claims that Hollywood films
adhere to a three-act structure, having the rough proportions of 1:2:1. In the first act
(25-30 minutes into a 2-hour film), a problem or conflict is established. The second
act, running about an hour, develops that conflict to a peak of intensity. The final half
hour or so constitutes a climax and denouement. Field translated this structure into a
screenplay's page counts, with each page counting as roughly a minute of screen time.
This plot anatomy has been taken virtually as gospel in the U.S. film industry,
with producers expecting submitted screenplays to adhere to it. It is as fundamental
to screenwriting as the 12-bar blues structure is to pop music. The three-act template
has been endlessly tweaked, recast, and filled out. With scholastic zeal, although
seldom with scholastic acuity, commentators have discussed what kind of action is
appropriate for each act, such as "backstory" during the first act and resolution in
the last. Most writers agree that the end of the second act should be the "darkest
moment," the point at which things seem to be utterly hopeless for the protagonist.
Yet getting there can pose problems; "the desert of the second act" is the toughest
stretch, most writers agree. An alternative to the three-act template was proposed by
the distinguished screenwriter Frank Daniel. He taught that the plot can be analyzed
into eight sequences, each running about 15 minutes. Still, this isn't a drastic chal-
lenge to conventional wisdom, because these sequences can easily be slotted into the
broader pattern of three acts. 24
Within film studies, and specifically within the research tradition of poetics, the
most salient revision has been proposed by Kristin Thompson. 25 She argues that since
the late 1910s, an American feature film tends to be constructed in 20-30-minute
chunks, each marking a distinct phase in the plot. The parts are defined not only by
running time but also by the formulation, redefinition, and achievement of goals by
the protagonist. According to Thompson, the film's Setup section endows the pro-
tagonist with a set of goals. The following section is the Complicating Action. This
recasts or even cancels the initial goals and ends with a new set of circumstances
governing the action. This situation may serve as a "counter-Setup," reversing the
conditions that governed the first part. Thompson calls the next section the Develop-
ment, launched at approximately the midpoint of the film, in which efforts to achieve
the goals are thwarted. Although there may be some forward movement in the main
action, some portions are likely to be rather static, emphasizing subplots, character
revelation, or simple delays. Characteristically, the Development ends with a piece
of action that puts the achievement of goals into a crisis. The plot's final section con-
stitutes the Climax, in which the protagonist definitely achieves or doesn't achieve
106 Poetics of Cinema
the goals. The Climax is often followed by an Epilogue, which asserts that a stable
situation has been achieved.
Thompson's account, tested and refined in relation to many films, is an induc-
tive generalization, and as such it usefully refines the three-act template. Instead of
describing the events that lead into the next act as incidents that spin the action in
a new direction," the most common formula for a "plot point," she is able to specify
that the principal character will define or change the relevant goals. Her model also
allows that not all portions of the plot will be proportional. Indeed, it turns out that
the Climax of a film is seldom exactly as long as the Setup. The Setup usually runs
25-30 minutes, but climaxes tend to be 20 minutes or so. Thompson recognizes as
well that a film may not run exactly two hours, a problem for the three-act template.
She suggests that a shorter film may display the four basic parts, or it may possess
only three, deleting either the Complicating Action or the Development. Likewise,
a longer film may have two Complicating Actions, two Developments, or even, as
in In Cold Blood, two Climaxes. In all these respects, Thompson's account is a func-
tionalist one, based on major changes that take place within the plot action, and not
simply on external measurement of minutes or page lengths.
Few would deny that You've Got Mail (1998) is a pretty formulaic movie, but
studying its structure along these lines helps sharpen our sense of how the formula
works. Running 115 minutes, the film fits Thompson's model snugly. The Setup intro-
duces the classic Hollywood dual plotline: a line of action devoted to work and a line
devoted to romance, each of which will affect the other. Joe Fox and Kathleen Kelly
correspond on e-mail without having met, and we can tell that they're falling in love.
This will cause problems later because each is living with another lover. Just as impor-
tant, Kathleen runs a cozy children's bookshop, whereas Joe is scion to a bookstore
chain that is expanding into the neighborhood. This first section also establishes Joe's
father, Joe's grandfather, and their two young kids, who are technically Joe's aunt and
brother. As often happens, the Setup has its own midpoint, a scene in the Fox offices
where the firm's expansion plans are announced. The Setup ends when Joe takes the
kids to Kathleen's bookstore and he becomes attracted to her, but he's torn by the
realization that his family's superstore will wipe out this hospitable family business.
The two protagonists meet at the 25-minute point (though neither knows the other s
cyber-identity), and the Setup ends about three minutes later.
The following 63 minutes consume what traditional screenwriting practice calls
the second act, but Thompson's layout allows the finer grain of the plot mechanics
to be revealed. At the start of what she labels the Complicating Action, the Fox store
opens and Kathleen's business slumps. When the two couples meet at a neighborhood
party, Joe's girlfriend Patricia and Kathleen's boyfriend Frank are attracted to one
another. At the same party, Kathleen learns that Joe is her competitor and a squabble
ensues. In later scenes, whenever they meet face to face, they quarrel, but as anony-
mous correspondents they confide their hopes and fears to each another. Joe urges
Kathleen to fight back (not knowing that he's her opponent), and she takes his advice,
asking Frank to write a news story about her shop. At first Kathleen had thought that
her store could live peacefully alongside the Fox behemoth, but she changes her goal,
Three Dimensions of Film Narrative 107
and this marks a turning point in the business-based line of action. But the romance
doesn't develop until the online correspondence leads Joe and Kathleen to make a
date to meet. Just before he encounters her, and before she sees him, he recognizes
that his correspondent is his adversary. He comes to the rendezvous not as her cor-
respondent but as the Joe Fox she loathes, and he doesn't admit to being her e-pal.
This pivotal moment occurs roughly at the midpoint (60 minutes), and constitutes a
major complication.
The Development section is based on the narrational division of knowledge. Joe
knows that Kathleen is his correspondent, whereas Kathleen still harbors romantic
ideals about the unknown man to whom she writes. But after her Galahad stands her
up, her hopes are dashed. She confides her feelings to him in e-mail, and Joe dithers—
at first not replying, then blurting out apologies. He feels guilty not just for standing
her up but also for putting her bookshop in jeopardy. As this series of scenes illustrates,
stretches of the Development may mark time, creating a fairly static situation that
allows characters to reveal themselves. On the business front, Kathleen drifts into the
Fox megastore and helps a customer find a children's book, a scene that Joe observes
with remorse. When Joe and Patricia are trapped in an elevator, he realizes he has no
one to love. The development also allows subplots to play out. Frank and Kathleen split
up, and so do Patricia and Joe. Kathleen's plight gets increasingly serious, and falling
business drives her to close her shop. The section ends, at the 91-minute mark, with Joe
deeply unhappy and Kathleen wandering through her hollowed-out shop, taking a last
look. Screenwriters would say that this is the darkest moment.
The Climax, as often, is a relatively short section, running about 20 minutes. Joe's
father, who's oblivious to his own empty life, says that he's never found anyone to fill
his days with joy, and this impels Joe to visit Kathleen. Her attitude is softening, but
she resists because she admits she's in love with her e-pal. We also learn that she is
writing a children's book—a hint as to the resolution of the workplace line of action.
Joe suggests that she arrange a meeting with her correspondent, and over several days
gently but teasingly continues to court her. Eventually he admits that if things were
otherwise, he would have tried to marry her. She heads to her e-mail rendezvous and
discovers that Joe is her correspondent. "I wanted it to be you so badly," she says, and
the plot is resolved with a kiss. There is no Epilogue; none needed.
The four-part plot structure articulates phases of the action. Joe and Kathleen meet
at the end of the Setup, Kathleen launches her struggle against the superstore at the
end of the Complicating Action, and she loses the battle at the end of the Develop-
ment. The film's narration also fits itself to the four-part structure, putting us ahead
of Joe until the midpoint, when he learns his e-pal's identity, and ahead of Kathleen
until the climax. The film is stuffed with secondary characters, motifs, montage
sequences, whiffs of pop tunes, and comic bits—all characteristic features of classical
construction. Yet part of our sense that this movie plays by the numbers comes from
its adherence to a proven plot structure.
Is picking a romantic comedy like You've Got Mail shooting fish in a barrel? Let's
take an example that might seem less formulaic. The first 29 minutes of Boyz N the
Hood (1991) are devoted to the childhood of Tre Styles, a boy who lives with his father
108 Poetics of Cinema
and falls among aimless neighborhood kids. This portion shows the father, Furious
Styles, trying to keep Tre straight in the face of crime, poverty, drugs, gang strife,
and police hostility. At the end of the Setup, after an encounter with a gang of bullies,
Tre sees his pals Doughboy and Ricky arrested for shoplifting. The title "Seven Years
Later" makes the Complicating Action into a countersetup, establishing variations
on the initial premises. Tre is working as a salesman, Ricky is a football star, and
Doughboy flits in and out of prison. New goals are established. Tre, Ricky, and Tre's
girlfriend Brandi hope to go to college. Ricky is recruited by the University of Southern
California, Tre tries to convince Brandi to have sex, and Tre is offered the choice of
coming back to live with his mother, who's now prosperous. The section ends at the
first hour mark with Tre, Ricky, and Brandi taking a college entrance exam.
With the new premises in place, the Development pushes some forward and
leaves others hanging. As is common in this phase, there are delaying maneuvers.
Tre's mother continues to press him to return to her. In a 4-minute interlude, Furious
lectures his son and others on the need to keep black neighborhoods whole and to
resist drugs and guns. Alongside these fairly static situations, Tre's romance with
Brandi develops. Most crucial is a tense confrontation with a street gang (whose
members appeared in the Setup and the Complicating Action). The Development,
running about 28 minutes, ends with Ricky's death at the gang's hands—a turning
point that forces Tre to make a choice. In the Climax, he abandons Brandi and his
father to join his pals on a mission of revenge. The narration intercuts their search
for the gang, Furious waiting anxiously at home, and Ricky's grieving mother learn-
ing that he passed his college entrance tests. But Tre has a change of heart and leaves
his friends, who go on to wipe out the gang members. Next morning, Doughboy
cracks his tough facade to confess to Tre his loneliness, fear, and despair at the cycle
of violence. Titles provide an epilogue. Doughboy will be murdered 2 weeks later,
presumably in revenge, whereas in the fall Tre will attend Morehouse College "with
Brandi across the way at Spelman College." The protagonist's twin goals—striving for
a better life and achieving romantic union—have been achieved.
Boyz N the Hood gives human weight to abstractions about youth, crime, drugs,
family ties, and hope within black urban communities, and it does so through a plot
that follows the four-part template as faithfully as does You've Got Mail. This tradi-
tional structure can smoothly absorb a variety of subjects and thematic materials.
Someone might argue that these models of plotting invite you to read in what you
expect to see. Because you expect something important to happen around minute
25, you'll tend to exaggerate the importance of whatever happens at that juncture.
You're looking for three acts or four parts, and you massage the film to fit it, but some-
one else could plausibly claim that the film consisted of 7, or 17, parts. Aren't these
measures just ad hoc? I don't think so. Although all events in a plot may contribute to
the overall progression, some intuitively stand out as significant moments, and others
are clearly secondary. There's a lot of agreement among us as to what those moments
are, and they occur, with a frequency greater than chance, at the points and with the
consequences that Thompson's model predicts. No one would argue that the visit of
a college recruiter to Ricky's home isn't significant for the action of Boyz N the Hood,
Three Dimensions of Film Narrative 109
but it isn't as central to all the characters' fates as Ricky's death is, and that's the event
that arrives at a canonical juncture in the running time. Granted, analysts can dis-
agree about a particular film's structure; Thompson and I don't divide The Godfather
in precisely the same way. But such disagreements are common within any critical
tradition. Musicologists may disagree about the most perspicuous way to analyze a
particular melody, but all accept the premises of phrasing and harmonic progression
that give the tune its identity.
Thompson's layout is a helpful tool for analyzing films made along classical lines,
laying bare constructional principles that seem widely used. Yet it raises some intrigu-
ing problems about the explanatory power of a poetics. Do viewers recognize these
distinct parts? No, because people are usually surprised when told of them. It seems
that this architecture achieves its effects without the audience's conscious aware-
ness; only experts detect the armature. This fact need not count against Thompson's
account, because listeners with no musical training can react properly to a song or
symphony without being aware of the mechanics of harmonic modulation, retrograde
inversion, and other techniques.
What, though, about the practitioners? How could they obey rules that they don't
consciously know? For example, if the three-act structure is the formula guiding
filmmakers, how could they have embraced a four-part structure? I can imagine
several possibilities. First, nearly all writers acknowledge that the lengthy second
act is difficult to write. It wouldn't be surprising that scenarists made this stretch
tractable by tacitly breaking it into two roughly equal chunks. Secondly, at least one
writer (speaking after Thompson's first study was published) has acknowledged that
the three-act structure is best thought of as harboring four parts. Akiva Goldsman
remarks that a screenplay consists of "four acts, or really three acts, but the second
act is really two acts, so we might as well call it four acts, and they're generally
30 pages long."26 From this angle, splitting the second act would be a craft habit that
just doesn't rise to the level of awareness. We shouldn't assume that all creators have
an engineer's grasp of what they're doing. Usually they're just following a tradition
whose features they've intuitively grasped, and the tradition gets replicated without
a lot of self-conscious reflection.
Just as interesting, screenwriting manuals recommend that scripts have the three-
act structure, whereas they derive their timing recommendations from finished films.
We know, however, that scripts are constantly modified in the production process, and
the film as shot can be recut in many ways. It would be a miracle if everyone involved,
from scriptwriter to editor, tacitly subscribed to a canonical structure without being
aware of it. Yet both the script gurus and Thompson show that the finished films
display this structure to a plausible degree. Unless we're hallucinating, somehow the
miracle does happen. Perhaps general principles of balance, plus ingrained habit
and an intuitive urge for symmetry, plus a culture-wide idea that entertainments are
digestible in 20-30-minute chunks all contribute to this tacit architecture. But the
question is far from settled.
It would be still more surprising if a comparable model reigned outside the
Hollywood tradition. Francis Vanoye has suggested that films by Claude Autant-Lara,
 Poetics of Cinema
110
Francois Truffaut, and Andre Techine adhere to the three-act structure, though he
offers no explanation of how this American paradigm found its way to France. 27
Michel Chion finds it in Mizoguchi's Sansho the Bailiff (1954).:28 The success of Field's
book in translation has probably led filmmakers all over the world to try following his
paradigm. 29 So even if the model fails for older Hollywood films, it might become the
basis of many films, from the United States and from other countries, in the future.
Apart from its heuristic value in bringing out the macrostructure of many films
in a major tradition, the act-based model of construction nicely lets us distinguish
between narration and plot structure. Consider Memento (2000). Narrationally that
is, in terms of the strategic regulation of fabula information—major sections of the
syuzhet present story events in reverse order. Yet Memento's syuzhet obeys the three-
act / four-part template, with turning points at the proper proportions. 30 Odd as it
sounds, even telling the story backward can respect canonical plot architecture.
T h e Narrative World
Most books introducing narratology start with discussions of the fabula, that
spatio-temporal realm in which the action unfolds in chronological order. Then
the author goes on to discuss how that world is rendered through patterns of narra-
31
tion—restricted point of view, flashback construction, and the like. This expository
strategy makes for clarity, but it's a little misleading. It runs athwart the obvious fact
that we have access to the fabula only by means of narration. Narration isnt simply
a window through which we watch a preexisting story that we might see from else-
where. By telling the lawyer joke at the start, I coaxed you into creating the story
world by virtue of our shared stock of stereotyped knowledge. Narration, the inter-
action of the syuzhet arrangement and the stylistic patterning of the film, is the very
force that conjures the fabula into being.
The demiurgic power of narration is especially hard to grant with respect to
cinema. Literary texts create worlds from mere words, but film presents us with a
rich array of images and sounds that immediately conjures up a dense realm. So it s
easy to succumb to what used to be called the "referential illusion," the sense that a
tangible world in some sense lies behind the screen, and that storytelling is simply
a matter of highlighting this or that moment in the world's unfolding. To some degree
this conception holds good for documentary narratives; it's indeed one presupposi-
tion of documentary as a mode. But a fiction film is narrated through and through.
Not just camera position but also the arrangement of figures in space, not just cutting
but also the movements executed by the actors, and not just zoom shots but also lines
of dialogue—everything, including the solid environment and behaviors we detect,
is produced by the film's narration. That's all we have to go on; we have no independent
access to the world portrayed on the screen. As I indicated earlier in my hypothetical
example of a phone conversation, to present an event is inevitably to choose among
ways of presenting it, and those ways constitute narration.
So something very strange is going on. The narration asks us to infer a world that
seems divorced from its representation. Once that cleavage is made, we can then con-
Three Dimensions of Film Narrative 111
sider how that world is treated (via point of view, time juggling, and so on), as if it
were a freestanding realm to which the narration merely gave us access. That is, once
we construct a world from its representation, we treat that world as re-represented via
particular narrational tactics (those very tactics that prompted us in the first place).
I'm far from offering an adequate account of exactly how this happens, but I'd
suggest that we start by recognizing how fast and easily we construct a recognizable
world populated with agents performing actions. It would be virtually impossible for
our minds to build it up piecemeal from scratch, so it's most likely that we project
onto what we see and hear a body of taken-for-granted premises. Surely some of
those premises, such as the idea that a movie's opening images will introduce relevant
information, come from our experience of films and artworks in other media. But the
bulk of those premises, it seems to me, must derive from automatic mechanisms we
use to make sense of the physical world we live in. There is just too much information
onscreen that would call for too much dedicated processing otherwise.
If the visual and sonic display onscreen conforms on the whole to our everyday
experience, we can build up a coherent story world very quickly. In effect, the default
would be as follows: In the absence of other information, assume that what you see
and hear is basically like what you would see and hear in your nonfilmic experience.
Our perceptual and cognitive capacities deliver a fast, more or less veridical grasp of
the areas of action portrayed in the image and evoked on the soundtrack. "Reading
for gist," we furnish a spatial and temporal environment for the agents. There's much
to be said about how that happens, in terms of perception and comprehension, and
elsewhere I've proposed some ideas along these lines.32
I'm suggesting, then, that narrative film calls upon the perceptual capacities I dis-
cussed in the first essay in this collection. But even as we construct the physical param-
eters of the story world, we are probing it more deeply. We ascribe effort and intentions
to the things moving on the screen. We assign agency, we trace causes and effects,
and we identify goals. Again, such activities are activated automatically in everyday
life, through a variety of means: dedicated neural circuitry, the machinery of intuitive
judgments, quick top-down deliberations, and the like. Again, the speed with which
we reconstruct the forces traversing this world suggests that cinematic narration has
fitted itself to mechanisms that we use all the time. These mechanisms, evidently, take
precedence over any explicit recognition of the processes of the representation itself.
That is, as viewers we treat the presentational vehicle (the medium and its patterning)
as secondary; we "look through" the how and concentrate on the what.33
Take an analogy. An orange looks much the same color in sunshine and in shade,
but by photometric measurement it will send off very different wavelengths under
those different conditions. We don't normally notice the fluctuating patterns of
illumination that are objectively in our environment; instead, our vision favors the
recognition of objects. By the same token, once the representational processes of film
have delivered us a recognizable world, the fine grain of those processes becomes
secondary to aspects of the "primary theory" Horton speaks of in Chapter 2.1 realize
that this position risks being called "naive realism" or "illusionism," but it isn't.
Sophisticated artifice is responsible for these effects of easy inference. It's just that
112 Poetics of Cinema
we can t pay attention to everything, so we fasten on what's salient in our everyday
mental life, such as spatial arrays and the action taking place within them. I don't see
any more plausible way of explaining the fact that we grasp novels and paintings and
movies easily, but it takes training and skill to notice the strategies informing literary
composition, representational painting, and filmic storytelling. It takes an artist's eye
to see the orange as subtly different in sunlight and shade.
My analogy is to visual perception, but I hasten to add that the mechanisms that
lock onto the film are no less attuned to social representations. We make inferences
about which characters are friends, relatives, and strangers; and who enjoys higher
status, greater strength or beauty, or more brains. We watch for signs of emotion and
thought. We bring, that is, all our perceptual and cognitive skills from the real world
to the task of figuring out the social dimension of this story, and we import anything
that we deem relevant to it.
At least until we're told otherwise. Marie-Laure Ryan has pointed out that we tend
to construe story worlds by the Principle of Minimal Departure. "We will project upon
these worlds everything we know about reality, and we will make only the adjust-
ments dictated by the text."34 The story world's departures from real-world schemas
will be signaled by the text, so if we meet a giant turtle in a film, we'll presume that the
monster will have the biology of a turtle, unless we're told otherwise. This principle
allows us to take for granted a great deal, and so the narration can piggyback on all
our real-world presuppositions. What we assume about bars, lawyers, argumentation
while drinking, and humans' sense of shame is brought to bear on my initial law-
yer joke. Even fantasy derives from our stock of everyday experience. To take Ryan's
example: If a story tells us that Babar the elephant enters a restaurant, we infer that he
is hungry. Why? Because even talking elephants who can be kings presumably have
the appetites all elephants have.35 Ryan is concerned with a slightly different problem
than mine, because she wants to understand the ontology of fictional worlds and I'm
more concerned with the folk psychology of narrative. Nevertheless, the Principle of
Minimal Departure offers one promising explanation for the rapid, unreflective way
we construct story worlds.
Most people couldn't imagine a story without characters, those person-like entities
that make things happen in the story, so I'll concentrate, in the rest of my survey here,
on characterization. To start, let me note that Seymour Chatman has pointed out
that we fill out characters through implication and inference, just as we do with story
lines. 36 In this, he agrees with Gustav Freytag:
The poet understands the secret of suggesting; of inciting the hearer, through
his work, to follow the poet's processes and create after him. For the power to
understand and enjoy a character is attained only by the self-activity of the
receptive spectator, meeting the creating artist helpfully and vigorously. What
the poet and the actor actually give is, in itself, only single strokes; but out of
these grows an apparently richly gotten-up picture, in which we divine and sup-
pose a fullness of characteristic life, because the poet and the actor compel the
37
excited imagination of the hearer to cooperate with them, creating for itself.
Three Dimensions of Film Narrative 113
Still, our imaginative activity runs along lines we can make more explicit. We
construct the characters within their narrative world as persons, and it seems to me
that we employ a schematic prototype for personhood. A person possesses a body,
presumed to be unified and singular (and thus gendered). A person perceives and is
self-aware; entertains thoughts, including beliefs and desires; feels emotions', possesses
traits, or attributes; and can launch self-impelled actions. In addition, complementary
to the concept of a person is the idea that any person can play various social roles.
This isn't a rigorously philosophical account, but rather an intuitive sketch of our folk
psychology. 38 That psychology doesn't arrive all at once, but there is strong evidence
that we're disposed to acquire this sort of information about others. We are born
pretuned to see people as people, not inert objects, and equipped with faces, insides,
and even minds. If our environment confirms these predispositions, we can go on to
learn a host of other things about our fellow creatures. As we grow, we can apply that
knowledge to understanding stories. 39
Presented with a narrative agent, we tend to project the whole cluster of schematic
features onto him or her or it. This is Ryan's Principle of Minimal Departure at work
again, because we expect the agents we encounter in the world to come supplied with
all the aspects of the schema I've outlined. So the narrative must tell us if the agent
lacks any of the critical features. In many science fiction films, we're informed that an
intelligent robot can't feel emotions; such, apparently, is the case with HAL. In many
cases, we'll ascribe characters' actions to beliefs, desires, traits, or social roles on rather
slender evidence, We assume that characters have all the person-like attributes, and
such assumptions allow us to fill gaps and inventively extrapolate. Of course many of
these extrapolations will be quick and dirty, guided by social stereotyping. My lawyer
joke relies on two conventional premises: Lawyers are scoundrels, and people don't
want to be considered assholes. (The denial of the second provides the joke.) When
Ryan, the protagonist of Cellular (2004), carjacks a Porsche, the narration depicts the
victim in quick strokes. He's a lawyer, and he's characterized in a way compatible with
our opening joke: In his cell phone conversation, he's rude, lewd, loud, arrogant, and
generally assholish. Our inferences about his personality are reinforced by the sight of
his face (aggressively beaverish) and his personalized plate (WL SU YOU 2), all sup-
ported by ethnic stereotyping (he's evidently Jewish).
The look, demeanor, and voice of the lawyer in Cellular remind us that, contrary
to literature, films present characters with distinct and identifiable bodies, and these
play a crucial role in cueing us to construct personal features for them. From the way
Sean Thornton stands and speaks in The Quiet Man (1952), we can believe he's been a
boxer. And whereas in theatrical performance the same role can be occupied by dif-
ferent actors' bodies, films tend to identify the character with the singular physical
presence of an actor. Once the actor has played other roles and become famous,
a star persona builds up, passing beyond the body and voice to other features of
personhood. Our conception that Humphrey Bogart is cynical, insolent, and worldly
wise informs both his private life and his screen characters.
Few films contain only one character, and the story world we build up is popu-
lated by an ensemble of persons, which we distinguish from each other along at least
114 Poetics of Cinema
two dimensions. We intuitively grasp a hierarchy of characters, making some more
important than others, and we do this partly because of the degree to which their
narrative functions activate aspects of the person schema. A hotel clerk may exist
solely to check our hero into a room, and thus only his body, his social role, and his
capacity for voluntary action are relevant to narrative causality. But the narrative can
characterize the clerk more fully by endowing him with superciliousness (a trait),
exasperation (an emotion), or suspicions about the heros identity (thoughts). As a
more vivid individual, the clerk will be more salient than other functionaries who
flit through the story world, as the lawyer in Cellular stands out from the other, more
anonymous drivers whom Ryan tries to flag down. If the clerk s attributes provide
causal impetus for the action, then he will move up in the hierarchy of characters. In
Cellular, the lawyer reappears in comic terms, quarreling with a policewoman, before
Ryan swipes his Porsche a second time. He is promoted to greater importance as a
more distinctive individual and as a causal factor in prolonging the action.
Apart from ranking characters in their relative importance, we quickly liken and
contrast them, using the dimensions of personhood I've indicated. Classic opposi-
tions offer clear instances: The hero may be young and virtuous with an attractive
body, whereas the opponent may be old, vicious, and misshapen. The Cellular lawyer
is selfish and unfeeling, whereas Ryan sacrifices a lot out of sympathy for Jessica, the
kidnapped woman calling on his cell. Marc Vernet points out that narratives tend to
array their characters' most salient features along overlapping contrasts and affinities. 40
In a heist film, one crook may be greedy, good-looking, and nervous; another may be
greedy but average-looking and confident; a third may be self-sacrificing but ugly;
and so on. In Cellular, Ryan and the cop Mooney are both compassionate, because
both try to rescue Jessica, but they're otherwise quite different in social roles (one is a
surf dude, and the other is a cop), bodies (handsome young versus weathered middle-
aged), and traits (impetuous but resourceful versus prudent but dogged). Other
characters display a mix of these features, along with still others. As in the world,
we contrast the people around us along various axes of personhood; but in daily life
the contrasting features run on to indefinitely large numbers. A narrative simplifies
our task by displaying contrasts along a fairly small number of axes and stressing the
salient ones as the action unfolds.
As we grasp the film's hierarchy and the contrasting features of the characters,
we make inferences. Here our social intelligence may not follow strict deductive or
inductive rules. It's now well established that informal reasoning about others relies
on heuristics, fast and somewhat dirty conceptual short cuts. The classic instance is
the fundamental attribution error. We tend to see others' actions as caused by personal
traits rather than situational constraints, whereas we tend to see our own actions as
shaped by circumstances 4 1 If you're grumpy, it's because you have a sour disposition,
but if I'm grumpy it's because I've had a bad day. In the real world, such attributions
are mistaken, but narratives rely upon them all the time to secure fast uptake. Often
we're introduced to characters in ways that encourage us to ascribe their actions to
their personalities rather than to the situation. 42 Is this tendency elicited only by plot-
driven movies that have to announce the heroes and villains swiftly? Not necessarily.
Three Dimensions of Film Narrative 115
Michael Newman has shown that even the Sb-called character-centered films of
American independent cinema, such as Welcome to the Dollhouse (1995), encourage
us to explain characters' actions by plans, desires, and character traits rather than by
situational factors. 43
Narratives play on other folk-psychological shortcuts. The primacy effect that
I mentioned earlier—the power of first impressions to establish the conceptual ground
rules—is strengthened by "belief perseverance," our tendency to resist changing a
judgment, as well as "confirmation bias," our unwillingness to entertain evidence that
would countermand an initial impression. Narratives are designed to give strong and
accurate first impressions of their characters, and rarely is a narrative designed to
introduce evidence that would make us change our judgments. Likewise, people usu-
ally don't reason statistically, but rather on the basis of vivid examples. Buyers of lot-
tery tickets can imagine themselves winning or recall the winners they've seen on TV,
whereas it's much harder to concretely imagine the odds of 13 million to one. Murder
is far rarer than suicide, but people think it's more common because they have vivid
exemplars from popular media. Perhaps this "availability" heuristic undergirds our
willingness to accept that every walk down a darkened street is dangerous, or that
lovers will accidentally meet in dramatic circumstances; it's easier to imagine them
meeting than not meeting. 44
Our shortcomings in purely logical reasoning may well stem from evolutionary
biases toward acting in the here and now, particularly when operating in small groups.
Ecological scientist Bobbi Low has suggested that our "illogical associative thinking"
stems from self-protective strategies that evolved in the context of social situations.
We are logically inept, but socially adept. One experience at being cheated,
and we are likely to generalize to future interactions with individuals of that
category. One dangerous event witnessed, and we fear it ever afterward. We
remember and overestimate the occurrence of rare (especially dangerous or
socially harmful) events and conditions. When people lived in small groups
and interacted with the same people repeatedly, this may have been a reason-
able predictor.... Although we can certainly learn logic, we nonetheless typi-
cally solve problems, at least initially, in the context of our social history. 45
For such reasons, I'll try to show in some later essays, certain narrative strategies
exploit our social intelligence by simplifying the complex negotiations that we must
conduct in everyday life.
Chief among these negotiations is what has come to be known as social mind-
reading. In dealing with other people, we need to hazard good guesses about what they
think and feel. They may speak, but do their words reflect their beliefs and intentions?
As social animals, we're inclined to cooperate, especially if we derive some benefit,
but we also know that some people will play us for suckers. So we are prepared to look
for signs of sincerity, trustworthiness, and deceit. Beyond the words people speak, we
study their vocal intonations and especially their facial expressions. Newborn babies
monitor their mothers' gazes and respond to expressions, evidently because there are
specialized neuronal cells for processing faces.46 There is considerable evidence that
116 Poetics of Cinema
five to seven emotions, along with their characteristic facial expressions, are recog-
nizable across cultures—another piece of what Horton, whom I quoted in Chapter 2,
would call the primary theory that is held in common among humans. 47
A film's story world can dramatize the entire range of mind-reading, but from a
baseline: We tend to assume that the narration's presentation is trustworthy. Once
again, this could be Ryan's Principle of Minimal Departure at work. We tend to make
the same assumption about the people we encounter, as if sincerity functions as a
pragmatic ground rule. 48 In a movie, the trustworthiness assumption is supported
by harmonious information in various channels. What the character is saying, how
she's saying it, what she's doing with her body, and what she shows on her face all tend
to reinforce our inferences about what she's thinking and feeling. Sometimes other
characters are privy to that information, but often we're the only ones witnessing the
behavior on display—which only further confirms the authenticity of the emotion. In
most films, the performers' expressive baseline tends to be somewhat more exagger-
ated than its real-world prototypes, as Ed Tan has pointed out. 49 Acting, no matter
how restrained, tends to stylize normal facial expressions of emotion.
Because sincere representation of mental states is the filmic norm, when the nar-
ration wants to show a character deceiving others, the cues aimed at us have to be
pretty strong. People are notoriously bad at detecting liars, so narratives, particularly
those in visual media, must streamline and simplify ordinary behavior. It's hard to
play a Machiavel subtly, and the victims may seem too easily taken in. "Can't they
see through her?" audiences ask when the sinister babysitter deceives the parents in
The Hand That Rocks the Cradle (1992). "She's so obvious!"
Mind-reading arouses emotion, and nowhere more so than when we're watching
faces. Facial expressions, Carl Plantinga points out, not only reveal the characters'
mental states. They also invoke "emotional contagion" in us (when others are
laughing, we tend to laugh too) and "affective mimicry" (when we copy, perhaps in
weakened form, the expressions or gestures of those we're watching). Through facial
feedback, the capacity to feel an emotion when we give our face the appropriate
expression, perhaps we can "catch" the emotion we see on the screen. All these mech-
anisms, Plantinga argues, can increase empathy, especially if our inferences about a
character's mental states allow us to imagine ourselves in her situation. 50 Accordingly,
when a character adopts a neutral expression in a charged context, we have diffi-
culty either grasping her mental states (and thus anticipating her reactions to ongoing
story events) or feeling empathy or sympathy for her. We will have hesitant, probing
responses to the flat acting on display in Chantal Akerman's Jeanne Dielman, 23 Quai
du Commerce, 1080 Bruxelles (1976), or some films by Andy Warhol and Jean-Marie
Straub and Daniele Huillet. Because the face in repose isn't completely unemotional,
a blank expression is rather unnerving.
Once again, facial expressions, gestures, and other cues for mind-reading are
brought to us through narration, as is the larger pattern of activity in which the char-
acters participate. In mainstream cinema, that activity is defined through desires and
intentions: A character seeks to achieve a goal, finds that goal thwarted, and thus is
plunged into some form of conflict. This pattern of narrative action was laid out as
Three Dimensions of Film Narrative 117
a model by Ferdinand Brunetiere in 1894. Brunetiere suggested that whereas a novel
might center on characters who merely respond to external circumstances, stage
plays demand a character who vigorously pursues his or her desire. 51 This pattern was
picked up in cinema and became central to dramaturgy in Hollywood and other film
industries. We're tempted to say that it's a product of Western modes of thought, of
imposing human will upon the world. But goal-driven striving, triggering a conflict
and a resolution, evidently propels some narratives from all cultures. The pattern
very likely springs from the human inclination to seek intentions behind every action
and to recognize that society is riddled with clashes between individuals, all eager to
fulfill their own needs. 52
We saw in our study of narrative structure that the classical tradition of cinematic
storytelling spells out the characters' goals quite early. Although the goals may be
revised or refined, they are apparent throughout, and they allow us to grasp an overall
pattern of development toward a climax. But the narration can also suspend informa-
tion about characters' goals. Ozu Yasujiro's Early Summer (1951) sets up a persistent
mystery about what exactly the marriageable daughter Noriko wants, and her sudden
decision to wed a friend becomes comprehensible only in retrospect. 53 Before we start
to wonder about Noriko's aims, Early Summer has firmly established the routines and
relationships of her life. This is typical of situations in which we're denied informa-
tion about characters' purposes; other aspects of the story world tend to be pushed
to the foreground. In Lodge Kerrigan's Claire Dolan (1998), the call girl protagonist
is presented through her daily routines—picking up Johns, having sex, and meeting
her pimp—until she flees Manhattan for New Jersey and gets a job as a hairdresser.
The objective narration withholds her aims in life, as well as her past history with her
pimp, for quite some time. Only fairly late in the film does Claire articulate her hope
to pay off her debts and have a baby. At that point, earlier incidents, such as her kind
treatment of a little girl she meets on the street, retrospectively cohere into a pattern.
A more classical narrational strategy would have treated each trick she turns as a step
toward breaking free, but by concealing Claire's goals, the narration throws all the
emphasis on her daily highs and lows, which seem to be leading her nowhere. 54
Delayed exposition of the character's desires and plans can give a shape to the
action within the story world, but what if the character has no desires and plans, or at
least no definite ones? What if the character is more passive, reacting to others rather
than initiating action? There is a tradition of filmmaking, associated with the "art
cinema," that puts such characters to the forefront. In such instances we must con-
struct a less causally driven story world, one ruled by passivity, chance encounters,
and emblematic episodes that evoke psychological and social themes. The homeless
Mona is purportedly the central character of Agnes Varda's Vagabond (Sans toit ni loi;
1985), but we come to know her chiefly through her encounters with people living in
the countryside that she wanders through. As her life accidentally touches theirs, the
narration reveals a cross-section of the civilization she has fled, surveying day labor-
ers, housekeepers, yuppies, thieves, and professors. In the process, we come to know
these peripheral characters far better than we know Mona. She remains psychologi-
cally opaque, not least because she doesn't have any goal that will define her sense of
118 Poetics of Cinema
herself. In a way, she lacks that dimension of personhood we associate with beliefs
and desires; her willful solitude is impregnable. My later essay in this volume, "The
Art Cinema as a Mode of Film Practice," tries to clarify the ways in which films like
Vagabond contribute to a relatively distinct tradition of cinematic storytelling. 55
A great many aspects of our construction of a story world need more exploration,
but I'll close this gross mapping of the terrain by considering one more issue related
to characterization. Apart from the overall pattern of activities undertaken by a char-
acter in that world, we sometimes encounter cues for what we usually call character
change. This is a slippery notion, I think, and can cover several of the dimensions of
personhood I've mentioned. Characters can change their social roles (e.g., a cop can
enter the clergy), their perception of the world (a blind man can regain his sight), and
their emotional states (a frightened man can become calm). What we usually mean by
it, however, is that characters change their thoughts or their traits.
In a great many narratives, characters alter their beliefs, desires, attitudes, opinions,
and states of knowledge. Call this epistemic change. In The Birdcage (1996), parents
biased against homosexuality eventually learn to tolerate their future son-in-law's gay
parents. This sort of coming-to-realize-the-truth change is quite common and is par-
ticularly valued when it's a change of knowledge not about external affairs (as when
the detective dispels a mystery) but about internal states. A sophisticated narrative,
many people believe, forces a character to better understand the sort of person he or
she is. This dynamic takes on a particular shape in mass-art storytelling, whereby the
character faces up to a character flaw or mistaken judgment. Hollywood screenwriting
manuals strongly suggest that there be a "character arc," whereby a basically good
person comes to recognize that they have erred and try to improve. The skyscraper
siege in Die Hard (1988) gives its hero, John McClane, the chance to realize how much
he loves his wife and to regret that he wasn't "more supportive" when she wanted to
advance in her career. "In the most simplistic terms," says one screenwriter, "you want
every character to learn s o m e t h i n g . . . . Hollywood is sustained on the illusion that
h u m a n beings are capable of change." 56 From this angle, change amounts to modify-
ing a judgment, admitting a slip, or, as in the case of an erring spouse, realizing that
short-term pleasure was an unworthy goal.
The sort of change that many consider the essence of a high-quality narrative is
more radical, involving a change in fundamental traits. Epistemic change can fuel
some changes in personality, but to alter a trait is to become a different person. Having
learned his lesson, McClane will be a more tolerant man, but nothing that happens
in Die Hard will induce him to become a pacifist, in the way that Scrooge becomes
charitable and Oedipus becomes humble. It's one thing to change your mind, another
to change your heart.
"Any character, in any type of literature," Writes Lajos Egri, "which does not undergo
a basic change is a badly drawn character." 57 One of the enduring contributions of
Egri's book The Art of Dramatic Writing, first published in 1942, is to show how trait
change can mesh with the classic approaches to plotting summarized by Brunetiere.
"A character stands revealed through conflict; conflict begins with a decision.... No
man ever lived who could remain the same through a series of conflicts which affected
Three Dimensions of Film Narrative 119
his way of living."58 Egri points to Othello, Tartuffe, Hamlet, Willy Loman, and other
characters who change in the course of the drama. Many of the changes are altera-
tions in knowledge of the kind I've just indicated, but some are more radical. Egri's
prime example is Nora Helmer in A Doll's House, who starts as a superficial, coddled
wife and becomes a mature, rebellious woman. Through extensive quotation, Egri
traces how carefully Ibsen displays Nora's growing understanding of her situation,
which in turn allows her to develop traits we could scarcely suspect she had. The trick,
Egri shows, is to let the situations force the character to change step by step.
A parallel instance in film is Jezebel (1938). In antebellum New Orleans, Julie
Marsden conducts a tempestuous courtship with Pres Dillard. She's headstrong and
willful, always prepared to flout convention in her demands that he put her first. She
pushes her luck, however, in insisting that she can wear a scarlet dress to a society
ball. Unable to force her to obey, Pres gives in, but then, when she realizes how she's
spurned by everyone, she wants to leave the dance. Pres forces her to stay, then takes
her home and breaks off their engagement. Her self-confidence is shaken, and though
she insists he'll return to her, she chokes back tears.
A year passes, and Julie has changed. Pres has left for the North, and she has
become a recluse in her Aunt Belle's house. When Julie learns that he is returning,
she bursts with hope, determined to beg his forgiveness. "I was vicious and mean
and selfish. And I want to tell him I hated myself for being like I was." This is already
a considerable growth; the Julie we meet in the opening scenes would never have
humbled herself. Nonetheless, her fierce energy hasn't abated, and she throws herself
into preparing the plantation household for Pres' return. But she is shattered when
Pres arrives with his new wife, Amy. She vows to get him back and bends her energies
to the task: "I've got to think, to plan, to fight."
During the couple's stay, Julie lets loose an escalating string of maneuvers. She
tries to rouse Pres' jealousy by flirting with his old rival, Buck. Failing in that, she tries
to seduce Pres, and failing in that, she goads Buck into defending her honor. Hav-
ing provoked a duel between Buck and Pres' brother, she realizes that she's playing
with men's lives. Her self-assurance begins to crack, and she plunges into a hysterical
mood, manically leading the plantation slaves in a song. When Buck is shot dead, she
feigns indifference but can't keep from weeping. She has become, Aunt Belle remarks,
Jezebel, the wicked woman who made her man a puppet and whose plots brought her
to a violent end.
The final phase of Julie's change comes when Pres contracts yellow fever during a
trip back to New Orleans. She rushes to his bedside and nurses him through the night.
When Amy arrives, Julie asks to be allowed to go with Pres to the leper island that
houses fever victims. Her speech is a fine example of the sort of emotional transitions
that Egri finds convincing. Julie first points out that Amy doesn't know enough of
southern customs to keep Pres alive in such harrowing conditions. More important,
she tells Amy, "I'll make him live—because I know how to fight better than you."
Finally, she begs, "Help me make myself clean as you are clean." It's a rhetorically
effective case, but the shifts from practical knowledge to the need to expiate her
sinful behavior also show Julie's own growth. She has always had boldness, tenacity,
120 Poetics of Cinema
and imperious force. Now, instead of serving her whims and self-importance, these
traits will sustain the man she loves. The selfish Julie of the opening has become the
selfless Julie of the final image, a long tracking shot that shows her tending to Pres
in a wagon piled with the dying. The biblical Jezebel was flung into a pile of offal
and devoured by dogs, but the film suggests that on the lepers' island Julie will be
redeemed. In the wagon, she rides alongside a nun.
Character change is usually not as fundamental as it is in Jezebel. Often it's a rever-
sion to what one once was, or privately already is. The plot action may reawaken the
devotion to duty lying dormant in the world-weary cop or the coquettishness in the
shy dowager. If Joe Fox in You've Got Mail becomes less aggressive through his love
for Kathleen, it's no shock, because we've seen his sensitive side pour out in his con-
fessional e-mails. His negative traits seem to be a less essential part of his personality
than the positive ones, which, when he meets Kathleen, are put temporarily aside.
Elsewhere I trace this process through Jerry Maguire, whereby Jerry's latent idealism is
made to resurface under the guidance of a good woman. 59 Another alternative would
seem to be the coming-of-age movie, set at a critical period when the character's traits
are still in a process of development. Scout in To Kill a Mockingbird (1962) comes to
accept the retarded Boo Radley as a friend because she has seen a cascade of unhappy
events proceeding from ignorance and fear.
All of which isn't to say that deep-seated character change is impossible in cinema,
only that it's rarer than we might expect. Far more common is character consistency,
with the plot being driven by a clash of purposes; gradual character revelation,
achieved by delaying the exposition (as in Early Summer and Claire Dolan); or charac-
ter revelation, achieved by thrusting the character into situations that expose different
facets of her personality.
The screenwriter's remark that character change is an illusion may reflect not only
Hollywood cynicism but also the fruits of experience. How many of us know people
who have fundamentally changed their natures after age 30? Perhaps popular film's
most extreme option is the reform or redemption plot, whereby the coward becomes
brave or the bad egg goes straight. Such extensive change of character usually requires
the recognition of a higher purpose. The death of Mr. Roberts shocks Ensign Pulver
into becoming the new thorn in the captain's side, and the love of a good woman
has turned many a sinner into a citizen, from Regeneration (1915) to The Apartment
(1960). Then too there's always divine revelation, as when the selfish playboy in Mag-
nificent Obsession (1954) becomes an altruistic surgeon through the intercession of a
quasi-spiritual holy man.
From the standpoint of a poetics, there is a great deal yet to be understood about
how we build story realms, particularly with respect to the ways in which cinematic
characterization plays off and plays with our real-world experience. My focus here
has been on the ineradicable role played by narration in coaxing us to build, through
fast but not simple inferential elaboration, that fabula world that seems so solid and
freestanding. The makers of narratives coax us to imagine characters and actions
according to guidelines at once artificial and deeply rooted in our mature abilities to
understand life around us.
Three Dimensions of Film Narrative 121
Afterword: Narrators, Implied Authors,
and Other Superfluities
"To give every film a narrator or implied author is to indulge in an anthropomor-
phic fiction."60 The one-page brush-off in my 1985 study Narration in the Fiction
Film has probably gotten more notice than the claims I make in the rest of the book.
I didn't realize that several theorists of narrative are very strongly committed to
such constructs. What follows is an effort to make my case more plainly and to reply
to some objections.
Everyone agrees that films sometimes have narrators. A film can present character
narrators, when a character in the story world tells someone, or us, about events that
have transpired. There are also noncharacter narrators, such as the external narrating
voice presented in Jules et Jim (1961) or in many documentaries (sometimes known
as the Voice of God narrator). Both character and noncharacter narrators are given
a voice (either on the soundtrack or through intertitles) and sometimes a body, as in
character narrators in the story world, or the meneur dejeu figure in La Ronde (1950),
who strolls through the story world but speaks to us. The crucial claim is whether
these more or less tangible narrators, along with everything else in the film, proceed
from a more encompassing narrator who "tells the film."
This cinematic narrator is the equivalent of the narrating voice we encounter in
literature. In a literary text, we usually have a strong sense of being told something
by someone because of the linguistic texture (the use of pronouns and tense) and
the managing of point of view. The character narrator is obvious at the start of
Huckleberry Finn.
You don't know about me without you have read a book by the name of
The Adventures of Tom Sawyer; but that ain't no matter. That book was made
by Mr. Mark Twain, and he told the truth, mainly. There was things which he
stretched, but mainly he told the truth.
Crucially, however, a narrating voice can remain present when it's not personalized,
as in the opening of Pride and Prejudice.
It is a truth universally acknowledged, that a single man in possession of a good
fortune must be in want of a wife.
However little known the feelings or views of such a man may be on his first
entering a neighbourhood, this truth is so well fixed in the minds of the sur-
rounding families, that he is considered as the rightful property of some one or
other of their daughters.
"My dear Mr. Bennet," said his lady to him one day, "have you heard that
Nether field Park is let at last?"
Mr. Bennet replied that he had not.
122 Poetics of Cinema
The theorists I'm considering differ on important details, but I think it's fair to say
that they agree that in order to understand cinematic narration, we must postulate
some agent parallel to the speaking or writing voice that presents the events in a
literary text. Andre Gaudreault and Francois Jost call that agent the monstrator, and
Albert Laffay speaks of le montreur d'images, the image shower. Tom Gunning speaks
of the "narrator-system" of D. W. Griffith's films. 61 My basic claim is that the narrator,
whatever its status in literature, is an unnecessary and misleading personification of
the narrative dynamics of a film.
Let me start by restating two objections drawn from my 1985 remarks. First is an
appeal to Ockham's razor, the principle of theoretical parsimony. We ought not to
create new concepts unless they do work that can't be accomplished by our current
concepts. If it turns out that nothing we want to describe or explain about filmic nar-
ration is better handled by the notion of cinematic narrator or implied author, we
ought to stick with our existing stock of concepts. As I'll indicate below, it's hard to
show that these new ideas do anything more than label features that we can already
detect and explain adequately.
Another objection I raised in 1985 depends on the greater saliency and perva-
siveness of the literary narrator's voice. In my examples above, we are very aware
of a speaker addressing us. Huck calls the reader "you" and identifies himself,
whereas the impersonal narrator of Pride and Prejudice generalizes about bachelors
and marriage, and it judiciously chooses to report Mrs. Bennet's remark but not
Mr. Bennet's reply. Quotation and summary stand side by side, acknowledging
the presence of a narrator sifting information for impact. And these voices remain
present throughout each novel. Huck constantly judges and amplifies on what
he reports, whereas the narrator of Austen's novel does no less in the impersonal
mode. At many moments, each novel's narrator comes forward and projects a
certain attitude toward the action represented. 62 But in cinema, a speaker's "voice" is
seldom so explicit. Exceptions a speaker's "voice" might be the opening and closing
sequences, when we're sometimes aware of being directed to notice this or that
detail. But this just seems a case of self-conscious address, as the narration frankly
acknowledging its act of emphasizing an item.
Furthermore, even when confronted with such self-conscious passages, we don't
characteristically attribute them to a narrator. For ordinary audiences, the relevant
agent or agents are the filmmakers, commonly known as they. "At the start of a
movie," someone might say, "they always show something important to the plot." In
a memo, Darryl F. Zanuck sums up patrons' complaints about The Gunfighter: "Why
didn't they let him live at the finish? After all, he had been reformed. He could have
63
been wounded, if they wanted to shoot him."
We needn't of course take ordinary responses as wholly determining our theoreti-
cal concepts; many readers would identify the speaking voice in Pride and Prejudice as
that of the author, Jane Austen. Nevertheless, many other readers would understand
that the intruding narrative voice of that novel is not necessarily that belonging to the
author. Still, very few viewers would take, say, a bit of actors' business or a pattern of
Three Dimensions of Film Narrative 123
lighting as having its source in an intermediary, a cinematic narrator, rather than in
either "the film itself" or the creative individuals on the set.
I'm aware that not everybody shares my intuitions on this matter. I think that this
is largely because many theorists think that in explaining the logic of cinematic nar-
ration, we don't need to appeal to any psychological activity. They would claim that
even if no viewer ever registers the presence of a cinematic narrator or implied author,
any explanatory theory must posit such entities. Narrative as a concept, regardless of
medium, requires a narrator. My alternative proposal is that in cinema, narration as
a process encourages us to build up the story, including the voices and behaviors of
narrators, but no narrator comparable to those agents is logically required to give us
the narration as a whole. As I put it in 1985, "Such personified narrators are invari-
ably swallowed up in the overall narrational process of the film, which they do not
produce." 64 Who produces the narrational process? The filmmakers.
Let me explain my grounds for this view, and then I'll return to the case for a
narrator. Recall the mentalistic framework I presented in the opening essay. Seen
from this perspective, films are made by human beings to provide other people with
experiences. Call the second bunch viewers, even though they're also listeners. The
viewers are engaged in the experience by virtue of cues built into the film by the first
bunch, the makers. The cues are structured to encourage particular paths of percep-
tion, comprehension, and appropriation, all three of these clusters of activities being
also invested with emotion. The experience proceeds by means of the viewer's inferen-
tial elaborations, some of them very fast and mandatory (in the domain of perception),
and some more slow and deliberative (typically in the domain of appropriation).
The filmmakers are practical psychologists. They have been viewers themselves,
and they are more or less accomplished practitioners of their craft, so they have many
ideas about how to shape the cues to provide experiences of a particular sort. They
can fail, or succeed beyond their initial hopes, but they organize the film so as to
solicit a range of effects. Like all humans, filmmakers can't anticipate, let alone deter-
mine, all the effects that may arise from their endeavors. Particularly in the domain
of appropriation, the viewer has a freedom to seize upon certain cues and not others,
pull them into a range of projects, and use the film in ways that couldn't be foreseen
by the filmmakers.
How does this framework affect narrative? Perceiving, comprehending, and
appropriating narrative, as well as responding emotionally to it, depend on cues
sown through the film. Those cues ask us to grasp the narrative in certain ways. The
viewer constructs, according to the unfolding narration, a story world and a structure
of events within it. That construction becomes a source of emotional and cognitive
experiences. Ideally, viewers construct the narrative as the filmmakers hoped they
would, but things aren't always ideal. A viewer may fail to pick up narrative cues,
or a filmmaker may fail to make them sufficiently salient. There may be a mismatch
between the filmmaker's schemas and the viewer's. Cinematic traditions, however,
secure a considerable amount of convergence between what filmmakers know can
affect viewers and what viewers do experience, especially in the domains of perception
124 Poetics of Cinema
and comprehension. Narrative traditions exist partly to enable this sort of agreement
about how the story world is to be constructed and construed.
Odd as it sounds to say it, this framework doesn't mean that communication takes
place. If communication means the transmission of an idea or concept from one
mind to another by means of some physical vehicle, then that notion doesn't capture
the experiential dimension I'm positing. Suppose as an amusement park engineer,
I design a roller coaster. You get on at a certain point and undergo a suite of turns,
swoops, climbs, and dives. In what sense have I communicated something to you?
You've undergone a physical and emotional experience that I planned in advance, but
I haven't transmitted any idea or concept to you.
Someone might reply that a roller coaster isn't a good analogy because it doesn't
offer an experience of representations. So substitute a pictorial landscape, like a topiary
garden, and my point will be the same. Or consider the layout of a museum display,
in which curators arrange the order and position of the items according to principles
of what they want to link and highlight. As we stroll through the exhibition, we don't
posit an intermediary figure between human agents and the array that we see. We
may posit someprinciples that seem to have guided their decisions, just as we presume
that principles governing structure, materials, load, and other architectural proper-
ties governed the decisions of a roller-coaster designer; but those principles needn't be
described as a virtual being. Or take another instance: A map can represent a territory,
but understanding map representation doesn't demand a terrain presenter embedded
in the map. Again, simply attributing the relevant features of the representation to
human makers and their plausible intentions suffices to cover the case.
Films traffic in concepts and meanings, but these, I submit, are the result of the
inferential elaboration of cues presented by the design of the work. Just as film-
makers anticipate that viewers will draw narrative inferences, they often anticipate
that viewers will infer appropriate topics and themes. A narrative film prompts us to
assign meanings at many levels, but none is communicated in the sense that a message
passes from the filmmaker's mind to the spectator's. Rather, a lot of what some theo-
rists would call communication I'd call convergent inference making. The filmmaker
has gotten us to walk down the path she planned. If we figure out that Clarice Starling
is the protagonist of The Silence of the Lambs (1991), that she is inexperienced and
shaken by her past but still courageous and determined, that her efforts to identify
Buffalo Bill initiate the story action, and that she is opposed by several other char-
acters but that she wins out eventually... we've done pretty well. And while or after
making sense of all this, we can go on, thanks to many cues, to find Clarice's activi-
ties satisfying, moving, socially suspect, or whatever. The communication model
would say that something passed from the creators' mind to the movie and then to
the viewer. I would say that the creators designed an experience such that viewers
are coaxed to construe the film in ways that yield a certain experience more or less
accurately foreseen by the filmmakers. Looked at this way, a film becomes a tissue of
cues, and these cues can be quite fragmentary and varied. If I, the filmmaker, want
to prompt you to think or feel something, I can shamelessly use anything that can be
put into a movie. Any image or sound that gets the job done is a potential candidate,
Three Dimensions of Film Narrative 125
regardless of strict logical consistency. If I want you to feel sad about what's onscreen,
1 can insert sad music. Where does this music come from? I can motivate it by locat-
ing it in the story world, or I can simply add it as sourceless (nondiegetic) accompani-
ment. As far as the experience is concerned, its provenance is less important than the
effect of triggering a sad feeling in you.
I think that this conception of narrative engineering handles some tough cases.
For example, many films open with a voice-over commentary by the central character
explaining what led up to the events we'll be encountering. This commentary's role
is plainly to orient us toward the story world and the plot. That function is sufficient
unto itself. It doesn't necessarily raise such questions as "To whom is the charac-
ter talking?" or "When is this conversation taking place?" Jerry Maguire's opening
voice-over narration presents what follows as his story, but we see many scenes that
he doesn't witness; the narration is, as we say, omniscient. Yet it would be strange
to protest, "But Jerry's the narrator! He's telling the story. How could he know what
Dorothy told her sister? Did Dorothy confide in him after they were married?" Such
questions are as irrelevant as asking whether the giraffes and turtles in a topiary
garden could survive in the same ecosystem.
Or consider the anomalies harbored by another common device, the flashback that
dramatizes what a character narrator tells. At the start of Leave Her to Heaven (1945),
Dick Harland arrives at a dock and is greeted by an older family friend, Glen Robie,
before he paddles a canoe to a house further along the shore. With that concision
characteristic of classic Hollywood, the first 90 seconds inform us that Dick has spent
2 years in prison and a woman is waiting for him. But why did he go to prison? Whom
is he going to meet? Rather than attaching itself to Dick, the narration stays with
Robie and his companion on the pier. As they have coffee, Robie says, "Of all the
people involved, I suppose I'm the only one who knew the whole story." As we hear
Robie's voice saying, "They met on the train," we segue into the past. We see Dick
meet Ellen, and their love affair begins. It will take a twisted path, involving suicide,
mental anguish, and death by misadventure. Robie enters the story action at inter-
vals, but there are long stretches in which he isn't present to witness intimate scenes
between the couple. Nor can he see what the others do when they're alone. Robie is
Dick's friend and lawyer, so we might assume that Dick relayed some personal infor-
mation to him, but other incidents aren't in Dick's ken either, notably those involving
Ellen and Dick's disabled brother, Danny. So in some sense Robie can't know the
whole story, at least the one we see and hear. Yet according to the communication
model, Robie is recounting the story, and you can't recount what you don't know.
We break the impasse by recognizing the primary functions that the recounted-
flashback device seeks to fulfill. Leave Her to Heaven aims to build up curiosity and
suspense from the start, and one norm-sanctioned way to do so is to show a scene
after the main action has concluded. One way to justify and clarify the breakup of
chronology is to assign a character to tell another about what led up to the current
state of affairs. A scene showing the character launching on the tale prompts your
understanding that what follows is a flashback. It doesn't matter that nobody could
tell an event with the sort of detail we find in the images shown in the flashback.
126 Poetics of Cinema
Nor is there any scandal in the fact that the narrating character didn't witness the
events that we're going to see. All that matters is that a scene calls forth in us a mental
schema, people tell one another about an event that has occurred, and that triggers
only one relevant inference: A time shift is coming up. (That's not to say that other
effects couldn't ride along with this flashback, such as emotional colorings.)
Narrative films are full of such purpose-driven anomalies. Ten North Frederick
(1958) begins with the funeral of Joe Chapin. As his widow Edith greets the guests,
son Joby and daughter Ann retire upstairs. Joby is drunkenly railing against their
mother for slowly poisoning their father's life. He says, "Only five years ago—remem-
ber?" and the camera tracks in on Ann's face. The flashback begins. The story traces
Joe's political ambitions in relation to his wife's implacable hatred of him and the
erring ways of Joby and Ann. Several scenes present events, such as clandestine cam-
paign maneuvers and Edith's affair with the local district attorney, that neither Joe
nor Ann has witnessed. Then something odd happens. The narration attaches us to
Joe as he falls in love with Kate, a model, and they begin an affair. Eventually Kate
leaves him, and Joe descends into alcoholism. At the climax, Ann sees Joe in his cups
and he hears, in a purely subjective auditory flashback, Kate's voice repeating a line
from the past, "Good night, my love." As Joe collapses, the flashback ends and we
return to Ann and Joby after the funeral. In an epilogue, Ann serves as bridesmaid
for Kate's wedding, and just before the ceremony she realizes that Kate and her father
were lovers. She says, "Now I understand it all, Kate."
The flashback is framed as Ann's recollection of her father's life, but it would be
embarrassing to claim that she recalls all the events we see. Presumably she d i d n t
hear Kate's voice when Joe imagined it. There are many scenes that she couldn't know
about. Most strikingly, her "embedded" narration tells us something of which she's
utterly unaware: that her father is having an affair with Kate. She doesn't learn of this
until the final scene, but we learned of it in "her" flashback! Just as in real life you can't
communicate what you don't know, you can't recall what you never experienced. Yet
film narrative has no problem presenting such paradoxes. 65
The lesson is this. In principle, narrative is utterly opportunistic and promiscuous.
It mobilizes systems and partial systems from all areas of life. It seizes anything
that can serve its purpose, regardless of logical or ontological constraints, and slaps
together all manner of disparate cues. Bent on shaping our experience in time,
it draws upon whatever will do the job. Narrative invokes our schemas for following
conversations or understanding confessions or responding emotionally to music or
grasping shifts in time, and those schemas fulfill wholly strategic purposes. In place
of a logic of narrative, we should be seeking a folk psychology of it.
I hasten to add that this is all in principle. In practice, particular narrative tradi-
tions have made certain engineering principles more likely, or more motivated, than
others. For instance, in the Hollywood studio cinema, flashbacks tend to be cued
in certain ways—by suggesting that a character is recalling events (the Ten North
Frederick solution) or that one character is explaining the past to another (Leave
Her to Heaven). In other traditions, and in Hollywood films since the 1970s, flash-
backs no longer need these sorts of lead-in. But then other cues will tend to come
Three Dimensions of Film Narrative 127
forward to signal that a flashback is coming up (music, color shifts, intertitles, and
story world factors). Alternatively, the filmmaker may withhold such cues, but, then,
that's a strategy too: The narration thereby makes us uncertain about how the events
are arranged in time. So if narrative is promiscuous in principle, it's likely to justify
its wantonness in practice, thanks to local conventions of motivation. This isn't to
say, of course, that some conventions don't appear in different traditions. There may
be even some universally accepted stratagems, such as the tendency to accompany
visible action with music from unseen or unknown sources.
If narrative is as opportunistic as I've indicated, then we ought not to expect con-
sistent circuits of communication to be embedded in stories. Stories are told to us
all the time, in everyday discourse and in all manner of media. Filmmakers seize
upon certain features of these narrative interactions but not others. In a movie, we
witness character narrators telling things to other characters, as we overhear people
in the real world. But what those fictional character narrators tell exhibits a range of
knowledge and wealth of detail that no real person could have. In a movie we hear a
narrating voice from outside the story recount what happened, as if we were listen-
ing to a storyteller at a campfire. But we don't have to worry whether there is really
a campfire, or any other concrete narrating situation, or an addressee on the same
logical level as the speaker. The impresario of La Ronde can sometimes talk to us as if
he were on a stage addressing an audience, but he isn't; he's on a sound stage talking
to a camera. Sometimes he's outside the fictional world addressing us; sometimes he's
inside the world as a walk-on character.
In such cases, one or two aspects of a narrating schema are appropriated and
collaged with the other components of the narration for purely strategic purposes. As
I put it in Narration in the Fiction Film, this condition presupposes a perceiver—you
and me—but no message sender. "The narrational process may sometimes mimic the
communication situation more or less fully. A text's narration may emit cues that
suggest a narrator or a narratee, or it may not." 66
By contrast, arguments for the necessity of a cinematic narrator rely more or less
explicitly on a communication model. The most cogent layout of the assumptions here
comes, as we'd expect, from one of the most meticulous narratologists of film, Seymour
Chatman. He proposes that we need two more constructs to explain the logic of filmic
narration: a cinematic narrator that is not as visible or audible as character narrators
are, and an "implied author" that is even more intangible. Both constructs are neces-
sary to complete the chain of communication that Chatman sums up in a diagram
(see Figure 3.6).67 The narrational process consists of story information passed among
a series of agents, some embedded in the text and some not. Every agency emitting
narration has its counterpart in an agency that receives it. The process moves from
real author to implied author to cinematic narrator to character narrators (if any) to
character narratees (if any) to cinematic narratee to implied reader to real reader.
Let me leave the issue of implied author-implied reader aside for the moment. On
what grounds does Chatman postulate a cinematic narrator? He offers both logical
and pragmatic reasons. Logically, he says, the very concept of narrative entails a
narrator. "Every narrative is by definition narrated—that is, narratively presented—
128 Poetics of Cinema
Narrative text
Real i Real
author '"T reader
Figure 3.6 Seymour Chatman's diagram of the communication process in a narrative text.
Source: Adapted from Seymour Chatman, Story and Discourse: Narrative Structure in Fiction
and Film (Ithaca, N.Y.: Cornell University Press, 1978), 151.
and that narration, narrative presentation, entails an a g e n t . . . . Agency is marked
etymologically by the -er/-or suffix attached to the verbs 'present' or 'narrate.'" 68 But
this claim secures only the fact that as an artifact, a narrative owes its existence to
an artificer (or several of them). No one disputes this premise. But it is no help to
an argument that we need the concept of a textual narrator distinct from the actual
novelist or filmmaker. Of course, we don't think that narratives fall from the skies.
They are created by humans. But the relevant agents in this context are real people,
not the postulated agents that Chatman argues for. To undergo the experience of a
roller-coaster ride, I don't have to imagine a ghostly intelligence standing between
the engineer and me, shaping the thrills and nausea I feel. The same holds true for
the topiary gardener or the mapmaker or the curator designing a museum display.
The very concept of a storyteller doesn't entail a virtual storyteller of the sort that
Chatman proposes.
In other places, Chatman strays from defending the textual cinematic narrator
and reminds us that real agents make texts. He objects, for instance, to my claim that
narratives are "organized" for perceivers but not "sent" as part of a communication.
"Surely," he writes, "the film—already 'organized'—somehow gets to the theater and
gets projected; something gets sent." He says that it would be uncomfortable to have
"a communication with no communicator—indeed a creation with no creator."69
I agree that movies get created and shipped out to theaters, but cinematic narrators
aren't splicing the footage or filling out FedEx forms.
In sum, Chatman hasn't convinced me that a postulated narrator, as opposed
to a living and breathing filmmaker, is necessitated on logical grounds. Perhaps,
though, conceiving of a cinematic narrator offers pragmatic rewards, helping us see
new things in narrative films or offering conceptual solutions to problems thrown
up by films. To appraise this prospect, we need to ask how we concretely recognize
the cinematic narrator.
Chatman maintains that the term doesn't commit him to a language-based concep-
tion of cinema. In a film (and presumably a ballet, a mime act, or a wordless cartoon),
the narrator isn't literally a teller; it's also a shower or, in Chatman's terms, a "presenter."
This need not be a "recognizably human agency." "I argue that human personality is
not a sine qua non for narratorhood." 70 So what is the equivalent of the speaking or
writing voice we encounter in literature? The cinematic narrator, Chatman explains,
is "the composite of a large and complex variety of communicating devices."71 What
devices? The list is open-ended and includes auditory elements (speech, noise, music)
Three Dimensions of Film Narrative 129
and image-based ones (mise-en-scene, editing, cinematography, etc.). These are all
gathered up by "the overall agent that does the showing." 72
No one will disagree that these elements are resources that filmmakers have at
their disposal. In a film, these techniques represent the narrative. But this list of
features is somewhat of a letdown after several pages of theoretical argument for the
utility of positing a cinematic narrator. All these techniques of representation are
just as easy to analyze by speaking of the film's form and style tout court, along with
the effects we propose that these features aim to produce. Critics and analysts have
been appealing intelligibly to these concepts for decades without assigning them to
a narrator. We need never invoke an extra intelligence that is bending them all to its
will (apart, again, from a real filmmaker or set of filmmakers). Chatman's cinematic
narrator looks like simply a label for the systematic formal and stylistic properties
we can detect in any narrative film. So the pragmatic utility of the narrator concept
seems questionable.
Chatman suggests that thinking of the narrator can be helpful in certain problem-
atic cases, as when we try to track unreliable narration. When the image track contra-
dicts the soundtrack, as in Badlands (1973), we have "a conflict between two mutually
contradictory components of the cinematic narrator." 73 Again, however, what have
we gained by postulating this extra agent and then saying that two "components" of
it clash? Why not simply say that we encounter an organized disparity of image and
sound? From the standpoint of theoretical parsimony, what more does the virtual
figure of the narrator add?
The communication model holds that for every sender, there's a receiver. So if
there's a cinematic narrator, there must be a narratee. Not the real viewer, nor the
"implied viewer," but a pickup agent at the other end of the narrator's communiques.
But most theorists holding this position tiptoe around the narratee, because such a
creature doesn't possess even the gossamer presence of the cinematic narrator. The
narrator is at least visible and audible via technical devices, the equivalent of the words
of the literary narrator. But where does the text provide signs of the narratee? And
what properties can be attributed to him or her? We can call Huck's narration plain-
spoken and the third-person narrator of Pride and Prejudice wryly judicious, but what
attributes can we ascribe to the literary narratee, let alone its cinematic counterpart?
Chatman's discussion of the narratee in his 1978 book focuses principally on literary
character narratees, those dramatis personae who attend to what character narrators
say. These are uncontroversial cases, because the literary texts are representing some-
one in the story world telling the tale to someone else in the story world. Chatman
also considers diary narratives (the writer becomes his or her own narratee) and the
sort of "Dear reader" entity that is sometimes signaled by an impersonal narrator of
the Austen sort. There is no discussion of the cinematic narratee. 74 His 1990 defense
of the concept of the cinematic narrator, consuming 14 pages, never mentions the
narratee. I suspect that this is because there is almost nothing to be said about it. The
concept does no theoretical work. All we can say is that some posited entity is picking
up the significance of every shot, line of dialogue, piece of performance, and so on
emitted by the cinematic narrator—and then relaying that information to the implied
130 Poetics of Cinema
reader-viewer, who then relays it to flesh-and-blood viewers. Positing so many ghosts
in the textual machine suggests once more that the communication model isn't the
most fruitful way to understand narration.
There's a general point at issue here. Marie-Laure Ryan proposes that there are
three positions to be taken on narrative across media. 75 One can hold that narrative
exists only in verbal media; few currently take this line. Or one can take narrative as
a fuzzy set of features, but hold that narrative is most fully implemented in language,
and thus the parameters of verbal language must be present in other media too. The
theorist will accordingly look for parallels to fictional voice, literary point of view,
the narrator-narratee relation, and so on. It seems evident that Chatman holds this
second view. He presupposes that language-based narrative contains the components
necessary to define or describe narratives in other media. That is, in order to character-
ize cinematic narrative, we must recast concepts derived from literature (specifically,
concepts based in communication). Chatman's overall taxonomy suggests that there
are no narrative techniques possessed by cinema that cannot be found in literature,
though cinema can actualize those techniques in strikingly different ways.
Chatman says at one point,
It is awkward to a general theory of narrative to say that some texts include the
component "narrator" and others do not. As Sarah Kozloff puts it, simply but
incisively, "Because narrative films are narrative, someone must be narrating.
Or if not necessarily someone, something. 76
Putting aside the slide from the author or a character (someone narrating) to the
cinematic narrator (something narrating), I still don't see why it s awkward to say that
cinematic narration, conceived as the process whereby the film guides the spectator s
construction of a story out of cues, has no narrator in the virtual sense Chatman
proposes. Narration has a narrator, or narrators, in the concrete sense that real agents
have presented this story to us. A filmmaker or group of filmmakers created the system
of cues we are to follow, and as real agents ourselves we engage with those cues. End
of story. I prefer to bite this bullet than to follow the logic of Chatman and Kozloff,
whereby after postulating a cinematic narrator, we must postulate a drama narrator
for every play, a dance narrator for each ballet, comic strip narrators for the funnies,
and so on—with each one turning out to be no more than the assembly of all expres-
sive techniques available in each medium. In sum, I'd rather be counterintuitive than
uninformative. And maybe the notion isn't that counterintuitive. Over the last decade
or so, other writers haven't found the idea of narratives without narrators of this sort
hard to swallow.77
Ryan sketches a third position on cross-media narratives as follows.
Narrative is a medium-independent phenomenon, and though no medium is
better suited than language to make explicit the logical structure of narrative,
it is possible to study narrative in its nonverbal manifestations without applying
the communicative model of verbal narration. 78
Three Dimensions of Film Narrative 131
This seems to me a satisfactory position. I'd go further, though. As indicated above,
cinematic narrative—and, for all I know, any form of narrative—is able to borrow
certain aspects of the communication process without buying the whole package.
We can have character narrators without character narratees (who is listening to the
protagonist at the beginning of Rebecca [1940]?) and character narrators who recount
things of which they have no knowledge (an impossibility if we stick to the commu-
nication diagram).
It may be that the communication model works well for literature because verbal
narrative mimics many aspects of everyday interchanges. In any case, to rely wholly
on verbal models for narratives in all media creates conceptual contortions, fails to
cover common cases, and may not tell us anything we don't already know. By the
principle of parsimony, we don't need to build a cinematic narrator into our general
theory of narrative. At a less general level, however, a film may signal that we are to
infer various sorts of narrators, through cues ingredient to the film or its tradition.
One more aspect of my account needs explanation. I've argued that we can describe
narration with terms like suppressive, self-conscious, and the like. This, some other
theorists object, contradicts my belief in an impersonal narrating process, for are not
these terms we ascribe to agents?79 My response in the 1985 book remains: These terms
are shorthand metaphors and constitute merely a fagon deparler. So to call a stretch of
narration "suppressive" is an elliptical way of saying that the representational process
fails to provide cues that would yield knowledge of relevant information about the
situation, relative to the filmic norms in force. A self-conscious narration provides
cues that prompt the viewer to acknowledge some artificial dimension of the narra-
tion itself, relative to the filmic norms in force. Each adjective I use can be cashed in
without remainder.
If the cinematic narrator as a general concept seems untenable, what of the implied
author? Again, Chatman makes the most extensive and detailed case. He claims that
the literary narrator can be located as a voice, but the implied author nowhere speaks.
Whereas the narrator is a presenter, the implied author is a creator. It is "the principle
within the text to which we assign the inventional tasks."80 More specifically, the
implied author is "the agency within the narrative fiction itself which guides any read-
ing of it. Every fiction contains such an agency."81 The implied author's counterpart is
the implied reader, an idealized pickup of the implied author's design and message.
Chatman offers several further characterizations of the implied author, not all
of which seem compatible. It is "the sense of a narrative text's whole structure of
meaning," "the unified invention and intent of the text," and "a sense of purpose
reconstructable from the text that we read, watch, and/or hear." 82 Chatman also says
that once the real author's creative activity is made tangible, "the text is itself the
implied author." 83 Principle, invention, sense of purpose, and text in itself—these
conceptions seem relevant but still distinct from one another. Yet one can see why all
seem attractive when we consider a particular example.
In Ring Lardner's short story "Haircut," the entire action is told, in first person, by
a garrulous barber cutting a customer's hair. Whitey is the narrator, and the narratee
is the customer in the chair, an unnamed stranger in town. In chatting about a local
132 Poetics of Cinema
scandal, Whitey misunderstands the import of everything he reports. He misjudges
the character of Jim Kendall and isn't aware of what led up to Jim s death. Because
the customer never speaks, we can't assume that he gets the point either. There is no
authorial commentary, but we are meant to infer that Jim was a bounder and got his
comeuppance from one of his victims. We judge Whitey's account unreliable, and so
we construe the actions he reports quite differently than he does. In Chatman s terms,
the implied author, not the narrator, is communicating the truth of the situation.
But why not simply claim that the accurate judgments on the action of "Haircut"
can be traced to Ring Lardner, the author? Because, Chatman would argue, there is no
dependable way to identify what Ring Lardner thought about the story action apart
from the text he has left us. A great many real authors aren't around to tell us what
they meant, and even if we could ask the living ones, they can lie, or forget, or play
the fool. We still have to decide on the basis of the text, which will provide the most
convincing evidence. Moreover, sometimes authors write better than they know. It s
possible that Mark Twain saw the last chapters of Huckleberry Finn, in which the
slave Jim becomes the butt of an elaborate prank, as a bit of good fun. But the implied
author, many critics would suggest, makes Huck and Tom look shallow and cruel,
casting a shadow over the friendship that Huck and Jim have shared in the bulk of the
story. This example is apt, because in the history of American criticism, one impulse
behind theorists' creation of the implied author is the need to account for unreliable
narration while avoiding what many consider the "intentional fallacy. 84 As Chatman
puts it, the implied author yields "a way of naming and analyzing the textual intent
of narrative fictions under a single term, but without recourse to biographism. This is
particularly important for texts that state one thing and imply another." 85
If the grounds for the narrator are both logically necessary and pragmatically
useful, Chatman claims only pragmatic utility for the concept of the implied author.
And some theorists who don't embrace the concept of a cinematic narrator do
accept the implied author as operative in both literature and film. My own response,
though, is a skeptical one. If the implied author is mainly a solution to the problem of
unreliability, I would suggest that the problem be solved differently.
We could put it this way. The text prompts the reader to construct the story action
a particular way, and that construction includes recognizing the gaps and short-
comings of the narration, given the norms in force. We judge a literary narrator to
be unreliable through inferential elaboration of the cues she or he presents, and that
elaboration may be at odds with the inferences drawn by the narrator. In "Haircut,"
we judge Whitey to be unreliable not because an invisible figure is signaling us behind
his back, but because Whitey's judgment of Jim Kendall's character, on the evidence
he presents, is ill-founded, according to our norms of behavior. He thinks Jim is a
card; we infer that Jim is a bounder.
In this respect, literary fiction is no different from real-life reportage or trial
testimony. Whatever a speaker says, we balance the information conveyed and the
trustworthiness of the source against standards of behavior and judgment. When a
reporter or trial witness presents information, we don't infer an "implied recounter"
or "implied testifier" backstage strategically shaping what we hear. Likewise, in
Three Dimensions of Film Narrative 133
film, we are guided to make inferences about the narration we encounter, regardless
of whether the information is recounted by characters or presented by the overall
organization of the film. Of course, those inferences may fit together smoothly or
they may contradict one another, just as in life. Naturally, in narratives, the fit or
the contradictions are largely created by the makers, in order to take us through a
particular experience, whereas life has no such artificer in the wings. In any event, we
don't need to personify an agent hovering over the text that is transmitting the truth
of the situation. If the implied narrator is the set of overarching principles of design
governing the film, we can simply talk about said principles, even, or especially, when
they create problems of unreliability for the spectator.
Further, it may be that the communication model creates the very need for an
implied author. Chatman argues that because Anne Frank never intended her diary
to be seen, the real Anne Frank can't be speaking to us. Still, "we read the diary as if it
addresses us," so "it can only be the implied author of the Diary who addresses us."86
By this logic, every diary that's read by somebody other than the diarist has an implied
author, whereas those that aren't so read don't. The implied author becomes the read-
er's projection, not the author's creation. So why do we consider it part of a process
of communication at all? Is a diary in fact an instance of communication? It seems
to me that reading a diary is best understood on the model of overhearing someone
talking to himself. Because we aren't the addressee, we don't need to posit an agency
that is shaping the monologue for our (or an implied hearer's) uptake. Several types
of solitary writing—grocery lists, Post-It notes to yourself—don't presuppose implied
agents of this sort. When we find self-addressed writing, we just hear or see the words
and draw our inferences accordingly, under whatever norms we think relevant. Like-
wise with cinema: A film is made so as to elicit inferential elaboration. Invoking the
implied author would seem to add nothing to our recognition of the principles under
which the film operates.
Finally, the issue of whether a cinematic narrator is pragmatically helpful or logi-
cally necessary raises a distinction I urged earlier. It's useful to distinguish, however
roughly, between a theoretical poetics, which aims to understand the conditions
of cinematic representation on a broad canvas, and historical poetics, an empirical
inquiry into particular ways of making. And I'd reiterate that we should build the
former as inductively as we can, tracing out commonalities among traditions that we
study in detail. When we try to be purely deductive, we start with intuitively salient
models, like that of literary communication, but we may assume that they're more
prevalent than they are. We also miss the fact that narratives, created by people for
other people, need not be built out of principles that are logically consistent. The
promiscuity of narrative construction reflects the quick and dirty reasoning charac-
teristic of minds attuned to social, not ontological, meanings.
r
I

h.

Cognition and

Comprehension

Viewing and Forgetting in

Mildred Pierce

By and large, audiences understand the films they see. They can answer questions about
a movie's plot, imagine alternative outcomes ("What if the monster hadn't found the
couple... ?"), and discuss the film with someone else who has seen it. This brute fact
of comprehension, Christian Metz asserted in the mid-1960s, could ground semiotic
film theory: "The fact that must be understood is that films are understood." 1
As semiotic research expanded in France, Britain, and the United States, the search
for explanations of filmic intelligibility took theorists toward comparisons with
i
language, toward methodological analogies with linguistic inquiry, and across several
disciplines. At the same time, though, theorists increasingly abandoned the search for
principles governing intelligibility. They turned their attention to understanding the
sources of cinematic pleasure, chiefly by defining "spectatorship" within theories of
ideology and psychoanalysis. The conceptual weaknesses and empirical shortcom-
ings of the latter doctrines have become increasingly evident in recent years. 2 It seems
fair to say that interest in them has waned considerably, and most French partisans of

135
136 Poetics of Cinema
psychoanalysis have returned to the "classic" structuralist semiotics of the 1960s and
3
early 1970s, or even to traditional film aesthetics.
The current "cognitivist" trend in film studies has gone back to Metz's point of
departure, asking, What enables films—particularly narrative films—to be under-
stood? But the hypotheses that have been proposed recently differ sharply from those
involved in semiotic research. The emerging cognitivist paradigm suggests that it's
unlikely that spectators apply a set of "codes" to a film in order to make sense of it.
Rather, spectators participate in a complex process of actively elaborating what the
film sets forth. They "go beyond the information given," in Jerome Bruner's phrase. 4
This doesn't entail that each spectator's understanding of the film becomes utterly
unique, for several patterns of elaboration are shared by many spectators. 5
For example, you are driving down the highway. You spot a car with a flat tire; a
man is just opening up the car's trunk. Wholly without conscious deliberation, you
expect that he is the driver, and that he will draw out a tool or a spare tire or both.
How we're able to grasp such a prosaic action is still largely a mystery, but it seems
unlikely that it happens by virtue of a code. In a strict sense, a code is an arbitrary
system of alternatives. It's governed by rules of succession or substitution, and it's
learned more or less explicitly. The system of traffic lights is a code: Red, green, and
amber are correlated with distinct meanings (stop, go, and proceed with caution), and
drivers must learn them through a mixture of exposure and tutelage. Yet there's no
code for understanding tire-changing behavior. Now imagine a film scene showing
our man opening up the trunk of his car. When you see the action onscreen, and in
the absence of prior information to the contrary (say, an earlier scene showing the
driver depositing a corpse in the trunk), you would conjure up the same expectation
as in real life: In opening the trunk, he's looking for a tool or a spare or both. In real
life or in a movie, no appeal to a code seems necessary.
This example suggests that the process of understanding many things in films
is likely to draw upon ordinary, informal reasoning procedures. Contrary to much
film theory of the 1970s and 1980s, we need not ascribe this activity to the Freudian
or Lacanian unconscious. Just as you did not learn a code for tire changing, so is
there no reason for your expectation to be ascribed to repressed childhood memories
purportedly harbored in your unconscious. Presented with a set of circumstances
(flat tire, man opening trunk), you categorize it (driver changing flat tire) and draw
an informal, probabilistic conclusion, based on a structured piece of knowledge
about what is normally involved in the activity. You aren't aware of doing so—it's
a nonconscious activity—but there seems no need to invoke the drive-and-defense
model of the unconscious.
This isn't to say that only real-world knowledge is relevant to understanding films.
Obviously in real life it would be unlikely that a space alien would pop out of the car's
trunk, but if the film is in a certain genre, and a prior scene had shown said alien
creeping around the man's garage, that might be an alternative. Likewise, certain
technical choices, such as slow motion or fragmentary editing, require experience of
movies in order to be intelligible to viewers. But the point would be that even genre-
based or stylistic conventions are learned and applied through processes exercised
Cognition and Comprehension 137
in ordinary thinking. No special instruction, parallel to that of learning a code like
language or even semaphore, is necessary to pick up the conventions of horror films
or slow-motion violence.
Looked at from this perspective, understanding narrative films can be seen as largely
a matter of "cognizing." Going beyond the information given involves categorizing,
drawing on prior knowledge, making informal, provisional inferences, and hypoth-
esizing what is likely to happen next. To be a skilled spectator is to know how to execute
these tacit but determining acts. The goal, as story comprehension researchers have
indicated, is at least partly the extraction of "gist."6 When confronted with a narrative,
perceivers seek to grasp the crux or fundamental features of the event. Transforming
a scene into gist—the basic action that occurs, and its consequences for the characters
and the ensuing action—becomes a basis for more complex inferential elaboration.
This perspective has implications for how we look at the films as well. Rather than
searching for a "language" of film, we ought to look for the ways in which films are
designed to elicit the sorts of cognizing activities that will lead to comprehension
(as well as other effects). Put another way: Not all spectators are filmmakers, but
all filmmakers are spectators. It's not implausible to posit that they have gained an
intuitive, hands-on knowledge of how to elicit the sort of activities that will create the
experience they want the spectator to have. True, the design may misfire, or specta-
tors may choose to pursue alternative strategies of sense making. But as a first step
in a research program, it makes sense to postulate that filmmakers—scriptwriters,
producers, directors, editors, and other artisans of the screen—build their films in
ways that will coax most of their spectators to follow the same inferential pathways.
How, then, can a cognitive perspective help us analyze a film's narrative design?
Before tackling a particular example, I need to spell out my theoretical frame of refer-
ence a little more.
Narrative Norms
Let's assume that a film displays systematic patterns of narrative, themes, style, and the
like. The patterns can be located historically with respect to wider sets of customary
practices, which I'll call norms. For example, it's a norm of Hollywood studio film-
making since the mid-1910s that dramatic action takes place in a coherently unified
space—such as a bedroom, a street, or the deck of an ocean liner. That space is
portrayed through such means as continuity editing, constancy of items of setting,
roughly consistent sound ambience, and so on.
We can think of norm-driven subsystems as supplying cues to the spectator.
The cues initiate the process of elaboration, resulting eventually in inferences and
hypotheses. The spectator brings to the cues various bodies of relevant knowledge,
most notably the sort known to cognitive theorists as schema-based knowledge. A
schema is a knowledge structure that enables the perceiver to extrapolate beyond the
information given.7 Our schema for car breakdowns enables us to fill in what is not
immediately evident in the flat-tire situation; we go beyond the immediate picture of
a breakdown to extrapolate the driver's plan for getting going again.
138 Poetics of Cinema Cognition and Comprehension 139

Understanding a film calls upon cues and schemas constantly. For example, a
series of shots showing characters positioned and framed in particular ways usually
cues the viewer to infer that these characters are located in a particular locale. A
scene that begins with a detail shot of a table lamp may prompt the spectator to frame
hypotheses to the effect that the scene will take place in a living room or parlor. These
inferences and hypotheses couldn't get off the ground without schemas. The spec-
tator of a Hollywood film is able to understand that a space is coherent because at
some level of mental activity, she or he possesses a schema for typical locales, such
as living rooms or pool halls. Similarly, in the spectator s search for gist, she or he
must possess some rudimentary notion of narrative structure that permits certain
information to be taken for granted and other information to be understood as, say,
exposition or an important revelation. When we see a character leave one locale and
enter another, we effortlessly assume that the second scene follows the first chrono-
logically and that what happened in the suppressed interval isn't of consequence for isn t supplied. The film thus poses the question of who killed Archer, and this creates
the story action. (In some films, such as Fritz Lang's You Only Live Once [1937] and one strand in the overall mystery plot.
Otto Preminger's Fallen Angel [1945], such ellipses are later revisited and reveal that A second normative option is exemplified at the very start of The Letter (1940).
the narration skipped over important information.) Finally, I suggest that all these Here the shooting of a colonialist is plainly committed by the Bette Davis character
factors vary historically and culturally. We ought to expect that different filmmak- (Figure 4.2). The question posed is now that of why she killed him. What, if any,
ing traditions, in various times and places, will develop particular norms, schemas, circumstances justify the crime?
and cues. Correspondingly, the inferences and hypotheses available to spectators will
The first two scenes of Mildred Pierce, however, offer a more complex case. In a
vary as well. lonely beach house at night, with a car idling outside, a man is shot by an unseen
My outline is very skeletal, so I'll try to put some flesh on the bones by considering
assailant. As he dies, he murmurs, "Mildred." We glimpse a woman driving off in
a concrete case. My specimen is Mildred Pierce, an instantiation of that vast body of
the car. In the next scene, our protagonist, Mildred Pierce, is seen wandering along a
norms known as the classical Hollywood cinema. 8 I'll be concentrating on its system
deserted pier.
of narration, which involves not only its construction of a plot and a diegetic world,
We couldn t ask for a better example of a film that lures us down inferential
but also its use of film technique.
pathways. In a remarkably brief time—the murder scene lasts only 40 seconds—the
First, I'll try to show that the film utilizes norms of narration so as to encourage
spectator has accomplished a great deal. She has perceptually constructed a diegetic
not one but two avenues of inference and hypothesis testing; both of these would
world—a beach house at night, peopled by two characters. Further, she infers that a
seem to have been available to contemporary audiences. Second, I want to show
homicide has taken place; that gist is central to understanding this narrative. Only a
that the film assumes that in the viewer's effort after gist, she or he will ignore or
little less probable is the inference that the killer has fled by car. And the viewer may
forget certain stylistic norms. That is, Hollywood norms posit a hierarchy of impor-
also have inferred that the murderer is the woman named in the film's title.
tance, with narrative gist at the top and local stylistic manipulations subordinated
Yet such inferences are not one-time-only products. They form the basis of hypoth-
to that. In Mildred Pierce, this hierarchy allows the filmmakers to conceal crucial
eses, which lead in turn to further inference making. As Meir Sternberg points out,
narrational deceptions.
narrative ineluctably leads us to frame hypotheses about the past (what he calls curiosity
hypotheses) and about the future (suspense hypotheses).9 Here, the spectator will expect
Two Methods of Murder that there are prior reasons for the murder of Monte and that the film will reveal them
Because Mildred Pierce opens with a murder, it's profitable to start our inquiry with in its progressive unfolding. As a mystery film, Mildred Pierce will, so to speak, create
a norm-based question. What kinds of options were open to filmmakers in the 1940s suspense hypotheses about how curiosity hypotheses will be confirmed.
who wished to launch their plot with such a scene? We can specify two primary inference chains that this opening prompts. One is
In the early 1940s, the options were essentially two. One is exemplified by the that Mildred is the killer. Most critics have assumed that the average spectator comes
second scene of The Maltese Falcon (1941). Here, the murder of Sam Spade's partner, to this conclusion, and they characteristically take the opening as carefully directing
Miles Archer, is rendered in a way that conceals the killer's identity (Figure 4.1). We us to form this inference. First, like the Maltese Falcon sequence, the scene does not
see the victim from over the killer's shoulder, but a reverse-shot view of the murderer show who fires at Monte; this poses the question of the murderer's identity. Moreover,
140 Poetics of Cinema
Mildred is implied to be the killer on the basis of certain cues: the word Mildred,
which Monte murmurs before he dies; the smooth transition from the murder to
Mildred walking along the pier; the next scene, in which she tries to frame Wally for
the crime; and the still later scene in which her ex-husband, Bert, steps forward to
claim, implausibly, that he committed the murder, presumably to protect Mildred.
But at the film's climax, we'll learn that Mildred is not the killer. The film's opening
narration has misdirected us. By suppressing the identity of the killer, and by using
tight linkages between scenes, the narration leads the spectator to false curiosity
hypotheses. One critic puts the point this way: The film shifts from asking, "Who
killed Monte?" to asking, "Why did Mildred kill him?" 10 Indeed, the film couldn't
mislead us if we weren't undertaking a process of hypothesis formation and revision.
Still, a second line of inference is available. The blatant suppression of the
murderer's identity might lead the viewer to ask, If Mildred did it, why does the film not
show her in the act, as the opening of The Letter shows its heroine killing her victim?
One plausible reason for the film's equivocation was offered by a contemporary critic:
We are tempted to suspect the murderer is the woman on the bridge, especially
when we learn her name is Mildred.
But naturally, being familiar with the conventions of mystery stories that appear-
ances deceive and circumstantial evidence is not all, we are wary; indeed we feel
that somehow we had better not assume that Mildred Pierce Berargon [sic] has
11
just killed the man we duly learn is her second husband.
Under this construal, all the narrative feints I itemized above, the tight scene linkages
and the strategic actions taken by Mildred and her ex-husband, will be seen as so
many red herrings, tricky but "fair" in the way that misdirection is in, say, an Agatha
Christie novel.
We commonly believe that not all spectators make exactly the same inferences,
but this film builds such divergences into its structure by creating a pair of alterna-
tive pathways for the viewer. One path is signposted for the trusting spectator, who
assumes that Mildred is the killer and who will watch what follows looking for answers
to why she did it. There is also a pathway for the skeptical viewer, who will not take her
guilt for granted. This spectator will scan the ensuing film for other factors that could
plausibly account for the circumstances of the killing. And needless to say, it would be
possible for a particular viewer to switch between these alternative hypotheses, or to
rank one as more probable than the other. If the goal of the inferential process is that
extraction of gist, the ongoing construction of the story, then the filmmakers set for
themselves the task of building a system of cues that can be used in both frameworks,
the trusting one and the skeptical one.
Across the whole film, hypothesis forming and testing will be guided by cues of
various sorts and subordinated to various sorts of schemas. As a first approxima-
tion, let's distinguish between two principal varieties of schemas. Some schemas will
enable the spectator to assimilate and order cues on the basis of patterns of action;
call these action-based schemas. The story comprehension research literature offers
Cognition and Comprehension 141
many particular instances, such as the canonical macrostructure proposed by Jean
Mandler and her colleagues. She proposes that a traditional story opens by defining
a setting and then presenting a series of episodes. Each episode shows a character
responding to an initial condition, and the response is often that of forming a goal
to do something about that condition. The result is a goal path that informs future
episodes, whereby the character tries to reach the goal and either succeeds or fails.12
This is a very general account, but that doesn't make it hopelessly vague. In our film,
it seems clear that both the trusting and the skeptical spectator will test hypotheses
according to the ways that events fill various slots in Mandler's macrostructure. For
example, the viewer could take the scene that follows the murder as Mildred's com-
plex reaction to having committed the crime: She attempts suicide. Thwarted in that,
she formulates a new goal: to implicate the lubricious Wally in the crime. Luring him
to the beach house and locking him in can be seen as serving this larger purpose.
Each episode can be seen as springing from a reaction to prior events and leading
to a formulation of goals that initiate further action. Each one offers further support
for the trusting construal, but none definitively disconfirms the skeptical construal.
So the potential uncertainty about the murderer is maintained across the film.
Another general collection of schemas is relevant as well, one that we can label
agent-based schemas. It is significant, I think, that Mandler's canonical story reduces
character identity and activity to plot functions (reaction, goal formation, and so on).
In this respect, it resembles structuralist work in narratology, such as the studies by
Propp, Greimas, and Barthes. Yet one can recognize that characters are constructs
without acknowledging that they are wholly reducible to more fundamental semantic
or structural features. This would seem a necessary move to make if you're study-
ing cinema, because here, as opposed to literature, characters are usually embodied.
A novel's character may be, as Roland Barthes puts it, no more than a collection of
semes, or semantic features, gathered under a proper name. 13 In cinema, however,
the character has a palpable body, and actions seem naturally to flow from it. A
reaction or a goal is attached to a face and frame. Thus the fact that Monte is not only
a victim in the murder scene but also a concrete individual, likely to be important
in the narrative to come, must count for a good deal if we are to execute the process
of inference and hypothesis casting. Similarly, that Mildred happens to occupy Joan
Crawford's body—rather than that of, say, Lucille Ball or even Bette Davis—is not
a matter of indifference. More generally, it seems clear that in understanding any
film, our hypotheses involve not only courses of action but also the qualities of the
characters, not only action-based schemas but also agent-based ones.
Simplifying things, I'd suggest that in any narrative in any medium, characters
are built up by the perceiver by virtue of two sorts of agent-based schemas. One sort
comprises a set of institutional roles (e.g., teacher, father, or boss). Another sort of
agent-based schema is that afforded by the concept of the person, a prototype possess-
ing a cluster of several default features: a human body, perceptual activity, thoughts,
feelings, traits, and a capacity to plan and execute action. 14 Roughly, then, a character
consists of some person-like features plus the social roles that she or he fills. This
 Poetics of Cinema
142
distinction would seem to be constant across cultures, even if the substantive concep-
tions of agent and role vary.15
Aided by role schemas and the person schema, the spectator can build up the
narrative's agents to various degrees of individuality. Mildred can be taken as a self-
sacrificing mother, as a heedless wife, as a ruthless business owner, and so on. At
various points, each of these conceptual constructions is important in making sense
of the plot, which in turn reveals new aspects of Mildred's character. (At the climax,
we'll learn that she's a very self-sacrificing mother.) And each conception of Mildred
can coexist with the trusting construal (the reasons why Mildred would kill Monte
are rooted in her personality and motives) and with the skeptical construal (even if
such characterizations are accurate, they may not actually lead to the murder we more
or less witness). And we should note that this construction of Mildred as a charac-
ter—person plus roles—constitutes no less an effort after gist than does the construal
of the action around the murder scene. The viewer plays down or omits concrete
details of character action in order to construct a psychic identity and agency of broad
import, capable of being integrated into hypotheses about upcoming or past action.
Such hypotheses are, of course, constrained in the overall course of the film. After
Mildred has lured Wally into being found at the scene of the murder, she is taken in
for questioning. As she tells her story to the police in a series of flashbacks, the film
breaks into two large-scale portions, and both action- and agent-based schemas are
involved in each. The first part consists of the lengthy flashback showing us Mildred's
rise to business success. One purpose of this is to establish that her former husband,
Bert, has a motive for killing Monte. This long flashback ends with Bert's granting
Mildred a divorce and insultingly knocking the whiskey out of Monte's hand. In the
framing story, the police inspector argues that this confirms Bert's guilt. And indeed
Bert's willingness to take the blame initially confirms that he is shielding Mildred.
Once again, though, this permits two alternative readings of the action. Our trusting
viewer takes Bert's confession as confirming that Mildred is guilty. The more suspi-
cious viewer, aware of genre conventions that manipulate this sort of information, is
likely to suspect that such an obvious foil for Mildred may conceal more than this.
That is, just as Mildred has been a red herring for the real culprit, Bert is a red herring
once removed, delaying the revelation of the real killer.
At the end of this framing portion in the police station, Mildred confesses to the
crime. This switches attention away from Bert and back to her. But her confession
creates a problem in motivation. At the end of the first flashback, Mildred is portrayed
as being completely in love with Monte. The task of the next long flashback is to show
how she could become capable of murdering him.
The flashback traces her gradual realization that Monte is deeply immoral. He's
lazy, evasive, and not above seducing his stepdaughter. The flashback also reveals that
Mildred is capable of murder. Here the crucial scene is her high-pitched quarrel with
Veda, in which Mildred orders her to leave: "Get out before I kill you." The crisis of
this stretch of the film comes when Mildred learns that Monte has destroyed her
business on the very night of Veda's birthday. Mildred takes out a revolver and goes to
Monte's beach house. This puts her firmly on the scene of the crime.
Cognition and Comprehension 143
Confirming that Mildred committed the murder would clinch the trusting viewer's
long-range hypothesis, based on action-driven schemas. Killing Monte becomes
Mildred's means to the goal of protecting her daughter, a goal she has held through-
out her life. The resolution would also invoke person-plus-role schemas: Mildred
remains the self-sacrificing mother to the end. But this resolution is invoked only to
be dispelled.
Once more we return to the present, and the inspector announces that the police
have captured the real murderer. Veda is brought in and, believing that Mildred has
implicated her, blurts out a confession. And Mildred's recitation of the events now
leads to the final flashback, which we enter with knowledge of the killer's identity.
As in The Letter, the interest now falls upon what circumstances triggered the murder
and how those vary from our initial impression.
The final flashback, recounted by Mildred, shows her arrival at the beach house and
her discovery that Monte and Veda are lovers. She pulls the pistol, but Monte dissuades
her and she drops it. Mildred walks outside, and Veda learns from Monte that he
no longer loves her. As Mildred is about to drive away, Veda shoots Monte. Mildred
hurries in and discovers the crime, but through a mixture of lies and cajoling, Veda
convinces her not to call the police.
This flashback, the real climax, confirms skepticism about Mildred's culpability,
and we learn the reason why the narration withheld the killer's identity. Moreover, all
of Mildred's subsequent behavior—trying to frame Wally and later confessing to the
crime—is consistent with the fact that Veda killed Monte. Veda's act of fury triggers
the same motherly sacrifice that has defined Mildred as agent throughout. Everything
that we saw at the start of the film is retrospectively justified by Mildred's acting as
Veda's accomplice.
Again, to arrive at this concluding set of inferences is to continue our effort after
gist. This ending reminds us that the filmmakers are practical cognitive psycholo-
gists. They know, for instance, the importance of default assumptions. One purpose
of the murder scene is to make us assume that only one person is in or around the
cottage when Monte is killed. This premise is crucial because even if we are not shown
who pulls the trigger, the viewer must not suspect Veda at all. If her presence is even
hinted at, the redundant and obvious clues pointing to Mildred will be seen immedi-
ately for the red herrings they are.
T h e Partial Replay
Reading, notes Barthes in S/Z, involves forgetting. 16 So does viewing. The ending
of Mildred Pierce is instructive partly because the film is so made as to exploit our
likely inability to remember anything but the material made salient by our ongoing
inference making and hypothesis testing. As practical psychologists, our filmmakers
know that we'll construct a diegetic world chiefly through landmarks, not fine details
of setting. They know that we'll move rapidly from items of appearance and behavior
to inferences about character beliefs and traits. And they know that under pressure of
144 Poetics of Cinema Cognition and Comprehension 145

TABLE 4.1 Mildred Pierce: The Opening Scene and Its Replay
Opening Shots (A)
1. 5 seconds: (extreme long shot) A beach
house at night; a car is visible alongside
(Figure 4.3). Dissolve to:
2. 4 seconds: (long shot) House and car.
Two pistol shots are heard (Figure 4.4).
3. 8 seconds: (medium-long shot) Monte
facing camera, looking off left (Figure 4.5).
Third and fourth pistol shots hit a mirror.
Monte is hit, staggers forward (Figure 4.6),
and falls to the floor. A pistol is tossed into
the frame (Figure 4.7).
4. 13 seconds: (medium shot) Monte
wobbles his head, opens his eyes, and says,
"Mildred" (Figure 4.8). Pan up to mirror;
sound of door slamming (Figure 4.9).
5. (long shot) Empty parlor, with Monte's
corpse in the firelight. Doorway is empty
(Figure 4.10).
6. (long shot) Car outside pulls off
(Figure 4.11).
Replay Shots (B)
1. 12 seconds: (medium shot) Mildred goes
to the car and tries to start the engine
(Figure 4.12).
2. 4 seconds: (medium close-up):
Mildred slumped over the steering wheel.
Two shots are heard (Figure 4.13).
3. 5 seconds: (medium shot) Veda fires four
times (Figure 4.14).
Figure 4.3 Mildred Pierce (1945): the opening Figure 4.4 Mildred Pierce: the opening scene,
scene, shot 1. shot 2.
4. 6 seconds: (medium shot) Monte is
staggering forward and falls to the floor
(Figure 4.15). A pistol is tossed into the
frame (Figure 4.16). Monte wobbles his
head, with his eyes open, and says,
"Mildred" (Figure 4.17).
5. (long shot) Pan with Mildred coming in;
sound of door slamming (Figure 4.18).
She meets Veda in the parlor
(Figure 4.19).
6-10. In the parlor, Veda tells Mildred lies Figure 4.5 Mildred Pierce: the opening scene, Figure 4.6 Mildred Pierce: the opening scene,
about shooting Monte. Near the shot 3. shot 3 (continued).
doorway, she begs Mildred to protect
her (Figure 4.20).

the clock, we're likely to overlook stylistic features. This last aspect is especially critical
in Mildred Pierce.
I compared the film to a mystery novel in its use of red herrings, but the film
compels us to recognize that certain features of cinema as a medium shape our infer-
ential activity too. Although few mystery readers may dutifully page back to check
a fact or appreciate how they were misled, they all have the option of doing so. A
book is in hand all at once, and you may scan, skim, or skip back at will. This isn't an
Figure 4.7 Mildred Pierce: the opening scene, Figure 4.8 Mildred Pierce: the opening scene,
option for the ordinary film viewer (at least, until the arrival of home video, and even shot 3 (concluded). shot 4.
then the exact pairwise comparison of passages is difficult). The classical Hollywood
cinema paced its narration for maximum legibility during projection. Accordingly,
filmmakers have learned that, for perceivers who can't stop and go back, cues must be
Many actions are reiterated in the second version, and the redundancies suggest
highly redundant. But in learning this, filmmakers have also learned how to prompt
that we are seeing a straight replay. In the opening scene, over the second long shot
mzsremembering. Given our effort after gist and our inability to turn back to check a
of the beach house (A2, Figure 4.4), we hear two gunshots. The cut inside to Monte
point (especially one made 90 minutes earlier), the film can introduce both redundant
facing the killer (A3, Figure 4.5) comes right on the third pistol shot. In the flashback
cues and highly nonredundant, even contradictory, ones.
version, the cut from Mildred in the car (B2, Figure 4.13) comes at exactly the same
Table 4.1 aligns the two sequences, the opening murder, (labeled A) and the
point. The next shot (B3, Figure 4.14) replaces the image of Monte with that of Veda
climactic replay of the shooting (B).
firing the revolver. No time can be said to be omitted here. More subtly, the screen time
 Poetics of Cinema Cognition and Comprehension 147
146

Figure 4.15 Mildred Pierce: the final flash- Figure 4.16 Mildred Pierce: thefinalflash-
Figure 4.10 Mildred Pierce: the opening scene,
Figure 4.9 Mildred Pierce: the opening scene, back, shot 4. back, shot 4 (continued).
shot 5.
shot 4 (concluded).

It would seem, however, that the narration profits from so many redundancies in
order to introduce some significant disparities. True, some are just minor. In the initial
scene (A1 and A2), there is no sound of the car ignition cranking as Mildred tries to start
it. (Perhaps Max Steiner's score smothers it.) There is, furthermore, no indication that
Mildred is in the car in the first scene (Figures 4.3-4.4). (True, she is slumped over the
steering wheel in the later version, but scrutiny of the first scene shows that the driver's
seat is empty.)17 These tiny disparities show again the perceptual saliency of causal,
event-centered information, especially as prepared by prior knowledge. On our first
Figure 4.12 Mildred Pierce: thefinalflash- view of the first scene, the apparent emptiness of the car suggests that the important
Figure 4.11 Mildred Pierce: the opening scene,
back, shot 1. action occurs inside the house. If anyone should recall that scene 100 minutes later, the
shot 6.
later shot of Mildred bent over the steering wheel (B2, Figure 4.13) suffices as a rough
explanation of why the car looked empty. In the absence of a chance to go back and com-
pare, the spectator can easily accept the later scene as consistent with the earlier one.
Other variations in the two scenes reveal that the filmmakers are exploiting the
viewer's inability to recall certain details. In the first version, when Monte is shot
(A3), he falls to the floor and rolls over on his back as the gun is tossed into the shot
(Figures 4.5-4.7). There is a pause. Cut to a closer view of his face (A4, Figure 4.8).
As his head wobbles, he opens his eyes, looking left as he murmurs, "Mildred," and
expires. The close-up emphasizes his expression and the word he utters, marking the
event for us to notice and recall. It may also suggest that he dies looking at his killer
Figure 4.14 Mildred Pierce: thefinalflash- and speaking her name.
Figure 4.13 Mildred Pierce: thefinalflash-
back, shot 3.
back, shot 2. But in the second version, the event is treated differently; or rather, it is no longer
the same event. Monte is shot and tumbles to the floor (B4, Figures 4.15-4.16). But
now he utters Mildred's name just as he starts to roll onto his back (Figure 4.17).
that elapses between the third gunshot and Monte's dying word is virtually identical
There is no close-up, and no pause either. He says nothing when he is in the position he
in both versions (9 seconds and 10 seconds respectively). Finally, the slamming door
assumed in the earlier scene (just as earlier, he said nothing when he rolled over). The
we hear in the opening scene (A4, over Figure 4.9) is revealed to be not the killer leav-
second version produces a different effect. By speaking when he is not looking toward
ing, as we initially inferred, but rather the sound of Mildred entering to find Veda in
his killer, he no longer seems to be naming the culprit but rather recalling Mildred.
the living room (B5, Figure 4.18). These are what the mystery novelist might consider
In this flashback, she no longer seems guilty. The narration gets two distinct cues out
fair misdirections of the spectator's attention. They suggest that the second version is
of the two versions, and it is able to do so because it counts on our remembering only
identical with the first, except that the former fills in certain details of the latter.
 Poetics of Cinema
148
Figure 4.17 Mildred Pierce: thefinalflash- Figure 4.18 Mildred Pierce: thefinalflash-
back, shot 4 (concluded). back, shot 5.
Figure 4.19 Mildred Pierce: thefinalflash- Figure 4.20 Mildred Pierce: thefinalflash-
back, shot 10
back, shot 5 (concluded).
that Monte said, "Mildred," not exactly when and how he said it. We recall the salient
features marked out for us earlier, but not the details of each situation.18
Even more striking than the reconstitution of Monte's dying word is the disparity
in the handling of the murder's aftermath. In the first scene (A4, Figures 4.8-4.9), a
camera movement carries us from Monte's face to the bullet-pocked mirror, which
shows a doorway opening onto the hall. We hear footsteps and a slamming door. Cut
to a long shot (A5, Figure 4.10) of Monte lying in the empty parlor. Cut outside to the
car pulling away, the driver dimly visible (A6, Figure 4.11). But the second version
follows Monte's death and the slamming door by Mildred's lengthy and intense
confrontation with Veda (B5-B9, Figures 4.18-4.20). And this encounter is played
out exactly in the doorway that is shown empty in the first version's fifth shot of the
empty parlor (Figure 4.10)! Moreover, because the second version never completes the
scene between Mildred and Veda, there is no depiction of either one driving off after
their interchange. (Indeed, we never learn their arrangements about leaving. If Veda
took the car, how did Mildred get to the pier?)
If we try to make the two versions compatible, we must posit that in the first
version, there is an ellipsis of several minutes between the end of the mirror shot
(A4, Figure 4.9) and the beginning of A5 (Figure 4.10), which presumably depicts
Monte lying dead in the room after Mildred and Veda have gone their ways. This
Cognition and Comprehension 149
ellipsis is, of course, not marked at all. Indeed, one overriding default assumption
of the classical film is that a cut within a defined locale is taken to convey continu-
ous duration unless there are stylistic or contextual indications to the contrary (e.g.,
a dissolve or some drastic change of costume or furnishings). Alternatively, it is
possible in retrospect to construe shot A5, the long-shot framing of Monte's corpse
stretched out (Figure 4.10), as simply a false image, provided to mislead us. Either way,
the opening scene's narration has concealed the crucial point that two women were
present, and it has cued the viewer to infer the gist of the situation—a man was killed,
and a woman fled the scene—in such a fashion that the details can't be recalled.
We can be fairly confident that this memory lapse is widespread. First-time specta-
tors seem not to notice the disparities between the two versions, and critics who have
written on the film have not mentioned them. Indeed, critics have proven especially
vulnerable to remembering gist and forgetting detail. One writer, describing the first
scene shot by shot, omits the crucially misleading shot of Monte's corpse by the fire
(A5, Figure 4.10).19 Another critic claims that in the second murder scene, the shot of
Veda firing the pistol (Figure 4.14) follows the shot of Monte (Figure 4.15).20 If critics
who have the luxury of reexamining the film can err in such ways, should we be
surprised that a writer in 1947, relying on mere memory, fleshed out what he saw
in unsupportable ways? He cites "the sequence of camera shots in which we see the
outside of the house, the woman's figure (or was it two figures separately?) leaving it,
her ride in the auto."21 And if you feel a need to check my claims to confirm your own
recollections, you realize that I'm not condemning these critics. They're doing what
we all do, "making sense," and they are making it along the lines laid down by a very
powerful system of norms and cues. It's not just that the film encourages us to deceive
ourselves. It deceives us blatantly but helps us overlook the deception. It accomplishes
this because narrative comprehension demands that we go beyond the data, jump to
conclusions—in short, make inferences and frame hypotheses.
Secrets and Lies, and Narration
A lone example can't prove a case, but I hope that this examination of Mildred Pierce
has illustrated how the cognitive perspective might tie together assumptions about
comprehension with concrete observations about a film's structure and style. The
result is a significantly new picture of a film and its viewer.
Instead of a "pure" text, understandable "in itself," we have a text that gains its
effects only in relation to a body of norms, a set of schemas, and the processes that
the spectator initiates. Instead of a communication model, which treats meaning as
dropped in upstream to be fished out by the spectator, we have a constructive model
that treats meaning as an expanding elaboration of cues located in the text. This shift
implies as well that, armed with certain schemas and knowledge of certain norms, the
spectator could "go beyond the information given" in ways unforeseen by the film-
makers. What makes a film understandable is not necessarily exhausted by what the
filmmakers deliberately put in to be understood. But, then, this is true of all human
activities; every action has unintended consequences, and so it's hardly surprising
150 Poetics of Cinema

that viewers appropriate movies in idiosyncratic ways. But the process of that appro-
priation is also a matter of inferential elaboration, based on fresh schemas the specta-
tor brings to bear on the film's discriminable features.
In isolating comprehension as a central viewing activity, the cognitive perspective
is open to the charge that it ignores other aspects of the experience and of the film
itself. What, for instance, about emotion, surely a prime ingredient of the filmgoing
5.
experience? And what about interpretation, which seems to go even farther beyond
the information given and involve very high-level constructs?
These are important questions, and the cognitive frame of reference needs to T h e A r t Cinema as a
respond to them. Up to a point, setting emotion aside is a useful methodological
idealization: In principle, you can understand a film without discernibly having an
emotional reaction to it. More positively, studies by Noel Carroll, Murray Smith,
Ed Tan, and others suggest that a cognitive perspective can enrich our understanding
Mode of Film. Practice
of emotive qualities. 22 This research boldly proposes that many emotional responses
ride upon cognitive judgments.
As for interpretation, elsewhere I've tried to show that, as an intuitive but principled
activity, it's highly amenable to a cognitive explanation. When a critic posits Mildred as
the Castrating Mother or a symbol of the contradictions of entrepreneurial capitalism,
the critic is still seeking out cues, categorizing, applying schemas, and making infer-
ences that carry weight among a particular social group. 23 To interpret is to cognize.
Finally, as a murder mystery, Mildred Pierce may play too much into my hands.
Not every film poses a mystery at its start; is the cognitive perspective at risk of turn-
ing every film into a detective story? It's true that mystery films show the process of
hypothesis formation quite clearly, but the cognitive framework doesn't favor them.
La Strada (1954), 8 1/2 (1963), Wild Strawberries (1957), The Seventh Seal (1957),
In Narration in the Fiction Film, I try to show that the activity of inferential elabo-
Persona (1966), Ashes and Diamonds (1958), Jules et Jim (1962), Knife in the Water
ration is prompted by melodramas (In This Our Life, Say It With Songs), Westerns
(1962), Vivre sa vie (1962), Muriel (1963): Whatever else one can say about these films,
(Wild and Woolly), comedies (His Girl Friday), and straight dramas (Heaven Knows,
cultural fiat gives them a role altogether different from Rio Bravo (1959) on the one
Mr. Allison). Every narrative of any complexity withholds some story information
hand and Mothlight (1963) on the other. They are "art films," and, ignoring the tang of
from both viewers and characters. This creates gaps in our knowledge, disparities
snobbishness about the phrase, we can say that these and many other films constitute
among various characters' states of knowledge, and mismatches between a character's
a distinct branch of the cinematic institution. My purpose in this essay is to argue that
knowledge and the viewer's knowledge, all the while generating Meir Sternberg's
we can usefully consider the "art cinema" as a distinct mode of film practice, possess-
response trio of curiosity, suspense, and surprise. Every film's narration depends
ing a definite historical existence, a set of formal conventions, and implicit viewing
upon regulating the flow of information, and we don't have perfect information until
procedures. Given the compass of this paper, I can only suggest some lines of work,
the end (if then). In this respect, every narrative harbors secrets.
but I hope to show that constructing the category of the art cinema is both feasible
There is much more to understand about how viewers understand films. The line and illuminating.
of inquiry sketched here puts a priority on studying particular films in the light
It may seem perverse to propose that films produced in such variable cultural
of how narrational and stylistic processes are designed to elicit certain spectato-
contexts might share fundamentally similar features. Yet I think there are good
rial effects. In this research program, Mildred Pierce exemplifies key features of the
reasons for Relieving this, reasons which come from the films' place in history.
classical Holly wood film. There are, of course, other traditions that call on differ-
In the long run, the art cinema descends from the early film dart and such silent
ent sorts of narrational cues, schemas, and norms. 24 Comparative inquiry into these
national cinema schools as German Expressionism and Neue Sachlichkeit and French
traditions can contribute to that research program I've called a film poetics. By avoid-
Impressionism. 1 (A thorough account of its sources would have to include literary
ing misplaced conceptions of codes or slippery analogies between film and language,
modernism, from Proust and James to Faulkner and Camus.) More specifically, the
the cognitive perspective offers a robust account of the viewer's activity, one that can
art cinema as a distinct mode appears after World War II when the dominance of the
guide a historical poetics of cinema.

151
 Poetics of Cinema
152
Hollywood cinema was beginning to wane. In the United States, the courts' divorce-
ment decrees created a shortage of films for exhibition. Production firms needed
overseas markets and exhibitors needed to compete with television. In Europe, the
end of the war reestablished international commerce and facilitated film export and
coproductions. Thomas Guback has shown how, after 1954, films began to be made
for international audiences. 2 American firms sponsored foreign production, and
foreign films helped American exhibitors fill screen time. The later Neorealist films
may be considered the first postwar instances of the international art cinema, and
subsequent examples would include most works of the New Wave, Fellini, Resnais,
Bergman, De Sica, Kurosawa, Pasolini, et al. While the art cinema is of little economic
importance in the United States today, it evidently continues, as such international
productions as The Serpent's Egg (1977) and Stroszek (1977) show.
Identifying a mode of production/consumption does not exhaustively characterize
the art cinema, since the cinema also consists of formal traits and viewing conven-
tions. To say this, however, is to invite the criticism that the creators of such film are
too inherently different to be lumped together. Yet I shall try to show that whereas
stylistic devices and thematic motifs may differ from director to director, the overall
functions of style and theme remain remarkably constant in the art cinema as a whole.
The narrative and stylistic principles of the films constitute a logically coherent mode
of cinematic discourse.
Realism, A u t h o r s h i p , A m b i g u i t y
The classical narrative cinema—paradigmatically, studio feature filmmaking in
Hollywood since 1920—rests upon particular assumptions about narrative structure,
cinematic style, and spectatorial activity. While detailing those assumptions is a task
far from complete, 3 we can say that in the classical cinema, narrative form motivates
cinematic representation. Specifically, cause-effect logic and narrative parallelism
generate a narrative which projects its action through psychologically defined, goal
oriented characters. Narrative time and space are constructed to represent the cause-
effect chain. To this end, cinematic representation has recourse to fixed figures of
cutting (e.g., 180 continuity, crosscutting, "montage sequences"), mise-en-scene (e.g.,
three-point lighting, perspective sets), cinematography (e.g., a particular range of
camera distances and lens lengths), and sound (e.g., modulation, voice-over narra-
tion). More important than these devices themselves are their functions in advancing
the narrative. The viewer makes sense of the classical film through criteria of verisi-
militude (is x plausible?), of generic appropriateness (is x characteristic of this sort of
film?) and of compositional unity (does x advance the story?). Given this background
set, we can start to mark off some salient features of the art cinema.
First, the art cinema defines itself explicitly against the classical narrative mode,
and especially against the cause-effect linkage of events. These linkages become looser,
more tenuous in the art film. In LAvventura (1960), Anna is lost and never found; in
A bout de souffle (aka Breathless; 1960), the reasons for Patricia's betrayal of Michel
remain unknown; in Bicycle Thieves (1948), the future of Antonio and his son is not
^ he ./Vrt Cinema as a jN/Iode of Film Practice
153
revealed. It will not do, however, to characterize the art film solely by its loosening of
causal relations. We must ask what motivates that loosening, what particular modes
of unity follow from these motivations, what reading strategies the film demands, and
what contradictions exist in this order of cinematic discourse.
The art cinema motivates its narratives by two principles: realism and authorial
expressivity. On the one hand, the art cinema defines itself as a realistic cinema. It will
show us real locations (Neorealism, the New Wave) and real problems (contemporary
"alienation," "lack of communication," etc.). Part of this reality is sexual; the
aesthetics and commerce of the art cinema often depend upon an eroticism that
violates the production code of pre-1950 Hollywood. A Stranger Knocks (1959) and
And God Created Woman (1956) are no more typical of this than, say, Jules et Jim and
Persona (whereas one can see Le Mepris, 1963, as consciously working upon the very
problem of erotic spectacle in the art cinema). Most important, the art cinema uses
"realistic"—that is, psychologically complex—characters.
The art cinema is classical in its reliance upon psychological causation; characters
and their effects on one another remain central. But whereas the characters of the
classical narrative have clear-cut traits and objectives, the characters of the art cinema
lack defined desires and goals. Characters may act for inconsistent reasons (Marcello
in La Dolce Vita, 1960) or may question themselves about their goals (Borg in Wild
Strawberries and the Knight in The Seventh Seal). Choices are vague or nonexistent.
Hence a certain drifting episodic quality to the art film's narrative. Characters may
wander out and never reappear; events may lead to nothing. The Hollywood pro-
tagonist speeds directly toward the target; lacking a goal, the art-film character slides
passively from one situation to another.
The protagonist's itinerary is not completely random; it has a rough shape: a trip
(La Strada; Wild Strawberries; The Silence, 1963), an idyll (Jules et Jim; Elvira Madigan,
1967; Pierrot lefou, 1965), a search (L'Avventura; Blow-Up, 1966; High and Low, 1963),
even the making of a film (8 1/2; Le Mepris; The Clowns, 1971; Fellini Roma, 1972;
Day for Night, 1973; The Last Movie, 1971). Especially apt for the broken teleology of
the art film is the biography of the individual, in which events become pared down
toward a picaresque successivity (La Dolce Vita; Ray's Apu trilogy, 1955-1959; Alfie,
1966). If the classical protagonist struggles, the drifting protagonist traces an itiner-
ary, an encyclopedic survey of the film's world. Certain occupations (stockbroking in
L'Eclisse, 1962; journalism in La Dolce Vita and The Passenger, 1975; prostitution in
Vivre sa vie and Nights of Cabiria, 1957) favor a survey form of narrative. Thus the
art film's thematic of la condition humaine, its attempt to pronounce judgments on
"modern life" as a whole, proceeds from its formal needs: had the characters a goal,
life would np longer seem so meaningless.
What is essential to any such organizational scheme is that it be sufficiently loose in
its causation as to permit characters to express and explain their psychological states.
Slow to act, these characters tell all. The art cinema is less concerned with action than
reaction; it is a cinema of psychological effects in search of their causes. The dissec-
tion of feeling is often represented explicitly as therapy and cure (e.g., Through a Glass
Darkly, Persona), but even when it is not, the forward flow of causation is braked
154 Poetics of Cinema
and characters pause to seek the aetiology of their feelings. Characters often tell one
another stories: autobiographical events (especially from childhood), fantasies, and
dreams. (A recurring line: "I had a strange dream last night.") The hero becomes a
supersensitive individual, one of those people on whom nothing is lost. During the
film's survey of its world, the hero often shudders on the edge of breakdown. There
recurs the realization of the anguish of ordinary living, the discovery of unrelieved
misery: compare the heroines of Europa 51 (1952), LAvventura, Deserto rosso (1964),
and Unefemme mariee (1964). In some circumstances the characters must attribute
their feelings to social situations (as in Ikiru [1952], I Live in Fear [1955], and Shame).
In Europa 51, a communist tells Irene that individuals are not at fault: "If you must
blame something, blame our postwar society." Yet there is seldom analysis at the level
of groups or institutions; in the art cinema, social forces become significant insofar as
they impinge upon the psychologically sensitive individual.
A conception of realism also affects the film's spatial and temporal construc-
tion, but the art cinema's realism here encompasses a spectrum of possibilities. The
options range from documenting factuality (e.g., II Posto, 1961) to intense psycho-
logical subjectivity (Hiroshima mon amour, 1959). (When the two impulses meet in
the same film, the familiar "illusion-reality" dichotomy of the art cinema results.)
Thus room is left for two reading strategies. Violations of classical conceptions of
time and space are justified as the intrusion of an unpredictable and contingent daily
reality or as the subjective reality of complex characters. Plot manipulations of story
order (especially flashbacks) remain anchored to character subjectivity as in 81/2 and
Hiroshima mon amour. Manipulations of duration are justified realistically (e.g., the
temps morts of early New Wave films) or psychologically (the jump cuts of A bout de
souffle signaling a jittery lifestyle). By the same token, spatial representation will be
motivated as documentary realism (e.g., location shooting, available light), as char-
acter revelation, or in extreme cases as character subjectivity. Andre Bazin may be
considered the first major critic of the art cinema, not only because he praised a loose,
accidental narrative structure that resembled life but also because he pin-pointed
privileged stylistic devices for representing a realistic continuum of space and time
(deep-focus, deep space, the moving camera, and the long take). In brief, a commit-
ment to both objective and subjective verisimilitude distinguished the art cinema
from the classical narrative mode. 4
Yet at the same time, the art cinema foregrounds the author as a structure in the
film's system. Not that the author is represented as a biographical individual (although
some art films, e.g., Fellini's, Truffaut's, and Pasolini's, solicit confessional readings),
but rather the author becomes a formal component, the overriding intelligence orga-
nizing the film for our comprehension. Over this hovers a notion that the art-film
director has a creative freedom denied to her/his Hollywood counterpart. 5 Within
this frame of reference, the author is the textual force "who" communicates (what is
the film saying?) and "who" expresses (what is the artist's personal vision?). Lacking
identifiable stars and familiar genres, the art cinema uses a concept of authorship to
unify the text.
The Art Cinema as a Mode of Film Practice 155
Several conventions operate here. The competent viewer watches the film expect-
ing not order in the narrative but stylistic signatures in the narration: technical
touches (Truffaut's freeze frames, Antonioni's pans) and obsessive motifs (Bunuel's
anticlericalism, Fellini's shows, Bergman's character names). The film also offers
itself as a chapter in an oeuvre. This strategy becomes especially apparent in the con-
vention of the multi-film work (the Apu trilogy, Bergman's two trilogies, Rohmer's
"Moral Tales," and Truffaut's Doinel series). The initiated catch citations: references to
previous films by the director or to works by others (e.g., the New Wave homages).
A small industry is devoted to informing viewers of such authorial marks. Inter-
national film festivals, reviews and essays in the press, published scripts, film series,
career retrospectives, and film education all introduce viewers to authorial codes.
What is essential is that the art film be read as the work of an expressive individual.
It is no accident, then, that the politique des auteurs arose in the wake of the art
cinema, that Cahiers du cinema admired Bergman and Antonioni as much as Hawks
and Minnelli, that Robin Wood could esteem both Preminger and Satyajit Ray. As a
critical enterprise, auteur analysis of the 1950s and 1960s consisted of applying art-
cinema reading strategies to the classical Hollywood cinema. 6
How does the author come forward in the film? Recent work in Screen has shown
how narrational marks can betray the authorial code in the classical text, chiefly
through gaps in motivation. 7 In the art-cinema text, the authorial code manifests itself
as recurrent violations of the classical norm. Deviations from the classical canon—an
unusual angle, a stressed bit of cutting, a prohibited camera movement, an unrealistic
shift in lighting or setting—in short, any breakdown of the motivation of cinematic
space and time by cause-effect logic—can be read as "authorial commentary." The
credits for the film, as in Persona or Blow-Up, can announce the power of the author
to control what we see. Across the entire film, we must recognize and engage with
the shaping narrative intelligence. For example, in what Norman Holland calls the
"puzzling film,"8 the art cinema foregrounds the narrational act by posing enigmas.
In the classic detective tale, however, the puzzle is one of story: who did it? How? Why?
In the art cinema, the puzzle is one of plot: who is telling this story? How is this story
being told? Why is this story being told this way? Another example of such marking
of narration is the device of the flashforward—the plot's representation of a future
story action. The flashforward is unthinkable in the classical narrative cinema, which
seeks to retard the ending and efface the mode of narration. But in the art cinema, the
flashforward functions perfectly to stress authorial presence: we must notice how the
narrator teases us with knowledge that no character can have. Far from being isolated
or idiosyncratic, such instances typify the tendency of the art film to throw its weight
onto plot, not story; we play a game with the narrator.
Realism and authorial expressivity, then, will be the means whereby the art film
unifies itself. Yet these means now seem contradictory. Verisimilitude, objective or
subjective, is inconsistent with an intrusive author. The surest signs of authorial intel-
ligibility—the flashforward, the doubled scene in Persona, the color filters at the start
of Le Mepris—are the least capable of realistic justification. Contrariwise, to push the
156 Poetics of Cinema
realism of psychological uncertainty to its limit is to invite a haphazard text in which
the author's shaping hand would not be visible. In short, a realist aesthetic and an
expressionist aesthetic are hard to merge.
The art cinema seeks to solve the problem in a sophisticated way: by the device
of ambiguity. The art film is nonclassical in that it foregrounds deviations from the
classical norm—there are certain gaps and problems. But these very deviations are
placed, resituated as realism (in life things happen this way) or authorial commentary
(the ambiguity is symbolic). Thus the art film solicits a particular reading procedure:
whenever confronted with a problem in causation, temporality, or spatiality, we first
seek realistic motivation. (Is a character's mental state causing the uncertainty? Is life
just leaving loose ends?) If we're thwarted, we next seek authorial motivation. (What
is being "said" here? What significance justifies the violation of the norm?) Ideally, the
film hesitates, suggesting character subjectivity, life's untidiness, and author's vision.
Whatever is excessive in one category must belong to another. Uncertainties persist
but are understood as such, as obvious uncertainties, so to speak. Put crudely, the
slogan of art cinema might be "When in doubt, read for maximum ambiguity."
The drama of these tendencies can play across an entire film, as Giulietta degli
spiriti and Deserto rosso illustrate. Fellini's film shows how the foregrounding of
authorial narration can collapse before the attempt to represent character subjectivity.
In the hallucinations of Giulietta, the film surrenders to expressionism. Deserto rosso
keeps the elements in better balance. Putting aside the island fantasy, we can read any
scene's color scheme in two registers simultaneously: as psychological verisimilitude
(Giuliana sees her life as a desert) or as authorial commentary (Antonioni-as-narrator
says that this industrial landscape is a desert).
If the organizational scheme of the art film creates the occasion for maximizing
ambiguity, how to conclude the film? The solution is the open-ended narrative. Given
the film's episodic structure and the minimization of character goals, the story will
often lack a clear-cut resolution. Not only is Anna never found, but the ending of
L'Avventura refuses to specify the fate of the couple. At the close of Les 400 coups
(1959), the freeze frame becomes the very figure of narrative irresolution, as does the
car halted before the two roads at the end of Knife in the Water. At its limit, the art
cinema creates an 8 1/2 or a Persona, a film which, lacking a causally adequate ending,
seems to conclude several distinct times. A banal remark of the 1960s, that such films
make you leave the theater thinking, is not far from the mark: the ambiguity, the play
of thematic interpretation, must not be halted at the film's close. Furthermore, the
pensive ending acknowledges the author as a peculiarly humble intelligence; she or
he knows that life is more complex than art can ever be, and the only way to respect
this complexity is to leave causes dangling, questions unanswered. With the open and
arbitrary ending, the art film reasserts that ambiguity is the dominant principle of
intelligibility, that we are to watch less for the tale than the telling, that life lacks the
neatness of art and this art knows it.
The Art Cinema as a Mode of Film Practice 157
T h e A r t C i n e m a in History
The foregoing sketch of one mode of cinema needs more detailed examination, but in
conclusion it may be enough to suggest some avenues for future work.
We cannot construct the art cinema in isolation from other cinematic practices. The
art cinema has neighbors on each side, adjacent modes which define it. One such mode
is the classical narrative cinema (historically, the dominant mode). There also exists a
modernist cinema—that set of formal properties and viewing protocols that presents,
above all, the radical split of narrative structure from cinematic style, so that the film
constantly strains between the coherence of the fiction and the perceptual disjunctions
of cinematic representation. It is worth mentioning that the modernist cinema is not
ambiguous in the sense that the art cinema is; perceptual play, not thematic ambiva-
lence, is the chief viewing strategy. The modernist cinema seems to me manifested
(under various circumstances) in films like October (1928), La Passion de Jeanne d'Arc
(1928), Lancelot du Lac (1974), Play Time (1967), and An Autumn Afternoon (1963).
The art cinema can then be located in relation to such adjacent modes.
We must examine the complex historical relation of the art cinema to the classical
narrative cinema. The art film requires the classical background set because devia-
tions from the norm must be registered as such to be placed as realism or authorial
expression. Thus the art film acknowledges the classical cinema in many ways,
ranging from Antonioni's use of the detective story to explicit citations in New
Wave films. Conversely, the art cinema has had an impact on the classical cinema.
Just as the Hollywood silent cinema borrowed avant-garde devices but assimilated
them to narrative ends, so recent American filmmaking has appropriated art-film
devices. Yet such devices are bent to causally motivated functions—the jump cut for
violence or comedy, the sound bridge for continuity or shock effect, the elimination
of the dissolve, and the freeze frame for finality. (Compare the narrative irresolu-
tion of the freeze frame in Les 400 coups with its powerful closure in Butch Cassidy
and the Sundance Kid, 1969.) More interestingly, we have seen an art cinema emerge
in Hollywood. The open endings of 2001 (1968) and Five Easy Pieces (1970) and the
psychological ambiguity of The Conversation (1974), Klute (1971), and Three Women
(1977) testify to the assimilation of the conventions of the art film. (Simplifying
brusquely, we might consider The Godfather I [1972] as a classical narrative film and
The Godfather II [1974] as more of an art film.) Yet if Hollywood is adopting traits of
the art cinema, that process must be seen as not simple copying but complex trans-
formation. In particular, American film genres intervene to warp art-cinema conven-
tions in new directions (as the work of Altman and Coppola shows).9
It is also possible to see that certain classical filmmakers have had something
of the art cinema about them. Sirk, Ford, and Lang all come to mind here, but the
preeminent instance is Alfred Hitchcock. Hitchcock has created a textual persona
that is in every way equal to that of the art-cinema's author; of all classical films,
I would argue, Hitchcock's foreground the narrational process most strikingly. A
film like Psycho demonstrates how the classical text, with its psychological causality,
its protagonist/antagonist struggle, its detective story, and its continuous time and
158 Poetics of Cinema
homogenous space, can under pressure exhibit the very negation of the classical
system: psychology as inadequate explanation (the psychiatrist's account); character
as only a position, an empty space (the protagonist is successively three characters,
the antagonist is initially two, then two-as-one); and crucially stressed shifts in
point-of-view which raise the art-film problem of narrational attitude. It may be that
the attraction of Hitchcock's cinema for both mass audience and English literature
professor lies in its successful merger of classical narrative and art-film narration.
Seen from the other side, the art cinema represents the domestication of modern-
ist filmmaking. The art cinema softened modernism's attack on narrative causality by
creating mediating structures—"reality," character subjectivity, authorial vision—that
allowed a fresh coherence of meaning. Works of Rossellini, Eisenstein, Renoir, Dreyer,
and Ozu have proven assimilable to art cinema in its turn, an important point of depar-
ture. By the 1960s, the art cinema enabled certain filmmakers to define new possibilities.
In Gertrud (1964), Dreyer created a perceptual surface so attenuated that all ambigu-
ity drains away, leaving a narrative vacuum.10 In LAnnee derniere a Marienbad (1961),
Resnais dissolved causality altogether and used the very conventions of art cinema
to shatter the premise of character subjectivity. In Nicht versdhnt (1965), Straub and
Huillet tool the flashback structure and temps morts of the art cinema and orchestrated
empty intervals into a system irreducible to character psychology or authorial commen-
tary. Nagisha Oshima turned the fantasy-structures and the narrational marks of the
New Wave to political-analytical ends in The Ceremony (1971) and Death by Hanging
(1968). Most apparently, Godard, one of the figureheads of the 1960s art cinema, had by
1968 begun to question it. (Deux ou trois choses queje sais d'elle [1967] can be seen as a
critique of Deserto rosso, or even of Unefemme mariee.) Godard also reintroduced the
issue of montage, a process which enabled Tout va bien (1972) and subsequent works to
use Brechtian principles to analyze art-film assumptions about the unity of ideology.
If, as some claim, a historical-materialist order of cinema is now appearing, the art
cinema must be seen as its necessary background, and its adversary.
Afterword
The preceding is the oldest essay in the volume, published in 1979 and reprinted here
without revision. Like many early statements in a research tradition, it has a peremp-
tory tenor: This is this, that is that, no fine gradations allowed. To revise it would go
beyond mere updating; I'd want to query its overconfident generalizations. Some of
my claims (like the faith in an emerging "historical-materialist" cinema) and terms
(like "the narrator") no longer convince me. Many of the generalizations still seem to
me on the right track, but they would need much more nuancing and refinement, and
the result would be very different, and much longer.
Actually, some of the refinements have snuck into other things I've written. Never
expecting to reprint the piece, I cannibalized it twice. I used it to counterpoint a study
of classical Hollywood narrative (The Classical Hollywood Cinema, pp. 370-377), and
I expanded it in a discussion of modes of narration (Narration in the Fiction Film,
pp. 205-233). These are more informative, but several readers have told me that they
The Art Cinema as a Mode of Film Practice 159
prefer the cleaner outlines of the original, and it has found its way into anthologies
and course packets, so I bring it back one last time. As you might expect, though,
I can't refrain from making a few new remarks, if only to flag some points that could
be usefully rethought. Kristin Thompson and I have tried to offer a more systematic
and comprehensive discussion of some of these issues in our survey text, Film History:
An Introduction.11
Since I wrote the piece, some scholars have examined the art cinema as an institu-
tion in world film commerce. A great many fiscal mechanisms support production,
distribution, and exhibition on the European scene.12 The varied mix of funding
sources (private capital, national subsidy, and European Union programs) has
brought forth resourceful media players such as Marin Karmitz of Paris, who started
by owning theater screens and has become both a producer and distributor of major
films from Europe, Asia, and the Middle East. Somewhat surprisingly, American
investment and distribution have also helped sustain art cinema, from small compa-
nies supporting the 1950s efforts to current interest on the part of Sony and others in
financing Asian projects. 13
In any producing country, films assume many diverse shapes. There are always
genre pictures, particularly melodramas and comedies showcasing popular local
talent. (The farce featuring TV performers seems a cross-cultural constant.) Local
output also usually includes a few prestige items, often adapted from national literary
classics or based on memorable historical episodes. But Europe also promoted a
conception of creativity that was rare elsewhere: the auteur film. The idea of a direc-
tor expressing his (only rarely her) vision of life on film remains crucial to the art
cinema. The head of New Danish Screen, a funding scheme from the nation's film
institute, says, "We secure a place to develop a director's personal style without the
pressure of commercial success criteria."14 Yet personal style can have cultural and
financial implications. The idea of authorship can accommodate policies that demand
that local films reflect national culture (who was more French than Francois Truffaut,
more Bavarian than Rainer Werner Fassbinder?), while also providing a marketable
identity to films made with low budgets and relatively unknown stars. A sector of
world film commerce still depends on the auteur premise. Acknowledging a powerful
creator as the source of the film's formal and thematic complexity yields something
marketable internationally, a brand name that can carry over from project to project.
Pedro Almodovar, Lars von Trier, Michael Haneke, Roman Polanski, and a few others
are still guarantees of saleable cinema. Individualized branding is even more impor-
tant as creators become international directors, as Lone Sherfig moves from Denmark
to Scotland to make Wilbur Wants to Kill Himself (2002) or the German Tom Tykwer
allies with Miramax to m ake Heaven (2002). And, of course, the concept of authorship
spread outside Europe rather quickly, with Kurosawa Akira and Satyajit Ray becom-
ing celebrated as individual creators in the 1950s and 1960s.
Two institutions that I didn't mention have become ever more important in the
cultivation of art movies. The first, and less studied, is the film school. The USSR
founded a national film school in 1919, and European countries followed after
World War II. Film schools have multiplied since the 1960s, either in universities or
160 Poetics of Cinema
under the auspices of national film institutes. Apart from ensuring a flow of trained
professionals into the industry, film schools carry in their curricula and course assign-
ments certain presumptions about what constitutes aesthetically worthwhile cinema.
Judging from my limited experience, European film academies were in the 1990s still
quite oriented to the idea of individual expression—though my sense is that students
who were interested in TV production, where most of the jobs were, were less com-
mitted to auteur premises. It would be a big project, but someone should study the
policies, the taste structures, and the craft practices of non-U.S. film schools, and
analyze the films that result.
A second sort of institution is receiving more study just now. The filmmakers
and movements that defined the postwar art cinema earned much of their fame on
the festival circuit, from Rashomon (winner at Venice in 1951) through If (winner at
Cannes in 1969). When my essay was published in 1979, there were at most 75 principal
film festivals; today there are about 250, with hundreds more serving local, regional,
and specialist audiences. The development of low-budget independent cinemas, the
ease with which films can be submitted on video, and the huge variety of festival
themes (e.g., animation, science fiction, gay and lesbian, and film scores) have made
the scene overwhelming. There is even a trade magazine for festival planners. 15 Each
year hundreds of programmers are chasing the world's top three or four dozen films.
Everyone wants red carpet events, with major stars and directors turning up for the
press. If a festival isn't allowed by the international association to award prizes, the
organizers can still fly in three or four critics from the Federation Internationale de
la Presse Cinematographique (FIPRESCI) and establish a jury for a FIPRESCI prize.
Festivals enhance tourism and give even the smallest city a moment in the limelight.
As packaging events, they build an accumulating excitement around films that many
attendees wouldn't bother to see in regular theatrical runs.
At the same time, festivals are the world's alternative to Hollywood's theatrical
distribution system. A decentralized, informal network of programmers, gatekeepers,
and tastemakers brings to notice films of daring and ambition. 16 Festivals are the
major clearinghouse for art cinema, with prizes validating the year's top achievements.
To win at one of the big three—Berlin, Cannes, and Venice—or to be purchased at
Cannes, Toronto, or Sundance lifts a film above the thousands of other titles demand-
ing attention. The payoff goes beyond cinephilia: Taiwan and Iran have used victories
on the festival circuit to improve their cultural image.17 Hong Kong cinema would
not have gained its prestige in the West without the energetic proselytizing of festival
programmers and loyal journalists. 18
Not all movies screened at festivals are art films, but festivals sustain the formal
and stylistic conventions that my essay tried to isolate. Those conventions emerged
earliest, I still believe, in Western and Eastern European cinema, but the essay did
slight other cinematic traditions. For example, filmmakers in developing countries
like Turkey and Egypt were sensitive to developing art cinema trends, but I simply
didn't know enough about them. Nor did I know enough about South American film
to do justice to it. Italian neorealism had a strong influence there in the 1950s, and
a few filmmakers, notably Leopoldo Torre Nilsson in Argentina, quickly picked up
The Art Cinema as a Mode of Film Practice 161
on Bergman and Antonioni. Brazil's Cinema Novo and other trends criticized art
cinema traditions in ways roughly comparable to the politicized modernist cinema
of Europe.
Asia may have lagged somewhat, with the exception of Japan. Although lacking
exact counterparts to the standard-bearers of European art movies, Japan had an
experimental tradition in mainstream production, and there were many more
convention-busting directors at work than the essay suggests (such as Suzuki Seijun
and Wakamatsu Kojiro). As the 1980s unfolded, however, the other cinemas of Asia
were drawing heavily on the models I review here. Directors of the Fifth Generation
in China, the Hong Kong New Wave, and above all the New Taiwanese Cinema were
salient examples. Chen Kaige's neorealistic Yellow Earth (1984) and his more stylized
efforts like The Big Parade (1986) and King of the Children (1987); Edward Yang's That
Day, on the Beach (1983) and The Terrorizers (1986); Ann Hui's The Secret (1979); and
Patrick Tarn's Love Massacre (1981) and Nomad (1982)—these and many other works
attest to the emergence of a transnational Chinese art cinema. Wong Kar-wai's Days of
Being Wild (1991) brought Hong Kong art cinema to maturity, and his time-bending
lyricism, from Ashes of Time (1994) to 2046 (2004), has been indebted to Western
literary and cinematic models.19 In Taiwan, the earliest New Cinema films belong to
an autobiographical redrafting of neorealism, but several directors moved beyond
it. Edward Yang was strongly influenced by European cineastes, notably Antonioni,
and his masterpiece A Brighter Summer Day (1991) married local realism (the film is
based on a notorious murder) and self-conscious artifice. Hou Hsiao-hsien was no
cinephile, working instead in Taiwan's local industry, but after making triumphant
contributions to New Cinema realism, he widened his ambitions. He experimented
with decentered historical narrative (City of Sadness, 1989; The Puppetmaster, 1993),
reflexive construction (Good Men, Good Women, 1995; Three Times, 2005), extreme
technical restraint (Flowers of Shanghai, 1998), and self-conscious invocations of film
history (the Ozu homage Cafe Lumiere, 2004).20
The 1990s also saw the emergence of a new generation of Japanese directors, includ-
ing Kore-eda Hirokazu (Maborosi, 1995), Suwa Nobuhiro (M/Other, 1999), Aoyama
Shinji (Eureka, 2000), and Kitano Takeshi, who shifted between poetic genre films and
more abstract efforts like Dolls (2002). At the same period, South Korean directors
Hong Sang-soo (The Power ofKangwon Province, 1998), Lee Chang-dong (Peppermint
Candy, 2000), Kim Ki-duk (The Isle, 2000), and Park Chan-wook (Sympathy for
Mr. Vengeance, 2002) began winning festival acclaim. Mainland China has reinsti-
tuted art cinema as an export commodity, with films such as Tian Zhuangshuang's
Springtime in a Small Town (2002) and Jia Zhang-ke's The World (2004).
Many of these newer traditions, it seems, replay at an accelerated pace the trajectory
of European art cinema. An indigenous realist movement, somewhat comparable to
Italian neorealism, becomes more conscious of the conventions involved in realism,
and develops more abstract experiments in form. The emergence of Iranian cinema
is a remarkable instance. Budgets are bare-bones by Western standards, and by using
nonactors and locations, filmmakers have presented post-Shah Iranian culture to a
world that knew little of it. The humanistic strain of neorealism finds echoes in films
 Poetics of Cinema
Figure 5.1 Moment of Innocence (1996):
Reenacting Makhmalbaf's youthful assault,
several performers meet, along with their
props, in the final freeze frame.
like The Key (1987), The White Balloon (1995), The Apple (1998), The Child and the
Soldier (2000), and Blackboards (2000). At the same time, and often within the same
films, we find sophisticated games with cinematic technique. The Mirror (1997) starts
with a little girl's frustration with trying to cross a busy intersection, then shifts its
story action almost wholly to the soundtrack when she barricades herself behind her
household gate and refuses to meet the camera. Mohsen Makhmalbaf s Moment of
Innocence (1996) shows him staging a film based on a crime he committed in his
youth, and the result is a dizzying mise en abyme reminiscent of 8 1/2 (Figure 5.1).
The country seems immersed in cinephilia. When a laborer and film fan pretends
to be director Makhmalbaf, Abbas Kiarostami covers his trial and stages a meet-
ing between him and Makhmalbaf. He calls the result Close Up (1990). But then the
impostor justifies himself by making his own film, called Close-Up Long-Shot (1996).
Kiarostami himself—superb screenwriter, director of exemplary documentaries
and fiction films, and experimenter with portable video and Warholian recording
(Ten, 2002; Five Dedicated to Ozu, 2003)—stands as an emblem of a culture in love
with cinematic artifice but also compelled to bear witness to the lives of ordinary
people. Who in the West would have predicted that a great cinema, at once humanist
and formalist, would have come from Iran?
Not that the period proved unproductive elsewhere. Russia and Eastern
Europe contributed to the tradition of philosophically weighty works with Andrei
Tarkovsky's The Mirror (1975) and Nostalghia (1983) and Krzysztof Kieslowski s
coproductions, notably the Three Colors trilogy (1993-1994). Aleksandr Sokurov
created mournful, quasi-mystical works (The Second Circle, 1990; Whispering Pages,
1993) that paralleled the elegiac music pouring out of late Soviet and post-Soviet
composers like Artymov and Kancheli. In Hungary, Bela Tarr (Satanstango, 1994)
and Gyorgy Feher (Passion, 1998) created harsh, palpably grimy tales of rural life.
France continued to support Philippe Garrel, Claire Denis, and others of ambitious
bent, whereas Belgium sustained the regional realism of the Dardennes brothers,
Jean-Pierre and Luc. Denmark provided Europe's newest Cinema of Quality, with
well-carpentered scripts, thoughtful themes, and versatile actors, as well as, thanks
to the Dogme 95 impulse, some films pushing against the ethos of professionalism
The Art Cinema as a Mode of Film Practice 163
with rawer works. A film, said Lars von Trier, should be "like a pebble in your shoe."21
Yet quite outside the dominion of Dogme lay Christoffer Boe, whose Reconstruction
(2003) owes a good deal to Alain Resnais' polished time jumping.
American filmmakers have been assimilating art-film conventions for a long time,
as my essay suggests, but the process has been given a new force by the rise of the
independent film sector. Steven Soderbergh can remake an Andrey Tarkovsky film
(Solaris, 1972 and 2002), Paul Thomas Anderson can borrow sound devices from
Jacques Tati (Punch-Drunk Love, 2002), and Hal Hartley can absorb ideas from Jean-Luc
Godard and Robert Bresson.22 The burst of experimentation on display in films like
Memento (2000), Adaptation (2002), and Primer (2004) probably owes as much to
the European heritage as it does to U.S. traditions of film noir and fantasy. In many
respects, the U.S. indie cinema blends European art-cinema principles with premises
of classical Hollywood storytelling.23 Ahmad, the protagonist of Ramin Bahrani's Man
Push Cart (2005), has as firm a set of purposes as any Hollywood hero, but the first half-
hour of the film conceals them from us. Instead, the scenes concentrate on his daily
grind as he sells coffee and pastries from a wheeled stall. We get to know him by the
way he lugs his propane tank, fills the coffee roaster, unpacks doughnuts and Danishes,
and hauls his massive cart through midtown traffic. Suspending our awareness of the
protagonist's goals forces us to focus on minutiae of the story world.
Several books would be needed to do justice to this worldwide activity,24 so I'll
close by pointing out two areas that have intrigued me from the standpoint of a
poetics. First is a new stylistic trend that coalesced as I was writing my essay. As if
in rebuke to the 1960s reliance on montage and camera movement, several directors
cultivated an approach based on the static, fairly distant long take. In Europe, this
took the shape of what I've called the planimetric image. The shot is framed perpen-
dicular to a back wall or ground, with figures caught in frontal or profile positions,
as in police mug shots. We can find this emerging in the 1960s, with the new reliance
on long lenses, but it became a feature of much European staging of the 1970s and
1980s, and it was picked up in other national cinemas (Figures 5.2-5.3).25 This device
presents the scene as a more abstract configuration, perhaps distancing us from its
emotional tenor, and it can support those psychologically imbued temps morts that
are crucial to the realistic impulse of the mode. This visual schema can also display
some of the arresting boldness of an advertising layout, as in the cinema du look trend
of 1980s France (Figure 5.4). The planimetric image became quite common in world
cinema (Figure 5.5) and constitutes one of the art cinema's permanent contributions
to cinema's pictorial repertoire. As a substitute for orthodox shot/reverse-shot cutting,
it became a staple of deadpan humor in both art films like Kitano Takeshi's and cult
hits like Napoleon Dynamite (2004).
Even when not composed with planimetric flatness, the very long take also became
a prominent technical option. In Europe, Miklos Jancso was identified with a dynamic
and fluid choreography of camera and figures in shots running for many minutes. By
contrast, the "minimalist" films of Daniele Huillet and Jean-Marie Straub, such as
The Chronicle of Anna Magdalena Bach (1968), employed quite static shots with little
or no camera movement (Figure 5.6). Allied with the planimetric composition, the
 Poetics of Cinema The Art Cinema as a Mode of Film Practice 165
164

Figure 5.7 A Time to Live and a Time to Die Figure 5.8 A Time to Live and a Time to Die:
Figure 5.3 The flattened image suggests a (1985): With the family gathered, the griev- She is drawn up and back by her relations, her
family portrait gone wrong in Terence Davis' ing mother buries her face in the dead father's movement and the panning camera revealing
Distant Voices, Still Lives (1988). blankets. one of the sons against the back wall.
Figure 5.2 Manuel deOliveir a uses the plani-
metric image as a theatrical address to the
audience in The Cannibals (1988).

Figure 5.4 Leos Carax' Mauvais Sang (1986)
uses the planar image for surprise dramatic
effects as Lise, pursuing the hero, flings her-
self at a subway door.
Figure 5.9 A Time to Live and a Time to Die: Figure 5.10 In Peacock (GuChangwei, 2004,
The mother is pulled back into the adjacent China), the telephoto full shot recalls not only
room, and while she's visible in the distance, the planimetric image but also the staggered
Figure 5.5 In Maborosi (1995), Hirokazu
her sons kneel in the foreground. Hou Hsiao- staging arrangements of Hou Hsiao-hsien.
Kore-eda mutes the drama of a drunken hus-
hsien's long lens creates layers of space that
band returning home. The distant, perpen-
are dynamically blocked and revealed by
dicular framing accentuates the slight gesture
small movements of the characters.
of the wife quietly lowering her head.

Figure 5.6 Bach in a distant, static shot (The
Chronicle of Anna Magdalena Bach, 1968).
static long take became somewhat common in the 1970s. In the 1980s, apparently
unaware of the European developments, Hou Hsiao-hsien made the fixed-camera
long take central to his style, but his use of the long lens supports more dynamic
staging principles than we see in most European instances. 26 Hou's complex blocking
is often employed to highlight small actions taking place quite far from the viewer
(Figures 5.7-5.9). His fixed long takes appear to have influenced many other directors
in Asia, though few have matched his virtuosity in recessive staging and dense image
design (Figure 5.10; see also Figure 5.5). Today's festival films are likely to include
at least some striking long takes with minimal camera movement, and entire films
can be built out of them. The international currency of the device is evident with the
French film Un couple parfait (2006), directed by the Japanese long-take adept Suwa
Nobuhiro. Moreover, a director may decouple the fixed camera from the long take,
creating unusually strict constraints on camera placement. We now have movies con-
sisting wholly of conversations in automobiles, with the camera anchored outside the
windshield (Kiarostami's Ten) or angled from the dashboard (Figure 5.11).
A second aspect touched on in this essay became a concern of my book Narration
in the Fiction Film. The art cinema engages us not only by asking us to construct the
fabula action but also by teasing us to make sense of the ongoing narration. So how
is this slippery narration patterned across the length of the whole film? Taking as my
example Resnais' La Guerre estfinie (1966), I argued that many art films create a "game
of form." The film initially trains the viewer in its distinctive storytelling tactics, but
as the film proceeds, those tactics mutate in unforeseeable ways. In La Guerre est
finie, the key device—the hypothetical sequence, showing several alternative actions
the protagonist might take in the future—is announced quite early. At first it seems
difficult and disruptive, but through repetition it becomes stabilized. Then, however,
 Poetics of Cinema 1 He Art Cinema as a Mode of Film Practice
166 167
1
1
11
j

Figure 5.11 Although it contains nearly

1,000 shots, Simon Staho's Day and Night
(2004, Denmark) employs only two camera
Figure 5.12 ; Vagabond (aka Sans toit ni loi; Figure 5.13 Vagabond: The garage owner Figure 5.14 Vagabond: The
positions, one showing the car's driver, and
1985): The most peculiar of the initial "reverse begins his tale. ends his tale, annoyed.
the other showing the passenger seat. Any
shots" of the police investigation.
action taking place outside the car is viewed
from those fixed angles.

the narration renders the hypothetical sequences more indeterminate, introducing

uncertainty by mixing in flashbacks and abrupt transitions to new scenes. The
final section of the film is the most transparent, as the story action comes to the
fore, but there are still variations that make the premises of presentation somewhat
unpredictable. The finale leaves open both the consequences of story causality and the
rules governing the narration itself.27 Figure 5.15 Vagabond: Yolande halts her Figure 5.16 Vagabond: Yolande recalling see-
This game of form isn't present in every film that belongs to the mode, but it does housework. ing Mona and David.

reappear. 28 A striking instance is Agnes Varda's Vagabond (aka Sans toit ni loi; 1985).
The film has a characteristic art-cinema situation: an unexplained psychological crisis
that makes a young woman quit her job as secretary and take to an itinerant life in
Montpellier. Her meanderings are presented more or less in chronological order, as
she links up various people in the neighboring villages and sinks into drunkenness
and exhaustion. The film begins with the discovery of her frozen corpse and assumes
the time-honored structure of a posthumous inquiry. People who encountered Mona
introduce flashbacks showing the last weeks of her life. But to the inevitable question
What led up to her death? the film adds, To whom are the witnesses telling their stories?
At the death site, the police question the witnesses. The scene is conventional
and coherent, except for a curious framing that might give wary viewers some pause
(Figure 5.12). And once Mona's body is removed, a female voice, heard over shots of the
beach and sea, interjects itself. "People she had met recently had remembered h e r . . . .
They spoke of her, not knowing she had died. I didn't tell them." This quasi-documen-
tary commentary might lead us to expect to see who's speaking, but we never do. If it
is Varda's voice, as I think it is, we have the oscillation among objectivity, subjectivity
(how people recall Mona), and authorial intervention characteristic of art-cinema
narration. In announcing that "these witnesses helped me tell the last weeks of her
last winter," the commentary sets up some rather ambivalent ground rules. The people
who recall Mona might speak to one another, as in ordinary cinematic conversation,
and these would segue into flashbacks. This is the Citizen Kane option. Alternatively,
the witnesses might be testifying explicitly for the camera, responding to an offscreen
questioner in documentary fashion. This strategy is pursued in Hou's biographical
picture The Puppetmaster. Vagabond, it turns out, uses both devices, and it introduces
still other variants. The witnesses "help" the filmmaker "tell" Mona's story in several
disquieting ways.
Vagabond's first flashback is introduced by idling boys recalling seeing Mona on
the beach; we simply overhear their reference to her. Later, at a truck stop a driver tells
another about picking her up, and then a garage owner tells another man about hiring
her. But already there's a modulation. The man whom the truck driver addresses is
quite visible, but the first view of the garagiste talking of Mona comes in a reverse-
angle composition, so we see only the shoulders of his listener (Figure 5.13). When we
come out of the flashback, he is still talking, but to an unseen listener (Figure 5.14). In
an ordinary film, such variations might simply serve to underscore the garage owner's
testimony, but here they also glide into a pattern of narrational uncertainty. Shortly
afterward, the maid Yolande turns from her dusting toward us to describe how to
break into the chateau that her uncle guards (Figure 5.15). This monologue doesn't
tell us about Mona, and it is more properly addressed to Yolande's layabout boyfriend,
Paulo, than to anyone else, inside or outside the fictional world. In retrospect, we can
justify the moment as Yolande talking to Paulo, who may be offscreen (he slips in and
out of the household), but this is by no means definite. The fact that we never see to
whom she's speaking dramatically violates the internal norm set up so far. Very soon
after this, when Yolande meets Paulo at the chateau, she again addresses the camera,
and now she confides that she was moved by seeing Mona and another vagrant, David,
curled up sleeping together (Figure 5.16). This time, there can be no doubt that this is
a soliloquy. The objective portrayal of quite unfeeling male witnesses has shifted into
 Poetics of Cinema
168
Figure 5.17 Vagabond: David in the boxcar,
his gaze shifting from side to side.
Figure 5.18 Vagabond: The shepherd and
his wife: Is he addressing her or us?
Figure 5.19 Vagabond: Near the end ofher life, Figure 5.20 Vagabond: Assoun after kissing
Mona's gaze brushes across the camera axis. Mona's scarf; the witness is silent.
a subjective monologue, with a woman telling us spontaneously of her yearning for
love and companionship.
As if taking his cue from Yolande, David the hippie is soon given a recounting
scene too. He squats in a boxcar and tells how he enjoyed being with Mona when
she had grass (Figure 5.17). In story-world terms, his testimony undercuts Yolande's
romanticism; he and Mona's relationship was based on staying stoned. Narrationally,
all bets now seem off. Is he soliloquizing? Unlike Yolande, he doesn't look straight at
us; his eyes move sharply left and right. So is he speaking to other people? No other
characters are visible, and the train moves away while he's still speaking.
Across all these scenes, the narration has laid out contradictory cues, and these
block any consistent construal of the characters' circumstances in these recounting
scenes. As the film goes on, the inserted recollections continue to oscillate among
these possibilities, mixing in other variants. Sometimes the repetitions affirm earlier
inferences we might have made, as when Yolande soliloquizes to the camera one last
time before leaving the story action; she is the only character given these private
monologues. At other times, the characters recount Mona's behavior to the camera, or
past it, or to someone in the scene, or perhaps some combination of all (Figure 5.18).
Perhaps most disconcertingly, at certain moments Mona's eyes graze the camera,
sometimes pausing as if she were looking at us (Figure 5.19). The narration's deter-
mination to throw us off balance runs to the very end, when Assoun, the Tunisian
immigrant worker who was closest to Mona, simply kisses her scarf and looks out
at us silently (Figure 5.20). In this purely emblematic moment, he seems not to be
responding to anyone on the scene, either character or filmmaker.
The Art Cinema as a K4ode of Film Practice 169
I
Figure 5.21 Cache (Michael Haneke, 2005):
The central family's apartment house, seen
on video. But who is watching the tape?
Figure 5.22 Cache: A nearly subliminal shot
follows. Where does this image belong in the
story chronology?
At the close, the tale of Mona's wanderings retains its episodic quality, and her
motives remain enigmatic. We construct this art cinemafabula in and through a series
of incompatible judgments about her made by diverse characters. So far, so conven-
tional. But by equivocating about the situation in which the characters recount their
stories, Yarda also invites us to enter the typical game of formal hesitation. Perhaps
the ambivalences of the onlookers' glimpses of Mona can finally be attributed to the
supreme control of this narration over the primary conditions of the story. The voice-
over commentary that quite early announced the film's method also includes a telling
passage. "I know little about her myself," says the woman's voice, "but it seems to me
that she came from the sea." At this moment we see a shot of Mona striding out of the
ocean, watched by the biker boys. The backstory begins with an oscillation between
objective presentation and a frank gesture of authorial imagination. Perhaps every-
thing that follows, even the firm texture created by regional customs and dialects and
nonactors, is to be understood as subordinate to the creative energies of authorial
vision (not an implied author, or a "cinematic narrator," but the person who made
this movie for us).
This poignant film is crisscrossed by many other patterns, notably the plan i metrically
composed tracking shots that coax us to find another theme-and-variations structure. 29
A complete analysis would also show how the game of form in the various narrators'
accounts blocks our constructing a complete psychological profile of Mona. Instead, it
swerves us toward comparing the productive or unproductive lives of the characters she
touches. One of the film's thematic dilemmas is how utter freedom can be reconciled
with the danger of solitude, and this tension is played out in these onlookers who admire,
criticize, or worry about Mona, homeless and lawless. Still, our consideration of themes,
as abstractions from our experience of the film, should be balanced by considering how
we make sense of the narration moment by moment, registering its mixed signals and
following its-zigzag paths. A crucial part of Vagabond's respectful uneasiness about
Mona's choices is born of an ambiguous narration.
Such formal play constitutes one norm within art cinema narration. It's as appar-
ent in the disjunctive editing and misleading camera positions (Figures 5.21-5.22) of
Cache (2005) as in works from 40 years before. As the next two essays try to show,
the study of this tradition from the standpoint of poetics continues to bring new pos-
sibilities to light.
 6.
Film Futures

In Jorge Luis Borges' short story "The Garden of Forking Paths," a character discovers
that the sage Tsui Pen has devised a labyrinthine novel:

In all fictions, each time a man meets diverse alternatives, he chooses one and
eliminates the others; in the work of the virtually-impossible-to-disentangle
Tsui Pen, the character chooses—simultaneously—all of them. He creates,
thereby, several futures, several times, which themselves proliferate and f o r k . . . .
In Tsui Pen's novel, all the outcomes in fact occur: each is the starting point for
further bifurcations. Once in a while, the paths of that labyrinth converge: for
example, you come to this house, but in one of the possible pasts you are my
enemy, in another my friend. 1

Tsui Pen didn't shrink from the ultimate consequences of this:

He believed in an infinite series of times, a growing, dizzying web of divergent,

convergent, and parallel times. That fabric of times that approach one another,
fork, are snipped off, or are simply unknown for centuries, contains all possi-
bilities. In most of those times, we do not exist; in some, you exist but I do not;
in others, I do and you do not; in others still, we both do.2

Borges' conceit has its counterpart in quantum physics, which has played host to the
idea of parallel universes—an infinite array of possible worlds, each as real as the one
we apparently know.3
To this conception of time Gary Saul Morson objects, in his exacting and stimu-
lating study Narrative and Freedom. Morson is concerned to show that most novels
create a pattern of closed time and foregone choices, which in turn make a broader

171
 Poetics of Cinema
172
moral or ideological position seem inevitable. Conventional techniques like fore-
shadowing and "backshadowing," he suggests, build a grim determinism into a plot's
very architecture, denying temporal openness and the moral freedom that implies.
Morson thus finds Borges' parable quite disturbing. If all possibilities exist equally,
when "nothing that could have taken place fails to take place," then ethical action
is rendered impossible. "Because all choices are made somewhere, the totality of
good and evil in existence becomes a zero-sum g a m e . . . . What difference does it
make what I do, if I also do the opposite?"4 For Morson, the conception of alterna-
tive universes cannot ground a responsible conception of human action, let alone an
adequate scheme of narrative time.
Yet Morson need not worry, I think. Although he finds many examples in Dostoevsky
and Tolstoy of his preferred method of conjuring up alternative courses of action (he
calls it "sideshadowing"), narratives derived from the forking-path conception don't
really approach Borges' "growing, dizzying web." In fiction, alternative futures seem
pretty limited affairs. Folklore bequeaths us the two-doors problem (the lady or the
tiger?) and the motif of the three paths leading to three fates. A Christmas Carol offers
Scrooge merely a binary choice about his future, and in O. Henry's 1903 short story
"Roads of Destiny," the poet-hero faces only three futures: to take the road on the left,
to take the one on the right, or to return to town.
Recent cinema is becoming more experimental on several fronts, particularly in
relation to complicated uses of time and point of view, and so we shouldn't be sur-
prised to find forking-path plots turning up more often on our screens. Like "Roads of
Destiny," they tend to proceed from a fixed point—the fork—and purportedly present
mutually exclusive lines of action, leading to different futures. Consider Krzysztof
Kieslowski's Blind Chance (1987), which after a rather enigmatic prologue shows the
medical student Witek racing for the train that will take him on his sabbatical from
medical school. He leaps aboard just in time and is carried to a life as a Communist
functionary. But when he reaches a crisis in that life, the film cuts back to the railroad
station, and he is a young man again, once more racing for the train. Now he fails to
catch it, stays at home, and is given a brand-new future. That future will be altered
once more when the narration flashes back to his run along the platform and a new
chain of events starts. A similar pattern is enacted in the Hong Kong film Too Many
Ways to Be No. 1 (1997) and Tom Tykwer's Run Lola Run (1998). These films present
their futures seriatim, returning to the switchpoint after each trajectory is finished.
By contrast, Peter Howitt's Sliding Doors (1998) presents its alternative plotlines in
alternation, continually intercutting one future with the other.
None of these films hints at the radical possibilities opened up by Borges or the
physicists. Blind Chance and Run Lola Run present only three alternative worlds,
whereas Sliding Doors and Too Many Ways to Be No. 1 offer the minimum of two. Just
as important, all these plots hold the basic characters, situations, and locales quite
constant across stories. In both trajectories of Sliding Doors, Helen must cope with
losing her job and coming to terms with her partner, Jerry (who is having an affair
with Lydia). Too Many Ways to Be No. 1 centers on Wong, a petty triad who's offered
a chance to work with a Mainland gang trying to smuggle automobiles into China.
Film Futures 173
The plotlines pivot around his decision: Grab the cars and flee? Or walk away from the
deal? Run Lola Run concentrates on a crisis: Lola's boyfriend, Manni, has lost money
belonging to the gang boss, Ronnie, and she must come up with 100,000 marks before
noon, when Manni intends to rob a supermarket to make up the deficit. The plot
traces various consequences of her efforts to get the money to save Manni.
In Blind Chance, the situation facing the protagonist offers somewhat more diffuse
possibilities, but the action eventually revolves around how Witek will live after the
death of his father. If he catches the train, he winds up becoming a functionary in
the Communist government. If he misses the train, he either becomes an activist in
an underground Catholic youth movement, or stays behind and returns to medical
school, marrying a woman he met there. In Blind Chance the outcomes boil down to
thematically grounded alternatives: In Poland of the late 1970s, every choice turns out
to be political, even the apparently nonpolitical choice of being a doctor.
So instead of the infinite, radically diverse set of alternatives evoked by the parallel
universes conception, we have a set narrow both in number and in core conditions.
None of these plots confronts the ultimate and more disturbing alternative world
demands: Lola is never shown as Manni's sister in a rival universe, Matt does not
become Wong's enemy, Helen does not turn into her rival Lydia, and in no version
does the protagonist fail to exist at all. We have something far simpler, rescuing the
characters from Morson's zero-sum game but—and this is my major point—corre-
sponding to a more cognitively manageable conception of what forking paths would
be like in our own lives.
Far from representing a failure of nerve on the part of filmmakers, I think that
the strategy of narrow alternatives offers clues to the way forking-path narratives
actually work and work upon us. Narratives are built upon not philosophy or physics
but folk psychology, the ordinary processes we use to make sense of the world. Often,
particularly in media like film, perceptual skills that we've developed to give us
reliable information about the world are deployed no less commandingly in following
stories. These skills sometimes fail the most stringent deductive tests, as experiments
in everyday rationality suggest.5 Yet the shortcuts, stereotypes, faulty inferences, and
erroneous conclusions to which we are prone play a central role in narrative compre-
hension. In following a plot we reason from a single case, judge on first impressions,
and expect, against all probability, that the rescuer will arrive on time because we
want it that way. Granted, this is partly a matter of convention, built up over decades
of filmmaking; but the conventions depend partly upon the propensities of folk
psychology. Film characters are usually stereotypes, but we rely on stereotypes—not
necessarily harmful ones—to get through our lives. We expect crossing guards to like
kids and politicians to evade hard questions. These stereotypes are confirmed at rates
greater than chance in everyday life, so no wonder that they turn up in movies.
Because we bring folk psychology to bear on narratives all the time, why should
tales of parallel worlds be any different? We spin counterfactuals in everyday life. If
I had left the parking lot a minute or two later, I wouldn't have had the fender-bender
that became such a nuisance to me for the next month. This sort of homely reflection
on short-term outcomes, in which only small things change from version to version,
 Poetics of Cinema
174
seems the basis of Sliding Doors, Too Many Ways, and Run Lola Run. Occasionally, of
course, we also meditate on our life course. Here, for instance, is Brian Eno explain-
ing how he found his career:
As a result of going into a subway station and meeting Andy [Mackay], I joined
Roxy Music, and as a result of that I have a career in music I wouldn't have had
otherwise. If I'd walked ten yards further on the platform or missed that train
or been in the next carriage, I probably would have been an art teacher now.6
It's this sort of speculation that seems to be captured in Blind Chance, and even if the
cast may change more drastically than in short-term imaginings, we remain the hero
of our imagined future.
Likewise, at any moment we can easily imagine two or three alternative chains of
events, as Eno does, but not 20 or 60, let alone an infinite number. It may be relevant
that outstanding examples of forking-path tales in literature conform to similar con-
straints. A Christmas Carol and "Roads of Destiny" display the same limitations—a
very, very few options and no deep ontological differences between the futures dis-
played. Storytellers' well-entrenched strategies for manipulating time, space, causality,
point of view, and all the rest reflect what is perceptually and cognitively manageable
for their audiences, and the multiple worlds invoked by Borges and modern science
don'tfitthat condition. Add to this the canons and conventions of the film medium as
well, and these may work to limit the proliferation of forking paths. In cinema, pow-
erful storytelling traditions reshape such uncommonsensical ideas into something
far more familiar. This tendency has the additional payoff of setting to rest Morson's
worries about a nihilistic reduction of an action's ethical dimensions; by opening up
only two or three forking paths, these plots make certain choices and consequences—
about politics, crime, and love—more important than others.
My main purpose in what follows is to chart some key conventions on which four
forking-path films rely. This will let us see how the exfoliating tendrils of Borges'
potential futures have been trimmed back to cognitively manageable dimensions, by
means of strategies characteristic of certain traditions of cinematic storytelling. I hope
to show that these forking-path movies, calling forth folk psychological inferences
and designed for quick comprehension, have stretched and enriched some narrative
norms without subverting or demolishing them. Indeed, part of the pleasure of these
films stems from their reintroduction of viewer-friendly devices in the context of
what might seem to be ontologically or epistemically radical possibilities.7
Rules of t h e G a m e
Here are seven conventions of forking-path tales.
1. Forking paths are linear.
In principle, as Borges' Ts'ui Pen indicates, any instant at all could initiate a new
future. As Kieslowski remarks, "Every day we're always faced with a choice which could
end our entire life, yet of which we're completely unaware. We don't ever really know
Film Futures 175
8
where our fate lies." In Blind Chance and our other films, however, narrative pattern-
ing obligingly highlights a single crucial incident and traces out its inevitable implica-
tions. Each moment isn't pregnant with numerous futures. Instead, one event becomes
far more consequential than others, and those consequences will follow strictly from
it. Such linearity helps make these plots intelligible, yielding two or three stories that
illustrate, literally, alternative but integral courses of events—something fairly easy
to imagine in our own lives and to follow on the screen. "Of course the number of
parallel universes is really huge," remarks a physicist. "I like to say that some physicists
are comfortable with little huge numbers but not with big huge numbers." 9 As film
viewers, we like the number of parallel universes to be really little.
Moreover, in our films, each path, after it diverges, adheres to a strict line of
cause and effect. There is usually no later branching after the first fork, none of what
Borges calls "further bifurcation." After missing or catching her train, Helen in Slid-
ing Doors doesn't divide again, and although Wong in Too Many Ways and Witek in
Blind Chance must make further choices along each path, the plot doesn't split into
more proliferating consequences. The narratives assume that one moment of choice
or chance determines all that follows.
Still, forking-path plots offer some wiggle room. Although causality becomes strict
once certain processes are put into motion, they can be set in motion by felicities
of timing. One lesson of such films is that split-seconds matter. If Witek's hand had
clutched the train car's handrail at just the right moment, if Wong had decided to pay
his share of the dinner bill and walk out of the massage parlor, if Lola had not been
slowed down by this or that passerby, if Helen's path had not been blocked by a little
g i r l . . . things would be very different. Again, the films pivot around a folk psycho-
logical "if only": We are back with Eno on the tube platform, when Music for Airports
owes everything to a momentary encounter with Andy Mackay.
Sometimes one of these films does open up a new fork, but it tends to do so
retrospectively, by looping back to another moment of choice from a later point. Even
then, it will presuppose yet another linear trajectory stemming from that moment. 10
Sliding Doors concludes by showing the upshot of one story, in which Helen survives
a fall downstairs, breaks off with Jerry, and leaves the hospital just when James does.
At this very late point, the film starts to reenact a moment in the film's setup; that is, it
marks a switchpoint earlier than the one that launched the film's alternative futures.
The result is a neat closure device I'll discuss later. By contrast, Too Many Ways to Be
No. 1 creates a new choice point in order to generate a somewhat open ending. Initially
the paths fork when Wong, invited to meet with the Mainland triads, is asked to pay
the bill for a meal and entertainment at a bathhouse. In the first version, he doesn't
pay and robs the Mainland gang, leading his pals on a frantic race out of Hong Kong
and over the border. In the second version, he does pay, avoids a fight, and moves to
Taiwan. The epilogue returns us to the initial situation of the fortune-teller in the
epilogue, and again Wong's friend Bo invites him to dinner and the bathhouse, but
now Wong's reaction implies that he may not accept the offer. In effect, the epilogue
suggests that a new choice point has been opened: Instead of not paying and paying,
there's going to the meeting (with the grim outcomes we've seen) and not going. Staying
176 Poetics of Cinema
Figure 6.1 Run Lola Run (1998): Lola brushes
against a protesting woman as she races along.
in Hong Kong and avoiding Bo's scheme altogether becomes a third option for Wong,
one that fits into a broader theme suggesting that Hong Kong's future lies neither with
the Mainland nor with Taiwan.
The chief exception to my claims about causal linearity and timing in these tales
comes in the interpolated flashforward passages in Run Lola Run. These present
very quick montages of stills, prefaced by a title ("And Then . .."), which trace out
the futures of secondary characters. Most of these also adhere to a linear chain of
cause and effect, but in one instance, their relation to the plot's main causal momen-
tum is complicated. In each trajectory, running Lola bumps, or nearly bumps, the
same woman on the street (Figure 6.1), and the film provides a flash montage of
the woman's future. In each story she, like Lola, has a different future. But why
should the timing of Lola's passing create such varying futures for the nameless
woman? Bumping or not bumping hardly seems sufficient to launch radically
different outcomes for this woman's life. Tykwer's insert works well as a mockery of
the "butterfly effect," but I suspect that audiences would have difficulty understand-
ing an entire film based around divergent futures that don't spring from a web of
causally connected conditions.
2. The fork is signposted.
Tykwer's "And Then . . . " titles can stand as an emblem of the explicitness with
which forking paths must be marked. Within the story world, characters may com-
ment upon their divergent futures. During Blind Chance's second tale, Witek remarks
to the priest, "Imagine! If a month ago I hadn't missed a train, I wouldnt be here
with you now." In Sliding Doors, Helen explains that her mugging delayed her return
home: "If I had just caught that bloody train it'd never have happened"—to which
Jerry, relieved that he wasn't caught philandering, replies dismissively, "If only this
and what if t h a t . . . " These are if only and what if plots.
To reinforce such bald announcements, each film's narration sets up a pattern that
clearly indicates the branching points—a kind of highlighted "reset" button, usually
emphasizing matters of timing. Blind Chance uses a freeze frame, a reprise of the
same musical accompaniment, and the return of nearly identical footage of Witek
pelting through the station. Sliding Doors employs a rewind mechanism; Helen fails
to catch the train, but her movements are then reversed so that she strides backward
Film Futures 177
Figure 6.2 Wong's wristwatch: the opening Figure 6.3 Too Many Ways to Be No. 1:
shot of Too Many Ways to Be No. 1 (1997). Wong's cracked watch crystal after his death
in his first future.
Figure 6.4 Too Many Ways to Be No. 1:
The wristwatch now on the street as Wong
thrashes Bo, launching his second future.
up the stairs and, after another pause, comes down and does manage to hurry on
board. Run Lola Run replays the fall of Lola's bright red phone receiver and her racing
through her mother's room, down the stairs, and out into the street. In addition,
before each new future Tykwer provides a slow, red-tinted scene of Lola and Manni in
bed brooding on their love.
'The motif of timing is also made evident in the branching point of Too Many Ways
to Be No. 1. A close-up of Wong's wristwatch (Figure 6.2) opens the film and leads
directly to his session with the palm reader (played silent). Wong goes out to the street,
where his pal Bo begins to urge him to attend the meeting. At the end of the first story,
as Wong and his gang lie dead, we see his watch on the ground, its crystal shattered
(Figure 6.3). Then we cut back to the watch-—this time not at the palm reader's but
placed on the street, as Wong is revealed once more scuffling with his pal (Figure 6.4).
The epilogue will be built around a return to the watch at the palmist's, as we first saw
it, but this time with the soundtrack giving us full information about the forking-
path predictions. The close-up of the watch becomes a singularized device marking a
return to points at which the stories relaunch.
3. Forking paths intersect sooner or later.
When we think about forking paths in ordinary life, we tend not to populate our
scenarios with utterly different casts of characters. When I imagine what would have
happened if I hadn't come in to work today, I don't conjure up a new wife and a fresh
set of friends and neighbors. Along similar lines, forking-path tales tend to present
178 Poetics of Cinema Film Futures 179

small worlds in which our protagonists remeet the same people. In both trajectories
of Sliding Doors, four characters (Helen; her partner, Jerry; Jerry's lover, Lydia; and
James, Helen's potential new partner) carry the burden of the action, and secondary
characters recur as well. Run Lola Run works with the same ensemble in all three
lines: Lola, Manni, her mother, her father, her father's mistress, her father's business
associate, and the security guard at the bank.
In Blirid Chance and Too Many Ways to Be No. 1, there is less overlap of characters
across alternative futures, but these films do include some recurring figures: The dean Figure 6.5 Blind Chance (1981): Witek screams Figure 6.6 Blind Chance: The second shot:
in the first shot. Casualties in the hospital.
of Witek's medical school appears in all three stories, and his aunt recurs in two; and
in Too Many Ways, Wong's partner Matt is a constant presence. Yet both these films
find other ways to weave in characters we've already met. In Too Many Ways, the
hero's partners in crime are killed in the first story, set on the Chinese Mainland. In
the second story, centering on Wong and his partner Matt as they try to make money
as hitmen in Taiwan, the partners reappear as the men who committed a crime for
which Matt and Wong are blamed.
Blind Chance contains a prologue covering some early events in the hero's life,
and this serves to create familiarity further along. A pal from Witek's boyhood Figure 6.7 Blind Chance: The last shot,
linking to the first (Figure 6.5).
reappears in the second story, and in the third story, while Witek is standing on the
train platform, the plot reintroduces another medical student, a woman who has been
highlighted in the prologue as his lover. She has come to see him off—though she's
hinges on appointments (with the Mainland and Taiwanese gangs) and deadlines
not been shown in any of the replays of his race through the station—and in the third
(chiefly, in the second plotline, the one pushing Matt to kill rival triad bosses).
story they end up marrying. Finally, the three stories in Blind Chance are linked by
As we might expect, our "art movie" Blind Chance is somewhat looser at this
certain pervasive social conditions. In each future, Witek is involved directly or indi-
level, relying more on the sheer successiveness of events and leaving appointments
rectly with the unofficial student movement and their underground publications. In
and deadlines offscreen. In the second story, for instance, Witek's childhood friend
his Communist career, he turns a blind eye to the movement; in his Catholic career,
Daniel appears at a meeting of the underground students' organization, along with
he is an activist within it, helping print the leaflets; and during his medical career, he
his sister Vera. Witek's subsequent romance with Vera is shown in brief scenes of
must replace his mentor, who is fired because his son is involved with the movement.
them meeting on the street, or spending time together in his apartment. These scenes,
Recurring characters and background conditions render widely divergent futures
like Vera's departures by train, aren't set up by explicit appointments, though such
more cognitively coherent.
arrangements must have been made. Indeed, when the couple split up, it's because
4. Forking-path tales are unified by traditional cohesion devices. of not making an appointment (Witek is told she's gone to Lodz, but actually she's
By cohesion devices, I mean formal tactics that link passages at the local level— waited outside his apartment for hours before finally leaving). As I suggested in the
from scene to scene or from one group of scenes to another. The classical narrative previous essay, this loosening of causal and temporal bonds is characteristic of much
cinema of Hollywood and the narrational strategies characteristic of art cinema have ambitious filmmaking in Europe after World War II.
developed many such tactics to aid the viewer's comprehension. We find them in the Yet in Blind Chance, cohesion operates from another angle. The film opens with
forking-path tales as well, usually serving to tighten up linear cause and effect. an enigmatic prologue showing Witek sitting in a train or airplane seat, facing us and
Two primary cohesion devices of mainstream cinema are appointments and starting to scream (Figure 6.5). The credits unroll over his howl. After the credits,
deadlines, and our forking-path movies provide these aplenty. Run Lola Run is built we see an enigmatic image of a hospital emergency room, with a woman's leg in the
around a looming deadline: If Lola doesn't meet Manni by noon, he'll try to rob a foreground and a bloody corpse hauled away in the background (Figure 6.6). Only at
supermarket to get 100,000 marks. Sliding Doors is structured around a cascade of the end of the film will these images make sense: In the final story, Witek is aboard
appointments: in one line of action, the appointments necessary to find Helen a new a plane to Paris and it explodes in midair; this is the last image of the film, over
job; and in the other, the dates she makes with James, the man who attracts her after which the final credits appear (Figure 6.7). Now we can place the opening shot of his
she leaves Jerry. In Too Many Ways to Be No. 1, each of Wong's alternative futures shriek—presumably, his last moments—and we can understand that it is apparently
180 Poetics of Cinema
his body that is dragged through the emergency room. The film curls around on itself,
back to front.
Whether the cohesion devices are indebted to norms of classical filmmaking or art
cinema practices, they call upon skills we already possess, notably our ability to bind
sequences together in the most plausible way in terms of time, space, and causality.
5. Forking paths will often run parallel.
One consequence of sticking to a core situation, the same locales, and the same
cast of characters is that certain components emerge as vivid variants of one another.
Thus, in Blind Chance we're inclined to contrast the three women with whom Witek
gets involved: the politically committed Chyushka, the more ethereal Vera, and the
practical, somewhat anxious Olga. After the death of his father, Witek finds a replace-
ment figure in each future—the veteran Communist Werner, the sympathetic priest,
and his medical school dean. Similarly, Lola seems to have the power to restore life: to
herself at the end of the first trajectory, to Manni at the end of the second, and to the
security guard Schuster, whom she revives in the ambulance at the close of the third
tale. Sliding Doors brings out parallels even more sharply by intercutting its alterna-
tive futures rather than presenting them seriatim. In one scene, Helen is ministered
to by her friend Anna before she showers; in the following one, Jerry ministers to her
cut head before she takes a shower. The cleverest moments in this organization come
when the two futures converge on the same locale, so that in one scene, the bereft
Helen drinks woozily at a bar while at a nearby table the happily ignorant Helen dines
with the boyfriend who's cheating on her.
Two Many Ways to Be No. 1 handles parallels in a joking manner characteristic
of the whole film. The second, longer story takes Wong and Matt to Taiwan, where
Matt lets it be known that he's a contract killer. They fall in with an enormous, hirsute
triad boss named Blackie White, who hires Matt to wipe out his twin brother, Whitey
Black. Matt already has accepted a job from an unknown boss, who turns out to be
Whitey, asking him to shoot Blackie. The entire confusion comes to smash at a party
where the two brothers sit side by side in complementary outfits and Matt bursts in to
earn his money, but is unsure which gang boss to terminate. The symmetrical staging
makes the alternatives comically explicit (Figure 6.8). Too Many Ways can be taken as
a send-up of forking-path stories generally, and this hyperexplicit parallel parodies a
central convention of the form.
Most narratives contain parallel situations, characters, or actions, though the
parallels are not always very salient. Sharply profiled parallelisms, as we know from
Intolerance (1916) and The Three Ages (1923), are a long-running cinematic tradition
and have become fairly easy to follow. Forking-path plots can bring parallelisms to
our notice quite vividly, thereby calling forth well-practiced habits of sense making.
Because these films hold many elements constant in each variant, parallels become
easy to spot.
Film Futures 181
Figure 6.8 Flamboyantly symmetrical stag-
ing in Too Many Ways to Be No. 1.
6. All paths are not equal; the last one taken, or completed, presupposes
the others.
A narrative, in Meir Sternberg's formulation, amounts to telling in time, and as
a time-bound process, it calls upon a range of human psychological propensities. 11
What comes earlier shapes our expectations about what follows. What comes later
modifies our understanding of what went before; retrospection is often as important
as prospection.
Forking-path films thus tend to treat replays of earlier events elliptically. When
action leading up to a fork is presented a second or third time, the later version tends to
be more laconic. Witek's three runs for his train are rendered in ever-briefer versions
(88 seconds, 67 seconds, and 59 seconds). Similarly, the first stretch of Too Many Ways
shows Wong meeting Bo, going to a cafe with his pals to propose dealing with the
Mainlanders, and then meeting the Mainlanders at the bathhouse, where the fight
over the bill ensues. After the massacre on the Mainland, the narration jumps back
to the meeting with Bo, and the following cafe session is rendered in 42 seconds, as
opposed to the 2 minutes it took in the first version. Because we know what happened
there already, the scene can be presented more pointedly the second time around,
even though it is, in that trajectory, still happening for the first time.
More importantly, forking-path narratives tend to treat information that we learn
in one world as a background condition for what is shown later in another. Sometimes
this pattern is fairly tacit, yielding the sense that alternatives are being exhausted
one by one. The types of choices offered to Witek in Blind Chance have this cumula-
tive quality: What if I took the path of least resistance and joined the Communist
Party? What if I summoned up more strength and opposed the party? Because each
of these choices fails, it seems, only through an apolitical stance can one maintain
one's decency, and that option is enacted in the third alternative. Alternatively, the
earlier narrative can explicitly contribute certain conditions to this one. In O. Henry's
"Roads of Destiny," the first story introduces the choleric Marquis; the second story
elaborates on his plot to overthrow the king. The third variant can therefore be much
more laconic in telling us whose pistol was responsible for the hero's suicide. Similarly,
in the second tale shown in Too Many Ways, when Wong is reunited with his pals
in Taiwan, the deaths of Bo and another gang member are reported at just enough
length to indicate that the men met the same fates as they did in the first story.
182 Poetics of Cinema
Makers of forking-path plots seem tempted to contaminate each story line a little.
At one moment in Sliding Doors, the heroine has an inkling of what is happening
in the parallel story. Walking along the river with her friend Anna, Helen seems to
anticipate what's happening at the same moment in the other story, in which her
counterpart cheers on a crew team: "Fairly weird. I knew there'd be a boat race going
on in purple and white shirts." Shortly, I'll show how the film's resolution depends on
this kind of crosstalk between futures.
Most surprisingly of all, sometimes a film suggests that prior stories have taught
the protagonist a lesson that can be applied to this one—thereby flouting any sense
that parallel worlds are sealed off from one another. One critic has noted that Witek
in Blind Chance seems to become more reflective from future to future, as if he were
cautiously exploring his "trilemma." 12 The first story of Too Many Ways presents
Wong as comically inept at nearly everything he tries; in the second story he is more
self-possessed, whereas Matt is the one who fouls things up. It's as if dying through
bungling in the first plotline has made Wong wiser. And if the epilogue of Too Many
Ways does suggest that Wong is considering not meeting the gang tonight, that
hesitation might depend partly on his intuiting, through means we cannot divine,
what happened in his first and second futures.
The clearest example of this tactic comes in Run Lola Run. Lola not only seems
to push the reset button at the start of each trajectory, but also learns to control the
chance that ruined her previous futures. During the lead-in, when Manni phones
to beg for help, Lola screams in frustration, and her screech explodes bottles sitting
on her TV monitor. In the first story, when Lola's father asks her to explain why she
needs the money, the pressure of time and anxiety triggers another scream, this time
bursting the glass on a clock face. But in the third alternative future, Lola tries to win
money at a casino, where she bets on a spin of a roulette wheel. She calculatedly emits
another scream, and this one not only breaks glass but also guides the ball into the
winning slot. It's as if she has learned to tame what was initially a sheer expression of
desperation, turning it to her purposes.
Due to the exigencies of telling in time, we might say, it's difficult for parallel futures
to receive equal weighting. The future shown first supplies some preconditions for
later ones, always for the audience and sometimes for the character. Psychologically,
the primacy effect treats the first future as a benchmark setting down the conditions
that will be repeated, varied, omitted, or negated in subsequent versions. Moreover,
given the fact that the hero or heroine is a constant presence in all these futures,
our entrenched expectations about character change—modification of personality,
or growing knowledge—alert us to any cue that, contrary to the laws of nature, the
protagonist may register and even learn from her or his alternative fates. This may be
a vestige of the supernatural and time travel versions of the parallel universes tale.
Scrooge retains psychological continuity in visiting different futures, and he's become
fully aware of all his options.
Film Futures 183
7. All paths are not equal; the last one taken, or completed, is the least hypo-
thetical one.
If something like a primacy effect establishes the first future as a benchmark, the
"recency effect" privileges the final future we see. Endings are weightier than most
other points in the narrative, and forking-path tales tend to make the early stories
preconditions for the last one. So these plots suggest that the last future is the final
draft, the one that "really" happened; or at least it reduces the others to fainter
possibilities. If the protagonist seems to have learned from the events shown earlier,
the ending may gain still more prominence as the truest, most satisfying one. This
suggests another reply to Morson's worries about forking-path plots. In principle,
multiple futures make all choices equiprobable and thus morally equal; but narra-
tive unfolds in time. By weighting certain futures through all the resources of order,
delay, point-of-view switches, and the like, the plot's design makes some options more
significant than others, both structurally and morally.
I've already suggested how, at the close of Too Many Ways, Wong might be said
to have assimilated what happens in his other futures, but the sense of "getting the
future right" is much more evident at the end of Run Lola Run. Manni has recovered
the stolen cash and returned it to Ronnie, whereas Lola has won big at the casino and
now has 100,000 marks for both of them. In a classic happy ending, they walk off
together. Manni asks, "What's in the bag?" and Lola smiles. The upbeat coda plays off
against the grim consequences of the previous two futures (Lola shot and Manni run
down) and renders them lesser options. A carefree ending is more in keeping with the
ludic tone established from the start, when the bank guard Schuster introduced the
action to come as a vast game. Tykwer goes even farther, seeing the last future as a
consequence of the other two:
At the end, the viewers must have the impression that Lola has done everything
that we've just seen (and not just one part, a third of it). She has lived it all—she
has died for this man, he has died, and everything that was destined to happen
has happened. She has all that behind her, and at the end, she's rewarded. 13
I think this corresponds with the intuitions of many viewers that Lola has somehow
lived through, and learned from, all the futures we've witnessed.
Blind Chance privileges its third future by the swallowing-the-tail strategy I've
already mentioned. The film's final shot of the plane exploding links neatly to the
prologue, showing Witek starting to scream in a plane seat, which also explains
the second shot of casualties in an emergency room. That gory image, moreover, is
glimpsed again in the beginning of the second story; in retrospect, we can see it as
adding more weight to the death-by-air outcome. Just as important, the ending is
given saliency by the fact that in the other two lines of action, Witek has planned to
take the plane to Paris but for one reason or another doesn't do so. In only the third
story does he catch his flight, and only the midair explosion shown in the final tale
explains the images that open the film.
184 Poetics of Cinema
Sliding Doors offers a fresh, equally ingenious way to weight the last plotline. Recall
that in one plotline, Helen misses her train, arrives home late, and so for a long time
remains unaware that Jerry is conducting an affair with Lydia. In the course of this
path, Helen picks up day jobs as a waitress and food courier to support Jerry while he
purportedly writes his novel. This line of action highlights the love triangle ofJerry-
Helen-Lydia, making James virtually absent, and it adheres fairly closely to those
conventions of deceit, superior knowledge, and abrupt emotional turns (including
Helen's eventual discovery of Jerry's affair) that are characteristic of film melodrama.
In the alternative plotline, Helen catches her train, meets James, and discovers Jerry s
infidelity. As a result, she leaves Jerry, gains confidence, falls in love with James, and
sets up her own public relations firm. This pathway highlights the love triangle of
Jerry-Helen-James. Lydia plays a secondary role, and thanks to James' stream of
patter and a generally lighter tone, this line sketches out a typical romantic comedy.
And because of the parallel worlds conceit, the melodrama plot and the romantic
comedy plot are intercut.
Both futures climax in Helen's being taken to the hospital near death (through
a fall downstairs and through being hit by a truck). In one plotline, she dies; in the
other, she lives. Remarkably, however, she dies in the romantic-comedy plot, and she
lives in the melodrama plot. So the problem is, How to end the film? If we conclude
with Helen's death, this would arbitrarily chop off the romance and punish someone
who has not wronged anybody. As in Lola, there is a presumption in favor of a happy
ending, preferably one in which Helen is united with James. But in the plotline in
which Helen survives, she doesn't even know who James is! How to arrange a con-
summatory ending?
Early in Sliding Doors, before Helen's paths fork on the tube platform, James runs
into her in an elevator, when she drops her earring and he picks it up. At the start of
the romantic comedy plotline, Helen meets him again on the tube, but she's so dis-
traught from having been fired that she can't accept his cheerful flirtation. Later in
the romantic comedy plot, it's established that James' mother is ill and must be taken
to the hospital. So at the end of the melodrama plot, after breaking up with Jerry,
Helen is discharged from the hospital. She enters an elevator; James, leaving his ill
mother, steps into the same elevator car. Again, Helen drops her earring; again, he
picks it up. Like Blind Chance, Sliding Doors lets its epilogue fold back on its prologue,
but instead of dooming the protagonist it allows the romantic comedy plot to restart.
And this time it starts properly: Helen is already wised up to Jerry's unfaithfulness
and can appreciate James. Helen also gets another disquieting glimpse of her parallel
life, for she is able to answer James' question with the tag he has used throughout the
romantic comedy line of action ("Nobody expects the Spanish Inquisition").
Here again, the last future we encounter is privileged by its absorption of the lessons
learned in an earlier one. Instead of calling these forking-path plots, we might better
describe them as multiple-draft narratives, with the last version presenting itself as
the fullest, most satisfying revision. Once more, this conforms to our propensity to
weight the last ending, to treat it as the culmination of what went before i t . . . even if
what went before couldn't really have come before.
Film Futures 185
S o m e Sources
As in any study of genre conventions, mine has had to play down major differences
among the films. The techno rush of Run Lola Run sharply contrasts with the sober,
philosophical pacing of Blind Chance. Moreover, I haven't gone on to examine other
forking-path films, such as Iwai Shunji's Fireworks, Should We See It From the Side
or the Bottom? (Uchiage hanabi, shita kara Miruka? Yoko cara miruka? 1993) and
Ventura Pons' To Die (or Not) (Morir [o no]; 2000). By the turn of the millennium, the
conventions of such films seem so well-known that new movies can play ofFthem. The
plot may initially set out two parallel futures but then concentrate on just one, bring-
ing the other one onstage at intervals (Me Myself I [1999] and The Family Man [2000]).
Or the plot may add one or two more switchpoints. Cesc Gay and Daniel Gimelberg's
Hotel Room (1999) starts by showing two different occupants taking the same room
in alternative futures. Within one of those futures, the plot branches out to two pos-
sibilities of further action, before moving back to a very early moment in the film and
opening a new fork there as a finale.
To confine my case just to cinema overlooks comparable experiments elsewhere
in popular culture. For example, J. B. Priestley's play Dangerous Corner (1932) begins
with upper-class idlers listening to a radio play at a cocktail party, then shutting it off.
The drama builds to sordid revelations, and one partygoer dashes out to kill himself
(just as in the radio play, at the moment they turned it off). The plot jumps back to
the opening situation, but instead of turning off the radio, the idlers let it continue.
On a bigger scale, Allen Drury's cycle of middlebrow Washington novels leads up to
an assassination at the climax of Preserve and Protect (1968), but the novel refuses
to specify which of two presidential candidates is killed. Come Ninevah, Come Tyre
(1973) shows the liberal, Commie-loving candidate surviving and becoming presi-
dent, but in the parallel sequel, The Promise of Joy (1975), the conservative candidate
takes the oval office and trumps the Russkies. Alternative-worlds plotting in today's
media can get remarkably complicated. Luxuriant binary branching is on display
in Alan Ayckbourn's eight-play cycle Intimate Exchanges (1982), from which Alain
Resnais drew his duplex films Smoking/No Smoking (1993).14 An episode of the TV
series Malcolm in the Middle gives two brothers parallel world adventures, sometimes
running the plotlines simultaneously via split screen.15 Chris Ware's ornate multi-
frame graphic novels sometimes harbor several branching paths, often within the
design of a single page.
The neatness of most parallel world plots is thrown eerily out of kilter in Stephen
King's duplex novels Desperation and The Regulators (both 1996). Characters from
one tale reappear in the other with only partly recurring attributes (same name, but
different body; or same name and body, but different personal histories and fates).
Robert Anton Wilson pursues this track more methodically, even monomaniacally,
in the 1979 novels making up the Schrddinger's Cat trilogy. The film that comes closest
to this massively recombinant strategy seems to be Mohsen Makhmalbaf's Time of
Love (Nobate Asheghi; 1991), in which across three episodes four actors swap roles as
husband, wife, lover, and onlooker, each episode yielding a different outcome. At an
186 Poetics of Cinema
Figure 6.9 Doc explains forking paths in
Back to the Future II (1989).
extreme, the fiction can merely raise the possibility of parallel worlds, without pro-
viding a consistent frame or definitive fork leading to linear paths. I don t know of any
cinematic examples, but a literary one that comes to mind is Robert Coover s short
16
story "Quenby and Ola, Swede and Carl."
We could also trace the strategy back through film history. The simplest option, it
seems, was to follow A Christmas Carol and postulate a supernatural agency that pro-
vides the protagonist an extended vision of one alternative future. The most famous
example on film is It's a Wonderful Life (1946), but it had at least one predecessor in
Turn Back the Clock (1933), in which a hero who'd like to relive his unhappy life is
struck by a car and miraculously given another chance. Essentially the same device
is revived in Mr. Destiny (1990) and Guido Manuli's short cartoon +1-1 (1998). The
strategy of presenting more than one parallel life course seems to be rarer, but in early
instances it is handled through the device of rival prophecies. In Eyes of Youth (1919),
a young woman consults a fortune-teller who offers her three futures, all with calami-
tous outcomes. The heroine resolves to avoid them all by picking a fourth option, mar-
rying the man she loves. This particular plot was recycled at least twice in the silent
era.17 (Too Many Ways to Be No. 1 obliquely revives the fortune-telling device.) A com-
parable frame can be provided by fantasy, as in the alternative methods of homicide
plotted by daydreamers in Unfaithfully Yours (1948) and Murder, Czech Style (1967),
or the cafe speculations on tragic and comic plotting that frame Melinda and Melinda
(2004). It's likely that the success of the Back to the Future movies (1985-1990) made
filmmakers realize that this template could be updated through fantasy and science
fiction conventions. The first installment of the trilogy is a straightforward time-
travel plot in which readjusting the past changes the present, but Back to the Future II
(1989) looks forward to the forking-path innovations of the 1990s. This installment
shuttles Marty McFly and other characters back and forth between past and future,
so that when Doc generously diagrams the various ways in which Marty's travels have
upset the space-time continuum, we're invited to see the events' alternative outcomes
as parallel worlds (Figure 6.9).18
Most of the early forking-path stories invoke supernatural, fantasy, or science
fiction premises. Thanks to mysterious forces, characters visit pasts they didn't have or
futures they haven't yet lived. But explicit acknowledgment that alternate worlds run
alongside that of the present has evidently had to wait for recent years. In physics the
Film Futures 187
idea was implicit in Schrodinger's cat-in-the-box thought experiment, and it received
an explicit working out in 1957, when Hugh Everett III posited what's come to be
called the "many-worlds" interpretation of quantum mechanics. 19 John Wyndham's
1961 short story "Random Quest" developed the idea in a science fiction framework;
but it seems to have had few immediate successors. 20 1 can only speculate on why the
1990s should see such a burst of parallel universe narratives in popular culture. Before
we confidently claim that something in our postmodern society impels us toward
them, though, I'd advise looking for more proximate causes. Audiences' familiarity
with video games, recalled fairly explicitly in Lola, would seem to be a major impe-
tus. Perhaps too the popularity of the Choose Your Own Adventure children's books
prepared young people to find such plots intriguing. Then there's the broader urge
toward narrative experiments of many sorts in contemporary film, both mainstream
and off-Hollywood. Forking-path plots have emerged in a competitive marketplace,
and nowadays films—especially "independent" movies—are encouraged to provide
experimental novelty. Like films with temporal loops (e.g., Donnie Darko, 2001) and
the network narratives I consider in the next essay, forking-path experiments are also
video-friendly, encouraging consumers to watch the movie many times over to enjoy
the meshing and divergence of parallel worlds.21
Whatever the proximate and remote causes, the concept of alternative futures
will probably be adapted to the demands of audience comprehension and particular
narrative traditions—pruning the number of options to those few that can be held in
mind, finding new uses for cohesion devices and repetition, and relying on schemas
for causality and time and space. It seems likely as well that the more radically that
the film evokes multiple times, the more constrained it must be on other fronts.
Smoking / No Smoking presenting two feature-length alternative futures, can permit
itself no more than two characters, always male and female, per scene. Groundhog
Day (1993) breaks with one of my conventions by proliferating a great many futures
for its repellant protagonist. To compensate, it presents those futures as very short-
term alternatives, and it multiplies redundancy by repeatedly signaling its forking
point (the clock radio's wake-up song) and the parallel events in the iterated day.22
If such a trade-off between innovation and norm seems to cramp the infinite
vistas opened up by Borges, we shouldn't underestimate the extent to which stretching
traditional narrative requires care. Stories are designed by human minds for human
minds. Stories bear the traces of not only local conventions of sense making but also
the constraints and biases of human perception and cognition. A film, although
moving inexorably for ward (we can't stop and go back), must manage several channels
of information (image, speech, noise, and music). It must therefore work particu-
larly hard to shape the spectator's attention, memory, and inference-making at each
instant. No wonder that moviemakers balance potentially confusing innovations like
the multiple-draft structure with heightened appeal to those forms and formulas that
viewers know very well. Artists are forever testing the limits of story comprehension,
but those very limits, and the conventions that accommodate them, remain essential
to our dynamic experience of narrative.
T
i

I
I

Mutual Friends and

Chronologies of Chance

Writers like me, who are not attracted to psychology, to the analysis of feelings,
or to introspection, are greeted by horizons no less broad than those dominated
by characters with clear-cut personalities, or those revealed to people who
explore the depths of the human mind. What interests me is the whole mosaic
in which man is set, the interplay of relationships, the design that emerges from
the squiggles on the carpet.... These human presences defined only by a system
of relationships, by a function, are the very ones that populate the world around
us in our everyday lives, good or bad as this situation might appear to us.

Italo Calvino1

Connections—we make them and are made of them.

Henry in Nine Lives (2005)

Inge takes.care of invalids and elderly people through home visits. She's unhappy, and
she attributes this to her unassertive ways. After seeking counseling, Inge decides to
change her life drastically. She quits her job, euthanizes her beloved dog Kiwi, and
joins a pole-sitting contest. Squatting on a pole in the ocean for days, Inge develops an
abrasive friendship with another contestant, the oafish Leif. Eventually, she outlasts
her competitors and goes for a world record. Inge breaks the record, but when she

189
190 Poetics of Cinema
sees Kiwi running along the beach, she dives off and tries to catch up with him. It is
of course a different dog, and she soon collapses in guilt and despair. Leif finds her
onshore and invites her to come off with him, offering her some hope.
Inge's search for happiness is traced in What's Wrong With This Picture? (Tid til
Forandring; literally, "Time for a Change"), a 2004 Danish film by Lotte Svendsen.
Actually, however, Inge's story consumes only a quarter of the 90-minute feature.
Interspersed with scenes from her life are episodes centering on other characters.
Svend is an elderly socialist still struggling to build support for the Third World. The
psychiatrist Erik has a drinking problem, and his wife Lea is a shopaholic bent upon
beautifying their home. Jens, an IT worker, is obsessed with a fashion model living
across from him. The film satirizes the petty concerns of the prosperous characters
and plays up the pathos of the weaker ones, like Svend and Inge.
In this survey of Danish urban life, the narration alternates among various lives
so that no one character emerges as a protagonist. What then connects these people,
apart from living at the same moment in Copenhagen? Inge is the link. She serves as
caretaker for Svend, who is an old friend of the once-liberal Erik. Erik is the profes-
sional counselor whom Inge visits while trying to redefine her life. And Inge is a former
classmate of the superficial but gorgeous model Gry, the object of Jens' surveillance.
Once these characters are introduced, we see their lives in alternation with Inge's.
Narrationally—that is, in terms of the flow of information about the action and
the story world—Inge serves to introduce us to the other characters, either directly
or indirectly. The credits give us a glimpse of Svend, and when Inge visits him in
his apartment, we learn of his project to hold a large dinner debate on Third World
problems. When Inge leaves Svend, however, we follow her. She encounters Gry in
the street and reminds her of their school days together, and Gry will eventually link
to Jens, her timid stalker. Later, when Inge visits Erik for counseling, we learn of
his alcoholism, and after she leaves we first see Lea's obsession with redecorating the
couple's home. Once Inge leads us into the film's world, though, she drops to lesser
prominence. While she squats grimly atop her pole, other characters act and react to
more changing circumstances.
The plotlines don't affect one another much. Once Inge decides to become more
assertive and take up pole sitting, she never encounters the other principals again.
Erik's descent into alcoholism is doubtless accelerated by Lea's urge to make their
home designer-quality, but she proceeds oblivious to his plight. Likewise, Jens' pursuit
of Gry isn't affected by the other characters' purposes. He eventually meets Gry by
chance and plays on her sympathies by having his disabled brother Jan pretend to
be an invalid. Eventually Gry dumps Jens and flees town, leaving him to take care
of her slightly retarded cousin Tina. The most crucial convergence comes when Lea,
after forcing an immigrant to sell her a rare sculpture, orders it winched down from a
window. It falls, killing Svend on his way home from his failed political dinner.
So A meets B, then C, then D. B and C already know each other. Through C we get
to meet E, and through D we get to meet F—even though A doesn't know them. The
film opens up a social structure of acquaintance, kinship, and friendship beyond any
one character's ken. The narration gradually reveals the array to us, attaching us to one
Mutual Friends and Chronologies of Chance 191
character, then another. And the actions springing from this social structure aren't
based on tight causality. The characters, however they're knit together, have diverging
purposes and projects, and these intersect only occasionally—often accidentally.
Talking of A's, B's, and C's is a fancy way of putting matters that viewers grasp
immediately. What's Wrong With This Picture? is easy to follow because it exemplifies
a common norm of storytelling in contemporary cinema. Many viewers are likely
to pick up the film's affinities with the n-degrees-of-separation template on display
in Nashville (1975), Short Cuts (1993), and Magnolia (1999). This sort of plot pattern
has been called "thread structure," 2 and the films have become known as tales of
"interlocking lives," "converging fates," and "the web of life." Variety seems to have
settled on "criss-crossers." Elsewhere I've called them network narratives, and I'll
mostly stick to that here, for reasons that I'll try to make evident. 3
Network narratives have emerged as an important alternative to the single- or
paired-protagonist plot. The network format has been cultivated by directors as
various as Robert Altman, John Sayles, Paul Thomas Anderson, Claude Lelouche,
Garcia Rodrigo, Otar Iosseliani, Joao Botelho, Alejandro Gonzalez Inarritu, and
Tanaka Hiroyuki (aka Sabu). Such stories have become remarkably common, with
nearly 150 films using the network principle released since 1990. (My list of candidates
appears at the end of this essay.) In 2005 alone, there appeared Tapas and Sud Express
from Spain, Chromophobia and Festival (England), Istanbul Tales (Turkey), Look
Both Ways (Australia), Year Zero and What a Wonderful Place (Israel), Who's Camus
Anyway? (Japan), Voisins Voisines (France), The Manual of Love (Italy), See You in
Space! (Hungary), Frozen Land (Finland), and Crash Test Dummies (Austria). In the
same year, filmmakers in the United States gave us Heights, Standing Still, Nine Lives,
Happy Endings, Loggerheads, American Gun, and Me and You and Everyone We Know.
Converging-fates plots seem to be for us what flashback tales were for the 1960s: the
dominant principle of offbeat storytelling.
This narrative strategy seems at once fresh and familiar, unusual and widespread.
Its presence in films from many different cultures poses intriguing problems for a
poetics of cinema. What conventions govern these network narratives? What enables
audiences to follow them? These sideways-shifting tales push us to reflect on how
filmic storytelling works. They make distinctive use of causality, chance, parallelism,
and narration. They force us to ask how we're able to follow action and pick out pro-
tagonists. Tapping into our social intelligence, they demand that we trace out a web of
personal relations among characters. Most generally, however the format gets specified
in local circumstances, it seems to be trading on a cluster of cross-cultural norms.
The emergence of these conventions also shows an artistic arms race in action.
Filmmakers on the festival and indie scenes are expected to innovate. Once the
innovations become canonized, though, anyone with little talent can draw on them.
So ambitious filmmakers will strive to take the norms farther—tweaking them,
revamping them, and testing their limits. But such ambitions carry a risk. Break
the bounds, and you may lose the audience. As with forking-path plots, the trick
is to mix novelty with familiar strategies in such a way that viewers don't become
confused or disengaged.
192 Poetics of Cinema
Protagonists a n d Projects
What, a skeptic might ask, isn't a network narrative? Put aside isolatos like Robinson
Crusoe and Pincher Martin. Any garden-variety narrative involves a group of char-
acters, and those characters enjoy social relations with one another. They're rela-
tives, friends, acquaintances, lovers, coworkers, and the like. Alexander Mackendrick
wisely noted that "in a well-told story, every fictional character functions within a
network or nexus, a cat's cradle of character interactions." 4 Aren't nearly all narratives
about social networks?
Yes. But narratives can highlight this social network to varying degrees. They do
this by building worlds, shaping structure, and treating narration in particular ways.
Because we're dealing with a concept having fuzzy borders, I'll start by pursuing
some clear-cut intuitions. I'll try to build on the points I made in Chapter 3 about the
criteria we apply to identifying the protagonist.
Uncontroversially, most stories have primary agents—protagonists. Many narra-
tives center on a single protagonist who acts to achieve a goal. Some center on a pair
of characters who have set themselves identical or complementary goals. There are
buddy movies like Butch Cassidy and the Sundance Kid (1969) and cop-crook pairings
like Heat (1995), and of course many melodramas and comedies feature a romantic
couple as dual protagonists. At times the dual-protagonist pattern is complicated by
setting two lovers against one another, as in His Girl Friday (1940) or It Happened One
Night (1934).5 In both single- and dual-protagonist plots, subsidiary characters tend
to participate in the action as helpers or as blocking or delaying factors. Even if the
secondary agents are given their own projects, as when two friends of the protago-
nists fall in love, those are likely to be shaped by the overall progress of the romance
between the principals. And sometimes a protagonist-based film can teem with lively
secondary characters, as in Cedric Klapisch's L'auherge espagnole (2002).
There are also multiple-protagonist narratives, featuring three or more primary
agents. In The Poseidon Adventure (1972), several characters survive the capsiz-
ing of a giant ocean liner, and they must find their way out of the wreckage. They
work together as a team pledged to a common goal, mutual survival. In Advise and
Consent (1962), the U.S. Senate must decide whether to confirm the president's
choice for secretary of state. The nominee himself and half a dozen senators are
delineated as they participate in this process of confirmation. However distinctive
each man's political strategy may be, all of the players are defined by their roles in
the overriding project.
In my central cases of network narrative, there are also several protagonists, but
their projects are largely decoupled from one another, or only contingently linked.
In Advise and Consent, the senators' political decisions are reactions to stratagems
launched by their opponents and allies. There is a thrust-counterthrust escalation
of conflict. By contrast, in What's Wrong With This Picture? Inge's unhappiness with
her life neither affects nor is affected by Lea's shopping binges or Jens' obsession with
the model Gry. Advise and Consent culminates in a senator's suicide, the result of an
attempt to blackmail him into voting for the president's nominee. But in What's Wrong,
Mutual Friends and Chronologies of Chance 193
Figure 7.1 Bug (2002): The credits sequence
presents a network mingling characters and
creators.
Lea's quest for a designer home causes Svend's death by sheer accident; he just
happens to be walking underneath the sculpture that plummets out of a window.
Granted, networked protagonists may influence one another in far-reaching ways.
Their projects can collide, and they may rethink their purposes through encounters
with their counterparts. One character may give another advice or assistance. Never-
theless, when paths cross meaningfully, they tend to remain distinct and of equal
prominence. In Beautiful People (1999), two men we've observed leading separate
lives come together on the battlefield of the Bosnian war. Jerry, a BBC reporter, starts
out quite self-assured, but once he's wounded in the leg he comes home angry and a
bit deranged. Griffin, a heroin-addicted skinhead, learns pity for war's victims and
returns to London to care for a boy blinded in battle. The men never meet again.
Their encounter in a field hospital changes both their lives, but their stories remain
largely independent.
When watching movies like this, we mentally construct not an overarching causal
project but an expanding social network. Any link can reveal further connections.
As Beautiful People unfolds, we discover distant affinities between Jerry and Griffin.
Griffin's father is principal of the school that Jerry's son attends, and in a hospital
Jerry rails at a nurse who had tended to Griffin earlier. In turn, the nurse also tends to
a pair of angry refugees from the war, and she works under Portia, a doctor in train-
ing who has fallen in love with a Croat exile. Eventually Beautiful People binds four
British families and several isolated people to the dissolution of Yugoslavia.
As the characters' activities diverge and converge, these crisscrossings eventually
settle into a more or less cogent pattern, Calvino's "squiggles in the carpet." No wonder
critics sometimes call these movies mosaics: When we back off from the tiny bits, we
discern a larger composition. Network narratives suggest geometry or choreography,
or boxes-and-arrows diagrams, or schematic circuits (Figure 7.1). As viewers we push
toward the bigger picture largely because we register the relative independence of
several protagonists' trajectories. We wouldn't get so strong a sense of a spreading web,
and we wouldn't discern the degrees of separation so vividly, if we were following the
sort of narrative that guides us to center on one or two protagonists and their goals.
194 Poetics of Cinema
T h i s Particular ^X/eb
The network principle has a proud place in Western fiction and drama. Middlemarch,
War and Peace, and other monumental 19th-century novels seem to anticipate most of
the formal devices we find in our movies. Our Mutual Friend (1864) creates a tapestry
of characters, some only remotely connected, and Dickens spells out the contingency
of some connections. When Lizzie Hexam joins John and Bella, "it fell out that she
became the unconscious means of bringing them together." Mann's Buddenbrooks
(1900) creates its network by tracing the family's marriages, friendships, and business
enterprises across 60-plus years and three generations.
Similar principles have underpinned fiction in the 20th century. The ties of family
and tradition are replaced by casual encounters and intersecting pathways in urban
novels like Manhattan Transfer (1925), though this tactic was at work in Balzac and
Dickens as well. Thornton Wilder's Bridge of San Luis Rey (1927) offered a converging-
fates plot that flashes back from three deaths on a Peruvian bridge to show what led
the characters to meet their maker. This template would inform dozens of later novels
and films. At about the same time, Vicki Baum synthesized the Grand Hotel format,
recycling it through a novel (1929), a play, and eventually the 1932 MGM movie.
Baum revived the formula in Shanghai '37 (1939). Since Baum's day, network narra-
tives have often shown up in novels both prestigious and low level. Camillo Jose Celas
The Hive (La Colmena, 1953) probably constitutes the record, at least in sweep, with
its 300 characters passing through a neighborhood cafe. (The 1971 film adaptation
reduced the cast to about 70.) More recently, Alexander McCall Smith revived the
daily newspaper novel by means of a network plot in 44 Scotland Street (2005). Crime
novels have made intriguing use of the device, ranging from the one-off intersection
in Charles Willeford's Sideswipe (1987); through Ed McBain's roundup of his entire
87th Precinct team in Hail, Hail, the Gang's All Here! (1971); to the precarious clusters
of kinship, alliances, grudges, and bad luck filling the Washington, D.C., sagas of
George P. Pelecanos. We find the principle in children's literature too, such as Diana
Wynne-Jones' Dalemark Quartet (1975-1993). More self-consciously literary efforts,
like Ron Loewinsohn's Magnetic Fields (1983), Alex Garland's The Tesseract (1999),
David Mitchell's Ghostwritten (2001), and Alice Mattison's In Case We're Separated:
Connected Stories (2005), treat social networks as ephemeral and sometimes mystical
or hallucinatory.
Central to network narratives in any medium is the fundamental tension between
realism (after all, we're all connected to each other) and artifice (order must be
imposed on all the potential connections we can find). Take the closed-environment
tale, which squeezes a network into a single locale. In film we have not only Grand
Hotel but also The VIPs (1963), which presents a life-changing few hours among
characters in a Heathrow waiting room, and more recently Km. 0 (2001), confining
itself to the wayward encounters of characters passing through a sunny Madrid plaza.
This spatial template has long been a source of literary experiment. Zola's Pot-Bouille
(1882) centers on a Parisian apartment building that houses five families plus servants
and single renters, and the novel traces out their ties to relatives, employers, mis-
Mutual Friends and Chronologies of Chance 195
tresses, friends, a priest, and a doctor. Michel Butor's Passage de Milan (1954) seems
a self-conscious rewriting of Pot-Bouille: Set in an apartment house, its 12 chapters
cover a 12-hour period in which characters are linked through not only acquain-
tance but also sheer spatial adjacency. No less rigorous is Georges Perec's Life: A User's
Manual (1978), which renders everyday activities in an apartment building as if they
were a series of moves on a 10 x 10 chessboard.
We don't lack commentators who find something distinctively novelistic in
network hookups. Their metaphors echo ours. George Eliot speaks of her task as
"unraveling certain human lots, and seeing how they were woven and interwoven,"
thereby revealing "this particular web."6 Margaret Anne Doody, reflecting on the
novels of Greek and Roman antiquity, finds letter writing a sign of the open-ended
flow of connections typical of the long prose form:
Epic characters tend to be present to each other or absent, living cleanly in a
now. Novelistic characters are in a web of entanglements and connections....
The possibility of letters means that nothing is quite ended, that private self and
public self will continue to shift and slide into one another, that each soul knows
relationships which are many, not one.7
The film Grand Hotel was adapted from a play that Baum based on her novel, and
this should remind us that there are stage precedents for the network schema as well. To
some extent, the films we're considering are descendants of the multiple-plot drama.
Roman New Comedy often created subplots involving baser characters, and many
plays in the English Renaissance presented three or four intersecting lines of action,
usually manifesting different tones and attitudes. 8 Likewise, romantic comedies in
the vein of Alfred de Musset's The Caprices of Marianne (1833) created romantic
triangles out of several pairs of lovers. Films like Cecil B. DeMille's Why Change Your
Wife? (1920) and Ernst Lubitsch's The Marriage Circle (1924) inherited a tradition
whereby casual flirtations shuffle together married couples and unattached friends.
Like other multiplot dramas, these films tend to be causally tight, with key incidents
in one plotline shaping the others. Vestiges of dovetailing romantic rivalries surface
in those network movies in which couples break up and reform in roundelay fashion
(Little City, 1997; Goldfish Memory, 2004).
Knowing literary conventions doubtless helps viewers tease out a film's web of
character relations, but there's a closer precedent in another medium. Television soap
operas and episodic shows like Hill Street Blues (1981-1987) made audiences adept
at keeping track of many characters and their interactions. 9 The affinities between
degrees-of-separation movies and TV narrative are nicely confirmed by the fact that
Paul Haggis' film Crash (2004) was originally conceived as a series. When the project
could find no broadcast buyers, Haggis and his cowriter turned it into a feature.
Ironically, after the film proved a success, a company purchased rights to adapt it as
a cable TV series.10
The network notion seems to seep into all domains of popular culture. During the
1960s a weekly Spanish comic strip, 13 rue de Percebe, presented a single large panel
showing a cutaway four-story apartment building, with several stories unrolling
196 Poetics of Cinema
among the people living in each flat.11 Chris Ware used a similar dollhouse device
in Building Stories, and another independent cartoonist, Daniel Clowes, created in
Ice Haven nearly 30 interconnected tales taking place in a single day in one town.12
Researchers can study social clustering by sieving through the family of Marvel comic
book superheroes. 13 Of course, the Simpsons' creators haven't been slow to seize on the
idea, notably in the episode called "22 Short Films About Springfield."
Beyond calling on our experience with narrative conventions across media,
network stories very likely tap some fundamental human capacities. As primates,
we're social animals and thus sensitive to hierarchy, status, and coalitions. It's very
important for us to know who's doing what with whom, so gossip comes naturally to
us. No one is immune from the tantalizing attraction of the latest news about their
familiars. Psychologist Robin Dunbar has suggested that language evolved as a parallel
to the mutual grooming found in bands of monkeys and apes. Dunbar speculates that
early humans used language primarily to exchange information about others. 14 Social
networks are salient for creatures like us, and we must constantly update our sense of
the group's dynamics.
One skill that helps us is mind reading. We are very good at guessing what others
think, and we practice this at several levels: Judy is convinced that Jack is lying to
her. She tells Vincent, who doubts it. Vincent shares his misgivings about Judy's beliefs
with Amy. A PowerPoint slide of this situation would show a cartoon Amy and her
thought bubble, with Vincent and his thought bubble inside Amy's, and Judy and her
bubble inside Vincent's, and so on. Put in language or pictures, this situation seems
complicated, but the understanding that it captures is transparent. Second-, third-,
and fourth-order mind reading comes easily to us as social animals. 15
Our skills in tracking social relationships and surmising what others think are
aroused by nearly all narratives, but we get a real workout in the sorts of tales I'm
considering. The unrestricted narration characteristic of network narratives exercises
our ability to compare what different characters know and don't know. Our skill in
mind reading can prime our deeper engagement with the unfolding story. In What's
Wrong With This Picture? we see the signs of Erik's alcoholism far sooner than
his wife Lea does, and our superior knowledge prepares us to watch each of their
conversations with keen interest. Will she finally discover what we know? We know
that Portia's parents in Beautiful People are stuffily upper class, and we can anticipate
their judgment of the na'ive Croatian boy she brings home to dinner.
These two films, like many network movies, are directed by women, and it seems
likely that the format encourages the exploration of mental and social skills tradi-
tionally aligned with femininity. Robin Dunbar, who considers women central to the
evolution of language because of their role in promoting social exchange, has found
that women are on average far better at higher-order mind reading than men.16 The
stereotyped association of women with gossip and soap operas may inadvertently
acknowledge a high sensitivity to the intricacies of social life. The primate researcher
Sarah Blaffer Hrdy has suggested that in our evolutionary past, mothers were the
chief mediators of what children learned about social relations, and that the asso-
ciations built up among female relatives favored a genetically based development of
Mutual Friends and Chronologies of Chance 197
brain regions enabling persons to size up social situations.17 It would be too simple
to trace a straight line from these capacities to network narratives. But perhaps such
narratives are one form in which certain cultures, for concrete historical reasons, give
local habitation to some gendered skills.
There are other enticements. Network narratives often sacrifice depth of charac-
terization to breadth of coverage. This isn't necessarily a drawback if we recognize
the satisfaction yielded by a narrative that offers a vivacious panorama of life. William
Empson pointed out long ago that the multiple-plot drama displays the "repetition of
a situation with new characters to show all its possibilities." It has
an obvious effect of making you feel that the play deals with life as a whole, with
anyone who comes onto the street the scene so often represents.... Just because
of this carelessness, much can be put into it.
He adds that in this structure, "queer connections can be insinuated powerfully and
unobtrusively," a phrase that captures the mysterious synchronizations evoked by a
film like Magnolia.18
However broad and deep the appeal of this format may be, we still have to con-
sider its historical fluctuations. Between Grand Hotel and the early 1990s, there
don't seem to be a lot of network films.19 It seems to me that the current vogue can
be dated to a batch of films from 1993 to 1994. Robert Altman's Short Cuts (1993)
presents a cross-section of Los Angeles society through contingent links (from traffic
accidents to unconnected characters merely striding past one another). The prologue
of Michael Haneke's 71 Fragments of a Chronology of Chance (1994) informs us of
the catastrophe in advance, so we watch in dread as the characters' fates slowly come
together. Atom Egoyan's Exotica (1994) relies on exposing secret obsessions shared
by its characters, and Wong Kar-wai's Chungking Express (1994) offhandedly splices
together two stories crossing at a couple of moments and teases us into hoping for
more connections. 20 Quentin Tarantino's Pulp Fiction (1994) offered a package of
devices that later filmmakers would retool: repeated scenes, titles that split the film
into chapters, and a covert reordering of time that makes the audience gasp when
they see the stories mesh.
We can't locate a single cause for this burst of network movies, but we can point
toward certain preconditions. In the United States, the broadening of the indepen-
dent film sector to wider audiences in the late 1980s was accompanied by a degree
of formal experiment. In the midst of this change, sex, lies, and videotape (1989),
Do the Right Thing (1989), and Slacker (1991) tilted toward the degrees-of-separation
format. Soon Tarantino and other directors realized that major actors could be
recruited to indie projects, and the network idea proved friendly. Big stars didn't have
to commit many days to an ensemble vehicle, and they didn't demand their usual
high salaries.21 In Europe, thanks to Jacques Rivette, Jacques Tati, Otar Iosseliani,
and Krzystof Kieslowski, the network narrative had been a minor tradition for some
time. As EU coproductions gained larger subsidies and wider distribution, it became
feasible to shoot actors in their native countries speaking the local language. This may
have encouraged filmmakers to create multiple storylines that cross borders. The new
198 Poetics of Cinema
arrangement also had the effect of making European unity and differences a common
theme of the films.
More generally, during the 1990s the rise of communitarian legal theory and a
deepening awareness of human connectivity on a global scale became recognized in
popular discourse. People were becoming aware of the Internet as well, and they were
using the word networking to describe meeting new people through mutual friends,
often with an eye to personal gain. At the same time a more formal theory of networks
and "small worlds" was emerging in several sciences. It had been pioneered in the
1950s and 1960s by the mathematician Paul Erdos and the political scientist Ithiel
de Sola Pool. In the late 1960s the social psychologist Stanley Milgram argued that
U.S. citizens were, on average, six steps away from one another. Later, mathematicians
studied how disease outbreaks hit critical levels depending on degrees of connectivity
among infected individuals, and sociologists suggested that social groups could be
understood by means of strong and weak ties among members. By the mid-1990s,
these lines of thought were energizing several disciplines.22
I'm not suggesting that Altman and Tarantino were curling up with treatises
in math, law, and sociology. Scientific and philosophical theories slip into popular
culture, often as slogans or vivid images. The notion of "six degrees of separation"
became a catchphrase thanks to John Guare's 1990 play and the 1993 film derived
from it. Another line of research at the time was chaos theory, which posits unex-
pected order in what appear to be random fluctuations. In the wake of James Gleick's
1987 book Chaos: Making a New Science, the theory came to be epitomized in the
image of the butterfly that softly flaps its wings and sets off a hurricane half a world
away. The notion that tiny actions in one person's life can trigger big consequences in
another's is constantly exploited in network movies, and the butterfly effect is cited
in the network movies Free Radicals (2003) and Happenstance (2000), whose original
French title translates as "The Beating of the Butterfly's Wings."23
M a j o r Players
What constructive principles govern network narratives, and how do they tease us
into some characteristic experiences? It would be comforting if formal network theory
could lay out a handy menu of possibilities, but it doesn't help us much. Most social
network theorists define a link as personal acquaintance. I'm linked to Bill Clinton
because I have a friend whose wife knows A1 Gore. A network theorist wouldn't say
that I was linked to Clinton if we were waiting at the same airport, or if I merely
attended a Democratic fundraiser, or if I bought one of Monica Lewinsky's gift ties
on eBay. Yet network movies trade on just such remote and fragile connections. For a
scientist studying networks, the hookups are concrete, prosaic, and well defined, but
our network narratives also evoke poetic linkages. A film's mottos tend to be both
"me and you and everyone we know" and "mind the gap"—gaps being the slender,
precarious affinities that can suggest subterranean forces bringing fates together.
Mutual Friends and Chronologies of Chance 199
We can most fruitfully chart the whys and hows of such tales with familiar con-
cepts of film analysis. Network narratives use any resource of narrative generally
(flashbacks, constrained time frames, and the like), but the form displays some
predilections of its own. Here's a preview of arguments about story world, plot struc-
ture, and narration that I'll be making in the rest of this essay.
With respect to the story world, a network narrative centers on several protagonists.
Some pursue discrete goals, whereas others may have no goals at all. Whether these
characters know one another or are strangers to one another, they inhabit more or less
the same space-time framework and can interact face-to-face in given conditions.
Their lines of action intersect, in one-on-one convergences or more inclusive relation-
ships. Sometimes characters plan intersections in advance (by setting appointments
or deadlines), but to a high degree the convergences are controlled by chance. In this
world, some initially unconnected characters typically meet by accident.
Most stories of any sort establish certain circumstances as bedrock. We must
understand, from the very beginning, some dependencies and affiliations among the
characters. Who's related to whom? Who's sleeping with whom? Who's working for
whom? By contrast, converging-fates narratives may conceal long-standing connec-
tions among the characters. Distracting us with a crowd of agents pursuing their own
projects, the narration may delay establishing how those agents are connected. This
is to say that the film may rely on what Meir Sternberg calls delayed and distributed
exposition. 24 Right up to the end we may be still learning about long-term connec-
tions among the characters.
The film gives great weight to the protagonists' constantly changing dependencies
and affiliations. To a large extent the movie's narrative structure rests upon the per-
petual commingling of characters, so the prospect of recombination vies with causal
logic as the impetus for the action. As story action becomes less goal directed, we're
asked to form an abstract sense of structure. If A and B have met, and B and C have
met, the logic of the network tale suggests the need for a scene in which A encounters
C—whatever the causal pretext that might bind them. They might merely pass in
the street. As in a kaleidoscope, the elements are fixed, but a new jiggle knocks them
into fresh configurations. Convergences, minor or major, can become as important as
the events in the separate story lines. The plot structure therefore must find ways to
isolate or combine characters in compelling patterns that will replace the usual arc of
goal-directed activity. The principal source of these patterns, as we'll see, is chance.
As important as real-world knowledge is in grounding our understanding, we
gain access to the story world only through narration. For one thing, in a multiple-
protagonist film several characters get a lot of screen time. We see Inge, the pole sitter
in What's Wrong With This Picture?, for a total of 22 minutes, about the same period
of time as we see the therapist Erik and his wife Lea. The most visible character is Jens
the IT worker, whose scenes consume 32 minutes in all, but even he is offscreen for
about two-thirds of the film's running time. More generally, the narration weaves the
network before our eyes, slowly or rapidly, creating firm knots or leaving great holes.
The narration can expose or obfuscate the affinities among the characters, tethering
us to a single mind or presenting a more unrestricted view of the doings of the film's
 Poetics of Cinema
200
population. It can give us clear-cut exposition delineating character relations, or it
can delay revealing crucial connections until quite late. All of these alternatives shape
the spectator's experience of the narrative.
Because of the shared-time principle, the narration tends toward omniscience. It's
showing us characters who might not meet, who might not know of each other's exis-
tence. Filmmakers have found several ingenious ways to handle this omniscience,
shaping our expectations in unusual ways and often cunningly withholding key
information. Within the overarching omniscience—signaled in some fairly explicit
ways—the narration may restrict us to what one or a few characters know at some
points, thus revealing unexpected links for the sake of surprise. Like a mystery film,
the network narrative frankly exposes the act of narration, inviting the viewer to
build inferences out of teases, hints, and gaps.
The film is also held together by privileged themes, such as the contrasts among
social groups or the hope that barriers can be transcended through commonali-
ties. These themes aren't simply "content" to be plugged into the network template.
They are plausible materials for these norms of story world, plot structure, and
narration. Filmmakers are well aware that their degrees-of-separation films should
hang together thematically. Rodrigo Garcia says that all the stories in Nine Lives are
about "a person trapped in a relationship with someone they could not escape." He
adds, "The common themes don't have to be evident consciously. But people have to
feel they belong together."25 That feeling is generated by the ways a film's narrative
strategies transform thematic materials already circulating through culture.
Familiars, Strangers, a n d R a n d o m NValks
Let's start with characters and their actions in the story world. Two broad options
seem to be open for network treatment. We might deny the characters any fixed goals.
This yields slice-of-life plots that follow characters' daily routines or their casual
meanderings, as in Wong Kar-wai's Fallen Angels (1995). Here a nameless paid killer
never meets Ho, but as they follow their routines they intersect with three women,
and Ho eventually hooks up with Killer's female agent. An extreme case of routine
and wayward convergences is offered by Jacques Tati's Play Time (1967). The film's
off-center comedy derives from interweaving several Parisians going about their ordi-
nary business in the course of a day, a night, and the following day. More commonly,
however, the network's plotlines are goal driven to various degrees. The Hong Kong
film Once Upon a Time in Triad Society 2 (1996) brings together three men, each with
a definite purpose. Dinosaur wants to become acknowledged as a tough Triad; Dagger
wants to rise in the Triads without risking his neck; and Dummy, a dutiful cop, wants
to make sure his wife safely gives birth to their baby. All three men's aims are tested
and achieved, or not, during a night when Triad gangs clash in a bloody street fight.
American independent cinema, often presented as an alternative to mainstream
storytelling, hasn't foresworn Hollywood's commitment to goal-driven characters.
Todd Solondz's highly transgressive Happiness (1998) is still square enough to give the
characters goals, albeit fairly lurid ones. A little boy wants to have an orgasm, a pudgy
Mutual Friends and Chronologies of Chance
201
cubicle drone fantasizes violent sex with his neighbor, and a pedophile dad plots to
rape his son's playmate. The title of Crash emphasizes the fortuitous, but the drama
is created by conflicting purposes. Director Paul Haggis points out, "My characters
had real needs. And each character was trying to make it through those 32 hours. One
was trying to save his marriage. Another was trying to save his career. Another was
trying to help his dad." He adds that the film can "keep it human" by being "grounded
in those needs."26
Goal driven or not, characters follow discrete trajectories, often acting in igno-
rance of their counterparts. The intersections of the strangers will provide the drama,
and the network. At the start of the Indian film Yuva (2004) by Mani Ratnam, a hired
killer shoots a student activist on a busy bridge, and a passing young man witnesses
the crime. This scene's role in splicing together their three lives is emphasized by
narrational construction. Before we learn the outcome of the attack, we flash back
to see the chain of events that has led Lallan, the killer, to the bridge. The shooting
is replayed, but Ratnam also fills in bits of information that were skipped over in
the prologue. Then comes a second large-scale section, which traces the events that
led the activist, Michael, to the bridge. After another replay, emphasizing new infor-
mation he has just before the attack, we get a third chunk that explains how Arjun
arrived at the bridge. When the shooting is run a fourth time, its results are finally
shown. Then the film moves forward chronologically, intercutting each young man's
life as it changes in response to the event. In the course of any network narrative,
some strangers will remain oblivious to their counterparts, but others are likely to
gain some awareness of the network to which they belong, and some strangers are
likely to join together intimately and permanently. As viewers, we tend to expect that
at least some vagrant souls will hook up by the conclusion of the film. Thus in the last
section of Yuva, the happy-go-lucky Arjun grows closer to Michael and his political
group, eventually finding fulfillment in running for public office.
If mutual strangers create gaps or weak links, at least initially, the strong ties in
the story world are provided by familiars. These characters are bound together by
long-term ties of kinship, love, friendship, or acquaintance. They belong to the same
family, or share a household, or work in the same business, or just hang out together.
As they follow their more or less independent paths, they will meet new people, and
those in turn may become familiars to the others. By centering on clusters of people
with strong ties, the film aids viewer comprehension. The story world is populated
with people bound by blood, love, desire, or common place of work—just like the
world we know. If the exposition signals these clear-cut affiliations immediately, we
can go on to track the more complex network of links that the film will build up. Of
course, the narration may also conceal the fact that in the story world, two apparent
strangers are in fact familiars; this forms the basis of the narrational surprise in
Heights (2005).
Families have often provided the basis for network narratives, as in Hannah
and Her Sisters, Hou Hsiao-hsien's City of Sadness (1989), Parenthood (1989), and
Happiness. Just as often, the familiars are couples. A great many network narratives,
in debt to literature's multiple-plot tradition in drama, involve troubled romances,
202 Poetics of Cinema
in either comic or melodramatic form. Love Actually (2003), billed as "the ultimate
romantic comedy," presents no fewer than 10 couples finding or failing to find love. As
a genre, romantic comedy relies on serendipitous encounters and repetitive situations,
and network movies take advantage of these contrivances. The prologue of Goldfish
Memory situates several couples in an upscale restaurant. Over several months the
couples break up and reform (in straight or gay mix), and an epilogue brings the new
combinations together in the same restaurant. The theme of romantic inconstancy
is associated with the purported scientific finding that goldfish remember things for
only 3 seconds.
The conventions of melodrama, another genre centered on romantic love, can make
use of network principles in order to shuffle couples together. In Things You Can Tell
Just by Looking at Her (2000), several women's lives are connected by their encounters
with the same trio of men. Cesc Gay's In the City (2003) presents four couples' love
lives as a series of overlapping triangles, spanning four seasons and culminating in
a doleful climax of group embarrassment. Somewhere between comedy and melo-
drama is the plot of Altman's A Wedding (1978), in which two extended families and
assorted hangers-on meet each other on the day of a couple's marriage.
Given a cast of familiars and strangers, groups and loners, what resources are
available to hook them up into a single, cogent structure? What conventions can supply
the sense that we're watching several interconnected stories that belong together? The
filmmaker might try linking protagonists without bringing them face-to-face. An
easy means of doing this is through mass media: Often characters are tuned in to
the same TV or radio show. In fact, a network narrative might be constructed wholly
out of telephone conversations, video conferencing, or e-mail exchanges, making the
characters' adjacency more virtual than physical. Full-fledged examples are hard to
find, but perhaps Anthony Wong's Top Banana Club (1996) comes close. Here a Hong
Kong radio call-in show provides a hub for three stories, each recounted by a caller
(and, weirdly, enacted by the same performers who play the DJs taking the calls).
A circulating object can also link characters at several removes. The device goes back
at least as far as Adventures of a 10-Mark Note (Die Abenteuer eines Zehnmarkscheines;
1926). Tales of Manhattan (1942) passes a coat of tails from owner to owner. In
Winchester 73 (1950), the object is a rifle, and in The Yellow Rolls-Royce (1964) it's
a luxury car. The people who possess the object need never meet, with the object
linking a series of discrete episodes as in The Red Violin (1998). But network nar-
ratives often make sure that the owners interact. The bill in Twenty Bucks (1993)
is tucked in a stripper's G-string, gets bloodied in a robbery, is used to buy bingo
cards, and even passes through the belly of a fish. As its adventures unfold, a groom
loses his bride, two thieves quarrel with fatal results, and a couple who are only one
degree apart finally meet. Through it all, a bag lady keeps turning up in other people's
story lines. Bresson's Au hasard Balthazar (1966) might be considered an unusually
sparse network narrative because its protagonists are as opaque as any in his films
centered on a single intelligence. Still, the donkey Balthazar, passed from family to
miller to smuggler to circus performer, functions both to differentiate plotlines and
to bring them together. Sometimes a circulating object won't determine the film's
Mutual Friends and Chronologies of Chance 203
entire structure but only a part of it, as counterfeit banknotes do in early sections of
Bresson's LArgent (1983).
Usually, though, the primary connections between characters occur more
directly. As in most narratives, the actions in network tales take place within a com-
mon milieu or time scheme. Sometimes the geographical stretch is very large, such
as a country or a region, but in the course of the action, the arena usually narrows
to permit face-to-face interactions. The most popular frameworks are provided by
a city or a neighborhood (Alila, 2003; Tapas, 2005). Single buildings are, of course,
popular spatial unifiers. Kieslowski's Decalogue (1989) sets 10 stories in an apart-
ment complex, and each episode includes one character who has appeared in another
of the series.27 The spatial frame can be confined to a single home, as in The Big Chill
(1983) and Gosford Park (2001), or even one room (Motel Cactus, 1997; Hotel Room,
1998). The time span of the action is typically quite circumscribed, sometimes con-
suming only an hour or so (Joki, 2001), more often a night or a day, or a weekend.
The tight time frame can yield a dramatic pressure, as the close timing of events
becomes a matter of seriousness or comedy. A broader span, such as a year, will often
be measured by seasonal shifts, as In the City and Free Radicals do. A family history,
like that of Claude Lelouche's Toute une vie (1974), asks us to compare generations as
either progress or decline.
So the recurring milieus of network narratives—hotels, apartment houses, cafes,
and the like—put protagonists within hailing distance. Interestingly, though, the
characters need not actually take notice of one another. One of the most satisfying
effects of the network narrative is having characters we've come to know in separate
plotlines brush past one another unawares. Grand Hotel extracts irony from the fact
that lowly Kringellein, spending the last nights of his life in the hotel, doesn't know
that the man beside him at the front desk is trying to change his own life through
love. Benedek Fliegauf's Forest (2003) carries this convention to a kind of limit. The
film begins with a scene in which several people pass through a lobby, oblivious to one
another. Then the film proceeds to show us several episodes, each of which highlights
one or two of the characters we've seen in the prologue. The film concludes with an
exact repetition of the first scene, but now we know more fully what has brought each
character to this crossroad.
A greater coherence is gained by what one screenwriting manual calls an event
frame—a wedding, a reunion, a funeral, a birthday party, a convention, a political
rally, a holiday, or any other occasion that can bring characters together. 28 The web of
relationships in Sunshine State (2002), for example, is revealed across a single weekend
during which the town is celebrating Buccaneer Days. Within the event frame of the
Christmas season, Love Actually takes us to a wedding, holiday parties, and a school
pageant, all sites of intertwining plotlines. Besides several meandering walk-ons, Tati's
Play Time (1967) includes two trajectories. While M. Hulot searches through Paris for
the businessman M. Giffard, a tour group from America is spending a day and a night
there. M. Hulot's sketchy project is eventually fulfilled (though we never learn why he
needed to find Giffard), but much of the film's structure derives from what we know
about tours. You arrive at the airport, you get on a bus, you're sent to a hotel, you're
204 Poetics of Cinema Mutual Friends and Chronologies of Chance 205

driven around landmarks before being given a chance to shop, you go out to dinner
en masse, and so on. All these typical events provide comic pretexts for characters,
major and minor, to crisscross. By restricting the arena of the action, an event frame
can sharpen our expectations about what can happen in the story world.

O n l y C o n n e c t , or at Least Collide

A very small cause that escapes our notice determines a considerable effect that Figure 7.3 Happenstance: She returns his
we cannot fail to see, and then we say that the effect is due to chance. Figure 7.2 Happenstance (2000): When Irene look. They won't meet again until the conclu-
gives her birthdate, Younes in the seat oppo- sion.
Henri Poincare 29 site glances abruptly at her.

Wittingly or unwittingly, sometimes under the umbrella of event frames, characters

converge and separate. Our strangers may meet by gathering at an appointed intersec-
tion, as when several characters head toward the same party (200 Cigarettes, 1999). More
striking and tantalizing is the sheerly accidental encounter. When the characters aren't
all familiars and they don't participate in a causal project, the action is usually triggered
by coincidence. In a plot populated by strangers, contingency replaces causality.
The French film Happenstance (2000) offers a nice anthology of various ways
chance can operate in these movies. As we'd expect, people often meet face-to-face Figure 7.4 Happenstance: Irene and Younes, Figure 7.5 Happenstance: The final shot's
having left the hospital separately, sit outside. symmetry announces the artifice of uniting
in shops, on the street, in the Metro, in cafes, and the like. Nearly as often, though,
the broken-nosed couple.
stray objects hook people up. An illegal immigrant hiding in a truck of vegetables
knocks off a head of lettuce, on which a drunken cyclist skids. A coffeemaker stolen
from a shop is passed from person to person until it's abandoned on a Metro platform,
near-miss. I had thought that Crash (2004), by lifting this conceit to virtual self-parody,
then glimpsed by yet another character from a passing train. A cookie discarded by
would purge it from our screens, but now I believe that it will never die. Not because
one character is nibbled by a pigeon; the pigeon flies away, and its dropping spatters
car crashes figure forth that lust for speed and perceptual overload characteristic of
a snapshot taken by another character. The photographer takes the snapshot into a
modernity, or because they embody that morbid fascination with spectacle character-
shop, where the clerk, wiping the photo clean, recognizes her old boyfriend. Wind,
istic of The Postmodern Moment—in other words, not because of some all-purpose
water, and other elemental forces can be important vehicles for chance in network
explanatory Zeitgeist. The car crash conceit will survive because as a device it snugly
narratives, and in Happenstance they bring together two people who do not speak
suits degrees-of-separation storytelling.
a word to one another in the film. A man gives a pebble to his mistress. She drops
For one thing, traffic accidents are plausible within a story world. We know that
it out the window. It hits a taxi windshield, and when the cab lurches, the passen-
they happen all too often. Moreover, they're the most obvious chance encounter that
ger Irene whacks her nose. Later the young man, Younes, given a yellow slicker by
a well-meaning passerby, is mistaken for the pickpocket who discarded the slicker can have grave consequences. Bump me with your shoulder, and we'll probably move
and winds up with a dislocated nose too. In the first scene, Irene and Younes had sat on and forget about it. Dent my car with yours, and we have to halt to sort things
opposite one another in a Metro car without speaking a word (Figures 7.2-7.3). Now out. Smash into my car, and our lives can change forever. Claude Lelouche's II y a des
in the final scene, they sit back to back outside the hospital, each unaware of the other jours ... et des lunes (1990) juggles several plotlines, some situated in a small town,
(Figure 7.4). The immigrant, having reached his family, has brought some sand from others in an airport, and others on the road. Throughout the bulk of the film, one
the desert, but a stray breeze blows the sand out the window and onto the two young character in particular doesn't directly link to the rest. Gerard is a short-tempered
people below. Turning to wipe their eyes, they recognize each other, each with identi- truck driver who quits his job and steals a sports car. He picks up a runaway bride,
cal nose splints (Figure 7.5). On this wryly symmetrical image of merged destinies, and-the two make their way toward the French Riviera. A 30-car pileup in a tunnel
the film ends. forces the'm to detour through the village, where they run into a traffic snarl created
The most common chance-based convergence, as conventional as a Main Street by an ambulance bearing victims of another car crash. Gerard pulls an evasive
shootout in a Western, is the traffic accident. It seems that a network movie can't do maneuver that makes his car block that of a doctor, who's driving his mistress from
without a traffic jam, smashup, fender-bender, felled pedestrian, or brake-squealing the airport. Each man refuses to give way. In a rage, the doctor stabs Gerard. As he
206 Poet ics of Cinema
collapses, most of the characters we've encountered in other plotlines assemble to
watch him die.
The gridlock in II y a des jours . . . is motivated within the story world—nearly all
the characters are motoring somewhere—but it has an architectonic function too,
tying together many story strands. 30 A traffic accident can be a structural climax, as
in Lelouche's film, or it can launch different story lines, as in Crash. A traffic mishap
yields arresting narrational possibilities as well. Although it may serve as a culmina-
tion of the fabula progression, it can be put anywhere in the syuzhet. Happy Endings
(2005) starts with a distraught young woman running down a hillside, only to be
struck by a car. There appears an intertitle promising, "She's not dead, and we flash
back 20 years. The rest of the film will bring us up to date on how she and several
strangers and familiars arrived at this point.
Likewise, the horrific collision that opens Alejandro Gonzalez Inarritu's Amores
perros (2000) is justified within the story world, because a pileup isn't unexpected in
populous Mexico City. Structurally, the crash climaxes one story, initiates another,
and has a more oblique tie to a third. To top things off, the narration reemphasizes
the accident's centrality by overt manipulations of time and character perspective.
After the first scene shows the car crash, Amores Perros flashes back to trace the
events that have sent Octavio and his pal hurtling through the intersection. Then
the crash is shown again, but this time the narration follows the story of the fashion
model Valeria, whose career is shattered by her injuries in the crash. She becomes
housebound and morose, and when her life seems to have reached stasis, the narra-
tion replays the smash-up once more. This time we follow the trajectory of El Chivo,
a passing street bum who rescues a dog from Octavio's car. Our impression of his
kindness is sharply altered when he is revealed to be a hired killer.
The film's narration is constantly whetting our interest. By presenting the car crash
first, it invites us to speculate on what led up to it. Obligingly, the first story shows us
the preconditions for the accident. But there's the danger that once we know how we
got to the opening burst of action, all further interest will slacken.31 So the narration
of the lead-up to the crash, titled "Octavio and Susana," is interrupted by brief scenes
introducing characters in the other two lines of action. Glimpses of them provoke us to
wonder how they relate to the crash. Likewise, the second tale, Daniel and Valeria, is
interrupted by intriguing vignettes showing El Chivo's surveillance of a young woman.
These will be explained only in the third stretch, "El Chivo and Marie. In order to
resolve the uncertainties set up in the first episode, his assassination scheme is intercut
with scenes showing the outcome of the Octavio-Susana plot. There's also an echo of
the second strand when near the close of the film, El Chivo passes a billboard banner
featuring Valeria. By means of flashbacks, crosscutting, and recurring motifs (dogs,
the banner), the central car crash arouses keen curiosity and suspense.
Of course, it's possible to treat a car crash as the pure convention that it is. The
Japanese director Sabu has made something of a career out of high-energy chases
that snowball through comic coincidence. His Hard Luck Hero (2003) ingeniously
crams a host of network conventions into 77 minutes. Six young men get embroiled
in a kick-boxing bout that leads to a shooting and a car chase. Sabu draws on the
Mutual Friends and Chronologies of Chance 207
back-and-fill time structure used by Gonzalez Inarritu, but exploits it for comic
surprises. As each pair of boys enters the chase from out of nowhere, the narration
flashes back to show us how each became involved in the overall plot. The flashbacks
unexpectedly reveal new aspects of the prizefight and shootout. Sometimes the
flashbacks interrupt themselves to carry us still further back in time. The inevitable
climactic three-way smash-up—shown twice, with two alternative outcomes—
emerges as a self-conscious acknowledgment that it's the only way to close an exercise
aiming to amplify cliches to the bursting point.
Traffic accidents are just the most extreme instance of dramatic but contingent
nodes, the unlikely points of connection within the film's network. A convergence
must be made salient for us, and a car collision is a vivid way to do that. But what-
ever the convergence, the syuzhefs overriding geometry must tame chance; even
a car crash is subsumed to a larger design. As Amores Perros and Hard Luck Hero
show us, the task of narration is to take us through this developing structure in a
way that yields a particular experience of the whole. The entire network is assembled
gradually, not only through the cascade of events and actions in the story world but
also through cinema's narrational resources. As ever, our curiosity, suspense, and
surprise are created by manipulating what we know and when we know it.
N a r r a t i o n Makes N e t w o r k s
Any film's narration, in coaxing us to build the story world a particular way, must
expose the relationships among the characters. In a network tale, the narration must
do this with an elaborateness seldom seen in the more ordinary movie. The narration
must reveal connections, anticipate connections, and conceal connections.
Most evidently, the film's narration must show how characters are separated by
a few links. The pigeon in Happenstance, nibbling one character's cookie and flying
off to shit on another character's photo, exists wholly to hook up characters. The
narration must also prime us to expect connections, thereby generating suspense.
When in Mind the Gap (2004), a man and a woman who don't know one another
both decide to go to a speed-dating party, we expect that they'll finally meet there.
(Interestingly, they don't.) Likewise, by bringing characters by chance into the same
locale, as often occurs in Happenstance, the narration teases us into wondering
whether they'll actually get acquainted.
The film can reveal and anticipate connections by employing unrestricted narra-
tion, skipping back and forth among people and places. The technique of crosscutting
among strangers is particularly useful. It can make their eventual encounters seem
less coincidental, for they've been connected for us, if not for each other, from the
start. Initially Edward Yang's Ihe Terrorizers (1986) binds three couples through
sheer crosscutting and spatial proximity. Accidents and plans eventually link the
couples, but crosscutting and spatial overlap have primed us to expect convergence
from the start. '
Most narratives strive to bring together their principals at the climax. From
the Odyssey to the final shootout in a Western, the ultimate confrontation tends to
208 Poetics of Cinema
assemble the protagonist, the antagonist, and surviving helpers and blocking figures.
Many network films do the same thing. But even when characters don t finally con-
verge in a single space, they may be brought together at the level of the narration.
Crosscutting can assure us that the widely separated resolutions of the characters
fates are somehow linked. Even plots that present only a single point of intersection
and trace each cluster of characters in distinct sections often resort to parallel editing
for the climax, as Amoves Perros does. The narration assembles the characters before
us one last time, even if the unfolding action doesn't. 200 Cigarettes cleverly deletes
the big party toward which all the plotlines run, but through crosscutting it shows
us who wakes up with whom on the morning after. During the credits, a series of
souvenir snapshots provides glimpses of the party we never saw.
Within the wide range of knowledge favored by network narration, more restricted
knowledge can be used quite shrewdly. Things You Can Tell Just by Looking at Her
(2000) is broken into sections centering on each of its main characters (headed by
chapter titles like "This Is Dr. Keener" and "Fantasies About Rebecca"). Each episode
brings one woman to center stage and concentrates on her range of knowledge, but
she will be a walk-on in an episode concentrating on another protagonist. In the City
centers on four couples, but the narration favors only one member of each, so our
curiosity and suspense bear on what each knows or doesn't know about the partner.
Whereas Things You Can Tell... sets each character's story side by side in large chunks,
In the City shuttles us frequently from one character's love life to another's. In either
case, the need to shift among several characters' limited knowledge leads naturally
toward the play of perspectives that often accompanies a network narrative.
The Terrorizer, In the City, and Things You Can Tell... are pretty reticent about the
characters' pasts, but other networks harbor secrets. Many tales therefore provide a
continuous exposition about the backstory. The movie advances on two fronts. As the
action unfolds, more characters become connected. As the narration unfolds, it not
only exhibits the newly forming connections but also illuminates prior connections
among the characters.
Thomas, a pet shop owner who smuggles rare eggs into Canada, is the first character
we encounter in Exotica (1994), but he's only remotely connected to the main plot. That
action centers on the nightclub Exotica, where several characters with shady pasts are
bound together. Gradually the middle-aged auditor Francis is drawn into the world
of Exotica, but we don't know his exact relation to Christina, the lovely young woman
with whom he spends time. Summoned to audit Thomas' books, Francis discovers his
smuggling scheme and blackmails Thomas into helping him kill the sadistic DJ Eric.
Not until the last scene do we learn why Francis hires Christina to go out with him.
Exotica s narration has given with one hand and taken away with the other. We're
introduced to Thomas long before he becomes necessary for Francis' murder plan, but
the narration has withheld from us crucial information about Francis' family and the
role Christina plays in his life.
So the narration may well hide links that exist in the story world, at least for a
while. In Premiere juillet, le film (2004), the event frame is provided by the fact that
July 1 is Montreal's traditional day to move into a new home. We are introduced to a
Mutual Friends and Chronologies of Chance 209
bevy of renters who need to leave before a new tenant arrives, or to find the very rare
apartment that hasn't been taken. But as the film unfolds we discover that nearly every
couple or family moves into a space vacated by other characters we've met. Network
films now routinely suppress certain information about the characters' relationships
for some time. By this point in the history of the form, we might expect that viewers
come to the movie ready to guess at who might already know whom.
Suppressive exposition is supported by the convention of shifting restrictiveness.
Once we get accustomed to being attached to one character and then another, we
accept that we're likely to learn only what each character knows. Watching Hard Luck
Hero, we eventually find out that all three pairs of young protagonists were present at
the fateful kick-boxing bout. But the first rendition of the scene concentrated on one
pair, who didn't know about the others, and so those weren't shown to us. Later a con-
densed flashback of the bout shows us the gunfight from the standpoint of a second
couple, unknown to the first pair. A third flashback shows how the third couple,
previously ignored by the two others, shapes the overall action. As often happens, the
shifts in point of view aren't there only to render certain connections more vivid; they
also conceal other connections.
Shifting our access among protagonists may oblige the narration to juggle time.
The flashbacks in Amores Perros, Hard Luck Hero, Happy Endings, and other network
narratives spring out of the switches from one limited perspective to another. When
the narration organizes the restricted section in large blocks, we can expect such a
reordering. Probably the most influential model is that on display in Pulp Fiction,
which famously creates three slightly overlapping plotlines, presented out of order
and from different attached points of view. Tarantino secures our understanding by
replaying bits of certain events—Vincent and Jules' murder of the college kids, and
the tail end of the robbers' love dialogue in the diner booth. These returns to previous
scenes, as in Amores Perros, Hard Luck Hero, and many other network narratives, cast
an event in a new light by revealing aspects of it we couldn't see before.
Because of its unrestricted narration, Pulp Fiction presents its flashbacks as frankly
unmotivated by anything but the simple decision to tell the story this way. In tra-
ditional Hollywood cinema, flashbacks are justified as character memories (even if
what happens doesn't always correspond to what the character could have known;
see p. 125). The change of story order is justified, at least at the start and close of the
flashback, as proceeding from a character's act of recalling or telling what happened
in the past. The rearranged event chains in Pulp Fiction, however, signal a very overt
narration, self-consciously addressing the viewer.
This overtness is a mark of network narratives generally. Consider for example
their reliance on intertitles. For one thing, titles aid comprehension, helping us to
chunk the movie into digestible parts. How easily could we follow Pulp Fiction or
Mind the Gap or Things You Can Tell Just by Looking at Her without those teasing but
cogent chapter, titles? At the same time, they serve as a self-conscious address to the
viewer. From Hannah and Her Sisters (1985) to Heights, chapter titles acknowledge
that the narration is guiding us, and we're expected to acknowledge the same thing—
that it is an artifact designed for us to respond to. The intertitles of Happy Endings,
 Poetics of Cinema
210
Nicky never ties.
He is not her son, if that's
what you're thinking.
But he does know him.
Figure 7.6 An external narrational commen-
tary; in Happy
r r / Endings
6 (2005).
Figure 7.7 Happenstance: To-camera address
motivated as a frontal shot/reverse shot.
sliding into the frame and taunting us with their superior knowledge, call attention to
the narration as well as to the events (Figure 7.6). "I like to tell the audience that this
is a story," says director Don Roos.32
Such self-conscious narration isn't a breakthrough. Hollywood movies often employ
it, and cinema since the 1960s has welcomed bursts of explicit artifice. Acknowledg-
ment of the audience takes on a particular saliency in network narratives, though,
because such tales already depend on overt narration at other levels. Titles oblige
us to ask, at certain moments, why we're being told something in a particular way.
The same question is posed by other tactics, such as the old device of glances to the
camera, which may be a one-off effect (Figure 7.7) or a systematic suite of variations,
as we'll see in Les Passagers (1999).
We're particularly aware of artifice when the narration openly suppresses informa-
tion we'd normally be given. The funeral scene opening the Finnish film Frozen Land
(2005) neglects to show us who's being buried. This creates curiosity about the story
world as we meet several characters in flashback who are candidates for the coffin, but
it also signals the overt narration conventional in art cinema. A film's narration can
also proceed with more stealth. Instead of providing us with a narrational mystery at
the outset and spurring our curiosity and suspense, a film can pretend that nothing
is amiss and lead us to think that we have full information. Later, when we recognize
that the narration has pulled a fast one, it becomes more overt.
The Indian American comedy Flavors (2003) begins with a wedding. As in Frozen
Land, we aren't shown a crucial component of the ceremony: in this case, the people
who are getting married. So far, so overt. Suddenly a cell phone rings, and several
young men grope for their mobiles. After one runs out of the church to take the call,
we flash back a few weeks. During the film's first half, we follow the trajectories of
several characters. Some of the men we've seen at the wedding are shown working in
the information technology industry We see others of the group involved in romantic
entanglements. But how they all connect remains unsettled.
Only after the halfway point does Flavors' narration gradually reveal that all the
men work for the same company and so know each other. Now, too, the narration
spells out who is dating or married to whom. The narration continues to dole out
basic background data until just before the wedding service starts. At the climax the
scene of the interrupting call is repeated, and the results of that call resolve the last
romantic plotline. The tying up of the plot coincides with the sort of full exposition of
Mutual Friends and Chronologies of Chance 211
character relationships that most films would provide at the start. Flavors uses all the
sell-conscious devices we've come to expect—chatty intertitles, crosscutting, and mon-
tages accompanied by a song ("It's all about connection")—but at certain moments, its
narrational manipulations are far from overt. Only in the second half of the film do we
realize how many links among the characters have been hidden from us.
Through intertitles, time juggling, or openly suppressive narration, network movies
gravitate toward a self-conscious address to the audience. The first scene of Lawrence
Wong's Cross-Harbour Tunnel (1999) introduces all its characters at the same bar,
drinking alone or in pairs. Suddenly the bartender closes the joint, and all the char-
acters leap into a frenzied, albeit awkward dance. This prologue is promptly forgotten
as we follow four successive stories involving the characters we've seen. Most tales are
linked by spatial overlap, but the fourth develops out of a replay of the first from a minor
character's standpoint. At the end of the fourth tale, in another display of communal
silliness, the entire cast launches into a cheerleading routine, complete with pom-poms.
Quarrels, enmities, and wounds are forgotten. Bouncing along the underground tunnel
they've passed through in various episodes, they beckon to us. The stylized prologue
and epilogue of Cross-Harbour Tunnel carry to a limit the frank acknowledgment of
artifice characteristic of network narratives. It's as if these characters know that their
fraught encounters have existed solely to create a beguiling mosaic.
Compare and Contrast
Like all formal options that have become conventional, the network narrative has
gravitated toward certain subjects and themes. But these materials aren't merely lumps
of content to be chopped up and jammed into the template. The film's formal process
develops given subjects and topics; it transmutes them into narrative and audiovisual
patterns. As viewers we grasp themes and subjects through our engagement with the
film's form. However familiar, even cliched, the materials might seem when taken out
of context, they become dynamized through their manifestations in the story world,
the plot structure, and the narration.
Central to our engagement with these films is our sense that characters, situations,
and activities tend to parallel one another. Criss-crossers are comparative. It's not just
that A knows B and may meet C. We are expected to register, if not consciously note,
that A, B, and C are held up as alternatives along some scale of judgment.
In a more causally driven film, such parallels are subordinate to the goal-driven
primary action. 33 In watching a network narrative, however, we're often coaxed to
notice how characters are sharply similar or different from one another.
Take Sylvia Chang's 20 30 40 (2004). It centers on three women who represent
the age ranges given in the title. Jie is a teenage Malaysian who comes to Taipei to
become a pop singer. Xiang is a flight attendant juggling unsatisfying love affairs. The
fortysomething Lily discovers that her husband has a mistress and son, divorces him,
and tries to find a new partner. If this were a more traditional movie, their stories
would depend on one another. Once separated from her husband, Lily might take the
other two into her large apartment, and then all three would help each other pursue
212 Poetics of Cinema
their goals (as in working-girl-roommate movies like How to Marry a Millionaire,
1953). The parallels would still be apparent, but our hypothetical plot's overall thrust
would come from the way each woman's life choices shape the others'. In 20 30 40,
however, after a prologue shows all three arriving at the airport, they go their separate
ways. Xiang and Lily are connected, but at several removes. (Lily starts dating a man
whose mistress is a friend of Xiang's.) There are a few spatial overlaps too, because the
women live in the same neighborhood. At one point Jie and her friend Yi are shown
in the background of a shot in which Lily drives to her flower shop. Across the film,
however, the women's lives run in parallel, and they never meet.
The effect of this construction is to emphasize not the women's influence on one
another but their independent decisions. Jie, lured by the promise of becoming a pop
star, learns that the world is more sordid than she'd realized, but before she goes home
she forges a tender, slightly erotic friendship with Yi, the girl hired as her singing
partner. Xiang, in search of wealth and excitement, comes to accept life with a humble
widower and his daughter. Lily doesn't find a partner but she struggles on, resilient
and good-humored. Like many network narratives, 20 30 40 engages the viewer by
asking that the protagonists be contrasted—in this case, according to the chances and
choices offered to women of different ages.
In most narratives, as causality slackens, parallelism fills the gap. By asking the
viewer to notice likenesses and differences among characters, network films are drawn
to certain traditional themes that depend on parallels. Perhaps most conventional of
all is the theme of contrasting romantic couples—different styles of loving, we might
say. In Yuva, the three men create a didactic counterpoint. The killer Lallan abuses
and ignores his wife, and she can save herself only by leaving him. By contrast, the
activist Michael enjoys a healthy equality with his lover Radhika, and the maturing
Arjun exercises a sobering influence on Mira the party girl. Altman's Short Cuts pres-
ents no fewer than eight male-female couples (throwing in a mother-daughter pair
for good measure). The couples' lives are shot through with pain, deceit, indifference,
and artificiality. It's hard not to see the film as a catalogue of ways that love can fail.
From very early in the history of cinema, films like The Kleptomaniac (1905) used
parallelism to draw class contrasts starkly, and What's Wrong With This Picture?
shows that the rich-poor duality remains a central theme. John Sayles' Sunshine
State portrays a moment of community crisis. While land speculators move in on
an undeveloped beachfront, real estate maneuvers reshape class relations. The film
introduces local residents of a traditionally black area, small-business owners,
corporate reps, a landscape architect, truckers and carpenters whose jobs depend on
development, a local politician succumbing to bribes, and a complacent chorus of
affluent senior citizen golfers. Besides bringing out racial and regional differences,
Sayles' network constitutes a cross-section of the town's economic strata.
The form is also suited to the theme we might call Great Events and Ordinary
People. Often an event frame will highlight the disparities between important cultural
rites and obscure private lives. The pope is visiting Berlin while people struggle to
survive and connect in Night Shapes (1999), but His Holiness offers little assistance.
Against the background of a holiday evoking national or religious traditions, the
Ivlutual Friends and Chronologies of Chance 213
film's action can remind us of daily travails. The ersatz historical celebration in
Sunshine State suggests that a bland ritual distracts the community from more press-
ing problems. More scathingly, Michael Haneke is fond of juxtaposing his characters'
lives with news footage showing wars and famines ravaging the wider world.
City life is a natural terrain for network narratives, as Balzac, Dickens, and Zola
realized long ago. Many films seem to assume that with the dissolution of traditional
rural communities, cities created aggregates of isolated individuals, each pursuing
his or her own aims. Accordingly, themes associated with urban solitude, unfulfilled
promise, the lonely crowd, and casual encounters in alien spaces will come to the fore.
13 Conversations About One Thing (2001) suggests that the city dweller must find con-
solation in a fleeting wave from a stranger in a passing subway train. Crash subjects
its characters to the wounds of racial division in the modern city. Often, though, the
city-based plot aims at overcoming urban alienation. Grand Canyon (1991) is pre-
mised on a raw urban fear—a white man marooned in a black neighborhood—but the
characters aim to reconcile race and class through respect and affection.
Elsewhere, the New Europe has become a recurring subject, and along with it
has come the theme of overcoming national barriers. Beautiful People is a typical
instance. The marriage between a wealthy nonconformist Brit and a charming immi-
grant is paralleled by a welter of whimsical or violent cross-national encounters in
London and the former Yugoslavia. Britain's own civil strife, for instance, is recalled
in a hospital room, where alongside a ceaselessly quarreling Bosnian and Serb lies a
dour Irishman who begrudgingly comes to accept both. The same sort of cautiously
affirmative treatment of the new continent is found in Happenstance (2000) and
Crash Test Dummies (2005). Haneke, predictably enough, offers a less upbeat vision
in Code Inconnu (2000). Here an illegal immigrant from the Bosnian civil war is beg-
ging on a Parisian street when a young man, Jean, contemptuously tosses a crumpled
bag into her lap. This enrages a young French African, Amadou, who demands that
Jean apologize. The altercation becomes a police affair, and the beggar woman, Maria,
is investigated and sent back home. The bulk of the film follows the lives of Jean's,
Amadou's, and Maria's families. The characters don't encounter one another again,
and in tracing the effects of this confrontation the film offers little hope for bringing
Europe's nativist and immigrant strains together.
Because unplanned encounters pull strands of action into a coherent whole, it's
not surprising to find that many network films thematically counterpose accident to
destiny. Multiple love stories are often built on the prospect that the couple who are
just right for each other will converge through a happy accident. Often characters
in the film state the theme explicitly. "There's luck and there's fate," declares the bag
lady in Twenty Bucks. In Bug (2002), one character insists that he's a Calvinist, but the
rest of the film suggests that predestination can't account for sheer coincidence. In
Frozen Land, one teacher explains to the class that events have an ABC linearity; later,
another teacher.assures them that everything is chaos.
The tension between fate and chance is so frequent in these films that we might
speculate that the idea holds a special appeal for audiences. Typically the film's form
shows that however powerful chance is, things sort themselves out satisfyingly at the
214 Poetics of Cinema
level of the whole film—the mosaic principle again. Along with the aesthetic pleasure
of seeing unconnected events fall into a pattern, many viewers may feel reassured that
Chance is just God's way of seeming anonymous. A social psychologist has suggested
that many people find the idea of "six degrees of separation" comforting because it
can be interpreted as a mysterious design, the sign of some spiritual order guiding our
lives.34 In this light, the rash of recent network movies from U.S. independents would
seem to be promoting a secular theology for bohemian kids in black.
But design need not conquer chance. At first Jacques Rivette's Paris nous appartient
(1960) suggests a vague but deadly purpose governing haphazard events. The young
woman Anne enters a network of intellectuals who are disturbed by the mysterious
death of a mutual friend, Juan. Apart from discovering obscure connections among
Juan's friends, Anne learns that the woman living next door to her was Juan's lover.
That woman and the American emigre Philip are convinced that Juan was killed by
a mysterious conspiracy, "the greatest conspiracy of all. It's on a worldwide scale....
They're ready to recapture power." Anne falls in love with Gerard, who is staging a
catch-as-catch-can production of Pericles. He grants that the play is convoluted and
puzzling but "it all ties in on another plane," and Anne agrees that "the world is less
absurd than it seems." Later events don't wholly support her view. Another woman,
Terry, assures Anne that the global conspiracy is purely a product of Philip's imagina-
tion, but at the very end, an ambiguous murder scene leaves things quite uncertain.
Gerard has said that Pericles "seems to fly off wildly, but it knows where it's going.
It just doesn't let us know." This seems a good epigraph for Rivette's inconclusive
ending, which refuses to choose between coincidence and design.
Of course, films that aren't network narratives can also examine different styles of
loving, disparities between rich and poor, contrasts between Great Events and ordi-
nary life, problems of urban anomie and national differences, and puzzles of chance
and fate. But these topoi are well suited to a form built out of degrees of separation
and converging characters. Filmmakers interested in the network tendency have
gravitated toward themes that give significance to the form's characteristic organiza-
tion and effects.
Criss-Crossers Cross O v e r
Assuming that a concept like "network narrative" is best understood as a core-periph-
ery one, I've confined myself mostly to examining clear-cut instances. But there are
borderline cases. Take Changing Lanes (2002). It has only two protagonists, but they
are pursuing distinct projects until a traffic sideswipe brings them together. The
accident has grave consequences for both men, and the bulk of the film shows each
one trying to achieve his goals but also interfering with the other out of an impulse
toward payback. Were we to add some more protagonists and make the plot moves
less tit-for-tat, we would have something closer to the core cases I've isolated. Still,
by relying on the traffic accident convention and by treating the two men's actions
through intercutting, the film exudes a little of the flavor of a network tale.
Mutual Friends and Chronologies of Chance 215
Like the traffic accident, the overlapping replay is a marked narrational ploy that can
move a film to the borders of the network mode. The money transfer in Jackie Brown
(1996) has this quality, especially because it stresses different points of view. More
elaborately, Go (1999) and Rave Fever (1999) echo the network approach by reiterating
events seen from different characters' angles and by expanding the roles assigned to
bit players. Rave Fever also uses a circulating filofax to connect its party animals.
Another borderline case, I think, is Slacker (1991). It opens with a lengthy dis-
quisition on parallel universes, followed by a sudden traffic accident. The situation,
a woman run down by her son, promises a complex chain of causation, but this
intriguing line of action aborts itself almost immediately. The film starts to unfold an
open-ended plot based on chance encounters, ephemeral conversations, and spatial
overlaps. We follow A until A comes into contact with B, whom we pick up for a while
until C enters, and so on. These don't create the sense of an expanding circuit of con-
nections. The story action doesn't bring its characters face-to-face, and the narration
doesn't lay bare unexpected connections among them. This isn't a network so much as
a wiggly, knotted string. If Slackers parade of minor perplexities and sidewalk rants
gives it a bit of the network flavor, perhaps it's partly because these motormouths are
locked into a geometrical pattern of which they are unaware. (If they knew, it might
feed their paranoid appetites.)
Certain formal choices can give a fairly conventional plot premise the air of a
network narrative. The prologue of The New Age of Living Together (Hong Kong,
1994) introduces three roommates, Eddie, Steven, and Coco. The film will explore
each one's love life. But instead of providing a plot like Three Coins in the Fountain,
in which the friends' romances intertwine, this film tries something jazzier. After a
prologue introducing all three, a long stretch of the film is devoted wholly to Eddie's
search for a girlfriend. Then a second episode takes up Steven's romance with an
older woman. The last episode focuses on Coco's romance with the yuppie Keung.
At first this section seems as self-contained as the others, but when Coco discovers
she's pregnant, Eddie and Steven take an active role (along with Coco's gay coworker).
The climax, in the delivery room, brings everyone together. The episodic segregation
in The New Age of Living Together gives an old-fashioned story template more of a
network narrative cast.
Network tactics can revivify genres. Crime novels have offered many opportuni-
ties for formal experimentation, and the same has held good in film. From the days
of film noir (e.g., the lying flashbacks of Crossfire, 1947) to our time (the reverse con-
struction of Memento, 2000), stories of mystery have been hospitable to plays with
temporal order and point of view.35 It's not surprising, then, that suspense films have
drawn upon network conventions. Claude Miller's Betty Fisher et autres histories
(2001), derived from a tightly plotted thriller by Ruth Rendell, opens the original
structure out through degrees-of-separation techniques. It provides chapter breaks
labeled as different characters' "stories," and it shifts among several fairly indepen-
dent lines of action and character attachments. It's basically Betty's movie, but as
the title indicates, some of the grifters and lowlifes whom we get to know are only
remotely connected to her.
216 Poetics of Cinema
A happy family is shattered when the father and two daughters are killed by a
hit-and-run driver. In an abyss of despair, the mother turns to drugs before another
man offers to help her find the driver. The two set out to get revenge. From this
pulp premise, Alejandro Gonzalez Inarritu constructs something approximating a
network film. 21 Grams (2003) shuttles among the wife, Chris; the helper, Paul; and
the driver, Jack. A fragmentary narration flashes to and fro across many stages of
the action, keeping the characters apart for long stretches and building up a sense
that three independent fates are merging. 36 Peripheral characters, such as the wives
of Jack and Paul, are fleshed out more than a straight thriller would. Paul enters the
action not as a witness but as beneficiary of the husband's transplanted heart. Then
there's that traffic accident, along with metaphysical speculations about chance and
necessity. It would be too harsh to say that Betty Fisher and 21 Grams use devices
of the network format to camouflage banal suspense plots, but clearly certain genre
situations can, with only a little tweaking, evoke a network.
Or take the Australian melodrama Look Both Ways (2005). A fatal train acci-
dent is witnessed by a woman artist. The plot structure and screen duration tend to
weight the romance that arises between her and a news photographer, a man who
awaits a doctor's verdict on whether he has cancer. Yet the accident victim's wife, the
photographer's editor, another reporter on the newspaper, and the engineer who was
driving the train all get somewhat more attention than they would in a more tradi-
tionally focused drama. These secondary characters are given scenes that trace how
the accident alters their lives. They don't become full-fledged protagonists, but they
do emerge as extended parallels to the main couple.
If the secondary characters' lives had a little more causal propulsion and more
screen time, and a few more links to one another, Look Both Ways would resemble Me
and You and Everyone We Know (2005). The latter film shows how flexible our sense
of a network movie should be, and how our criteria for determining the protagonist
can pull against one another. The title suggests equality among many protagonists.
The director, Miranda July, has said, "I was under pressure to cut down some of the
other characters. But I never thought of the movie as a love story. I wanted it to be
kaleidoscopic."37 The characters move on discrete tracks: an elderly woman dying,
adolescent girls exploring sex, a little girl collecting linen and appliances for her
future life as a wife. Their lives interlock through acquaintance. The teenage girls
have sex with a boy who is a friend of the little homemaker-to-be. She in turn buys
new shoes from the boy's father.
Yet despite July's claim, structure and narration focus her film on one romantic
couple. Christine is an aspiring artist, and Richard an awkward but pure-hearted shoe
salesman. Although most characters have ephemeral or vague purposes, Christine
tries to get to know Richard, and she submits her mixed-media artwork to a gallery.
These goals propel much of the film's action. Christine and Richard link all the char-
acters, and the biggest secret revealed at the end is an unexpected convergence, via
degrees of separation, between someone in her life and someone in his. Narrationally,
the film primes us by giving the first sequence to Christine and the second to Richard,
with her voice-over accompanying shots of him. The third and fourth sequences
Mutual Friends and Chronologies of Chance 217
attach us to Christine, the fifth attaches us to Richard, and the sixth shows them
meeting in the shoe store. The first 17 minutes of the film prompt us to link every new
character to the couple that seems to be forming. Christine and Richard remain at
center stage thereafter, occupying about two-thirds of the film's running time. 38 Star
power of a sort is a factor too, with Christine played by the director Miranda July, who
is also a notable performance artist. In all, there's reason to read the title differently
than I initially indicated. Me is the primary protagonist, a woman who makes auto-
biographical art. You are the man I see and want. Everyone we know is, well, our
networks, filled out more than in the usual movie romance.
Such mixed cases help me make the case for poetics. As network construction
becomes more common in our movies, we can feel new conventions bubbling to the
surface. If we want to analyze our intuitions, we can start by examining certain recur-
ring features in the films and in our experiences of them. So the poetician tries to
make explicit the principles that might underlie the regularities we detect. By using
our analytical tools, such as the roles of causality and convergence or the dimensions
along which protagonists can be distinguished, we can point out some conventions
guiding the regularities. Like every norm, though, this one is flexible. Our sense of
the norm is strongest in cases when we encounter films that exhibit several network
traits or highly salient ones. Other films furnish mixed cases (and aren't necessarily
better or worse for that). Some simply replicate the conventions, and others, like Betty
Fisher, 21 Grams, Look Both Ways, and Me and You and Everyone We Know, blend
them with tactics characteristic of thrillers, comedies, and dramas.
And some films exploit the resources of the category more fully or freshly than
others do. For instance, a film might slide from a protagonist-based structure to a
network one gradually, as Barbara Albert's Free Radicals (2003) does. The woman
Manu starts as the protagonist, and the first 20 minutes show her escaping a plane crash
and taking up a normal life with her husband and daughter. But quite unexpectedly
she's killed in a traffic accident, and the stories of her family and friends take over.
Manu becomes an absent center, abruptly withdrawn from the tissue of relations that
have oriented us.
Then there are the films that seem to fracture norms altogether. Across his career,
Jean Luc-Godard has created collages of film forms and techniques. His early films
casually appropriate conventions from the war film (Les Carabiniers, 1963), the musical
(Une femme est unefemme, 1961), and film noir (A bout de souffle, 1960; Alphaville,
1965; Made in USA, 1966). So it's not surprising that several of his later films snatch
up traits of the network narrative. One title in For Ever Mozart (1996) can translate as
36 Characters in Search of a Story. More thoroughly, in Sauve quipeut (la vie) (1980), a
plot centered on three characters acquires the texture of a network tale. Paul Godard
has left his wife and daughter to have an affair with Denise, who is moving to the
countryside. Their affair is interwoven with events centered on the prostitute Isabelle,
who meets them at intervals by chance. At the climax Paul is struck by a car driven by
a client of Isabelle's sister, who has just turned to prostitution. The random encoun-
ters, the parallels among characters who are all in the process of changing their lives,
the car accident, the impersonality of the city, some vignettes of other couples who
 Poetics of Cinema
218
might become important in the plot (but don t), and the narration that periodically
crosscuts among the principals—all these devices conjure up network construction
rather than a linear drama.
Godard's Detective (1985), set in and around a luxury hotel, employs the confined-
arena convention to intertwine the financial problems of a boxing manager with the
fading marriage of an airline pilot and his wife. In a parallel plot, the hotel detective
sets out to solve a mysterious death that occurred years before. The plots intersect, by
sheer accident, in a burst of violence. Even more tenuously, the two plotlines of Je vous
salue Marie (Hail Mary, 1985) merely take place in the same small city; their connec-
tion is chiefly thematic. Marie, the virgin girl who finds herself pregnant, reenacts the
New Testament story, whereas Eve and her lover, a professor who believes that aliens
guided the emergence of life on earth, evoke the Garden of Eden. The only action-
based link between the couples comes when the professor and Eve take a cab driven
by Marie's boyfriend, Joseph.
It's always misleading to write about Godard films as if the characters' identities
and relationships were transparent. Often we figure out the story's givens quite late,
or never. Soigne ta droite (1987) presents a central tale of Godard himself as the
Idiot, making a film to be released on the evening of the day he starts it. But most
scenes show recurring characters interacting in mystifying ways—splashing water,
crawling on tabletops, and posing for an offscreen photographer—yielding a sort of
absurdist version of Play Time. The film gestures toward a network without confirm-
ing its existence. By refusing the most elementary exposition, Godard's narration can
tease us with the possibility that the characters are linked, but we're hard-pressed
to say how. The mosaic before us is cracked and crumbling. Reliable story informa-
tion is replaced by snippets of soaring music and glimpses of landscapes, clouds, and
reflections in water. Godard's late films yank apart all manner of storytelling tradi-
tions, including principles of network plotting, and then fill the gaps with stuttering
bursts of pictorial and auditory lyricism.
Four Small Worlds
In teasing out the principles underpinning network narratives, I've downplayed their
variety, and I haven't charted at a finer grain the way these principles can be instanti-
ated in a single film; So I finish this survey by analyzing four important films in the
tradition. These mini-essays will try to show that network conventions permit innova-
tion at several levels. They'll also suggest that innovation in this format depends on
the filmmakers' commitment to wider traditions. Directors working with a classical
American conception of storytelling tend to explore the network form in ways different
from filmmakers working in what an earlier essay in this collection calls art cinema.
The classical narrative system, forged in the late 1910s and continuing to the
present, has exercised great power over both mainstream and "independent" cinema.
In the mainstream, the model has persisted throughout many changes right up to
the present. As for U.S. independent film, rather than challenging the classical tradi-
tion, it has developed deeper or "quirkier" (the favorite word) variants of tendencies
Mutual Friends and Chronologies of Chance 219
already seen in limited form during the studio years. So some films develop a more
"character-driven" version of classical plotting, as seen in melodramas like sex, lies, and
videotape (1989) and In the Bedroom (2001) and romantic comedies like The Opposite
of Sex (1998) and Garden State (2004). Other indie films experiment with subjectivity,
flashbacks, unreliable narration, and other formal strategies that became prominent
in Hollywood films of the 1940s and 1950s. Neo-noirs like The Usual Suspects (1995)
and Memento exemplify this tendency. Directors are driven to differentiate their low-
budget product through offbeat subjects or formal gambits, thus getting attention at
film festivals and in the press. But even the quirkiest indie movie tends to obey many
precepts of Hollywood plot structure and narration. 39
I've already said enough to indicate that many features of classical construction
find their way into degrees-of-separation narratives, whether made in Hollywood
or other countries. Character goals and conflicts, deadlines and appointments, and
other unifying devices can structure the intersecting plotlines. The Hollywood three-
act (actually, four-part) plot structure (pp. 104-110) can be fitted to a network tale.40
Richard Curtis says that Love Actually tried to create "ten good beginnings, ten good
middles, and ten good ends."41 Audiences have become used to "multi-thread" struc-
ture from television,42 but American filmmakers grant that there are limits. Steven
Gaghan, screenwriter for Traffic (2000) and director of Syriana (2005), admits that
too many plotlines can tax the audience. Hence the need for clarity. "At first every-
thing seems confusing and overwhelming, and you feel helpless to understand. Then
gradually you realize that the stories are actually quite simple and human." 43 As we've
seen in 21 Grams, Happy Endings, Me and You and Everyone We Know, and other
instances, the filmmaker can explain and humanize the network by means of genre
conventions (romantic comedy, melodrama, and thriller) and stylistic markers (voice-
over narration, musical motifs, and color coding of different strands of the network).
So in either the mainstream or independent movie, the interweaving of lives is likely
to be governed by norms familiar from decades of American moviemaking. Once we
trace the historical development, however, we see changes as well as continuity. The
premises of studio-era degrees-of-separation films like Grand Hotel were creatively
revised in Nashville (1975), and those were in turn recast by Magnolia (1999).
All the constructive principles governing art cinema can be deployed in network
tales too. These films can display loose causal connections, diffuse or abandoned
goals, interplays of fantasy and reality, unreliable and ambiguous narration, incon-
clusive endings, and innovations in visual and auditory technique. In the Belgian film
Any Way the Wind Blows (2003), across 32 hours, eight Antwerp bohemians and one
dancing wanderer known as Windman move through daily routines and seriocomic
twists of fate (Figure 7.8). A plague virus is stolen from a lab and gets exposed to air,
but the film is more concerned with a party that brings together the main characters.
The film dwells on images of randomness: A Frisbee spins in through a high-rise
window, the wind blows, and characters invent spasmodic dances on the spot. In a
way, the potential viral contagion has already been represented by the zigzag links
forged by the turtleneck-wearing Windman, who might exist on some other plane of
220 Poetics of Cinema
Figure 7.8 Windman is introduced at the
start of Any Way the Wind Blows (2003),
striding through a subway and dodging flying
credits and curls of white string as if he knows
he is part of a film.
Figure 7.9 Colin ponders correspondences
(Out 1: Noli Me Tangere, 1971).
existence than the characters (Figure 7.9). Although after the party several characters
seem to have found peace, overall the plot refuses closure.
The inconclusiveness of Barman's film is typical of European and Asian network
movies. In Ulrich Seidl's Dog Days (2001), some suburbanites are linked up, but many
are not, and some mysteries—car vandalism, a poisoned dog—remain unexplained.
Initially, Wong Kar-wai's Days of Being Wild (1990) centers its plot structure and nar-
ration on a strong protagonist, the pouting, self-absorbed Yuddy. He seduces women,
then abandons them when they resist his control. He also conceives a goal, that of
finding the mother who left him at birth. When he leaves Hong Kong to search for his
mother in the Philippines, however, the narration lingers on the girlfriends he leaves
behind and on two men drawn to them. By the end of the film, when he meets one of
the men by chance, Yuddy has become, like Manu in Free Radicals, a placeholder, a
hollow center in the web of relationships he has unwittingly created. In the very last
scene, Wong introduces a new character, a gigolo preening before a night on the town.
How might he connect with the network we've explored? Wong planned to make
a second film in which Yuddy's influence on other lives lingers after his death—a
startling variant of degrees-of-separation norms. 44
This tendency to leave many things dangling can be traced back to Paris nous
appartient. Rivette piles up suspicions and suppositions to suggest a conspiracy that
can never be verified, and ends with most matters unresolved and unexplained. 45 He
takes the concept to a perverse extreme in Out One (1973), which runs twelve hours
and forty minutes. Two theatre troupes are connected by mysterious forces and by
two outsiders, the con artist Frederique and the obsessive loner Colin, who believes
that a mysterious group named after Balzac's sinister 13 (Les Treize) is conjuring up
secret affinities among people and places. Colin finds uncanny correspondences all
around him, and he tussles with coded clues revealed in anonymous messages, street
names, and references to Alice in Wonderland (Figure 7.9). But the correspondences
he finds could be merely coincidences, and the real crimes we glimpse, including a
murder, aren't fully explained. A modernist reworking of 19th-century serial novels
Mutual Friends and Chronologies of Chance 221
and 1910s installment films like those of Louis Feuillade, Out One hints at a network
only to unravel it at the same time.
The effort to leave a network in bits and pieces recalls another art cinema ten-
dency. Peter Haneke's Code inconnu is subtitled Incomplete Story of Diverse Trips, and
black frames bracket the scenes, sometimes interrupting a piece of action. His earlier
converging-fates film, 71 Fragments of a Chronology of Chance, is built out of flat,
laconic scenes. Instead of developing lines of goal-directed action, these chunks of
time are often filled with banal activities, such as an old man watching television. The
fragmentary quality is enhanced by the bursts of black frames separating the scenes.
Unlike Crash, which uses mainstream dramaturgy to wring pity or shock out of every
conflict, Haneke's films tend to deflect, muffle, and abort dramatic development.
"We don't perceive the world as a whole," he remarks;
we have separate impressions and we only put them together in our heads . . . .
Although I find films like Magnolia and Short Cuts very well done, they use
aesthetic means to present an illusion of totality that does not exist. In reality,
our impressions are isolated. I present the fragments as they are.46
Stylistically, both Hollywood and indie films have been conservative, relying on
close-ups, continuity editing, and standard visual design. Network narratives in the
mainstream have innovated chiefly by finding new equivalents for traditional tactics
of structure and address; the pop-up titles of Happy Endings offer a good example.
Stylistic innovation in the art cinema mode has been somewhat more venturesome.
Code inconnu, for instance, renders most of its scenes in single takes, either static
or long tracking shots. Cutting takes place in a TV montage of photographs, and an
orthodox shot/reverse shot is "quoted" in a film-within-the film. This simple dual-
ity allows Haneke to posit different registers, that of his story and that of modern
media coverage, and the pattern is brought forcefully to our notice. When stylistic
choices change the status of narration in this way, our relation to the film is altered.
We may find ourselves appreciating geometrical structure for its own sake, as in Otar
losseliani's Favoris de la lune (1984). Or we may be pulled into a process in which
structural patterns get spoiled, obliging us to rethink the rules of the storytelling
game, as Jean-Claude Guiguet does in Les Passagers (1999).
Nashville (1975)
Nashville gave Robert Altman a congenial showcase for what was emerging as his
signature style of offhand performances, tentative zooms, thick sound mixes, and
digressive vignettes. Just as important, Nashville became an exemplar. Audacious and
ambitious, it became a critics' favorite, an instant classic, and an enduring prototype
of the network format. Released a year before the Bicentennial, almost simultane-
ously with the publication of E. L. Doctorow's novel Ragtime, Nashville signaled that
serious artists were now working on a big canvas, portraying American life in bold,
broad strokes.
222 Poetics of Cinema
Broad, especially. Nashville was influential partly because it brashly showcased
its expansive storytelling. With two dozen featured players, rough-hewn shooting
methods, and meandering structure, Nashville presented itself as a breakthrough.
"The damnedest thing you ever saw," the tagline boasted, and posters spread the
cast's portraits across the back of a denim jacket and the stripes of Old Glory. Altman
told an interviewer, "It's like working on a mural.' 47 Some found the exaggeration
too much, complaining that Altman and screenwriter Joan Tewksbury caricatured
country music and served up a gallery of cliches and a barrage of cheap shots. None-
theless, the film offered a splashy, uncompromising example of how the crisscross
structure could replace traditional plotting.
In one long weekend, from Friday through Tuesday, 24 individuated characters cir-
culate through Nashville. Some are country music stars. Haven Hamilton, his wife
Lady Pearl, and their son Bud are the local aristocracy. They're visited by their beloved
colleague, the frail singer Barbara Jean, accompanied by her manager-husband Barnett.
Also passing through are Barbara Jean's rival Connie White and black country singer
Tommy Brown. On a less exalted scale, businessman Delbert Reese lives with his wife
Linnea and their two sons. Further down the ladder, we find Mr. Green; his wife is
dying in the hospital when his niece L. A. Joan comes to visit. The waitress Sueleen
and the grillman Wade Cooley fill out the cross-section of Nashville social classes.
Then comes a motley procession of strangers. The country-folk-rock trio of Bill, Mary,
and Tom arrive for a recording session. The star-struck Albuquerque ditches her
husband and sets out on her own. The Tricycle Man passes through in a high-powered
motorcycle rig; Private Kelly, on leave from the army, trails his idol Barbara Jean; the
British reporter Opal collects interviews; and the subdued but courteous Kenny Fraser
rents a room from Mr. Green. Most proactive of the newcomers is John Triplette, a
Kennedyish advance man for Hal Philip Walker, the presidential candidate hyped in
posters, stickers, and the speeches blaring from a cruising van.
It isn't just that the movie is heavily populated. Nashville is dense in a way that
flaunts its multiple-protagonist premise. Tewkesbury got the idea for the structure by
visiting the city and spotting the same people again and again over several days. "The
whole thing," she remarks, "became a transcript of overlapping lives."48 Some scenes
bring all the characters together, but even putting those scenes aside, Nashville turns
out to be a very small town. On average, each character is linked in some way to ten
other characters. Only a few are connected by kinship and long-term acquaintance;
in this city, the characters make most of their contacts through brief conversations
or simply by occupying a spot in a common locale. The scene construction keeps the
mixture bubbling. Instead of holding his characters on separate paths for much of
the film, intercutting their lives until they converge, Altman assembles a batch in
one scene, then assembles another in the next, and another and another, each group
overlapping in membership with some of the others. With every shake of the kaleido-
scope, a new configuration tumbles together.
Some characters are more central to the pattern than others. Everyone wants to
see and meet Barbara Jean and the sexy crooner Tom, whereas Connie White is in
much less demand. Del Reese the fixer presses the flesh of many locals, introducing
Mutual Friends and Chronologies of Chance 223
Triplette to people who might help the Walker candidacy. Other figures are floaters.
The Tricycle Man, who never speaks, is likely to introduce a scene by roaring up on his
bike, and Sueleen, the waitress who wants to write and sing, can be glimpsed on the
fringes of the action. New to Nashville, the strangers move more or less awkwardly
among the familiars. Opal, the purported BBC correspondent, and Kenny, the some-
what hulking guy with the violin case, make casual contact with over a dozen other
people, and L. A. Joan switches boyfriends as quickly as she changes wigs.
In this dense network, the interaction of performers becomes crucial. Working in
documentary fashion, Altman staged an airport welcoming ceremony, several club
sessions, a luncheon buffet, and three stage shows, filming them with several cameras
and using radio microphones to pick up every line. He let the actors develop their
dialogue beyond the script, but warned, "If you bore me, I'll just cut away from you."49
He does just that, freely interrupting one exchange by another. Unconcerned with
the usual arc of rising conflict and propulsion into the next sequence, Altman builds
the ensemble scenes out of fragmentary interactions. These yielded what he called
escape hatches," moments that would enable him to hook sequences in editing. 50
Seldom is anyone seen alone, and we don't have access to any mind through flash-
backs, voice-overs, or restricted attachment. The ensemble scenes consist largely of
glimpses, gestures, and sound bites. Social display becomes the keynote, and the
revelations are embarrassing. Except for Triplette the Easterner, these people haven't
learned to hide their pretensions and vulnerabilities. Haven Hamilton greets Elliott
Gould, "Welcome to my lovely home." Opal tests her recorder by saying, "Un, deux,
trois, quatre." Albuquerque says that if her career fails she can sell trucks, but she's
shattered when Kenny tells her that nobody buys trucks from a girl. A club announcer
says a star is in the room tonight, and Haven starts to arrange his face into a gracious
smile before he realizes that it's Connie who will get the spotlight. A Jonsonian fresco
of ambition, vanity, sexual frustration, selfishness, corruption, and exploitation is
filled out by an agglomeration of social details.
Granted, there are solo moments, as when Albuquerque teeters across the road,
daftly oblivious to the car crash provoked by her miniskirt. There are a few brief duets
as well: Tom and his bedmates, Mary and Bill, and Lady Pearl recalling the Kennedy
boys to Opal. Perhaps the most emotionally wrenching duet comes when Barbara
Jean, crouched on her hospital bed and painting her toenails, becomes distraught
listening to a broadcast of Connie singing in her stead. She starts to break down,
and she's soothed and threatened by Barnett. "Don't tell me how to run your life.
I been doing pretty good with it." He departs to give Connie a present (which she will
politely disdain, in another socially precise gesture), leaving Barbara Jean in the dark.
Her moment of solitude equates her with the women who want to be her: the ditzy
rambler Albuquerque, and Sueleen, the only person given a sustained scene in private.
She cheerily studies her body in a mirror, stuffs her bra, and practices her song for her
nightclub debut, oblivious to her utter lack of talent.
On the whole, the ensemble scenes dominate, becoming nodes of character con-
vergence. There is the party at Haven's lovely home and the club scenes, taking place
at night, where a few characters gather to listen to or perform musical numbers. These
224 Poetics of Cinema
Figure 7.10 Connie White, present only as
an image, and candidate Walker represented
by a sticker, in the airport convergence of
Nashville (1975).
scenes often develop dramatic action among the listeners during the onstage perfor-
mance. The most famous instance takes place in the Exit Club, where Tom sings "I'm
Easy," while four women listen, each convinced that he's singing just for her. This
scene is crosscut with the Walker fundraising smoker, in which Sueleen's botched
singing turns into a humiliating striptease. The other midsize nodes are the two
lengthy concerts at Opryland auditorium and at Opry Bell, the fake showboat. Shot
like concert documentaries, these draw together a dozen or so principals and estab-
lish the split between the star entourages and the onlookers (Kenny, Opal, L. A. Joan,
and Private Kelly). This division will be crucial at the Parthenon climax.
In the two biggest convergences, Altman brings together all 24 characters, and
plenty more. The first convergence forms at Nashville airport and the highways
outside. Some characters, like Sueleen and Wade, work at the airport, whereas several
outsiders disembark from arriving planes. Meanwhile, Barbara Jean's private plane
taxis in and she's greeted by a bevy of rifle-twirling majorettes. Even Connie White,
who won't show up in person until quite a bit later, and Hal Philip Walker, who won't
ever be seen, are there by proxy, on a cardboard standup in the gift shop (Figure 7.10).
The action shifts to the highway, where a car crash creates a massive traffic jam. Those
few characters we didn't see at the airport are shown snarled in traffic along with the
others. Within the first 25 minutes, the vast cast has been identified, and the film has
laid out the principles of overlapping lives and incisive sound bites.
The closing sequence balances the airport-freeway gathering. At Triplette's politi-
cal rally at the Parthenon, all our principals assemble again, this time to be addressed
by the as-yet-unseen presidential hopeful Walker. At the close, long after the primary
action has ended and our protagonists have dispersed, Altman lingers on documen-
tary shots of the crowd singing. This epilogue opens out the film's dense network to
the population of the city as a whole.
The primacy of performance reminds us that Nashville is not only a political assassi-
nation film (a going concern in the mid-1970s) but also a musical. Both genres supply
some well-defined narrative conventions that keep the film from being quite as way-
ward and episodic as it might have been. The musical's familiar figure of the amateur
struggling toward stardom is incarnated in both Albuquerque and Sueleen. One gets
her big chance, and the other blows it. Romance, a staple of the musical, defines some
plotlines, notably the one centered on the womanizing balladeer Tom. The major
Mutual Friends and Chronologies of Chance 225
performance scenes typify another convention: the extended musical number that
slows the causal progression while we and some characters enjoy a song.
The plot is driven partly by political goals. Triplette's mission gives an overall shape
to the action, and he becomes the figure linking the two thematic centers, musical
performance and political maneuvering. Triplette enlists Del Reese to help persuade
country stars to endorse Walker's candidacy. The prize is Barbara Jean, but she is
shielded by her friend Haven and her prickly husband Barnett. So Triplette offers Haven
a chance to run for governor. Later, Barnett caves in after Barbara Jean breaks down at
the Opry Bell and he has to placate the crowd. This is the major turning point, which
leads directly to the climax at the Parthenon rally. Other characters are pulled into
Triplette's scheme, notably Sueleen, whom Reese enlists for the fundraising smoker,
and Bill, Tom, and Mary, two of whom wind up appearing onstage with Barbara Jean.
The rally that concludes the film is the culmination of Triplette's effort to win the
grassroots to Walker's side. Like The Parallax View (1974), Nashville climaxes with an
assassination at a political rally, though here the politician isn't the target.
I don't want to give the impression that the film is as tightly constructed as others
employing the network strategy. It does have a symmetrical prologue (in the studio, a
song about America) and epilogue (in the park, a song sung in front of Old Glory). Its
airport opening is balanced by its Parthenon conclusion. And it is filled with motivic
echoes and parallels (failed love relationships, Opal's monologues). Yet Nashville is
distinctly episodic. The plot leads up to the biggest scenic blocks, the two Opry concerts,
in a fairly haphazard way, providing few of the foreshadowing hints, appointments,
and deadlines that another film would employ. Altman exploits the dispersive side of
the network premise, along with the episodic tendencies of the musical genre, in order
to fray the cohesion that characterizes most classically constructed films.
The urge to present an untidy slice of life is most striking in Nashville's daring
inconclusiveness. Will Albuquerque's flailing performance at the close launch her on
a career? What will become of the character relations we've glimpsed—Tom and Mary,
Mr. Green and his niece, Del and his wife Linnea? More crucially, why did Kenny
shoot Barbara Jean? And will she survive? The ending presumes that living is messy
and no story can be wrapped up neatly—especially one consisting of hundreds of
connections. Altman's inconclusiveness lays down a challenge to his successors. They
can follow his teeming and ragged ways. Or they can tidy up the network and make
it conform to more coherent script structure. Or, and this is the largely European
solution, they can find a way to create patterned inconclusiveness. Create, that is, a
game of form that generates regularities and then breaks them, but then absorbs those
breaks into a larger pattern.
Nashville's reluctance to wrap up its story lines carries to the limit a tendency we
see throughout its narration. The crosscutting and parallels, the unconsummated
scenes hooked tenuously together, and the story's wide scope all permit rather ellipti-
cal narration. We can't be everywhere at once, so the narration often has to bring us
up-to-date on what's already happened elsewhere. When the Tricycle Man pulls up
to Haven's party carrying Opal, we realize that there was doubtless a moment when
she made contact with him—we just didn't see it. Every time we glimpse L. A. Joan
226 Poetics of Cinema Mutual Friends and Chronologies of Chance 227

she's with a different man, and sometimes we don't see her pick him up. The ellipses
can also affect characterization and backstory. On Sunday morning, the film presents
several characters at church services. Lady Pearl is at Mass, but her husband Haven
sings in a prosperous Protestant choir. Del Reese and his sons attend the Protestant
service, but his wife Linnea, a part-time gospel singer, is at an African American
Baptist church. Not only does this sequence delicately show us a piety not seen in
Figure 7.11 Nashville's credits, selling the
these people's everyday dealings, but it also sketches in a social truth we might have movie as an album on late-night TV.
forgotten: Some husbands and wives subscribe to different faiths.
The evocative, gappy narration becomes downright devious as we approach the
climactic assassination. The two most opaque loners are Private Kelly and Kenny whole contrivance framed by a whirl of fake albums cut by the characters. The jokes
Fraser. Tom taunts Kelly for being a soldier, and Kenny carries around a mysterious pile up. The lead-in has indicated that Nashville was financed by ABC Entertainment,
violin case. A series of scenes introduces an ominous tone. Opal chatters to Triplette
a television subsidiary, and the ad concludes with the announcer explaining that the
about political assassination, claiming that a gun-carrying culture encourages inno-
film will be brought to us "without commercial interruption." The ad quickly shrinks
cent but confused people to "pull the trigger." On that line, we cut to Kenny talking
to a pinpoint, as if a television were switched off. Even the scratched-up Paramount
to his mother on the phone. We've already seen Kenny pass the Walker campaign
logo at the very start recalls a weather-beaten 16mm print running in the small hours
headquarters; it s a throwaway, but it's distinctive for being one of the few occasions
on a local TV station. The hyperbolic, direct-address prologues of Bug and Magnolia
we see any character alone. Now, as Kenny explains where he is, he prevents L. A. Joan
from investigating his violin case. Then he cuts off his mother's call. All pretty sinister owe a lot to Altman's recognition that a network narrative, although apparently close
behavior. Later at the Opry Bell, we zoom in on Kenny raptly staring at Barbara Jean, to life in some respects, is no less an exercise in artifice than any other movie.
but the camera quickly pans to Opal, asking Private Kelly what he did in Vietnam. Altman claims he had long been tempted by the converging-fates format. In the
When the crowd starts to boo Barbara Jean after her breakdown onstage, Private 1960s, he imagined a film featuring four strangers whose lives intersect only in a
Kelly stands up in defiance of them. yoga class.54 He'd also thought about shooting two films based on Michener's Tales of
So the narration equivocates. Kenny is an obvious suspect, so perhaps the wor- the South Pacific, one centered on the nominal hero, and the second film putting the
shipful Private Kelly will turn out to be a crazed Vietnam vet? In fact, the narration subsidiary plotlines to the fore and making the first film's hero a minor character. 55
has tricked us. It's the target, not the shooter, who will come as a surprise. Altman Along these lines, he considered releasing two features, Nashville Red and Nashville
sustained this narrational ploy in his interviews. "We also condone political assas- Blue. Each would cover the same story but would favor different characters. Venture-
sinations," he remarked. "Of course we don't shoot our pop stars or sports heroes, some ideas like these merely confirm the evidence on the screen: Altman brought a
because they don't arouse divisive feelings."51 In his film, we do shoot our pop stars, new vitality to cinematic storytelling. One could argue that Nashville is more com-
for reasons he leaves to the viewer to decide. plicated than complex, that scrambling together lots of one-off encounters is easier
"Nashville's a musical, really," Altman notes.52 The film contains over an hour of
than sculpting everything into a coherent design. Yet the film's openness to accident
performances, and this isn't counting all the background tunes we hear in clubs or the
and digression, its urge to create a fresco of American life, and its counterweighting
bursts of nondiegetic music now and then. Most of the big numbers were shot as actual
of sharp social realism by stylization made it hugely influential. Just as influential was
shows with roving cameras and the 16-track recording system used for rock concerts, a
the realization that familiar conventions of genre and style could help audiences pick
major departure at the period. 53 As often happens in the genre, the songs comment on
their way through a teeming profusion of incident. Nashville pointed the way toward
the action, as when after Mary learns that Tom is bedding other women, she must sing,
a renewal of narrative form in American cinema.
"Since you've gone, my heart is broken—another time." The sign-off tune, "It Don't
Worry Me," threads through the film, heard in snatches during the traffic jam, at the
drag races, and in the Exit Cafe before its all-out, hypnotically repetitive treatment in IsAagnolia (1999)
the Parthenon epilogue makes it virtually a new national anthem.
For all the traces of documentary capture, Nashville also trades on the traditional About six years before the disappearance of Ambrose Small, Ambrose Bierce
musical's self-conscious stylization. This is most evident in the brilliantly odd open- had disappeared. Newspapers all over the world had made much of the mystery
ing credits, a parody of late-night TV pitches for compilation LP albums (Figure 7.11). of Ambrose Bierce. But what could the disappearance of one Ambrose, in Texas,
This rapid-fire pseudo-commercial introduces the music industry, the premise of the have to do with the disappearance of another Ambrose, in Canada? Was some-
ensemble cast, the faces of the principal players, and lines from the songs we'll hear, the body collecting Ambroses?...
228 Poetics of Cinema
In the explanation of coincidence there is much of laziness, and helplessness,
and response to an instinctive fear that a scientific dogma will be endangered.
Charles Fort, Wild Talents
The prodigal energies of Nashville steered American filmmakers in several directions.
Drive-in (1976), Citizens Band (aka Handle With Care; 1977), and Honky Tonk Freeway
(1981) simplified the narrative strategies Altman had broached. Ragtime (1981), which
Altman was originally slated to direct, reduced the sweep of the novel to a more inti-
mate network of fictitious families involved with the murder of Stanford White. Spike
Lee's Do the Right Thing (1989) used the day-in-the-life format for social criticism,
whereas John Sayles became identified with the politically charged network film,
perhaps most ambitiously in City of Hope (1991). Pulp Fiction (1994) wedded network
principles to film noir, and it scrambled time schemes and other formal devices in
ways that made all the ingredients seem fresh. There followed several similar efforts,
notably Two Days in the Valley (1996) and Todd Solondz's Happiness (1998).
So network storytelling was already a reinvigorated tradition when Paul Thomas
Anderson's Magnolia was released. Yet the attention Magnolia received is partly due
to its revisionist impulse. For all Anderson's reverence for Altman, his film shows
how the ornery diffuseness of Nashville can be tamed by conventions of Hollywood
storytelling. At the same time, Anderson acknowledges the artifice inherent in the
structure in fresh and arresting ways. Nashville displays the format's potential for
ragged, slice-of-life evocation, in the end remaining coolly objective. Magnolia relies
on histrionic acting, interior monologues, and immersive musical accompaniment
to sharpen the pathos of the characters' situations. Anderson admired not only
Nashville but also Network (1976), a cut-and-thrust satiric drama delineating power
plays in the television industry. Perhaps it's not too much to see Magnolia as blending
the two films. It wants at once to be traditional in structure and effect but also firmly
modern—ironic, self-conscious, and virtuoso—in its narrational maneuvers.
Magnolia bears the marks of the family melodrama. It centers on the relatives and
employees of Big Earl Partridge, a TV producer. Earl is dying of cancer, and his young
wife Linda is becoming increasingly frantic in dealing with his decline. Earl's son
Frank has disowned him because Earl abandoned him and his mother. Frank is now
a self-help guru, promoting an aggressive scheme for men to seduce women. Earl has
produced a long-running quiz show called What Do Kids Know? hosted by Jimmy
Gator. Jimmy is estranged from his daughter Claudia, who picks up men in bars to
support her cocaine habit. The quiz show is linked to two other characters: Stanley,
the withdrawn boy genius, and Donnie Smith, a famous contestant now grown up
and working at a low-end job. Touching on these characters' lives are two professional
helpers. The home-care nurse Phil Parma tends to the bedridden Earl, whereas the
policeman Jim Kurring calls on Claudia because of a noise complaint.
Far more thoroughly than Nashville, Magnolia is propelled by goals, and of an
intensely personal sort. Jimmy Gator, also dying of cancer, calls on Claudia hoping
for reconciliation. The semidelirious Earl asks his nurse Phil to find his son for him,
and Phil dutifully tries to contact Frank. Former quiz kid Donnie, needing money
Mutual Friends and Chronologies of Chance 229
to fix his teeth, conceives an amateurish robbery scheme. Above all, the characters
seek love. Linda turns to a doctor and a lawyer to find how to best deal with Earl's
death, for she has suddenly realized that she loves him. Early on, the lonely cop Jim
tells us he wants to do good on the job, but we also see him seeking a young woman
for companionship. This leads him to ask Claudia on a date. Donnie Smith cries out
to Brad, the bartender he adores, "I have love to give!" The other side of the coin is the
cold manipulativeness promoted by Frank's man-power program—sex only, exploita-
tion instead of affection—presented as the perverse result of his father's rejection of
his mother. The long-standing melodramatic convention of the search for love drives
most of the plotlines.
The obstacles to these goals are largely created by parent-child conflicts, another
constant of melodrama. Donnie claims that his parents cheated him of his quiz-show
earnings, and Stanley is pushed beyond endurance by his father. Indeed, fathers come
in for a drubbing. Earl abandoned Frank, and Jimmy may have molested Claudia.
Nearly all the major characters can be lined up on a Seven Ages of Man chart, from
the child Stanley through the youngish Phil, the slightly older Jim and Frank, the
middle-aged Donnie, and the aging Jimmy, to the elderly Earl. If a "woman's picture"
like Stella Dallas (1937) or Mildred Pierce (1945) often centers on tensions between
mothers and daughters, Magnolias gallery of miserable men creates a hyperbolic
variant of the dynastic melodrama. Through films like Cat on a Hot Tin Roof (1958),
The Long Hot Summer (1958), and Home from the Hill (1960) stalk monstrous
patriarchs, warped sons, ineffectual mothers, promiscuous wives and daughters, and
substitute sons (what Phil arguably becomes for Earl). As usual in family melodrama,
the past haunts the present, as more of the fathers' misdeeds come to light and the
young strive to break out of the self-destructive patterns etched during childhood.
The film's use of the sins-of-the-fathers theme becomes explicit in the repeated cross-
cutting between Earl, mumbling regrets for his profligate life, and scenes of Frank
urging men to exploit women or fabricating a false childhood in response to a female
interviewer. Frank tells her that he urges his men to forget the past, but soon he finds
out what characters assert at different points: "We may be through with the past, but
the past ain't through with us."
The script threads all these characters and their associated goals through about
24 hours. For the first two thirds of the film, the characters are linked largely through
crosscutting. Their paths don't intersect much. Instead, Magnolia works a cunning
variation on Hollywood plot structure. After a prologue and a quick introduction to
the protagonists, all their problems are established in a 30-minute section marked off
by a weather-report title. The plot takes the principals through a rising action lasting
nearly 100 minutes, with that broken into 20- to 30-minute chunks, many of them
marked by musical cues or turning points in the What Do Kids Know?--broadcast.
At the 2-hour mark, most of the characters go into a tailspin. Jimmy Gator has
collapsed during his show. Stanley has defected from the quiz, Jim has lost his service
pistol, and Frank, having learned of his dying father, finds his composure crumpling.
Hollywood screenwriters call the point when all seems lost the "darkest moment,"
and at this point, Anderson cuts among his sorry individuals to suggest that things
230 Poetics of Cinema
are at a stasis. (The moment is, however, made vivid by means of a stylized tactic
I'll consider shortly.) The final portion of the film suggests some salvation. A title
announces, "Rain clearing," and although that might seem misleading, given that a
flurry of frogs will burst from the sky, it actually promises some hope.
But first the various lines of action proceed to a nighttime climax. Donnie starts
his burglary, Frank arrives at Earl's bedside, and Linda attempts suicide. Jimmy Gator
confesses his infidelities to his wife, and he admits that he may have molested Claudia.
Claudia flees from Jim after their awkward restaurant date. Finally several actions flow
together, with Magnolia Boulevard as the point of convergence. The prostrate Linda
is in an ambulance, which passes Donnie on his way to rob the store, and he passes
Rose Gator on her way to find Claudia. Jim, driving from the restaurant, glimpses
Donnie's bungled effort to return the money he has stolen. At this moment, frogs fall
from the sky.
The rain of frogs becomes like the earthquake in Short Cuts or 20 30 40, an arbi-
trary natural event that abruptly changes several characters' fates. Jimmy Gator is
about to commit suicide, but a frog smashing through a skylight knocks the pistol
from his hand. Rose, her car windshield smeared by frogs, smashes into the ambu-
lance, yielding the obligatory car crash. Seeking shelter, Jim captures Donnie, whereas
Rose finds Claudia and the two women crouch in panic. As Earl dies, he recognizes
Frank, who starts to weep convulsively. The frogfall provides a very odd climax, but
structurally it does what it must: Push to a crisis all the lines of purpose-driven action
running through the film.
Classically constructed films tend to close with an epilogue, one that ties up the
unresolved issues and celebrates a new stability in the characters' lives. After the
amphibious rain clears, most of Magnolias principals have reason to hope. Jim's
pistol drops miraculously from the sky. Frank finds Linda at the hospital and begins a
reconciliation. Stanley stands up to his father. Jim helps Donnie return the money and
then lets him go, in the process becoming the raisonneur of the film's development:
"Sometimes people need to be forgiven. And sometimes they need to go to jail." How
do we judge? How do we transcend the cruelty and betrayals of childhood? "What can
we forgive?" Jim asks. One answer is suggested in the final image. Claudia lies in bed,
and her expression gradually brightens as offscreen Jim offers to take care of her. The
film that began with the tune "One Is the Loneliest Number of All" ends with "Save
Me." Although some characters' fates remain unknown (Jimmy Gator's and that of the
black family involved in a murder Jim discovers), the young people have hope for love.
A corny Hollywood ending? At least, a traditional one. But the final shot's naked
appeal to the theme of a good man redeeming a fallen woman also maintains an
innovative address to the viewer. Jim has provided a narrating voice-over from early
on, and that role is resumed just before this last scene. This pleasing symmetry makes
him an upbeat, slightly square center of consciousness, or perhaps conscience, for a
film that is steeped in pain and despair. At the same time, the optimism is muted.
Jim's voice isn't all that audible during the last shot, coming at us in soft scraps under-
neath the gentle Aimee Mann song. This sort of sound mix, in which the nondiegetic
Mutual Friends and Chronologies of Chance 231
Figure 7.12 Magnolia (1999): A suicide Figure 7.13 Magnolia: As Jim the cop drives
attempt and the portentous coiled hose. through the streets, a bus sign flashes through
the frame.
music lets only whiffs of dialogue come through, was heard during the film's exposi-
tion, where it announced music as a muscular force in the movie.
Just as important, this last shot is embedded within a more abstract epilogue that
frames the story world in a way that makes conventions of the network narrative
audaciously explicit. We've seen that a common theme of converging-fates tales is the
tension between chance and destiny. Anderson makes this convention overt, signal-
ing it in the design of the story world, the film's structure, and narrational patterning.
Magnolia reflects openly on how we are to understand the sort of coincidences that
propel network narratives.
At first glance, chance shows itself in freak natural events. Quiz Kid Donnie
Smith was struck by lightning years before, and the rain of frogs becomes an unex-
pected piece of weather. A kind of antinatural meteorology, frog rains are one of
many puzzling events collected by the eccentric Charles Fort, whose book Wild
Talents Stanley is reading in the library. Fort adduced thousands of coincidences
and paranormal phenomena as challenges to contemporary science; for him, they
pointed toward enigmatic pieces of wisdom he called "abstrusities." Stanley seems
to be a budding Fortean, absorbed as he is in the history of freemasonry, a manual
of forensic medicine, and a collection of real-life anomalies by magician Ricky Jay
(who supplies a voice-over narration and plays a minor role in the film). When the
frogs start to fall, Stanley looks up with a smile of satisfaction and remarks, "This is
something that happens."
But does it happen through chance or necessity? The black boy Dixon tells Jim at
the crime scene, "I'm the prophet," and ends his rap with the lines "When the sunshine
don't work / the Good Lord brings the rain in." Dixon's reference to the Lord suggests
that the film reconciles freak weather and childhood prophecy through echoes of
the Bible. When Pharaoh refuses to let the Israelites leave Egypt, one of the plagues
that God visits on the land is an infestation of frogs. The frogs are prefigured within
Magnolia's story world by oblique citations of Exodus 8:2 ("And if thou refuse to let
them go, behold I will smite all thy borders with frogs"). The figures 8 and 2 appear
throughout the prologue, most intricately in a coil of hose on a rooftop (Figure 7.12),
and before the deluge a warning can be glimpsed in a bus shelter (Figure 7.13). The
connotations are elusive—does the rain of frogs redeem the characters from their
232 Poetics of Cinema
psychological bondage?—but the citations link Fortean curiosities with Old Testament
morality. In this world, what appears to be chance secretly serves a grand design.
Anderson doesn't merely scatter hints through his characterization and mise-
en-scene. He opens the film with a prologue dramatizing three exemplary cases.
On Greenberry Hill, Sir Edmund Godfrey was murdered by Green, Berry, and Hill.
"I would like to think," muses the voice-over narrator, quoting Charles Fort, "that it
was only a matter of chance."56 Blackjack dealer Delmer Darien was killed in a freak
accident by a pilot who had lost money at his table just the day before. Also, perhaps, a
matter of chance. In Los Angeles, Sidney Berringer secretly loaded a shotgun that his
parents brandished when they quarreled. Then he jumped off the roof. The shotgun
discharged as Sidney dropped, killing him on the fly. The narrator reconstructs the
accident, and as the patterns are peeled back, he confesses his unease. "This was not
just 'something that h a p p e n e d ' . . . not 'just a matter of chance.' These strange things
happen all the time." The most vivid parable involves a child made miserable by his
parents, foreshadowing the family melodramatics we'll witness, but also preparing
for the propitious deluge—merely, Stanley reminds us as he stares out the library
window, "something that happens."
The epilogue takes up the same theme, now over imagery of stability and reconcili-
ation. As Earl's body is removed from the house, we hear,
There are stories of coincidence and chance and intersections and strange
things told. And which is which, and who only k n o w s . . . . And it is the humble
opinion of this narrator that strange things happen all the time. And so it goes
and so it goes. And the book says, "We may be through with the past, but the
past ain't through with us."
Now the strange things that happen aren't ghastly coincidences or enigmatic showers
of frogs, but rather those moments of love and forgiveness that override the damage
that parents inflict on their children. The story world is framed by the commentary
of a noncharacter narrator that glimpses a mysterious purpose behind the unlikeliest
happenings. The film's elaborately frank artifice pays tribute to the design governing
the most casual coincidence. And that design, impenetrable though it is, can work
for good or ill. If chance can trigger car crashes, why can't it also reconcile enemies
or unite lovers? Magnolia cleverly motivates its weird frogfall by appeal to the super-
natural, while also allowing serendipity to create a relatively sunny conclusion.
Jim's inner monologues and the voice-over commentary are the most obvious
instances of the role played by sound in sustaining the network. Just before the dark-
est moment, Earl's delirious rasps of contrition are heard over scenes of others' lives,
commenting on their problems. Music is a more pervasive sound ingredient, with
long stretches of Shostakovich-like ostinati binding the crosscut characters together.
(Magnolia iterates some melodies for as long as 18 minutes.) Earl's voice-over linkages
among story lines and the persistent music prepare us for a narrational device as bold
as the frog deluge. At the static "darkest moment," the characters, though far apart
and unaware of each other, sing along with Aimee Mann.
Mutual Friends and Chronologies of Chance 233
At first her song "Wise Up" is heard in Claudia's apartment, where other Mann
songs have been played, and Claudia sings with it as she inhales a line of cocaine. But
then the song is pried loose from its niche in story space. It is heard in various locales,
with every character, even the dying Earl, singing a line or two of the lyrics, each
phrase suggesting some relevance to the character's situation. The camera, moving
forward in many earlier shots of the film, coasts rhythmically up to each person. The
sequence is a bold stroke, as flagrant as the frog deluge. It brings to a culmination the
crosscutting of lines of action, the musical binding of one story to another, and each
protagonist's surrender to passivity. The end of the sequence is signaled by a reverse
camera movement that withdraws from Stanley's glowing face in the library as he
completes the verse, "So just give up." At this point the rain clears and the climactic
intersection of characters is launched.
Verisimilitude is flagrantly broken here, but the moment fits Magnolia's narra-
tional strategies. "Wise Up" flaunts the artifice of the network film while at the same
time asserting that the formal conventions do capture mysterious currents in our
lives. Straining to locate Earl's son, Phil says that many movies contain a scene in
which someone helps a stranger find someone else. He tells the voice on the other end
of the line, "This is that scene," and adds that movies contain these scenes because
they really happen. In the epilogue, Ricky Jay's narration remarks of the strange
things that happen all the time: "And we generally say: Well, if that was in a movie,
I wouldn't believe it. Someone's so-and-so met someone else's so-and-so, and so on."
Coincidences and chance events and intersections and "strange things told" are
presented as at once the stuff of fiction and the foundations of our world. Magnolia
has it all ways. It molds the network movie to classical canons of genre, plot, and
narration. At the same time, it admits that those canons are utterly artificial. And
yet it holds out the promise that those canons are somehow true to the abstrusities
coiled within our lives.
Favoris de la lune (1984)
Jacques Tati employed crisscrossed running gags in M. Hulot's Holiday (1953), in
which Hulot was at the center of seven days of comic mishaps in a seaside resort.
In the comic masterpiece Play Time (1967), Tati pushed further toward a dispersed
narrative structure by "democratizing" his comedy, making Hulot less prominent
(he is offscreen for several minutes at a stretch) and creating a tissue of casual encoun-
ters in the modern city. At the same time, Tati showed that network narrative could
sustain purely stylistic experimentation.
In Hulot and Play Time, he relied on very distant shots, decentering his principals
and creating complex interplay within the frame (Figure 7.14).57 This technique gains
great strength when allied to a network tale. For one thing, we are invited to see
the frame itself, not just the locale, as a point of intersection. It becomes an abstract
rectangle gathering together certain plot threads. The sustained long shot can also
favor the proliferation of story lines. When the image is crammed with several
characters, as in the dazzling restaurant sequence of Play Time, we must maintain an
234 Poetics of Cinema
Figure 7.14 Play Time (1967): In early por- Figure 7.15 Play Time: The dense restaurant
tions of the film, Jacques Tati stages his gags at scene packs several gags into a shot. Here the
an immense distance. Arriving in the airport, owner drinks what he thinks is a headache
Barbara glances back to see a woman stroking remedy in the middle ground while confu-
her bag and to hear a dog softly yelping. Still sion reigns in the distance.
farther back, the silhouettes of airline workers
are revealed as cardboard cutouts.
alert, Where's-Waldo frame of mind (Figure 7.15). We re obliged to identify a charac-
ter not by face or voice but by what we might call silhouette—characteristic clothing,
gait, and bearing. The strategy provides Tati a wonderful technique for his comedy of
social ineptitude. The long shot displays bystanders' attitudes of shock and bewilder-
ment and magnifies those silences that greet a gaffe.
When Otar Iosseliani left Soviet Georgia and took up residence in Paris, Tati
befriended him. As often happens, influence favored the prepared mind. Iosseliani
had already moved toward group-centered narratives in his Georgian films April
(1961) and Pastorale (1976). The latter film, about a chamber ensemble who spend a
summer in a village, is made almost entirely out of unremarkable routines as neigh-
bors quarrel, boys play pranks, and the players rehearse and relax. Iosseliani remarked,
"I always start from a concise phrase. For Pastorale it was: 'How people come together
58
and separate and never see one another again.'" In France Iosseliani gave up the
choppy, quasi-documentary look of Pastorale in favor of something more controlled
and austere. Tati showed that the long-shot technique mixed with unemphatic pacing
could give network structure a dry, slightly puzzling flavor. This is the direction
Iosseliani explored in Tesfavoris de la lune (Favorites of the Moon; 1984), La chasse
aux papillions (Chasing Butterflies; 1992), Brigands, Chapitre VII (1996), and Adieu,
plancher des vaches! (Farewell, Home Sweet Home; 1999). Granted, Tati's gentleness is
missing from these scathing comedies, but Iosseliani brilliantly shows that one can
lay bare the abstract geometry of a network story through a distinctive visual style.
In the characteristic Iosseliani scene, a distant shot makes everything unfold more
or less on the same plane (Figure 7.16). Seldom is there an aggressive foreground or
a dominating background. There's none of the casualness of the Altman long-lens
shot, and certainly none of the intensity Magnolia achieves by rhythmically track-
ing in to faces. The longish takes refuse to supply a close-up, the technique Iosse-
liani calls a "fake striptease."59 The camera stays back, standing resolutely outside the
action. Characters lose their inner lives, their pasts, and even their names. They gain
Mutual Friends and Chronologies of Chance 235
Figure 7.16 Favoris de la lune (1984): Less Figure 7.17 In Favoris, the characters are
extreme than Tati, Otar Iosseliani still lets identified as full-length figures, character-
his gags play out at a distance. ized through costumes and stances: The old
retiree greets the amiable hookers.
integrity and solidity, as unique bodies moving through a chunk of urban or rural
space (Figure 7.17). The discreet panning and slight zooms do not favor one body
over another. Instead, each figure becomes part of a spatial polyphony. Iosseliani has
emphasized the musical metaphor.
In my subconscious, I treasure the memory of counterpoint, the three-part
form of the rondeau, which permits me to sustain something that might fall
away. When you repeat the same theme, after a moment, the spectator rediscov-
ers the armature [squelette], the story structure, which keeps him from getting
lost. It's entirely formal. 60
The film that set the tone for losseliani's later work is Favoris de la lune. In Paris
there live two families, that of a police inspector and that of a wealthy arms dealer.
The criminal family helps some Arab terrorists obtain explosive devices (which often
fail). The bombs are manufactured by a man whose lover is conducting an affair with
the police inspector. The bomb maker's neighbor is a burglar who robs both families
and who seduces the arms dealer's wife. We meet some assorted cops, bums, prosti-
tutes, terrorists, children, and rock bands. By the end, all of the characters have been
brought into some conjunction with others, most often through spatial overlaps.
The pattern is further complicated by two circulating objects, a set of antique plates
and a painting of a nude woman. (In a prologue, Iosseliani mordantly sketches the
history of each treasure.) In an auction early in the film, the arms dealer's wife buys
the plates, and the painting is bought by the inspector. Both antiques are soon stolen
and pass through many owners, sometimes crossing the paths of family members. For
long stretches the objects are simply forgotten before they resurface and connect up
plotlines, sometimes trivially. In their adventures, the plates get broken and mended,
and the painting sliced into smaller and smaller pieces. At the end, in a metalworking
shop, the painting has become a pinup and one of the plates serves as an ashtray.
Convoluted though it is, my summary is easier to grasp than the film, which almost
completely lacks exposition of the normal sort. The first 15 minutes of Favoris de la
lune is virtually a silent movie. We see characters at a distance getting in and out of
 Poetics of Cinema Mutual Friends and Chronologies of Chance 237
236
cars or walking down the street, fighting over cabs and passing on staircases. The
relationships are presented visually, and the dialogue, muffled or murmured, isn't
much help. But unlike silent films, which can use intertitles to transmit information,
Favoris is in no hurry to tell us who's who. No domestic breakfast scenes introduce
us to the families, no suspenseful or sexy incidents set up the crooks and prostitutes.
A cluster of relationships emerges gradually out of daily routines and perfunctory
meetings, and throughout the movie we are still learning about simple givens. The
more important dramatic events, like sexual affairs, are suggested simply by stereo- Figure 7.18 The tourist Barbara receives a Figure 7.19 Play Time: Glancing out as her
typed bits of behavior—a man picking up a woman in his car, and a man presenting sprig of plastic flowers as a parting gift from bus moves off, she sees them as echoing the
a woman with a rose before taking her to a hotel. Romantic partners are shuffled, but M. Hulot (Play Time). shape of streetlights.
there is no sense of intimacy because only the most diagrammatic signs of affection
are given us. Iosseliani has called his characters marionettes, and as in a puppet show,
personal identity is shown in curt, schematic gestures.61 bedroom or the thieves? Instead, Iosseliani cuts to the machine shop where the bomb
Iosseliani, who reflected more theoretically on his chosen form than Tati did, maker starts his day, as the arms dealer's wife comes to give him his assignment.
seems to have understood that criss-crossing can create some fruitful uncertain- This rhythm of expanding adjacency functions as cryptic exposition. By register-
ties. It can hint at a relationship visually without pinning it down. (Is the hooker ing the isolated gestures and tiny intersections, we gradually—very gradually—come
a lover of one of the burglars, or just their confederate?) The throwaway dialogue, to grasp a set of circumstances and a narrative action. The dominant impression is
which we overhear rather than attend to, doesn't specify exactly how characters are
the fluid, ephemeral hookups among individuals. "What's important to me is this
linked. There's a more insidious effect as well. After two characters intersect, the next
possibility of treating the characters or the actions as [musical] themes, which permits
moment can attach us to either one, and then, when the one to whom we're attached
everything in the film to flow, all the streams to join and separate."62 The pattern that
brushes past someone else, that's another opportunity to deflect the narrative line.
finally results is as wayward as the protagonists. Iosseliani fills his networks with
String together several brief encounters like this, and we can be quite unsure exactly
clumsy characters and awkward accidents. The paths intertwine, he seems to be
whom the narration might follow next. We're thrown back on simply registering each
moment in itself, however opaque it might seem. This effect is magnified if no char- suggesting, but the figure they form has little overall order. The cross-confusions of
Favoris de la lune end, obscurely, with a sniper's bullet.
acters are favored through close shots and the visual emphasis always falls on the
ensemble. Iosseliani leads us through a branching maze that will eventually bring Tati, who was fascinated by cliches, concluded Play Time by investing kitsch with
us back, through character overlap or through crosscutting, to familiar terrain. The genuine emotion, matching a sprig of plastic flowers to curved street lights as the
effect is to treat the social world as a series of spreading and overlapping trajectories young woman leaves Paris (Figures 7.18-7.19). Iosseliani likewise makes common-
that we study from a distance. places the pretext for ingenious formal play. In Favoris de la Lune, we wander in
One short scene soon after the prologue furnishes a good example. The bomb an impersonal city (has Paris ever looked more drab?), and we register the familiar
maker has helped his blonde girlfriend get ready for work, and she goes out, passing tactic of cross-class comparisons: Both the bourgeoisie and the underclass live by
the prostitutes clustered at her doorway. She crosses the street to a waiting car, where theft. The painting and crockery pass from rich to poor and turn from objets d'art
a man (whom we'll later learn is the police inspector) greets her with a kiss. The car into bric-a-brac. This motif, Iosseliani tells us, implies that the search for possessions
drives past the camera, which follows it to pick up an old man starting to cross the is a vain thing. The cyclical exchanges and traffic among the characters invoke the
street. He greets the prostitutes, and one goes with him to the building opposite.
theme of the eternal return, "the old, banal notion, nothing new under the sun."63 As
In the next shot, as the old man and the hooker arrive at her landing, the bomb
usual, however, ingenious cinematic form can turn even overfamiliar materials into
maker passes them. He's the one we follow next as he descends to the street, where
invigorating exercises for the eye, ear, and mind.
he greets the hookers. Pan with him across the street and hold on another prostitute,
who is approached by a prospect. Then we cut back to show the hooker and the old
man entering her apartment. The two men we'll eventually come to recognize as the L,es Passagers (1999)
burglary team come out of their next-door apartment and give her keys, adding,
The prow of our tram moves through the streets and into a tunnel, and most of the
"Tonight." (The stingy Iosseliani won't supply a medium shot of this gesture, so the
credits emerge from the darkness. Once we pass back into daylight, minor incidents
actor must hold up the key chain for us to see it clearly.) The two men depart by
the staircase, and we're left with another fork in the road: Follow the couple in the develop. One rider has no ticket because the vending machines aren't working. A man
238 Poetics of Cinema
Figure 7.20 Les Passagers (1999): The Narra-
trice introduces the tramcar journey. Unlike
the heroine of Happenstance (Figure 7.7), no
one is listening to her in reverse shot.
with a dog gets off, and a youth suddenly dashes after him. An old lady demands to sit
in her accustomed seat. The tram passes a cemetery.
The opening moments of Jean-Claude Guiguet's Les Passagers don't provide
ordinary exposition, but they do seem to set up a stable situation. The title and the
camera's confinement to the tram suggest that we'll follow several travelers pursuing
their lives. Yet it's characteristic of the more ambitious films in the art-cinema tradi-
tion to establish a scheme only to break it, then absorb the break into a larger pattern
that retrospectively makes sense. This is what happens here, again and again. The
network's story lines are fractured (Guiguet calls the scenes fragments) 64 and so we
must constantly rethink the givens. How are we to construct a coherent story world?
What propels the plot? How are we to distinguish representations of inner states from
objective truth, reliable from unreliable narration? And what thematic evocations
issue from this game of form?
So the overture provides a few dissonant notes. As the tram glides along, we hear
a woman's voice-over commentary, and then a woman on the tram continues in the
same voice, but she looks out at us (Figure 7.20). She speculates on why a young man
is carrying a bouquet. A wedding? No, she tells us, the wedding comes later. (How
can she be so sure?) And when her seatmate complains about the ticket machines,
she agrees, first in voice-over and then by addressing us. It's an odd story world in
which a figure in the midst of others talks to the audience directly. Still, because she's
commenting on the action we see, we might take her as the sort of personified, all-
knowing narrator we find in films like Our Town (1940) and La Ronde (1950). The
final credits will identify her as the Narratrice, and her points seem helpful. When she
remarks that machines selling round-trip tickets don't understand that someday we
may not return, her words are appropriate to the cemetery that the train passes.
Despite the Narratrice's jarring habit of addressing us, the film seems to settle into
a stable pattern. For about the first half hour, the tram serves as the core of a small
and slight network. A traveler is shown, then we follow him or her a little bit after
getting off, and then we return to the tram and pick up another passenger. The young
man with the bouquet visits a cemetery, and through flashbacks and the Narratrice's
commentary we learn that his lover died of AIDS. Again, there is a bump—the nurse
Mutual Friends and Chronologies of Chance 239
reading to the dying lover is shown not reading at all but rather reciting to him—but
when we return to the tram, the Narratrice completes the young man's story. Then
comes a popular song, started by a woman on the tram but picked up by the other
passengers and eventually finished by a recording. We pass a wedding, perhaps the
one prophesied by the Narratrice. There follows a more comic episode, involving a
passenger who's a fetishist seeking the perfect female foot. He gets married (another
wedding), and we're back on the tram. A schoolgirl gets off and goes to the funeral of
an old woman who died alone. Each passenger's episode is little more than an anec-
dote, and we might start to ask about the time frame we're in—one trip or several?
trips in the past or the future?—but we recognize a predictable rhythm of structure
and narration, alternating between the tram ride and private lives.
A full-fledged story seems to start when Pierre, the man with the dog, begins to
chat with a woman, Christine. A bit of a narratrice herself, she says that when we
see strangers, they suggest stories. The situation hints that a romance might emerge,
particularly when their conversation is interrupted by shots of Pierre at home caress-
ing himself. But now the status of the insert is questionable. Other fragments, like
the scene of the bouquet bearer's dying lover and those presenting the foot fetishist's
quest, are evidently flashbacks. But the shots of Pierre's solitary sex aren't marked
as past or future. And instead of returning to the conversation with Christine, the
narration launches another story line. David alights from the tram and is followed
by Marco to his apartment, where they begin an affair. Only then do we return to
the tram, and more talk between Christine and Pierre, now sitting at the window.
The narration begins to braid the story lines. Back in David's apartment, after he and
Marco have made love, they talk of AIDS. Back on the tram, Christine watches Pierre
alight and meet a nurse, Anna. We then see Pierre and Anna attend a Bach concert.
In this portion of the film, Guiguet has complicated his structure. From single-track
story lines flowing out from and back to the tram, now characters' lives are intercut,
arid in a way that compares gay lovers and straight ones. The rules have changed,
though not disarmingly.
They're about to change again. We're back on the tram, this time with a business-
man who launches into a wild 6-minute monologue about sexual identity (Figure 7.21).
At first, his rant seems violently antigay ("I like women to be feminine and men to
be masculine"), but it turns into a celebration of polymorphous sexuality. He praises
women who while having sex with men imagine they're men too, and he acknowl-
edges their counterparts, "lesbian men." He asserts that everyone is bisexual, then
corrects himself to say we're trisexual, because masturbation should be ranked along-
side homo- and heterosexual activities, especially in the age of HIV. The semicomic
oration runs for a minute before a reverse-angle cut to a silent, watchful passenger
(Figure 7.22) anchors the speech as part of a one-sided conversation. Still, it's too
late to cast away all sense that the comments are aimed at the audience more or less
directly, not least because the man's slightly angular gaze has been used before, in the
Narratrice's ruminations to us (Figure 7.23). The film's first scenes of love and sex have
involved homosexual couples; one of the men dies of AIDS (and speaks directly to the
camera). So our earnest businessman might well be commenting on the action we've
 Poetics of Cinema
240
Figure 7.22 . . . a shot of the youth, appar-
Figure 7.21 Les Passagers: The ranting busi-
nessman, declaring that there are three sexes. ently watching him, perhaps the target of the
At first it seems to be another direct-address rant.
monologue, like that of the Narratrice, but
it's interrupted b y . . .
Figure 7.23 Les Passagers: On occasion,
although speaking to us, the Narratrice
looks away.
seen so far. The equation of gay and straight sex will continue throughout the film.
Guiguet has claimed that the monologue, placed nearly halfway through the running
time, encapsulates the film's sexual themes. 65
Thereafter, stories proliferate. From Pierre greeting Anna and taking her to the
Bach concert, we move sidewise to Anna with her friend Isabella, and then to another
monologue in which Isabella, in somewhat confused fashion, confesses her fantasies
of the perfect life. We see the Narratrice as a senior physician in the hospital, where
she mentors other women staff. After Christine meets Pierre on the train, we will
eventually follow her home and learn that she is the Narratrice's daughter. This
middle stretch of Les Passagers cracks any sense of a consistent time frame. No titles
or appointments break the film into clear-cut days, and shots of the city in both day
and night are intercut with the scenes. All these actions couldn't occur in a single day
or two, yet some characters wear the same clothes on every appearance.
As the characters' love affairs develop and time becomes indefinite, the tram
ceases to be a framing device and serves merely to punctuate the increasingly dis-
persed network. Between episodes, we no longer ride the tram with the characters,
but images of its travels and the cityscapes float up as lyrical interludes. Gradually,
shots of the tram's circuit are replaced by grim tower-block apartments, abandoned
factories, and other marks of industrial desolation. This introduces a third theme to
Mutual Friends and Chronologies of Chance 241
Figure 7.24 Les Passagers: In an abstract
space, the characters deliver apothegms to
the camera while dressed in monkish cloaks.
match those of death and sex: the degradation of modern work. As the film nears its
conclusion, even these interludes are curtailed and the narration cuts, unapologeti-
cally, from one story line to another. Some problems of love between Anna and Pierre
seem temporarily solved, and the Narratrice seems able to comfort an anxious hospi-
tal worker. If this seems like a victory for orthodox storytelling, it's only temporary.
Christine returns home and tells her mother that she's quit her job because of an
aggressive memo from her boss. The Narratrice commiserates, and Christine sits
down before the television, which is announcing workers' demonstrations pressing
for full employment. In a shot very unusual for the film, the camera tracks forward to
her anxious face. The rules change again, and we get a new sort of sequence. Against
abstract black backgrounds, people—nearly all of our characters, and many m o r e -
dressed in monklike robes address us (Figure 7.24). In a flurry of shots, they recite
platitudes ("Let's instill company spirit") or critiques ("Business despises its work-
force"). The chorus of severe commentators culminates in Christine's question to us:
"And man—when will he matter?" The Aimee Mann sing-along in Magnolia daringly
brings conventions of the musical into the introspective melodrama, but Guiguet
moves in a different direction. Taking place in no recognizable realm, the sequence
abandons characterization and causal logic for the sake of art-cinema ambiguity. Is it
a projection of Christine's worries as she sits before the television? Or a denunciation
of business amorality, issuing from the h 1 mmaker-auteur? The film invites us to
consider both prospects. The sequence ends with a survey of desolate industrial land-
scapes, accompanied by a plaintive popular song about loss, a passage that caps the
lyrical interludes that have been scattered through the film.
The chorus of ordinary people also brings to a new pitch one of the major strata-
gems in the film's game, the to-camera address. Introduced early enough for us to get
accustomed to it, the device is first reserved for the Narratrice before being assigned,
ambivalently, to the nurse Anna reading to her patient. (She's also seen against a black
background, foreshadowing the dark vacuum that houses the chorus at the close.)
There remains a margin of play throughout: The Narratrice may sometimes be talk-
ing to another character when her gaze becomes oblique, and we've already seen the
false lead set up by the businessman's monologue (Figures 7.21-7.22). Still later, when
242 Poetics of Cinema
Figure 7.25 Les Passagers: The Narratrice as Figure 7.26 The final scene of Les Passagers:
nurse, talking to another nurse offscreen, in The Narratrice bids goodnight to those sleep-
frontal shot/reverse shot like that of Figure 7.7. ing in their graves.
The more traditional device is introduced
after the unconventional usage, to revise the
film's narrational norm.
a woman refuses to give money to a beggar on the tram, she looks at us. During
Isabella's monologue, which presents her hopes as practically schizophrenic, there's a
full-blown uncertainty. Like some of the witnesses in Vagabond (pp. 166-69), Isabella
might be talking to us, or to someone offscreen (though we're never shown anyone else
present). Finally, in the hospital, the Narratrice is talking to the other women in the
story world, but the camera position is that of her earlier monologues (Figure 7.25).
The effect is rather like the double-layered effect of the businessman's harangue on
sexual identity. By the time we get to the montage of faces, the effect is unequivocal.
There's no doubt they're addressing us—now the problem revolves around who these
people are and from what spot they're speaking.
In an epilogue that rhymes with the opening, the tram slides through the night.
Only the Narratrice remains on board as it passes the cemetery. She looks out at the
gravestones and says, "Good night, my sleepers." The motif of death, prefigured from
the start as a one-way tram trip and linked to AIDS, closes the film, and the woman
looks out at her sleepers, not at us (Figure 7.26). She has become a character again, and
the story world has been made whole. If we watched this brief scene out of context,
we'd find nothing unusual about it. Yet after the interwoven stories and the lyrical
interludes and the variations of to-camera address, this ordinary sequence is thrown
into relief as an elegiac conclusion. One task of art-cinema narration is to place more
classical narration within a wider context of expressive possibilities. Once we've
strayed far from the customary path, a return to it can be invested with an emotional
simplicity that feels earned.
Sustaining the Network
I write this in the winter of 2005-2006, when it's clear that network films are flourish-,
ing. Syriana, for example, explicates a large-scale political process by following several
major players with parallel purposes. Already there are announcements of Mischief
Night from the United Kingdom (an ensemble cast in a multicultural English town)
Mutual Friends and Chronologies of Chance 243
and Sad Movie from South Korea ("a story of four couples whose fates intertwine"). 66
By the time you read this, you may have already seen Conversations on a Sunday
Afternoon (from South Africa), 2:37 from the U.S. (six high school students whose
lives intersect), the Iranian film Crossroads (several lives linked through . . . an auto
accident), All for Love (six romantic couples mingling in Seoul), Oktoberfest from
Germany (a long day at Munich's city wide beer party), Heavens Doors (a Moroccan
film about three lives linked by a revenge murder), No Sweat (seven characters meet
weekly in a sauna), Selon Charlie (seven males obscurely connected over three days),
and of course A Prairie Home Companion, which showed that three decades after
Nashville Altman could still practice his signature format.
Is the form getting worn out? In an ominous development, the New York Times has
discovered it.67 A review of the DVD release of Crash deprecates its "fashionable inter-
secting story-line structure." 68 In January 2005, Variety's most influential reviewer
castigated a new entry: "A protracted parade of woefully familiar motifs from the
Amerindie playbook, Happy Endings comes off like an undernourished Paul Thomas
Anderson wannabe." 69 In the summer, another Variety writer was more hopeful, but
she did acknowledge that things were getting stale. "Just when multicharacter criss-
crossers were threatening to become passe, along comes biting Brit black comedy
Festival, reinvigorating faith in the format." 70
My survey should indicate that the conventions shaping the crisscrosser are fairly
stable, and variations upon them have been rung for a very long time. But modern
artists are competitive, and many aren't content with mere imitation. So we should
expect that once the format is firmly established, filmmakers will seek new twists to
set their work apart. We can regard the blending and the crossover with melodrama
and romantic comedy as mild efforts toward overhaul. And the artistic arms race
can take more fancy forms. Each of the five interwoven strands in Istanbul Tales
(2005) is handled by a different director and cites a different fairy tale (Snow White,
Cinderella, Sleeping Beauty, etc.). In The Five Senses (2000), every major character
is identified with a different sense mode. At the limit, a filmmaker can invoke the
network premise only to dispel it. When a film presents independent story lines that
never meet, as in La Vie Moderne (2000), Four Shades of Brown (2003), and One Day
in Europe (2005), we're somewhat taken aback, as if we'd found a Western without
a gunfight. To wait alertly for connections that never come is an experience these
films want us to have. In sum, the demand for novelty built into commercial film-
making and the festival circuit will constantly press ambitious filmmakers to ring
new changes on familiar schemas.
One way to revivify the format is to go bigger. Network narrative is a signature
brand for Alejandro Gonzalez Inarritu, so it's not surprising that he has released Babel
(2006), a film of global convergence. Presenting four story strands in four languages
(Arabic, English, Spanish, and Japanese), the action takes place in four countries and
on three continents. Once more, the story develops the topos of crossing barriers.
"It will be like a cultural prism that [shows] how we are all connected," claims the
director. 71 But the expanded scope remains justified by conventional means. 72 Most
characters are linked by a rifle that moves from Japan to Morocco, and the action
244 Poetics of Cinema
there triggers events in California and Mexico. The director explains, as if someone
still hadn't heard, "It's this theory about the butterfly that leaves Tokyo and there's a
storm in New York."73
Expanding the story's scope in another way, Lucas Belvaux's Trilogy (2002) spreads
the network principle across three feature-length films. Three couples' lives inter-
twine in a single day, and each film restricts itself chiefly to one pair. The results are
a thriller, a domestic comedy, and a melodrama that can be watched in any order.
A smaller scale of modularity is on display in Greg Marks' 11:14 (2003). In its original
version, it finds a new wrinkle on the shifting-viewpoint and time-juggling conven-
tions. The five story blocks line up at exactly 11:14 p.m., but we are given them in
reverse chronological order. The last one we see takes place 20 minutes before the
first one starts. That's tricky enough, but when the film was released to video, Marks
decided to give his audience new points of entry into the network. The chaptering
of the DVD version allows the viewer to interrupt one line of action by switching to
another occurring at the same time. The story lines will still dovetail, but different
reorderings of the segments will yield different patterns of experience. We may expect
that other filmmakers will start to plan their criss-crossers to allow interactivity.
As ever, the temptation is to talk about all this in a zeitgeisty way. One filmmaker
does our work for us: "I think these movies reflect a sense of disconnectedness." 74 If so,
it's a sense that's been around for 150 years, as reflected in Balzac, Dickens, Eliot, and
others. More plausibly, these films bear the traces of an effort to refresh storytelling
devices in ways intelligible to broad audiences. The fact that we can find network-
based films in virtually every moviemaking culture suggests that the premises of this
sort of storytelling plug into contingently universal appeals and skills. The conven-
tions are easily picked up, recast, and given local significance in Helsinki and Bombay,
Quebec and Hong Kong. I'm not suggesting that there's some panhuman essence that
created network narratives. Rather, we have proclivities toward gathering and weigh-
ing information about social relations, and these are targeted by a type of storytelling
that surfaces as much in soap operas as in art movies.
No doubt, the recent burst of such films owes a lot to the cultural saliency of
Hollywood and Western European filmmaking. Major models like Short Cuts,
Pulp Fiction, and Magnolia probably encouraged filmmakers to try out the criss-
crosser format themselves. Before we treat this as cultural imperialism, though, let's
recall that not all our innovative films come from the First World. Edward Yang's
Terrorizers was made in 1984, and there may be other early non-Western instances.
There are, of course, more immediate conditions for the spread of this form of
filmmaking. The self-consciousness of the films' narration, shouting or whispering
more or less directly to the audience, goes along with filmmakers' growing interest in
sheer artifice. If gray is the new black, form is the new content. The unusual structure
of Pulp Fiction or Memento or The Usual Suspects helped generate buzz. Now, thanks
to film festivals and the Internet and DVD commentaries, directors and screenwriters
everywhere are aware of innovations. Infotainment television programs, fan maga-
zines, websites, and director commentaries all discuss film technique to a degree
Mutual Friends and Chronologies of Chance 245
unheard of even a decade ago. The fact that Nine Lives (2005) consisted of nine long
takes was spotlighted in the film's publicity and commented upon in reviews.
Today a screenwriter or director in almost any country can learn about storytell-
ing experiments elsewhere very fast, and that awareness feeds competition. After all,
films are jostling for international notice. Since the 1990s, the world has been produc-
ing about 4,000 features annually. In Europe, where about 1,000 films are produced
each year, the ambitious filmmaker needs something to make her work stand out. The
same is true in Asia, which produces about 2,000 films per year. And of course the
U.S. market is the most harshly competitive, despite its comparatively small output
of 400-600 domestic titles.75 A few network films like Pulp Fiction, Crash, and Love
Actually have found wide commercial success; some like Nashville and Short Cuts
have become contemporary classics; and many more have won acclaim on the festival
circuit. As the format crystallized in the 1980s and was revivified in the 1990s, film-
makers who embraced it had the choice of replicating its conventions or of revising
them. Like all forms, this one offers problems that can be solved in routine ways or
more ingenious ones. It's up to the filmmaker to decide how much novelty to r i s k -
always remembering that balancing it with familiarity on other fronts will help the
viewer grasp the innovations more confidently.
Network Narrative: A \^orking Filmography
I've been able to see only about four-fifths of the films on the list that follows, so I can't
be sure that every one fits the conception of network form I propose in this essay. Still,
it's probably better to be as comprehensive as I can.
The list runs through mid-2007. Most films have an international title in English,
so I've usually presented that first. I've also indicated directors. Further details on
titles, cast, and other personnel can be found on the Internet Movie Database (http://
www.imdb.com) and other standard sources.
11:14 (Greg Marks, 2003)
12 Storeys (Eric Khoo, 1997)
13 Conversations About One Thing (Jill Sprecher, 2001)
20 30 40 (Sylvia Chang, 2004)
2000 + 1 Shots (Dimitris Athanitis, 2001)
21 Grams (Alejandro Gonzalez Inarritu, 2003)
2:37 (Muralli K. Thalluri, 2006)
71 Fragments of a Chronology of Chance (Michael Haneke, 1994)
9 (Umit Unal, 2002)
Adieu, plancher des vaches (Farewell, Home Sweet Home; Otar Iosseliani, 1999)
Adrift in Manhattan (Alfredo de Villa, 2007)
Adventures of a 10-Mark Note (Die Abenteuer einses Zehnmarkscheines; Bertold
Viertel, 1926)
Alila (Amos Gitai, 2003)
All for Love (Min Gyu-dong, 2005)
246
American Gun (Aric Avelino, 2005)
Among Adults (Stephane Brise, 2007)
Amoresperros (Alejandro Gonzalez Inarritu, 2000)
Amourfou, L' (Jacques Rivette, 1969)
And Now.. . Ladies and Gentlemen (Claude Lelouche, 2002)
Any Way the Wind Blows (Tom Barman, 2003)
April (Otar Iosseliani, 1962)
Babel (Alejandro Gonzalez Inarritu, 2006)
Bangkok (Colin Drobnis, 2006)
Beautiful People (Jazmin Dizdar, 1999)
Before the Rain (Milcho Manchevski, 1994)
Belle Histoire, La (Claude Lelouche, 1992)
Betty Fisher et autres histoires (Alias Betty; Claude Miller, 2001)
Bobby (Emilio Estevez, 2006)
Broken Hearts (Rafael Montero, 2001)
Bug (Phil Hay and Matt Manfredi, 2002)
Buttoners (Petr Zelinka, 1997)
Canicula (Dog Days; Alvaro Garcia-Capelo, 2002)
Cape of Good Hope (Mark Banford, 2004)
Carnages (aka Carnage; Delphine Gleize, 2002)
Century's End, A (Song Neung-han, 1999)
Chasse aux papillons, La (Chasing Butterflies; Otar Iosseliani, 1992)
Chelsea Walls (Ethan Hawke, 2001)
Chromophobia (Martha Fiennes, 2005)
Chungking Express (Wong Kar-wai, 1994)
Citizens Band (aka Handle with Care; Jonathan Demme, 1977)
City of Hope (John Sayles, 1991)
City of Sadness (Hou Hsiao-hsien, 1989)
Club Havana (Edgar G. Ulmer, 1945)
Code inconnu (Michael Haneke, 2000)
Colmena, La (The Hive; Mario Camus, 1982)
Colossal Youth (Pedro Costa, 2006)
Colour of Happiness (Jozsef Pacskovszky, 2003)
Company, The (Robert Altman, 2003)
Conversations on a Sunday Afternoon (Khalo Matabane, 2005)
Crash (Paul Haggis, 2004)
Crash Test Dummies (Jorg Kalt, 2005)
Cross-Harbour Tunnel (Lawrence Wong, 1999)
Crossroads (Abol Hasan Davoudi, 2006)
Dangan Runner (Tanaka Hiroyuki, aka Sabu, 1996)
Dark Horse (Dagur Kari, 2005)
Day a Pig Fell in the Well, The (Hong Sang-soo, 1996)
Days Like This (Mikael Hafstrom, 2002)
Days of Being Wild (Wong Kar-wai, 1990)
Poetics of Cinema Mutual Friends and Chronologies of Chance 247
Dead Girl, The (Karen MoncriefF, 2006)
Decalogue, The (Krzysztof Kieslowski, 1989)
Distant Lights (Hans Christian Schmid, 2003)
Dodes'kaden (Kurosawa Akira, 1970)
Dog Days (Ulrich Seidl, 2001)
Do Over (Cheng Yu-chieh, 2006)
Drive-In (Rodney Amateau, 1976)
Earrings of Madame de..The (Max Ophuls, 1953)
Edge of Heaven (Fatih Akin, 2007)
Emergency Hospital (Lee Sholem, 1956)
Exotica (Atom Egoyan, 1994)
Fallen Angels (Wong Kar-wai, 1995)
Falling Sky (Gunnar Vikene, 2002)
Fast Food Nation (Richard Linklater, 2006)
Favoris de la lune (Otar Iosseliani, 1984)
Festival (Annie Griffin, 2005)
Fiasco (Ragnar Bragason, 2000)
Five Senses, The (Jeremy Podeswa, 1999)
Flavors (Raj Nidimoru and Krishna D. K., 2004)
Flowers of Shanghai (Hou Hsiao-hsien, 1998)
Forest, The (Benedick Fliegauf, 2003)
Free Radicals (Barbara Albert, 2003)
Frozen Land (Aku Louhimies, 2005)
Full Frontal (Steven Soderberg, 2003)
Go (DougLiman, 1999)
Goldfish Memory (Liz Gill, 2004)
Gosford Park (Robert Altman, 2001)
Great New Wonderful, The (Danny Leiner, 2006)
Guitar Mongoloid (Ruben Ostlund, 2004)
Gun (From 6 to 7:30 P.M.), The (Vladimir Alenikov, 2003)
Happenstance (Battement dailes dupapillon; Laurent Firode, 2000)
Happiness (Todd Solondz, 1998)
Happy Endings (Don Roos, 2005)
Happy New York (Janusz Zaorski, 1997)
Hard Luck Hero (Tanaka Hiroyuki, aka Sabu, 2003)
Hawaii, Oslo (Erik Poppe, 2004)
Health (Robert Altman, 1980)
Heart (Horst Johann Sczerba, 2001)
Heaven's Doors (Swel Noury and Imad Noury, 2006)
Hold Up Down (Tanaka Hiroyuki, aka Sabu, 2005)
Home (Matt Zoller Seitz, 2006)
Hommes femmes: Mode d'emploi (Claude Lelouche, 1996)
Honky Tonk Freeway (John Schlesinger, 1981)
I Am an Apartment Building (Lara Azzopardi, 2006)
248 Poetics of Cinema Mutual Friends and Chronologies of Chance 249

II y a des jours... et des lunes (Claude Lelouche, 1990) Private Fears in Public Places (Alain Resnais, 2006)
Ingorgo—Una storia impossibile, L' (Luigi Comencini, 1979) Pulp Fiction (Quentin Tarantino, 1994)
In the City (Cesc Gray, 2003) Rave Fever (Mak Siu-fai, 1999)
InterMission (John Crowley, 2004) Ready to Wear (aka Pret-a-porter; Robert Altman, 1994)
Istanbul Tales (Umit Unal, Kudret Sabanci, Selim Demirdelen, Yucel Yolcu, and Rush (Lee Sang-in, 1999)
Omur Atay, 2005) Sad Movie (Kwon Jong-kwan, 2005)
Joki (The River; Jarmo Lampela, 2001) Safety of Objects, The (Rose Troche, 2001)
Km. 0 (Yolanda Garcia Serrano and Juan Luis Iborra, 2000) Sauve quipeut (la vie) (Jean-Luc Godard, 1980)
Lawless Heart (Neil Hunter and Tom Hunsinger, 2001) Scenes of a Sexual Nature (Ed Blum, 2006)
Leave on Word of Honor (Karl Ritter, 1938) See You in Space! (Jozsef Pacskovszky, 2005)
Life as It Comes (Stefano Incerti, 2003) Selon Charlie (Nicole Garcia, 2006)
Loggerheads (Tim Kirkman, 2005) Sexual Life (Ken Kwapis, 2004)
Look Both Ways (Sarah Watt, 2005) Ship of Fools (Stanley Kramer, 1965)
Love Actually (Richard Curtis, 2003) Situation, The (Philip Haas, 2007)
Love and Happiness (Jordan Allan, 1995) Standing Still (Matthew Cole Weiss, 2005)
Magnolia (Paul Thomas Anderson, 1999) Strawberry Shortcakes (Yazaki Hitoshi, 2007)
Manual of Love, The (Giovanni Veronesi, 2005) Street of Shame (Mizoguchi Kenji, 1956)
Me and You and Everyone We Know (Miranda July, 2005) Sud Express (Chema de la Pena; Gabriel Velazquez, 2005)
Mind the Gap (Eric Schaeffer, 2004) Sunshine State (John Sayles, 2002)
Monsoon Wedding (Mira Nair, 2001) Tales of Manhattan (Julien Duvivier, 1942)
Motel Cactus (Park Ki-yong, 1997) Tapas (Jose Corbacho and Juan Cruz, 2005)
Nashville (Robert Altman, 1975) Terrorizers, The (Edward Yang, 1986)
New Age of Living Together, The (aka In Between; Samson Chiu, Yeung Fan, and They All Laughed (Peter Bogdanovich, 1981)
Sylvia Chang, 1994) Things to Do Before You're 30 (Simon Shore, 2004)
Nicotina (Hugo Rodriguez, 2003) Things You Can Tell Just by Looking at Her (Rodrigo Garcia, 2000)
Night of the Sunflowers (Jorge Sanchez-Cabezudo, 2006) Three Palm Trees (Joao Botelho, 1994)
Night Shapes (Andreas Dresen, 1999) Three Seasons (Tony Bui, 1999)
Nine Lives (Rodrigo Garcia, 2005) Tic Tac (Daniel Alfredson, 1997)
Nines, The (John August, 2007) Top Banana Club (Anthony Wong, 1996)
Noel (Chazz Palminteri, 2004) Tout qa ... pour gal (All That.. .for This?!; Claude Lelouche, 1993)
Nosotras (Women; Judith Colell, 2000) Toute une vie (And Now My Love; Claude Lelouche, 1974)
No Sweat (Eoin Moore, 2006) Trafico (Joao Botelho, 1998)
Oktoberfest (Johannes Brunner, 2006) Trilogy (Lucas Belvaux, 2002)
Once Upon a Time in Triad Society 2 (Cha Chuen-yee, 1996) Two Days in the Valley (John Herzfeld, 1996)
Opening (Rob Nilsson, 2006) Uchoten Hoteru, The (aka Suite Dreams; Mitani Koki, 2006)
Out 1: Noli me tangere (Jacques Rivette, 1970) Une pour toutes (Claude Lelouche, 1999)
Parisiens, Les (Claude Lelouche, 2004) Uns et les autres, Les (Claude Lelouche, 1981)
Paris nous appartient (Jacques Rivette, 1960) Up and Down (Jan Hrebejk, 2004)
Park (Pat Holden, 2006) Utopia—Nobody's Perfect in the Perfect Country (Morten Tyldrum et al., 2002)
Passagers, Les (Jean-Claude Guiguet, 1999) Voisins voisines (Malik Chabane, 2005)
Pastorale (Otar Iosseliani, 1976) Way Off Broadway (Daniel Kay, 2001)
Play Time (Jacques Tati, 1967) Weapons (Adam Bhala Lough, 2006)
Prairie Home Companion, A (Robert Altman, 2006) Wedding, A (Robert Altman, 1978)
Premiere juillet, le film (Moving; Philippe Gagnon, 2004) Weirdsville (Allan Moyle, 2007)
250 Poetics of Cinema

What a Wonderful Place (Eyal Halfon, 2005)

What's Wrong With This Picture? (Lotte Svendsen, 2004)
When a Man Falls in the Forest (Ryan Eslinger, 2007)
White Dress, The (Michal Kwiecinski, 2004)
Who's Camus, Anyway? (Yanagimachi Mitsuo, 2005)
Witnesses, The (Andre Techine, 2007)
III
Wonderland (Michael Winterbottom, 1999)
Yacoubian Building, The (Marwan Hamed, 2005)
Year Zero (Joseph Pitchade, 2005) Studies in Style
Yellow Rolls Royce, The (Anthony Asquith, 1964)
Yuva (Mani Ratnam, 2004)

251
 8.
Cinecerity

At a state college up the Hudson, in the fall of 1965, the English department hosted
a debate on contemporary American movies. The adversaries were Pauline Kael and
Andrew Sarris, both just coming into the national spotlight. One freshman came
early and sat with comic earnestness in the front row. In high school he had devoured
the "American Directors" issue of Film Culture but had also enjoyed Kael's I Lost It at
the Movies (and its devastating essay on The Group).1
Kael was witty and acerbic, tossing off judgments with a wave of her cigarette
holder. The Cincinnati Kid was the best of the current crop; Norman Jewison showed
great promise. Sarris, in a crumpled suit, was resolutely uncharismatic, looking
mildly unhappy to be dragged blinking out of the Thalia and shipped upstate. He
talked fast, interrupted himself, and, finding few recent movies to praise, celebrated
Max Ophuls and Jean Renoir. He delivered enigmatic observations like "All movies
should probably be in color."
At evenings end, I knew which camp I belonged in. I obtained an appropriately
nerdy autograph: "Cinecerely yours, Andrew Sarris." More important, I exulted in a
sense that my almost grim obsession with movies had been validated. Not for some
time would I realize that I had enlisted, to put it melodramatically, in a fight for
American film culture.2 The battle lines were drawn more sharply in Manhattan than
in Albany, but everywhere one thing was clear. Kael was clearly the standard bearer
for them, and Sarris was ours. They were the hip intellectuals who enjoyed a night at
the movies, but who wanted a critic to be a little superior to what she criticized. At
faculty parties throughout the 1970s, I had to check myself when professors of lit or
law asked me if I didn't find Kael's review that week intoxicating. And she writes so
well! As chief New Yorker critic, Kael became the grand tastemaker, a person even

253
254 Poetics of Cinema
filmmakers courted. But who would try to curry favor with the guy who wrote for the
Village Voice? It puts you in mind of the joke about the starlet so dumb she slept with
the screenwriter.
Of course, like all acolytes we overplayed the differences. Both Kael and Sarris
loved classic studio cinema, the performers and scripts as much as the directorial
touches. Both critics bemoaned Soviet montage and what they saw as its descen-
dant, the overbusy technique of the 1960s. Both deplored stylistic aggressiveness (the
Tony Richardson syndrome), and both idolized Renoir. Both loved lyricism. Yet we
understood that in talking about contemporary cinema, Kael wrote for people who
admired Bogie and found Jean-Paul Belmondo cool. She was, in both a good and a
bad way, a sensationalist, and her most provocative judgments have not worn well.
She could not see that Bernardo Bertolucci was trying to be Jean-Luc Godard, Alain
Resnais, and Pier Paolo Pasolini all in one, and, failing, became the very exemplar of
a metteur en scene. She wrote for those who dug movies. Sarris wrote for those who
loved cinema—the medium itself, or rather an exalted idea of the medium's potential,
an ideal form of expression at once dramatic, poetic, pictorial, and musical. He saw
each film as bearing witness to the promise of what cinema might be, and he looked
in even the tawdriest products for something approximating his dreams. Despite his
enthusiasms, he tried for detachment, viewing the latest movie from some unpredict-
able historical perspective. Kael famously declared that she watched a new film only
once, because that was the way her readers would consume it. But who, we thought,
would want to write about a movie after seeing it only once? If it was a good movie,
who wouldn't want to see it more than once? In that same freshman semester, when
there were still first-run double features, I sat through The Glory Guys twice in order
to watch Help! three times. And that wasn't a very good movie.
Kael looked for faults; Sarris, for beauties. Kael made each weak film seem like a
blow to her intelligence. Sarris forgave. He taught that it was better to leave a door
open than to write someone off—even Bertolucci, even Ingmar Bergman, whom
some of us would never learn to like until Persona, and some not until Fanny and
Alexander. Sarris subordinated his personality to that of the movie and its director,
which made him seem less fiery than his uptown counterpart, but his attitude suited
our own somewhat adolescent amorphousness of character. The arrogant certainty
of our tastes was, we thought, born of a passionate humility, a sense of serving wise
masters named Carl Theodor Dreyer, F. W. Murnau, and Mizoguchi Kenji.
Sarris'lessons were many. His taste was uncannily accurate. The American Cinema,
the book derived from the Film Culture issue, remains the best guide to what is lasting
in Hollywood studio filmmaking. 3 More surprising, the man who wrote the gospel
on auteurism assiduously read the source novels or plays for most of the movies he
reviewed. It was as if by sizing up the raw material, he could better come to grips with
how the director had transmuted it. And not least, in Sarris we found someone willing
to treat popular movies as a pictorial art.
This side of his work hasn't been fully appreciated, perhaps because auteurism
quickly became a code word for recurring themes. After all, doesn't the auteur theory
hold that a director communicates a "vision of the world " ? And wouldn't this entail that
Cinecerity 255
the auteur's world—the stories, the characters' attitudes and actions—take primacy?
And wouldn't that mean that the critic's task is to concentrate on interpreting plot
and character? Certainly that chain of faulty induction became what the "auteur-
structuralists" took from Sarris. They built large-scale thematic complexes out of his
suggestions about continuities of directorial vision. 4 And I once noted Sarris' 1956
essay on Citizen Kane as a key source of what have become well-practiced routines of
film interpretation. 5 1 think I was right up to a point, because Sarris' readings power-
fully influenced those of his successors. Nonetheless, I didn't sufficiently emphasize
the ways in which other aspects of his work revealed a sensitivity to visual style that is
still rare in writing about movies.
Sarris puts style on the agenda because he believes, after all the qualifications are
granted, that most of the best films show the dominance of the director's individual
sensibility, and they show it through mise-en-scene, which Sarris defined as
all the means available to a director to express his attitude toward his subject.
This takes in cutting, camera movement, pacing, the direction of players and
their placement in the decor, the angle and distance of the camera, and even the
content of the shot. 6
We can quibble about the term—I'd say that Sarris is here talking about visual style
in toto—but the premise itself is fertile. It won't explain everything in all American
studio movies, but it remains the best way to understand why so many of them are extra-
ordinary. Here we come to what I believe to have been Sarris' most important insight.
Style and Film History
Where does cinematic beauty come from? Sarris' answer has been straightforward: It
comes from expressive style. He has argued that such a notion of style could define
cinema as an art, mark out its most worthy achievements, discriminate among
artists, and trace historical patterns of continuity and change. He proposes, that is, a
conception of cinematic style congruent with that proven successful in the fine arts.
He also offers a sophisticated account of how a mass-produced popular art could
display distinctive ways of using its medium. To understand beauty in the movies,
Sarris insisted, we must be sensitive to style.
Stated so generally, this is a pretty old argument. The best early commentators on
cinema strove to define the techniques of the new medium, particularly as it differed
from theater. By the end of the silent era, theorists and critics had generated a rich
vocabulary for talking about camera placement, composition, and cutting. Many of
these ideas can be found conveniently summarized in Rudolf Arnheim's Film (1933)
and Raymond Spottiswoode's A Grammar of the Film (1935).7
Sarris has objected to this line of thinking for its uncritical celebration of silent
montage and its contempt for the mainstream movie as an enacted story aimed at
mass audiences. Still, this tradition has given him one of his most punishing polemi-
cal weapons, a tacit notion of cinematic specificity. He scorns "literary" critics who
refuse to pay attention to what's on the screen. "The vertical camera movement down
256 Poetics of Cinema
a tower after Arkadin's death supersedes the script and quite naturally escapes the
attention of our literary film critics."8 And when he upbraids directors who forget pic-
torial expressivity, Sarris can come on like any anti-talkie proponent of visual values.
"As one of the breed of writer-directors, [Richard] Brooks has a bad habit of saying
what he means without showing what he feels."9
Sarris, however, would probably protest that any ideas he borrowed from the
earliest theorists of style are far less important than his debt to Andre Bazin. Bazin and
his contemporaries revised and corrected silent-era ideas by arguing for the centrality
of staging, camera movement, and other aspects of mise-en-scene. These techniques
became salient with the coming of sound, and Bazin and his colleagues revealed them
to be as rich in expressive possibilities as stylized photography and complex cutting.
Undertaking his examination of American cinema after 1929 (a period that almost
exactly coincides with his own lifetime), Sarris was understandably drawn to a critical
perspective that saw talkies not as a betrayal of cinema's essence but a development of
some of the medium's richest possibilities.
Each school of thought conceived of cinematic style differently, but both of them
converged on the belief in a universal stylistic history of the medium as such. Many
silent film critics treated cinema as a historical actor in its own right, striding inexo-
rably toward the full revelation of its distinctive artistic essence. The critics' emphasis
on style led to an idea that cinema was evolving toward a predetermined goal, often
conceived as the full exploitation of the "essentially cinematic" qualities of montage.
Bazin was as aesthetically catholic as he was spiritually Catholic, but he too fell in with
this habit of thinking, replacing the old teleology (development toward "pure cinema")
with a new one (development toward "total cinema," the most realistic of all).10
In setting out to write the history of the American sound film—Sarris' stated
purpose in formulating the auteur theory—he has had no patience with either version
of the "evolution of film language." He has criticized the very idea as "pyramid
history," with each creator patiently putting into place his or her contribution to the
ever-ascending edifice of Cinema as it tapers toward ultimate realism or pure artifice.11
Sarris offers instead an inverted pyramid, "opening outward to accommodate the
unpredictable range and diversity of individual directors."12 For Sarris, a genuine
history of the cinema will not be the tale of a mystical rise and fall of the medium
independent of the artists who use it. "Griffith, Murnau, and Eisenstein had differing
visions of the world, and their technical 'contributions' can never be divorced from
their personalities." 13 As film history exfoliates from fairly simple and unitary begin-
nings, we will find that the careers of the creators diverge, overlap, and interlace in
unexpected ways. We have something closer to a network than to linear progress;
not a tree but a bush. In effect, Sarris champions a version of what historians call
methodological individualism. History is made by persons; even institutions, group
processes, and "impersonal forces" must be explained, finally, through the activities
and temperaments of people.
True, there will be some broader trends to notice, but even these will manifest
through the interplay of directorial personalities. For example, 1930s Hollywood can
be emblematically represented in the image of Leo McCarey and Frank Capra playfully
Cinecerity 257
tussling over an Oscar, "a very brief moment before the surge of John Ford and Orson
Welles."14 Stylistically, the 1930s can be summed up in Ernst Lubitsch's unnoticeable
editing, whereas the 1940s belong to "the Wellesian resurrection of Murnau's porten-
tous camera angles. The decade of plots gave way to a decade of themes."15 And, "As
Lubitsch was the unobtrusive cutting of the twenties and thirties, Preminger is the
camera movement and long take of the fifties and sixties."16 It is not just the master-
works—The Awful Truth, Mr. Smith Goes to Washington, Stagecoach, How Green Was
My Valley, Citizen Kane, The Marriage Circle, One Hour With You, Laura, and Fallen
Angel—that make history; in the passages I've cited, Sarris doesn't mention any titles.
The filmmakers' names come to mark the historical turning points, and the names
in turn suggest entire bodies of work and distinctive personalities. Sarris' move looks
even more audacious when you consider that the careers ran in counterpoint. John
Ford started making films in 1917, but his "surge" comes in the late 1930s; Lubitsch
keeps making films into the 1940s, but as a personality he is irrevocably tied to an
earlier time. That doesn't mean that Lubitsch's later films are negligible, of course; it is
just that if his 1940s works preserve his authorial sensibility—his civilized grace, and
his amused tolerance of self-deception—they do so within a new milieu.
The paradoxical result of Sarris' conception of directorial individuality is the fact
that he has never produced an orthodox survey of American film history.17 The lists
and rankings tabulated in The American Cinema, he has conceded, sought both to
map an unknown region and to create a subject worth talking about.18 They display
the open-textured, contrapuntal nature of his conception of film history. These
categories and chronologies constitute only the raw materials of a broad-based film
history. Neither argument nor narrative, they do not of themselves solve the riddle of
how American cinema as a whole manifests large-scale continuity and change.
Still, this is probably not the riddle that really preoccupies Sarris. In my view,
The American Cinema represents not a study in film history so much as an influen-
tial reference book (the forerunner of today's fan books) and a work of historically
informed criticism. That is, instead of deriving a broad historical argument from
critical analysis and appraisal, Sarris has been a connoisseur, using his vast historical
knowledge to shape his scrutiny of particular movies and oeuvres. Although I doubt
that we can build a film history by aligning filmographies, surely thinking of
Hollywood cinema as displaying stylistic continuity and change within individual
careers can produce excellent criticism. Sarris' writings are dazzling proof.
Style and the Critic
"The unconceptualized eye is at the beginning and the end of all visual appreciation."19
This line from 1973 sounds a bit sixtiesish to us now, but its Ruskinian trust in inno-
cent perception is characteristic of a critic who always insisted that film criticism had
to attend to what was happening on the screen. One of Sarris' great gifts to American
film culture has been to show how visual style is central to film as an art.
This emphasis is particularly intriguing coming from a writer determined to
explore the history of sound cinema. With the coming of the talkies, most critics
 '8 Poetics of Cinema
Figure 8.1 The Searchers (1956). Drinking Figure 8.2 The Searchers: His point of view
his coffee, the Ward Bond character glances reveals her gently stroking the uniform of her
toward the wife of the household. husband's brother.
lamented the death of pictorial values. The expressive compositions and dynamic
cutting of the silent classics seemed ill-suited to a dialogue-based cinema. Bazin,
however, saw that the sound film required a more unobtrusive but no less supple
visual style than had developed in the official classics of the 1910s and 1920s. Bazin's
godsons among the Cahiers gang did not for the most part pursue this line of thought.
If the auteur theory indeed constituted a renewal rather than a brand-new position,
it was partly because Sarris could show that expressive uses of film technique had
not withered after 1928.
For one thing, he could show that the director, as the creator of the visual track,
shaped the overall context that would govern how the script was taken. The dialogue
and the plot situation could be reweighted through performances, camera angle,
camera movement, and cutting. Moreover, a producer was less likely to object to a
director's shots than to a writer's words, so the director could inflect, nuance, or even
work against the script. Similarly, Sarris realized that what silent-era aestheticians
castigated as the talkies' betrayal of film's plastic essence could be considered as "an
extraordinary economy of expression."20 For example, sound encouraged a density of
audiovisual texture. Alfred Hitchcock proved himself "alive to the expressive poten-
tialities of every encounter. His cutting is the means by which he contradicts what
people say by what they do."21
Sarris refines his account by noting that each director is free to discover distinctive
strategies of stylistic economy. Howard Hawks' style calmly renders every twist in the
plot, embodying his clean-lined pragmatism; he doesn't use technique as a "reflective
commentary" on the action. 22 Ford, however, adds grace notes. "He could always spare
a shot or two for a mood that belonged to him and not to the plot."23 The key phrase
is "a shot or two." In the brilliant scene in The Searchers in which Ward Bond chival-
rously averts his eyes from Wayne's sister-in-law as she strokes Wayne's cavalry coat,
Ford knows just how far to go (Figures 8.1-8.3). "If it had taken him any longer than
three shots and a few seconds to establish this insight into the Bond character, the
point would not be worth making." 24 The sound cinema imposes a discipline on all
directors, and the best ones find idiosyncratic ways to be at once crisp and evocative.
We can see Sarris' perspective taking shape as early as the notorious "Notes on
the Auteur Theory in 1962."25 Although sketchy and programmatic, as a manifesto
Cinecerity 259
Figure 8.3 The Searchers: Cut back to Bond
as the wife and her brother-in-law meet for
a tender farewell. Bond stares off into space.
"Nothing on earth," writes Sarris, "would
ever force this man to reveal what he had
seen. There is a deep, subtle chivalry at work
here, and in most of Ford's films, but it is
never obtrusive enough to interfere with the
flow of the narrative."
Source: The American Cinema (Chicago: Uni-
versity of Chicago Press, 1985), 47.
should be, this essay lays out some key premises for the study of authorial style in
the cinema. I want to consider them briefly, because although they are well-known,
Sarris' points remain worth mining today.
Sarris envisions authorial expression as a series of concentric circles. The central
circle represents the premise that a director's "technical competence" is a fundamental
criterion of value. "A great director has to be at least a good director."26 In 2001, for
instance, Stanley Kubrick fails "to tell a story on the screen with coherence and a
consistent point of view. Kubrick's tragedy may have been that he was hailed as a
great artist before he had become a competent craftsman." 27 By contrast, the classical
period of Hollywood cinema was characterized by a "relaxed craftsmanship" sadly
absent from the screen of the 1960s.28
The criterion of craft is not such an obvious point of departure as it might seem. By
taking technical skill as a baseline of inquiry, Sarris obliges the critic to understand
what collective norms are operating in the directors milieu. The critic must watch a
lot of movies to gain a sense of what is minimally competent. Although Sarris did not
pursue this avenue, the quest for tacit craft knowledge can become a legitimate pursuit
for the student of style. We can shed much light on how movies are put together by
studying a period's prevailing practices—the rules of thumb, the taken-for-granted
procedures. Alexandre Astruc outlined the possibility as early as 1946:
On the sound stages of Hollywood there was passed along a sort of empirical
grammar formed from the long experience of highly devoted artisans. They
knew for example that near the end of a film it was better to increase the num-
ber of close-ups in order to raise the degree of emotion. They also knew that the
plan americain [the knees-up framing] was the most efficient shot, permitting
the greatest economy of editing.
 Poetics of Cinema
260
This technique may have lacked ambition, but it was faultless and sure. Today it
29
would still be interesting to analyze its smallest details.
More recently, some scholars have taken up the Astruc-Sarris challenge, investigat-
ing what constituted technical competence in different filmmaking milieus. 30
To halt our inquiry at this circle, though, risks turning us into "forest critics," those
generalizers interested only in collective convergences and not individual idiosyncrasies.
Hence the need for a second concentric circle, correlated to the first. This criterion of
value is "the distinguishable personality of the director."31 Sarris notes that the director
is known chiefly through the films (and not, say, the lived life), but just as important,
the personality is grasped in large part through the films' style. "Over a group of films
a director must exhibit certain recurring characteristics of style which serve as his
signature."32 And this style is, as we have seen, primarily visual: "The way a film looks
and moves should have some relationship to the way a director thinks and feels."33
There is much to say about this premise, but to belabor Sarris for using the term
personality, as some academic critics have, is probably not fruitful. The chief point
is this: Of all the hands working on a film, directors have the best opportunity to
blend a film's various ingredients, and visual style is one obvious way in which this
takes place.34 Once this is granted, who would deny that a director's habitual ways
of orchestrating the diverse materials of the medium may reflect the same sorts of
qualities we use to characterize people we know? It's hard not to say that Ozu's work
is sensitive, that Dreyer's films exhibit outrage in the face of religious intolerance, and
that D. W. Griffith's films prize innocence. Perhaps to some extent, directorial signa-
tures allow us to know the people who made movies; but they are primarily ways of
knowing the movies more intimately. And by seeking personal signatures, we can
confirm our sense of diversity expanding across film history, the "inverted pyramid"
that is actually a bush exfoliating outward from its roots.
One useful way to sharpen the critic's sense of stylistic differences is to find a
common problem that two directors solve in difFerent ways. Sarris astutely contrasts
Charlie Chaplin's and Buster Keaton's treatments of a rudimentary gag involving
statues, and he distinguishes Lewis Milestone from King Vidor through their
handling of the stock situation of enemy soldiers confronting one another on the
battlefield. 35 Craft conditions create shared problems; directorial personalities solve
them in ways that reveal subtle differences of purpose and attitude. Once more,
style—a rich ensemble of concrete choices about camerawork and lighting, perfor-
mance and cutting—becomes the center of concern.
A distinguishable personal style is a necessary but not sufficient condition for
being an auteur. "Visual style is never an end in itself.... Any visual style can be
mechanically reproduced." 36 There must also be a distinct expressive quality arising
from that style. Here, at Sarris' outermost circle, is where we encounter the premise
that aroused so much controversy. What does it mean to say that a true auteur's work
manifests "interior meaning"?
I find Sarris' account of interior meaning rather obscure. The term has been taken
to be something like "the deepest significance" or "the most abstract meaning" we can
Cinecerity
261
37
assign to a film or oeuvre. But Sarris is quite insistent that interior meaning can't
be paraphrased. "It is ambiguous in any literary sense because part of it is imbedded
in the stuff of cinema and cannot be rendered in non-cinematic terms." 38 So it can't
be just a vision of the world or a director's attitude toward life, for these we can try to
render, perhaps even eloquently, in language.
It seems to me that interior meaning is best understood as an expressive quality
that arises from the differences we can recognize among directorial personalities.
And although this expressive tone may pervade an entire work, as nostalgic melan-
choly suffuses Ophuls' films, it is just as likely to show up in privileged moments.
Sarris' example is the moment when, in La regie dujeu (1939), Octave turns and hops
in response to Lisette's call, then continues "his bearishly shambling journey" to
Christine. 39 No other director but Renoir, Sarris suggests, would have Octave break
the scene's rhythm in exactly this way (and in a manner, Sarris' bear metaphor hints,
that anticipates Octave's costume for the climactic party). The nonchalant, slightly
awkward interruption epitomizes Renoir's entire style of filmmaking, which makes
everyday gestures unexpectedly graceful. Octave's hop, and its resonance for Renoir's
film and his oeuvre, cannot be adequately described; it is "imbedded in the stuff of
cinema." The critic can merely point to those privileged moments where "interior
meaning" shines forth.
Elusive though such a notion is, the idea that certain stylistic events can crystal-
lize the creator's unique sensibility captures one important way in which discern-
ing viewers watch movies. We can praise large-scale formal effects, like smooth plot
construction or consistency of performance. But what both cinephiles and ordinary
viewers tend to notice and remember are those luminous instants that change our skin
temperature. We come out of a film satisfied if a few epiphanies have flashed upon us.
Interior meaning is the product of craft and personality, but it can't be reduced to
them. It is often a fugitive beauty, all the more precious because it can only be seen,
not described.
Envoi
Sarris' ideas are far from fully recognized today. Granted, Variety now uses auteur
as comfortably as it does blockbuster. Producers swear that they want to support a
director's "vision." Largely because of Sarris, the study of American cinema is now
recognized as a somewhat worthy intellectual pursuit. Auteurism made director
studies respectable enough for publishers to support books on Preston Sturges,
Keaton, and—endlessly—Hitchcock. Still, much hasn't changed.
In the university, the study of film has largely moved toward a bland interdisci-
plinarity in which movies matter less than theories. In 1973, Sarris wrote, "The time
has come to acknowledge film scholarship as an end in itself and not merely as a
means to revitalize the lesson plans of other disciplines."40 He could not have foreseen
that a consumer technology would turn all humanities courses into movie courses
(or, rather, movies-on-video courses), and convert many class meetings into amateur
sociology and wishful political thinking.
262 Poet ics of Cinema

In the literary culture, the caricature of Hollywood remains largely the same as that
circulating in the 1930s: Producers are monsters, writers are martyrs, and directors
are philistine opportunists. The readers of the New York Review of Books have all
their preconceptions about the venality of studio filmmaking regularly reaffirmed by
the waspish reminiscences of Gore Vidal and John Gregory Dunne. For the educated
9.
public, the American cinema remains of interest chiefly as a reflection of national
folly. To the condescensions of pop, camp, and trivia, about which Sarris warned
us long ago, the academy has added a patronizing postmodernism and a populist
cultural studies. As for auteurism, everybody knows the word, but today's film nerds Taking Things to Extremes
often prefer auteurs closer to Tony Richardson than to Samuel Fuller. Sarris taught
us that the director mattered, but he didn't expect that directors could trumpet their
personal vision as a path to celebrity. When directors became superstars, like most Hallucinations Courtesy
superstars they proved vacuous.
The British fan magazine Empire (a slightly more knowledgeable version of the
American Premiere) recently asked its readers to rank the 20 greatest directors of all of Robert Reinert
time. 41 The poll put Steven Spielberg at the top, with Martin Scorsese, Kubrick, Ridley
Scott, Peter Jackson, and Woody Allen in the top 10. Hitchcock won second place,
and Kurosawa Akira and Orson Welles slipped in too. In the next tier of 10, Ford
and Billy Wilder squeezed in, but below Clint Eastwood, David Lean, Joel and Ethan
Coen, James Cameron, Quentin Tarantino, and Oliver Stone. Hawks, Fritz Lang,
and Chaplin popped up in the top 40, alongside middlebrow masters like Anthony
Minghella and Peter Weir. Absent were Griffith, Keaton, Harold Lloyd, Murnau,
Lubitsch, McCarey, Josef von Sternberg, Renoir, Godard, Jacques Tati, Robert Bresson,
Jean-Pierre Melville, Mizoguchi, Ozu, Naruse Mikio, Vincente Minnelli, William
Wyler, George Cukor, Otto Preminger, Samuel Fuller . . . but why go on? The question
isn't what I think of the ranking (though I'm appalled). Rather, the readers' presup-
positions about what counts as a great movie reveal that film history before 1980 is If anyone doubts that progress in film studies is possible, we need only point to our
hardly a living presence for them. increasing understanding of the extraordinary tradition of staging in European films
We shouldn't be surprised. Today's tastes are the legacy of the 1970s. Directors of the 1910s. The first historians of silent film, smitten by montage, criticized the
attuned to psychological and pictorial subtlety couldn't make much headway then, staging tradition as too theatrical and set it against the more progressive and properly
when Hollywood, groping for a new business model, lurched between megapictures "cinematic" cutting of D.W. Griffith. Bazin in turn speculated that the films of this
and heavy-handed "personal" films. Neither The Poseidon Adventure nor Five Easy period harbored a "primitive" deep focus that was eventually to find fulfillment in
Pieces has room for moments like that of The Searchers. The midrange program pic- the work of Jean Renoir, Orson Welles, and William Wyler.1 Across the 1990s, a host
ture was the arena of so many of Hollywood's greatest creators, and its elimination set of studies showed that this tradition constituted a robust aesthetic in its own right. 2
the menu for most popular film consumption today, the choice between the relentless Certainly the 1910s filmmakers' art remained "theatrical" in many respects, but it was
steamrolling of the blockbusters and the preening quirkiness of independent cinema. also "cinematic" by virtue of its reliance on the perspectival qualities of the camera.
For those of us who learned from Sarris, then, there remains much to do. In partic- Thanks to projective optics, the playing space of any shot consists of a pyramid with
ular, although "style" as sensory bombast has become part of a movie's packaging and the apex resting at the lens. The filmmaker can guide our attention within the depths
marketing (X-Men and JFK order up Technique in 40-gallon drums), viewers remain of that fanned-out space by centering figures or props, moving figures to and from the
almost completely unaware of style in any rigorous sense, let alone of its nuances. camera, turning their faces toward and away from us, creating discrepancies of size or
"The auteurists," Sarris remarked over 25 years ago, "are still fighting an uphill battle texture or illumination, and choreographing performers so that they block and reveal
to make movie audiences conscious of style."42 La lutte continue, as the French used crucial actions at the proper moment. 3
to say.

263
264 Poetics of Cinema Taking Things to Extremes 265

Figure 9.1 The Traitress (1912): The Marquis
talks with Lieutenant von Mallwitz, whose
head catches our attention by being tilted to
occupy the center of the picture format.
Figure 9.2 The Traitress: Von Mallwitz lifts
his head to clear a space for the Marquis'
servant, who enters and announces Yvonne.
Figure 9.4 The Traitress: After a moment
of hesitation when she sees von Mallwitz,
Yvonne comes forward. Blocking the servant,
she starts tugging at her father's sleeve to get
him to leave.
Figure 9.5 The Traitress: Von Mallwitz takes
a step toward her, and she nods nervously. Her
realignment reveals the servant, still standing
in the rear; but so as not to call undue atten-
tion to himself, the actor playing the servant
lowers his eyes.

Figure 9.3 The Traitress: As the servant

steps aside, the Lieutenant rises and goes
to frame right, clearing still more space for Figure 9.6 The Traitress: After a final glance
Yvonne's entrance and establishing a remote- at von Mallwitz, she leaves, the servant bowing
ness between them that will have dramatic as she passes.
consequences.

Take for example the 1912 German production The Traitress (Die Verraterin),
directed by Urban Gad and starring Asta Nielsen. In one scene, Lieutenant von Mallwitz
calls on the Marquis de Bougival; the Marquis' daughter Yvonne comes in to join them,
but then leaves coldly in response to von Mallwitz. The drama could hardly be more rudi-
mentary, but Gad stages it elegantly The Lieutenant's head moves aside to clear a path,
first for the servant and then for Yvonne (Figures 9.1-9.3). In her turn, Yvonne masks
and then reveals the servant as she tugs at her father's sleeve (Figures 9.4-9.5). Her exit is
highlighted by the ensemble performance, with all turned to watch her go (Figure 9.6).
Gad has used no cut-ins to stress the dramatic moments, but his choreography high-
lights what we need to see at any moment. A modern viewer is surprised, I think, by
the kind of depth we see in such 1910s shots. Instead of the "big foregrounds" of 1940s
U.S. cinema, we have "distant depth," with small changes taking place in layers quite far
from the camera. Gad steers our attention by centering, movement, the salience of faces
and hands, and the judicious rhythm of blocking, then revealing, scenic elements. On
these simple principles, Louis Feuillade, Evgenii Bauer, Franz Hofer, Georg af Klercker,
Victor Sjostrom, and many other directors worked powerful variations.
The practice of depth staging hung on a little longer than we might expect. In 1917,
Hollywood films like Wild and Woolly and The Clodhopper were presenting rapidly
cut scenes built out of master shots, reverse angles, and close views of objects, faces,
and hands. Yet most European filmmakers persisted in shooting scenes in lengthy
takes with considerable depth. It may seem odd that directors did not rush to embrace
Hollywood-style editing for its production advantages alone; but perhaps we shouldn't
expect filmmakers to abandon the craft that they had long practiced and that had
proved a supple, subtle means of expression. Also, during the war years, filmmakers
in some countries could not see American movies, so it is not surprising that late 1910s
films from several countries—notably, Germany—look stylistically "backward."4
At first, depth-based filmmakers incorporated editing into their scenes in conser-
vative ways. The most common device was the axial cut-in, a straightforward enlarge-
ment of one or two figures at a climactic moment. Bauer's scenes typically include
266 Poetics of Cinema
Figure 9.7 In a later scene of The Traitress, Figure 9.8 The Traitress: Her movement
is carried over into a slightly more distant
Yvonne and her father are dining, and she
starts to rise from the table. framing, taken from the same angle as the
previous shot.
a few such shots, often at emotional high points, and Feuillade uses some remark-
able cut-ins. 5 Our specimen The Traitress uses axial cuts of this sort at certain points,
sometimes with a smooth match on the player's action (Figures 9.7-9.8). Even when
1910s European films employed intrascene editing such as this, they tend to subordi-
nate it to the depth aesthetic. On the whole, these filmmakers prefer the gradation of
emphasis achieved within a single shot to the abrupt shifts of stress yielded by cuts.
But at least one director of the period found a weirder way to tell his stories visually.
Revelations at the Cinema Verdi
The Giornate del Cinema Muto, the annual silent film festival held at Pordenone,
Italy, has helped us refine our understanding of the 1910s along a great many dimen-
sions. In 1997 the Giornate screened two films by Robert Reinert, Opium (1919) and
the rediscovered Nerven (1919). Beautiful prints shown at the huge Cinema Verdi
were supplemented by an orienting essay by Jan-Christopher Horak, then curator of
the Munich Film Museum. 6 Horak's essay shows that Reinert, never tracked on film
historians' radar screens, fell into neglect as early as the 1920s. He was remembered
chiefly as the scriptwriter of Homunculus (1916). On the rare occasions that Opium
made it into historical accounts, it was as one of the sensational films that purportedly
reflected the decadence of Germany's postrevolutionary interregnum. 7 Horak reveals
that Reinert had written several scripts, supervised productions at Deutsche Bioscop,
and directed a popular series of melodramas at prodigious speed (ten released in
1916 and 1917). He produced Opium and Nerven under the auspices of his own firm,
Robert Reinert Monumental-Film, and after these projects he released an even more
grandiose production, the two-part Sterbende Volker (Dying Peoples, 1921). But with
the failure of that picture and his next, Die vier letzten Sekunden des Quidam Uhl
(The Four Last Seconds of Quidam Uhl, 1924), Reinert returned to scriptwriting and
overseeing productions. Of Reinert's output, only Opium and Nerven are currently
accessible and relatively intact. 8
Taking Things to Extremes 267
What filled the Verdi screen were two spectacular phantasmagorias. Opium tells
of one Professor Gesellius, who lives in China. He rescues the beautiful girl, Sin, from
her husband, Nung-Tschang (played with leering relish by Werner Krauss), and takes
her back to Germany, pursued by Nung-Tschang's warning, "You shall not escape my
revenge!" Sin becomes a nurse in Gesellius' clinic for opium addicts. Gesellius' wife,
Maria, is secretly in love with the spindly Dr. Richard Armstrong (Conrad Veidt).
Richard, tormented by his deception of Gesellius, tries to commit suicide by reck-
less horseback riding, but he survives and winds up hospitalized. Maria comes to his
sickbed, where Gesellius discovers them in an embrace. Now, as everyone is torn by
remorse, Nung-Tschang reappears, ready to prey on Gesellius. Richard is stricken
mute, then commits suicide by taking a poison he had asked Gesellius to mix for him.
In the meantime, the Professor decides to experiment with opium himself.
Maria and Richard's father (a reforming opium addict) blame Gesellius for Rich-
ard's death, so he flees to India with the devoted Sin and plunges into the opium
dens. Hoping to clear Gesellius' conscience, Sin claims that she poisoned Richard,
but this only makes him furious with her. Again, Nung-Tschang is on hand to twist
the knife, tempting Sin to return to him and telling an Indian count that Gesellius
has flirted with the countess. As the city burns (the fire has been set by—who else?—
Nung-Tschang), the count's men capture Gesellius. They dump him in a lions' den,
but Sin rescues him. Gesellius and Sin return to Europe, where Richard's father is now
in charge of the clinic. Again, Sin claims she murdered Richard and is taken away by
the police. While Gesellius undergoes treatment for his addiction, Nung-Tschang is
shot trying to free Sin from prison. Maria realizes that Sin has confessed out of love
for Gesellius, and she arranges for the girl to be released. But by then Gesellius has
swallowed poison before settling down to his final opium pipe.
Wild as this plot is, Nerven outdoes it. The prologue shows a neurotic man stran-
gling his wife. Then, noticing that the family's caged bird lacks water, he lovingly
refills the cup. The plot proper begins in a frenzy. While crowds rage (presumably
a reference to the civil disturbances following the Armistice), the neurasthenic-
looking Marja cries out that the earth is trembling on the eve of her wedding. A
young man bursts into the street and hacks at passersby with an axe; he is summar-
ily dragged to a wall and shot. Having presented a world in chaos, the film devel-
ops the conflict between Marja's brother, RolofF, a powerful industrialist opening a
new factory in town, and Johannes, the charismatic preacher of brotherhood who
lives with his blind sister. (It is Johannes who is throttling his wife in the prologue;
presumably, he has since found enlightenment.) Marja is infatuated with Johannes
and calls off her wedding. This leads Roloff to believe that Johannes has raped her.
Johannes is arrested, and as his blind sister wanders despairingly through the raging
crowds, Marja, pressed by her mother, confirms Roloff's fantasy. Johannes is tried
for rape and convicted. Meanwhile, Marja has a change of heart and tells Roloff
that Johannes is innocent. Roloff refuses to believe it; has he not seen Johannes
attack her? As he starts to strangle Marja, he realizes that he is in the grip of a
hallucination. Johannes is freed, and Marja leaves her family on a mission to carry
his doctrines to the wider world.
 Poetics of Cinema
268
Figure 9.9 Nerven (1919).
Near madness, Roloff seeks help from a "nerve specialist" and visits Johannes,
begging forgiveness. The men become friendly, but Roloff's morbid fantasies return
even more threateningly. He confesses to Johannes that he feels a powerful impulse to
kill his wife, Elisabeth. Roloff begs Johannes to give him poison, and after an agoniz-
ing hesitation, Johannes does so. As in Opium, several characters feel guilt. Johannes
torments himself for assisting Roloff's suicide, and Elisabeth pines for Johannes as
Marja had. Every character's desire is blocked. In a suicidal fit Elisabeth sets fire to
the Roloff mansion, and Johannes rescues her—but his blind sister is left behind to
die with her faithful dog. After a decent interval in a convent, Elisabeth leaves with
Johannes to start a new life.
Such are the pleasures of revisionist history, Pordenone style. What bald synop-
ses can't convey is the delirious impression left by these cascades of outrageous con-
trivances. Opium and Nerven offer a stew of 19th- and early 20th-century imagery:
a Romantic interest in extreme psychic states (suicide, drug visions, and madness),
cliches of academic painting, and conservative spins on avant-garde motifs. Opium
recycles Orientalist fantasy and the iconography of sensational fiction, while Nerven
situates itself in vulgarized expressionism. The giant machines of Roloff the industri-
alist are set against the teeming mob for whom Johannes claims to speak, and Roloff's
sickness is at once personal—a monstrous fear of women's sexual betrayal that shifts
from his sister to his wife—and a symptom of modern society: "In my own nerves
I recognize the nerves of the world." As in Lang's Metropolis, though, the psychosocial
themes from expressionist theater fade to the background as the characters work out
their melodramatic destinies through misunderstandings, guilty secrets, hidden sui-
cide notes, and heroic sacrifices for love.
Reinert has filled both films with symbolic mental imagery. By smoking opium,
Gesellius enters a world of languid, nymph-filled countrysides redolent of the arcadias
of salon painting; his dying dream shows him bidding farewell to this bacchanal.
Roloff's neurotic visions in Nerven are more expressionistic. He is haunted by bodies
in death throes, twisted landscapes, and contorted imagery of murderous hands
(Figure 9.9). A character's fantasy may uncannily echo plot events. In Gesellius' first
opium dream, he kills Richard; when Sin rouses him to report that Richard is dead,
Taking Things to Extremes 269
this convinces Gesellius that he is the young man's killer. More disturbingly, Nerven
often encourages us to take its characters' hallucinations for reality. After a series
of flashbacks in the first reel, Roloff claims that Johannes has seduced Marja, and
another flashback portrays it; because the other flashbacks have been trustworthy,
we are not quite ready to question this one—especially because Johannes was once a
lustful youth and Marja is initially portrayed as the unstable one in the family. Later,
Roloff is told that Johannes has hanged himself in prison, and he wanders outside in
confusion. Johannes, now apparently alive, strides up to him and says, "You killed
me, Roloff!" We are ready to take Johannes as an apparition, but it will emerge that
Roloff has in fact imagined the report of Johannes' death. I confess that I find some
incidents in Nerven obscure, and perhaps crucial footage or titles are still missing.
Even with fuller explanations, though, it seems likely that The Cabinet of Dr. Caligari
would have looked less innovative if generations of historians had seen Nerven,
released two months earlier.
Close-Up Depth
To a greater extent than Caligari, however, Reinert's films rely heavily upon the prem-
ises of 1910s European depth staging. German filmmakers of the period were still
placing the camera quite far back from the "front line," the nearest point that actors
might occupy,9 and Reinert was perfectly capable of staging extended scenes at a
middle distance. A striking example occurs in Opium, at the climactic moment when
Gesellius discovers the affair between his wife and the ailing Richard. The scene is
played out horizontally, with characters occupying parallel rows that lie perpendicu-
lar to the lens. Richard and Maria's embrace (Figure 9.10) is interrupted by the arrival
of Gesellius in the rear, centered and frontal so we cannot miss him (Figure 9.11). As
the lovers separate, Gesellius steps in between them and Sin enters on screen right
(Figure 9.12). As Maria rises, Gesellius comes forward (Figure 9.13) and Sin steps
into the gap created at the center (Figure 9.14). But then Gesellius steps still closer to
Maria, closing the gap, so Sin must compensate by moving to frame left, to the head
of Richard's bed, from which she can launch her final bit of business (Figure 9.15).
Gesellius turns away in sorrow (Figure 9.16), and Sin falls to the edge of Richard's
bed as Maria leaves. The shot ends with only the guilt-ridden face of Richard visible
(Figure 9.17). The characters' approach to the camera and retreat from it frame a
series of subtle shifts in lateral positions. In all, the scene is a fine specimen of late
1910s choreographic norms.
Yet very often, Reinert pushes the norms to violent li mits in order to intensify his
manic plots and performances. Instead of providing a smooth scale of emphasis, he
subjects his figures to harsh contrasts of scale and position. He shoves them abruptly
in and out of the shot, squeezes them up against the frame edge, and thrusts them
into the viewer's face. As Opium continues, we realize that the measured horizontal
treatment of Gesellius' discovery of the affair (Figures 9.10-9.17) is merely a prelude
to the tumultuous moment in which Maria confronts her husband after Richard is
found dead. Now Gesellius advances to the camera in medium shot (Figure 9.18),
 Poetics of Cinema Taking Things to Extremes 271
270

Figure 9.10 Opium (1919). Figure 9.11 Opium.

IIIK^BIS
Figure 9.13 Opium. Figure 9.20 Opium.
Figure 9.12 Opium.

and Maria presses in to accuse him of murdering Richard (Figure 9.19). As the
foreground becomes crowded with faces, the dead man is virtually hidden from
view (Figure 9.20).
The packing of the playing space in this scene is characteristic of many Reinert
compositions. Like Hofer and af Klercker, he favors sets that create a dense array
of masses through which characters pass. Early in Opium Gesellius sits idly by the
street when a palanquin carries Sin past him. Reinert stages the encounter within a
cluttered frame. Among the welter of plants and lanterns, Gesellius is a centered black
Figure 9.15 Opium. figure, and the palanquin emerges in the distance (Figure 9.21). He ducks out of the
Figure 9.14 Opium.
foreground (Figure 9.22) so that the palanquin can enter (Figure 9.23) and stop quite
close to the camera. Suddenly Gesellius' face pops in at the middle right of the frame
to talk with Sin—the woman whom he will eventually rescue from Nung-Tschang
(Figure 9.24). The surprise effect, achieved through framing Gesellius in the palanquin
window, is doubled when Nung-Tschang himself slides out of the shrubbery on the
left like a snake to watch the two (Figure 9.25).
Not only does Nung-Tschang's foray into the left foreground counterbalance the
pocket of attention established on the other side of the frame, but it also creates a
line that joins the three heads, however disparate they are in size. This tactic tends to
be Reinert's alternative to the more orthodox horizon-line isocephaly of the hospital
Figure 9.16 Opium. Figure 9.17 Opium. scene in Figures 9.10-9.16. He lines up the heads in considerable depth and clusters
 Poetics of Cinema
272
Figure 9.25 Opium.
them around distant nodes. Often these faces serve to reiterate the reactions of second-
ary figures to the main action, a kind of visual echo (Figure 9.26), but the device can be
used for protagon ists too. When the old Dr. Armstrong accuses Gesellius of killing his
son, we get an inten se impacting of the four characters' faces in a single zone of screen
space (Figure 9.27). In Nerven, Reinert pushes further. He crowds the foreground with
two faces—one frontal, the other in profile (Figure 9.28)—and sometimes strings
cheeks, jaws, and single eyes together, even in the distance (Figure 9.29).
Several of the scenes already mentioned display Reinert's reliance on bringing
players from the background to the foreground. We've already seen that means of
Taking Things to Extremes 273
Figure 9.28 Opium. Figure 9.29 Opium.
guiding the viewer's eye on display in The Traitress. One measure of Reinert's stylistic
exaggeration is the way in which this convention is subjected to considerable pressure.
When after young Armstrong's death Gesellius flees to India, one scene begins with an
extreme long shot of an opium den, furnished with dancing girls seen through an arch.
A woman stands in seductive silhouette under it. Gesellius, now an addict, arrives in
the distance (Figure 9.30) and walks through the arch past her (Figure 9.31). We might
expect the shot, having served to establish the locale, to end there and be followed by
a cut-in to the dialogue between the woman and Gesellius. Instead, he advances to
the foreground, halting in a tight medium-shot framing (Figure 9.32). Unexpectedly, a
woman (possibly the one seen in the shadows on the left in the early phase of the shot;
see Figure 9.31) appears beside him to tempt him with a pipe (Figure 9.33). As so often
happens in 1910s cinema, the shot's climax takes place in the front plane, but here the
effect is intensified by Gesellius' sudden emergence from the dark foreground and the
equally abrupt entry of the woman. Reinert's shot employs to-camera movement not
to attract attention but to create a kind of spatial suspense, delaying our full view of
Gesellius, and, as usual in Reinert, to fill the screen with the faces of his suffering souls.
Horak notes that critics praised the striking depth of field achieved by Reinert's
cinematographer, Helmar Lerski (Figure 9.34).10 In these two films Reinert sets his
characters closer to the viewer more consistently than any other 1910s depth-oriented
director I know. In the shots in the opium den and in the lecture hall, the foreground
 Poetics of Cinema Taking Things to Extremes 275
274

Figure 9.31 Opium. Figure 9.36 Nerven (1919). Figure 9.37 Nerven.
Figure 9.30 Opium.

Figure 9.32 Opium. Figure 9.33 Opium.

Figure 9.35 Opium. Figure 9.40 Nerven.

Figure 9.34 Opium.

figures are slightly out of focus, and elsewhere Reinert seems willing to sacrifice sharp
focus, in one plane or another, for the sake of aggressive foregrounds (Figure 9.35).
Nevertheless, the range of focal depth in the image remains extraordinary. Many
shots seem to rely on a 35mm lens rather than the customary 50mm one, and the
brilliantly sunlit sets customary in Germany at the time allowed Lerski to keep both
close and distant planes reasonably sharp. According to a contemporary technical
manual, if a director focused on a player at the four-meter line, a 35mm lens set at f19
would yield acceptable focus from about three feet to the back wall of a set. A director
ready to focus on something much closer, at one meter, could obtain, with the same
lens and diaphragm setting, acceptable focus from two feet to eleven feet.11
Given an abundance of light, a tolerance for a slightly fuzzy plane now and again,
and a cinematographer of Lerski's skill, Reinert could confidently bring his front-
most figures four to six feet from the lens. The looming foregrounds that result
are exploited throughout Nerven. Sometimes the effect is of sculptural stolidity
(Figures 9.36-9.37); at other times, we get a frantic thrusting against the confines of
the shot. In the pulsating first scene, a young man, brooding beside his parents, leaps
to his feet (Figure 9.38) and rushes out of the frame (Figure 9.39), with his mother
and father straining against the lens to watch him (Figure 9.40). When Roloff's sister,
Marja, learns of the verdict at Johannes' trial, she is stricken by guilt and recoils
from a painting of the Crucifixion; soon she will recant her charge against Johannes.
276 Poetics of Cinema
Figure 9.41 Nerven. Figure 9.42 Nerven.
Figure 9.43 Nerven. Figure 9.44 Nerven.
Reinert shows Marja restrained by her mother (Figure 9.41), twisting away in agony
(Figure 9.42), then hurling herself to the left foreground (Figure 9.43) and ducking
out of the shot (Figure 9.44). Later, Johannes returns home from prison, and the
high-angle shot begins with a fairly distant view of his blind sister waiting at the
window (Figure 9.45). By the standards of 1910s cinema, he is likely to appear in the
background, but instead she turns to us and comes forward (Figure 9.46). Johannes
bursts in from the right corner (Figure 9.47), embraces her, and swings her around
in a tight medium shot (Figure 9.48).
Problems and Solutions
All this results, undeniably, in greater dramatic intensity. The nervousness of the
characters is amplified by a style that turns the players' conventional advance to
the foreground into a crescendo of passions and anxieties, then lets the characters
assail us with their overheated expressions and gestures. Yet in the effort to create
overwhelming emotional effects within the 1910s depth style, Reinert also goes some
way toward undermining that style. He seems to presume, for instance, that if the
shot's climax is a powerful close-up, it is anticlimactic to end the scene as most of his
contemporaries would, by letting the character retreat from the lens and exit through
a distant doorway. How much more dynamic—and suitable for a film in which
Taking Things to Extremes 277
everyone seems on edge—to let the character simply spring out of the foreground
and vanish. Similarly, the placid rhythm of the sister's patient waiting is broken by
Johannes' abrupt foreground entrance and passionate embrace.
Yet by moving several characters upstage, close to the camera, Reinert creates a
new problem for himself. Because cinema's playing space is pyramidal, with the apex
at the lens, the closer that actors get to the camera, the more frame space they occupy
and the more background areas they blot out. As characters crowd the foreground,
the background gets lost. A depth strategy has wound up canceling depth.
How, then, can you place two or three characters in semi-close-ups and still pre-
serve depth? Reinert finds one answer in limiting the rear playing space to small
slots. We see this in the scene of Johannes' return, when a shot very slightly varies
the composition examined already (Figures 9.45-9.48) to allow us a view of brother
and sister reunited, with one middle-ground zone reserved for the family dog, who
scrambles through the window in the background (Figure 9.49). Another possibility
is to move foreground figures aside in the course of the shot, revealing distant areas
more fully. This, however, is hard to do with very close figures like Reinert's. They
take up so much of the narrowing wedge of the playing space that they must pivot a
lot, or slide over to the frame edge, as we have already seen (Figures 9.28, 9.43, and
9.47-9.48). Another solution is to bring into the foreground the sort of tightly timed
ensemble playing and minute blocking and revealing characteristic of the sort of
278 Poetics of Cinema Taking Things to Extremes
279

Figure 9.49 Nerven. Figure 9.50 Nerven. Kgure951

distant views exemplified by our Traitress scene. The primary instance of this strategy
in the two films comes at a climax in Nerven.
Roloff, now intimate with Johannes, has come to him and confessed his dream
of killing his wife. He begs Johannes to give him poison. Johannes goes to fetch it,
agonizing over the decision, while his blind sister, unknown to him, stands nearby,
sensing what he is doing. This situation might have been handled by a more tradi-
tional depth shot, placing Johannes in plan-americain and the sister some feet away
in the background. Instead, in his search for intensity, Reinert places Johannes in
medium shot as he opens the cabinet (Figure 9.50), moves him forward as he looks at
Figure 9.52 Nerven. Figure 9.53 Nerven.
the bottle (Figure 9.51), and halts him remarkably close to the camera. Johannes fills
half the frame, and the opened cabinet accounts for about a third of the lower half
(Figure 9.52). As he lifts his head, his sister appears in a slot of middle-ground space,
between his arm and the cabinet door (Figure 9.53). His decision made, Johannes
rushes out of the shot, leaving her behind (Figure 9.54).
An out-of-focus foreground figure stretched taut in psychic torment, a packed set
out of which a blind woman's face peeps uncomprehendingly, a jarring frame exit
exaggerated by proximity to us and the accelerating effects of the wide-angle lens: This
is pure Reinert, a kind of paroxysmic summation of the abnormalities of his mise-en-
scene. But the shot, in its isolation and magnification of faces and its almost complete
blockage of the surrounding space, turns itself into an exercise in close-up filmmak- Figure 9.54 Nerven.
ing. Here is the paradoxical limit of Reinert's method. Working within a tradition that
avoids the cut-in close-up, he pushes the tradition's strategies to an extreme and comes
door through which Johannes must have come in Figure 9.47? Reinert's urge to inten-
up with an image consisting of what are in effect close-ups . . . without editing.
sify the diagonal thrusts of the 1910s approach led him to an in-between aesthetic
An important consequence of this strategy is to remove some of those cues for
position: His typical depth shot is neither a tableau composition that surveys the
spatial stability that had been prized by the 1910s directors. Feet and legs are cut off,
entire locale nor a part of an American-style editing pattern that builds a consistent
even in establishing shots. Similarly, Feuillade, Bauer, Hofer, and most other directors
overall space out of several complementary camera positions. A Reinert "full shot"
of the time use rear doors or entryways to let us know how characters enter or leave
may present a relatively small patch of the scene's space, and we may never be properly
the scene. With such close foregrounds as Reinert's, however, characters may leave
introduced to the overall arena of action. As European directors were adopting
the frame but not necessarily the locale, and this creates an uncertainty we seldom
Hollywood decoupage, Reinert comes up with a fragmentary scenography that looks
find in the work of his peers. Where does Marja go when she darts out of view in
ahead to the strategically incomplete establishing shots employed by Robert Bresson,
Figure 9.44—to another part of the room, or to another room? Where is the cottage Jean-Marie Straub and Daniele Huillet, and Hal Hartley.
 CinemaScope

The ISAodern ISAiracle You See

In this context, perhaps the most conventionally hallucinatory images that
haunt Nerven take on a new significance. Shots like Figure 9.55, or the delirious
superimpositions in which Roloff imagines being pursued by Johannes' corpse
Without Glasses
(Figure 9.56), clamp something in close-up against a distant view. The depth is purely
phantasmagoric, of course, but the imagery is often only one step beyond the startling
compositions that we find in the characters' "ordinary" lives. Reinert pushes a norm
to the limit in order to raise its expressive power, but he ends up abolishing the norm's
raison d'etre. Instead of the integral space calmly mapped out by the 1910s masters,
cinematic depth becomes a void out of which spectral figures rise, then freeze or
slowly swivel before plunging out of view again. In its perverse way, Reinert's cinema
relies as much on magnified faces and hands as does Hollywood's. But these hands
twist in despair, and the looming faces warn of dreadful dreams.

The brand name conjures up a time when Hollywood bet on overstuffed spectacle—
harem cuties, beefcakes wrapped in togas, and hoofers with Ipana smiles, all splashed
across screens the size of billboards. For film buffs, the very word (double-capped)
spurs misty nostalgia and murmurs about mise-en-scene. But you could reasonably
ask, What's all the fuss? Despite Twentieth Century-Fox's aspirations, the wide-
screen process known as CinemaScope never dominated the industry's output. Some
studios ignored it; others abandoned it quite quickly. Major older directors like Alfred
Hitchcock and Cecil B. DeMille never worked with it, and those who tried it, like
Howard Hawks (Land of the Pharaohs, 1955) and John Ford (The Long Gray Line,
1955; Mister Roberts, 1955), weren't enthusiastic. In Jean-Luc Godard's Contempt
(1963), Fritz Lang, who made Moonfleet (1955) in Scope, famously pronounced it as
good only for filming snakes and funerals.
Nor was Scope a proven money spinner. Granted, the first 1953 features, especially
The Robe and How to Marry a Millionaire, caught fire at the box office. But soon each
year's top five hits included only one or two Scope titles. Of the ten top-grossing
movies of the 1950s, only three (Lady and the Tramp, 1955; The Robe; and The Bridge
on the River Kwai, 1957) were in Scope.1 Scope pictures claimed a share of Academy

281
282 Poetics of Cinema
Awards, mostly technical ones, but only two films ever won Best Picture honors. Most
remarkably, CinemaScope had a short life. Introduced in late 1953 as a prestigious
brand, it was generally out of favor five years later. Ten years after it appeared, it was
most widely used in low-end output like Take Off Your Clothes and Live (1963) and
Prehistoric Women (1967).
So why still speak of it? Why devote book chapters to a technology that flourished
about as long as laserdiscs? Why does the prospect of seeing a Scope print in 35mm
draw cinephiles to repertory houses and art centers? Why do so many devotees, mostly
middle-aged men, fill lovely websites with arcana about this defunct format? Why do
judicious historians speak of a CinemaScope revolution? And why are students of film
style fascinated by the look of Scope movies?
Some answers are apparent. CinemaScope didn't catch on as quickly as sound or as
widely as color, but the emergence of the format signaled that widescreen film was here
to stay. Once Scope was announced, all the major studios and production companies
abandoned the 4:3 aspect ratio that had been in place since the silent era. The immedi-
ate stimulus to the switch was the 1952 success of Cinerama, a vast three-panel pro-
cess for specialized venues, but the simpler Scope technology demonstrated that any
movie could swell to awesome proportions. Most films would be made in still cheaper
formats, usually yielding less overbearing visuals, but Fox's all-out push for Scope
surely accelerated the changeover to widescreen cinema as an industry standard.
Although Scope faded fairly quickly, its physical premise, anamorphic optics, has
remained an important filmmaking resource. Scope's innovations were the basis of the
more robust and versatile widescreen system established by the Panavision company.
Today many films are designed to be seen in the stretched proportions established by
CinemaScope; the very wide ratio is considered a cool way for images to look. And
cinematographers still casually call any image-squeezing system a "scope" format.
Granted, claims for a CinemaScope revolution were oversold at the time. Fox
President Spyros Skouras called his new gadget "the greatest medium of improve-
ment to the screen to date" and "one of the most remarkable feats in all the annals
of industrial and artistic endeavor."2 No less modestly, Fox producer Jerry Wald
considered it "the greatest boost the picture business has gotten since it discovered
sex."3 But calmer minds have argued that Scope did significantly change cinema. John
Belton, the foremost CinemaScope historian, suggests that Scope was "a reinvention
of sorts of the cinema," returning it to its original state of overpowering visual specta-
cle. The peepshow and fairground gave cinema its initial appeal, which was sustained
on increasing scale in the 1920s picture palaces. For Belton, CinemaScope becomes
the last installment in film's effort "to recapture, through the novelty of its mode of
4
presentation, its original ability to excite spectators."
There are plenty of other reasons to study Scope. It's hard to understand the
auteur theory as it developed in Paris and London without understanding the grip
that Scope films had on young critics. The format—at once deep and flat, dense with
realistic detail and yet as geometrically stylized as a frieze—epitomized the artistic
possibilities of the contemporary cinema. After Andre Bazin had taught the younger
generation the virtues of pictorial depth in the work of Jean Renoir, Orson Welles, and
CinemaScope 283
Figure 10.1 In situations when some actors Figure 10.2 Godard seems to take this
were sitting and some were standing, Cinema- schema as the visual premise for this shot
Scope close views posed problems, and some from Pierrot lefou (1965), which features one
directors opted to chop out a body, leaving head at far left and a clone (actually a sculpted
a head at the lower frame line (The Hunters, head) at far right.
1957).
William Wyler, the Young Turks of Cahiers du cinema ventured to differ from their
mentor. They sang the praises of width. Jacques Rivette discovered that the greatest
directors of the past had laid the foundations of widescreen imagery by seeking
a perfect perpendicular to the spectator's look. From The Birth of a Nation to Le
Carrosse dor, from the Murnau of Tabu to the Lang of Rancho Notorious, this
extreme use of the breadth of the screen, the physical separation of the charac-
ters, empty spaces distended by fear or desire, like lateral units, all seems to me
to be—much more than depth—the language of true filmmakers, and the sign
of maturity and mastery. 5
In a parallel gesture, Francois Truffaut's and Godard's anamorphic work can be seen
as creative responses to the American Scope films they admired (Figures 10.1-10.2).6
Although a ponderous movie feels elephantine in Scope, many excellent films were
made in the format, and it enhanced their quality. If some leading directors resisted
it, others explored it. George Cukor, Vincente Minnelli, Douglas Sirk, Samuel Fuller,
Elia Kazan, and Nicholas Ray, along with less celebrated filmmakers like Richard
Fleischer, Delmer Daves, John Sturges, Joshua Logan, and Jack Webb, made superb
use of it. (It seems that directors who began their careers in sound filming did better
with Scope than those who started in the silent era.) We can learn a great deal about
cinematic technique, particularly staging and composition, by studying how talented
directors managed this distended image.
Just as important, Scope cinema illustrates how stylistic continuity and change
can interact during a period of technological overhaul. When a new tool is intro-
duced into U.S. studio filmmaking, it's usually shaped to fit existing routines. Film-
makers try to exploit the new device's unique features while still integrating it into
standard work practices and stylistic functions. For example, when synchronized
sound was innovated, it was quickly absorbed into the overarching system of spatial
and temporal continuity that we call "classical" stylistics. The problems of filming
sync sound—camera noise, unselective microphones, and breaking a scene into
shots—were solved by an interim tactic, that of multiple-camera shooting. By filming
with several cameras poised at distant spots, the director could retain some editing
284 Poetics of Cinema
options. Improvements in sound recording soon permitted a return to single-camera
shooting, but in the meantime classical scene dissection was preserved.
The ripple effects following from sync sound affected visual style in subtle ways.
The dollies and cranes designed for shifting the heavy camera around the set enabled
filmmakers to use more traveling shots. Multiple-camera shooting relied on refram-
ings, those slight nudges of the frame left or right, and even after single-camera shoot-
ing returned, reframings achieved a new prominence. Eventually the experience of
multiple-camera shooting proved valuable for television. Today, most sitcoms and
soaps are shot with three cameras, a routine stemming from compromises of the early
sound era. Still, we shouldn't expect that a new technology promotes steady improve-
ment, because each new benefit exacts a cost. The strengths of orthochromatic film
stock were lost because filmmakers switched to panchromatic, which was better suited
to the types of illumination required for sound recording. Nonetheless, the dynamic
of innovation, recovery, and discovery allows new technical devices to be adjusted to
traditional visual schemas, even while they yield unanticipated payoffs.
Like 1920s sound recording, CinemaScope challenged some established methods
of making movies. We might say that there were both technological and aesthetic
problems, but it turns out that in general many aesthetic problems spring from tech-
nological ones. One lesson of the Scope era is that the physical constraints of a new
technology have stylistic consequences. At the same time, problems don't admit
of only one solution. The "classical style" isn't an iron rule but a set of principled
options, adaptable to different situations. By spelling out the range of craft choices
that CinemaScope yielded, we can better understand how directors used the new
format for storytelling purposes.
The Big Picture
Efforts at widescreen film date back to the earliest years of cinema, but it wasn't until
the 1950s in the United States that the wide image became more or less standardized,
yielding the formats we know today.7 All widescreen systems alter the "aspect ratio"
of the image. Most silent film images fill a 4:3 rectangle, yielding a 1.33:1 ratio. After
the coming of sound, the U.S. ratio was standardized at very close to this (1.37:1, to
leave room for an optical soundtrack). From 1954 onward, though, most U.S. films
were designed to be shown wider than 1.33.8
There are three basic ways to widen the traditional film image. The least common
is the multiprojection system, seen most famously in Cinerama. The system employed
three side-by-side cameras to record a wide view, with three synchronized projec-
tors being required to show the film (at an aspect ratio of 2.59:1). Cinerama made
filmmakers appreciate the potential of widescreen cinema, but the process remained a
novelty confined to few theaters. Although most Cinerama features were travelogues,
two fiction features were shot in the format, The Wonderful World of the Brothers
Grimm (1962) and How the West Was Won (1962).9 Multiprojector systems survive
today in theme park attractions such as the 360-degree wraparound screen in Florida's
Walt Disney World.
CinemaScope 285
Another way to create a wide image is to employ a film gauge wider than the
normal 35mm. Though several formats were tried, 70mm came to be the standard
wide gauge. (The film would actually be shot in 65mm, with the extra width used to
accommodate the soundtrack on the release prints.) The first successful 70mm process
was Todd-AO, launched with Oklahoma! (1955). This remarkable system employed
unusually wide-angle lenses for shooting and ran the film through the camera at 30
frames per second (as opposed to the usual 24). The result was an image of stunning
sharpness, with an aspect ratio of about 2.2:1. The other major wide-gauge system
of the era was Super Panavision 70 (first used on The Big Fisherman, 1959). 70mm
virtually vanished as a capture medium after 1970; only the USSR's Sovscope 70 kept
it alive into the 1980s.10 American films shot in 35mm continued to be released in
70mm blowups partly because of the superior sound quality offered by the format.
Before their demise in the 1990s, 70mm release prints were usually meant to be shown
at an aspect ratio of 2.0:1. Wide film survives principally in the IMAX format, which
uses the 70mm gauge and the squarish aspect ratio of 1.435:1.
The usual way to create a widescreen image is by manipulating the image on tradi-
tional 35mm film. Most simply, the picture area can be masked. If it's masked during
filming or during printing, the result is a letterboxed image on the film strip, with
black bars at top and bottom. Or the image may be shot and printed full-frame, in
which case it's up to the theater projectionist to crop the picture by slotting the correct
aperture plate into the projector. In screening full-frame prints, projectionists have to
watch out for microphones, incomplete sets, and other intrusions. In the unmasked
35mm frames of The Godfather Part II (1974), you can see the actors' marks laid out
in tape on the floor.
When an image is masked, the aspect ratio can vary. In the early 1950s different
studios and producers opted for various proportions, but eventually 1.85:1 became
the more or less standard "Academy ratio." It isn't always honored. My local multiplex
has apparently decided to show its prints at 2:1. Overseas filmmakers continue to
employ 1.66:1 and 1.75:1 ratios as well.
Less common than masking is the use of anamorphic lenses to widen the image.
During filming, the anamorphic lens squeezes a wide field of view onto the film strip;
the result is a squashed image, showing abnormally skinny people. A corresponding
lens attached to the projector unsqueezes the picture. CinemaScope was the most
famous anamorphic system, but kindred systems were developed in France, Sweden,
England, Italy, Russia, japan, and Hong Kong. American variants included Naturama
(developed at Republic), Vistarama (at Warner Bros.), and WarnerScope. 11
Initially the CinemaScope aspect ratio was planned to be 2.66:1, exactly twice the
width of the standard image, but engineering considerations reduced it to 2.55:1.12
In adopting this ratio, the Fox staff sought to maximize the picture area by eliminat-
ing the optical soundtrack and putting stereophonic sound information on magnetic
striping running along both edges of the film. But most theater owners didn't want
to install stereo playback equipment, so some CinemaScope prints began to include
optical monaural soundtracks. This meant sacrificing more picture area, resulting in
an aspect ratio of 2.35:1. In 1956, all Scope prints began to be released in "magoptical,"
 Poetics of Cinema
286
containing both types of track, thereby making the narrower ratio the CinemaScope
standard. The 2.35:1 proportion remained the default for other anamorphic systems,
most notably that of Panavision. In the 1970s, a thicker splice lowered the height a bit,
changing the standard anamorphic aspect ratio to 2.40:1.
The other widescreen systems that emerged in the 1950s offered variants of the basic
possibilities. The wide gauge was combined with anamorphic optics in CinemaScope
55 (premiered in Carousel, 1956) and MGM Camera 65 (Raintree County, 1957), the
latter of which became Ultra Panavision 70 (Mutiny on the Bounty, 1962). Some of
these systems had aspect ratios as wide as 2.76:1. VistaVision, created at Paramount,
relied on running the filmstrip horizontally through the camera, making each image
larger than the conventional frame. The greater frame area offered excellent sharp-
ness, and VistaVision could yield prints in ratios from 1.33 to 2.1. The Technicolor
company combined a horizontal camera path with anamorphic optics to create
Technirama (The Monte Carlo Story, 1957) and printed the image in a wider gauge for
Super Technirama 70 (Sleeping Beauty, 1959). For Techniscope, developed in Italy, the
traditional frame was split into two horizontal strips during filming, each only two
perforations high. Each wide frame was printed anamorphically as a single image at
the standard four-perforation height, then unsqueezed in projection. 13
Wider movies needed mammoth screens. In a period when many theaters housed
screens no bigger than 16 by 20 feet, Cinerama's three-projector system induced shock
and awe. Its minimum screen area was 3,000 square feet, and a width of 75 feet was
common. CinemaScope aimed at an impressive scale as well; 24 feet by 64 feet was the
recommended size for its high-reflectance "Miracle Mirror" screen. Even the compro-
mise formats like 1.85 looked more imposing on bigger screens (although blowing up
the standard image introduced new problems of illumination and graininess). With
the new widescreen systems, studios and exhibitors offered a cinematic fresco that
made the living room TV monitor look minuscule.
Cinerama was not the only technical innovation steering producers toward wide-
screen systems. The success of the cheaply made Bwana Devil (1952) briefly per-
suaded many studios that 3-D was the next big thing. The 3-D boom fizzled in less
than a year, but the idea of immersing the audience was promoted by the backers
of widescreen systems too. The screens designed for Cinerama and Todd-AO were
deeply curved, and viewers found their peripheral vision stirred by these enveloping
images. The CinemaScope screen was curved less pronouncedly, but Fox's publicity
encouraged the impression that its images somehow attained high relief. "From its
panoramic screen . . . actors seem to walk into the audience, ships appear to sail into
the first rows."14 Trailers and newspaper ads announced "The Modern Miracle You
See without Glasses!" Fox soon gave up this fiction, but to this day, if a film isn't shot
"scope" (i.e., anamorphically), cinematographers say it's shot "flat," an echo of a time
when CinemaScope was felt to compete with 3-D in its power to engulf the audience.
CinemaScope
287
Hollywood Cadillac
The industry's eagerness to embrace 3-D, Cinerama, Scope, and other novelties
stemmed from immediate pressures. In the late 1940s the Supreme Court declared
the U.S. film industry guilty of monopolistic practices. The studios could no longer
own theaters, which were important not only for their box office receipts but also for
the metropolitan real estate they occupied. Nor could studios rent films in blocks,
demanding that exhibitors take the middling items along with the sure-fire ones.
Now every film would have to be sold on its own. Meanwhile, war-weary consumers
discovered cars, bowling, barbecues, suburban child-rearing, and television. In 1946,
90 million Americans went to the movies each week, but by 1952 weekly attendance
had plummeted to 51 million. This translated into $300-400 million in lost ticket
sales annually. 15 As operating costs rose, the studios' profits were sinking by 50 to
75%.16 Producers were convinced that the industry needed fresh attractions to win
back moviegoers. The success of This Is Cinerama in September 1952 suggested that
big-screen spectacle was worth gambling on.
Even before the postwar crisis, producers had been seeking ways to enhance pre-
sentation. From the early 1940s, studios dramatically increased their commitment
to color film production, while also researching stereophonic sound, wide film, and
television broadcasts direct to theaters. 17 But Cinerama's success tipped the balance.
In December 1952, Twentieth Century-Fox president Spyros Skouras acquired Henri
Chretien's anamorphic lens system. The first tests of the lens convinced Darryl F.
Zanuck, head of Fox production, to adopt the system immediately. Chretien's best
lens was assigned to The Robe, already in production, and a second lens went to How
to Marry a Millionaire. Bausch & Lomb quickly revised the Chretien design, and in
the spring more lenses were available for three other productions: Beneath the 12-Mile
Reef, King of the Khyber Rifles, and Knights of the Round Table (MGM). Remarkably,
all were ready for release in the last four months of 1953.18
Skouras financed CinemaScope boldly, borrowing heavily from banks and mort-
gaging the studio, the backlot, and Fox real estate holdings.19 The firm launched a
massive publicity campaign. In the spring, demonstration footage was screened for
the industry and short films toured Europe. 20 Studios, convinced that the future lay
with widescreen, scrambled to release their remaining 1953 titles in masked versions.21
Zanuck announced that all Fox-produced films would be in color and Scope, and in a
memo to studio staff he declared that for the next year and a half, "intimate comedies or
small-scale, domestic stories should be put aside." Every film would contain elements
that "take full advantage of scope, size, and physical action."22 This policy led Zanuck
to withdraw his commitment to make On the Waterfront, a decision he regretted even
before it won an armful of Academy Awards. Soon he conceded that the success of
Three Coins in the Fountain (1954), a more or less "intimate" story shot in Scope, had
changed his mind about what scripts were suitable for the widescreen. 23
CinemaScope might have gone the way of 3-D if The Robe had flopped. It did
not. Opening in Manhattan's Roxy theater in September 1953, the film took in over
$3.5 million in its first 12 days, a New York record. Eventually The Robe garnered
 Poet ics of Cinema
288
Figure 10.3 Richard Egan suffers Cinema-
Scope mumps in Love Me Tender (1956).
$25 million worldwide, making it the top grosser of the year and one of the highest-
earning films of the decade. The other CinemaScope pictures released in late 1953
also did very well. Although Scope was estimated to add an average quarter of a
million dollars to a production budget, producers came to believe that the expense
was worth it. The no-star adventure Beneath the 12-Mile Reef took in almost
$6 million internationally. By the end of 1954, all studios except Paramount (home of
VistaVision) had licensed the format from Fox. 24
The Robe opened in friendly territory, for the Roxy was a flagship venue of National
Theatres, Inc. This chain of some 500 houses, headed by Skouras' brother Charles, had
once been Fox's exhibition arm, and it had invested in the Scope system. 25
Although
some exhibitors resisted the conversion to Scope, most circuits signed on. By 1955, when
Scope was available in over half of U.S. theaters, it seemed likely to become the high-
26
end industry standard. "We want the public to say there never was a bad CinemaScope
picture," Skouras declared, "just like they'd say there was never a bad Cadillac."27
For a couple of years, Scope enjoyed fairly broad support from studios. In 1955
Scope films made up nearly 20% of the majors' feature releases. Columbia released
8 titles, Warner Bros. 13, United Artists 17, and MGM 18. Many of these—Mister
Roberts, Battle Cry, 20,000 Leagues Under the Sea, and The Seven Year Itch—scored big
box office returns. The number of annual Scope releases hit a peak (about a hundred)
in 1957. But problems were already emerging.
Some were technical. The earliest Chretien lens had been mounted in front of the
prime camera lens, and various Bausch & Lomb improvements, including housing
both lenses in a single rather monstrous unit, didn't alter that arrangement. This
severely cut down on light-gathering power. In addition, the "squeeze ratio" of
Chretien's lens design varied across the horizontal axis.28 These optical tics created
distortions and patches of soft focus. The most embarrassing flaw, created by faults
in magnification and the uneven compression of the visual field, made central figures
look oddly bloated. In close-ups, the result was "CinemaScope mumps" (Figure 10.3).
Not all of the films credited to Scope were shot with Bausch & Lomb lenses, but other
brands of anamorphic lenses tended to cause the same problems.
Enter Panavision, which began as a supplier of anamorphic projection lenses.
Panavision's engineers solved the mumps problem by using counterrotating cylinders
that adjusted image compression smoothly. Although anamorphic optics in them-
selves weren't patentable—hence the several Scope clones—Panavision did patent
CinemaScope
289
29
its method of correcting lens astigmatism. In addition, for some focal lengths,
Panavision lenses had the anamorphosing element placed behind, rather than in
front of, the prime lens, an arrangement that increased sharpness and light-gathering
power.30 First developed for MGM's wide-gauge Camera 65, Panavision's optical
system crept into other projects. Most of MGM's anamorphic releases of the late 1950s
were shot with Panavision lenses, although the credits still bore the CinemaScope
trademark and Panavision was not always credited as a supplier. After Panavision's
energetic 1958 marketing campaign, other studios took up the system.31 In 1959, the
Auto Panatar photographic lens won an Academy Scientific and Technical Award, and
the staggering success of Ben-Hur (1959), shot in anamorphic 70mm with Panavision
lenses, secured the company's reputation. By 1961, Panavision anamorphic lenses
were said to be employed on a third of all the films made in Hollywood. 32
Even before Panavision surpassed Scope, rival formats had won important market
shares. In many venues, major releases looked better in VistaVision or Panavision.
The emerging roadshow market, with its luxuriously outfitted theaters and steep
ticket prices, favored the sharp, luminous images that VistaVision, Technirama, and
Todd-AO could deliver. The Robe's grosses were outstripped by returns for The Ten
Commandments (1956, VistaVision) and Around the World in Eighty Days (1956,
Todd-AO). And the box office returns of From Here to Eternity (1953), The Caine
Mutiny (1954), and Giant (1956) proved that serious drama in the flat format could
earn more than virtually any Scope extravaganza. Did producers really need Scope
to bring in customers?
Fox was in a weak position to recover the initiative. The studio was plagued by
financial problems, and in 1956 a discouraged Zanuck left. His successor as head of
production, Buddy Adler, cut expenses drastically. Adler forbade location shooting,
permitted directors to print only one take, and insisted that producers reuse sets rather
than build new ones. One of his economies rescinded Zanuck's commitment to color:
now Scope films could be in black and white. Adler contracted with independent
producer Robert Lippert's Regal Pictures to turn out cheap films shot with Bausch &
Lomb lenses.33 Lippert released 20 RegalScope films in 1957 alone. Adler's new policy
also attracted Universal and marginal independents looking to add a touch of class to
routine product. The 1957 uptick in Scope usage is largely attributable to the diffusion
of black-and-white Scope.
Like Sony with Betamax videotape, Fox suffered early-mover disadvantage. Skouras
and Zanuck had shown the way toward bigger screens and anamorphic image displays,
but now Fox was saddled with an attraction that was no longer anything special; the
Cadillac had become a Ford. By going down-market with black-and-white films and
the RegalScope line, Fox further cheapened its brand. Most major studios withdrew
their support. In 1958 Columbia released five Scope titles, United Artists merely two,
and Warner none. Only two Scope films released after 1956 (River Kwai and Peyton
Place, 1957) earned slots in the 60 top-grossing films of the decade. 34 By the time
that films in Scope won Best Picture Oscars (with River Kwai and Gigi, 1958), it was
a dying format. 35
 Poetics of Cinema
290
Figure 10.4 The Last Action Hero (1993): A
micro-close-up gag that could never have been
accomplished in traditional CinemaScope.
For the industry as a whole, 3-D, widescreen, roadshows, and stereophonic sound
amounted merely to holding actions. Throughout the 1950s and 1960s attendance and
box office receipts continued to plummet, eventually leveling off at around 20 million
viewers a week. Those exhibitors who remained in business chopped up their theaters,
offering patrons tiny screens, dim projection, and monaural sound. Most studios suf-
fered financial crises, despite the cash coming in from TV production and the sales of
film rights to broadcast networks. Hemorrhaging money, and bought and mismanaged
by conglomerates, the Hollywood studios were beached behemoths by the early 1970s.
They were resuscitated by tax breaks and a new generation of filmgoers and film-
makers. Steven Spielberg, Francis Ford Coppola, and George Lucas had grown up on
the overwhelming spectacle of the waning studio years. Sharing a gearhead sensibil-
ity, the Film Brats yearned for movies on a colossal scale. They embraced anamorphic
imagery, 70mm presentation, and multitrack sound. Jaws (1975), Star Wars (1977),
Close Encounters of the Third Kind (1977), and other box office triumphs paved the way
for the multiplex of the 1980s and 1990s, shrines that fulfilled the showmen's ambitions
of three decades before, but in a package fitted to the tastes of suburban adolescents.
A Lack of Scope
Today's anamorphic movies give us wildly canted angles, complicated tracking shots,
and extreme close-ups (Figure 10.4). It wasn't always so. The Robe, How to Marry a
Millionaire, and innumerable other Scope items look lumbering and archaic, largely
because of constraints built into the first wave of the technology. The films looked
just as stiff to professionals of the time, and they eyed the new format with suspicion.
Delmer Daves recalled a panicky meeting of directors called by Zanuck at which
CinemaScope was unveiled. Daves had deep reservations.
Was this the end of the close shot or the two shot? What could you do about
all of that out-of-focus space when you're on someone two feet away from the
camera? Was all the intimacy of filmmaking going to be lost? Darryl didn't have
any answers. 36
There was no hiding the optical drawbacks of the system, especially on The Robe.
It was shot with only one lens, a 50mm prime that had to be focused separately from
CinemaScope
291
Figure 10.5 Steps that should be parallel are Figure 10.6 CinemaScope made classic
bulging in this shot from The Young Lions columns look bloated, although some set
(1958).
designers tried to disguise the distortion by
adjusting the curve on the set (Alexander the
Great, 1956).
Figure 10.7 The Man Who Never Was (1956):
The officer on the right is made abnormally
thin by the anamorphic distortion.
the anamorphic attachment in front. Director Henry Koster recalled that in looking
at the rushes, about half the shots showed actors out of focus. Immobility was the best
solution: "If we kept the actors in the same spot, the focus was all right."37 Scope sets
were bigger than usual, and the Chretien lenses, even modified by Bausch & Lomb,
were poor at gathering light, so cinematographer Leon Shamroy had to flood the Robe
sets with intense arc illumination.
Early anamorphic lenses offered very limited depth of field (that is, the range within
which objects would appear well focused), and they were at their sharpest when film-
ing from far back. 38 Directors were advised to put the camera no closer than seven feet
from the subject. Worse, the picture yielded some startling distortions. The central
horizon line might appear straight, but other horizontals were bowed (Figure 10.5).
On the vertical axis, columns, walls, and fence posts bulged (Figure 10.6).39 In close-
ups, faces in the center of the frame contracted Scope mumps, whereas in long shots,
figures on the sides were pinched rail-thin (Figure 10.7). Areas that should have been
in focus proved not to be.40 For Brigadoon (1954), Joseph Ruttenberg used two men
just to handle focus. 41 Because no U.S. studio cameras had reflex viewing, operators
had traditionally lined up their shots with viewfinders mounted on the side of the
camera. These suffered from parallax problems, especially at close distances: What
the cameraman saw was not exactly what the lens took in. Scope made parallax prob-
lems far more severe, so cinematographers were advised not to track forward or back
because the viewfinder couldn't "toe in" or "toe out" sufficiently to show what the lens
was centered on.42
 Poetics of Cinema
292
Bausch & Lomb improved Chretien's design in 1954, particularly with respect
to focus,43 but the system remained inferior to the spherical lenses used for "flat"
cinematography. Most anamorphic lenses remained subject to distortions and
mysterious dropouts. As late as 1956, a cinematographer was advising Scope
filmmakers to avoid horizontals, to block off verticals on the screen edges, and to
minimize close-ups and wide-angle shots.44
Audiences didn't seem to mind the flaws, but the professional community boiled
with complaints about anamorphic widescreen. Cinematographers hated it, and
several directors found its proportions ridiculous. "If the CinemaScope size had been
any good," Hawks remarked, "painters would have used it more—they've been at it
a lot longer than we have."45 Directors who used Scope skillfully, like Minnelli and
Cukor, admitted a dislike for it 4 6 Even those reconciled to the format complained
about having to fill in the stretches around the actors, especially in close-ups. 47
Scope
films included jokes about their slightly freakish dimensions. In Gentlemen Marry
Brunettes (1955), Jeanne Crain wakes up wailing from "nightmares in CinemaScope,"
and when Jane Russell tips her head, from her point of view we see the Eiffel Tower
fitted sideways into the frame. The prologue of Frank Tashlin's The Girl Can't Help
It (1956) mocks the ratio, and so does a musical number in Silk Stockings (1957), in
which Fred Astaire and Janis Paige assure us that to attract moviegoers, "You've gotta
have Glorious Technicolor, Breathtaking CinemaScope, and Stereophonic Sound."
Fox anticipated complaints early on. Before The Robe's premiere, the studio
launched a publicity campaign spearheaded by directors and cinematographers
who worked on the earliest Scope films. For a 1953 promotional book, New Screen
Techniques, a string of articles signed by the craftsmen (but probably authored by
publicists) sought to turn the system's limitations into advantages. Henry Koster
didn't confess that he'd had to fasten his Robe stars into place; indeed, an article
bearing his name claims that in Scope the director "has an unparalleled chance to
demonstrate his ability to move actors logically and dramatically."48 Are close-ups
of single players impossible? Yes, usually, but Koster's essay notes that the big screen
provides constant close-ups—"and close-ups not of a single person, but of two, three,
or half a dozen simultaneously."49 Are camera movements restricted? Yes, but now
they're unnecessary. "Instead of moving the camera in to the actor to get a close-
up, I stage their movements so that they walk into the close-up."50 Do Scope films
minimize cutting, as Delmer Daves feared? Yes, and that's a good thing. An article
signed by Jean Negulesco claimed that now directors can't hide behind flashy cuts
and must learn to dramatize good dialogue and performances more honestly.51 In
shooting The Robe, cinematographer Leon Shamroy discovered, even action scenes
can be handled in "one smoothly flowing, life-like scene [i.e., shot]."52 Although shots
would run longer in Scope, Shamroy judged that "this won't be apparent to most audi-
ences because any well-edited film seems like one long uninterrupted strip of film
anyway."53 The campaign succeeded with some film critics and theorists, who argued
that CinemaScope fostered cinematic realism by minimizing the need for editing and
by emphasizing what happened within the shot.54
CinemaScope 293
Figure 10.8 Massive warping of space in
the opening of Around the World in 80 Days
(1956). The aberrations were less noticeable
on the huge, steeply curved Todd-AO screen.
As filmmakers began to discover the format's drawbacks, it needed more defending.
For a 1955 issue of American Cinematographer, the erudite director of photography
Charles G. Clarke provided a guide for shooting in Scope. Fox republished the piece
as a pamphlet to be given to workers at other studios. 55 Acknowledging that people
have voiced reservations, Clarke's essay tries to revise the official line. He points out
that the equipment has improved; there are now single-unit lenses in five focal lengths
(from 35mm to 152mm). Longer lenses make close-ups more feasible, because the
camera doesn't need to be moved close to the actors, but (perhaps granting the focus
problems with the long lens) Clarke recommends other options. Instead of big close-
ups, two shots are quite adequate in Scope: "The figure size of the Two-shot' is larger
than was the 'big head' on the older, smaller screen."56 If you feel the need for a close-
up, an over-the-shoulder (OTS) shot favoring the character will do the trick.
Clarke goes on to recommend how to shoot a typical scene, moving from establish-
ing shot to medium shot, with the characters maneuvered so that the person with the
most important dialogue is seen to best advantage. Instead of depth, use breadth.
No longer must we confine the actors to areas forward and backward from the
camera, but may now also use lateral movement. Spreading out of the action is
what is done in stage productions, and indeed CinemaScope technique is like
that of the theatre. 57
Nonetheless, some untheatrical effects, such as views straight ahead from moving
vehicles, can heighten the sense of "participation" (echoes of Cinerama and 3-D
again). Clarke reiterates the 1953 line about cutting as well. Because the big picture
approximates human vision, scenes can be staged with minimal editing. "I believe
that it is more comfortable, interesting, and natural to the spectator if scenes [i.e.,
shots] are sustained and a minimum number of cuts are made."58 In arguing for tech-
nological innovations, Hollywood's artisans have often recommended best practices,
and Clarke's article provided a reassuring message that Scope could easily fit into
established work routines.
Scope caught most of the blame for shortcomings in widescreen technology
generally, with critics overlooking the fact that, for instance, Todd-AO provided
distortions more warped than anything in Scope (Figure 10.8). In addition, as John
294 Poetics of Cinema
Belton points out, Scope began as an effort to imitate the shock-and-awe effects of
Cinerama, and in that enterprise close views and cutting were of less concern than
immersive spectacle.59 Probably Scope's technical problems were exaggerated as well.
As we'll see, close-ups and a degree of depth weren't completely off-limits in Scope,
and late 1950s work shows a distinct lessening of edge distortion. The multichannel
sound could assign dialogue, music, and effects to different areas of the image, so
that the words spoken by a character on screen left would come from the left speaker
behind the screen. This sound localization, quite alien to us today, could direct the
audience's attention within the wide frame. 60 Nonetheless, in 1955 the Fox team took
a conservative approach. It's likely that they weren't laying down an ironclad set of
rules for shooting Scope but rather suggesting the safest approach; if people followed
their recommendations, they wouldn't encounter great problems.
Undeniably, however, the new restrictions seemed to take away some essential
tools. Directors and producers valued the freedom to track the camera into and out
from the set, to use the crab dolly to turn in short arcs. Such "fluid camera" shots
added production values, and if efficiently executed, they could save shooting time,
replacing separate setups. Filmmakers didn't like being told to restrict themselves
to certain movements, such as panning shots (careful ones) and diagonally tracking
back with walking actors. Close-ups were an even bigger issue. Since the silent era, all
directors wanted facial close-ups in order to provide an emotional accent, to punch
up a drab scene, or to cover continuity gaps. Producers wanted close-ups because they
showed off the cast and allowed scenes to be recut in postproduction. Actors wanted
close-ups because they were actors. The industry remained skeptical of a camera pro-
cess that couldn't get within 7 feet of a star. Who wanted "close-ups" of several actors
at once?
One might expect that filmmakers would have been more receptive to Fox's cham-
pioning of lengthy shots, because the 1940s had seen a long-take vogue. Although
most films remained within U.S. sound cinema's traditional 8-11-second range, some
relied on extended takes virtually without parallel over the previous 2 decades. At all
levels of production, it isn't hard to find 1940s films with average shot lengths (ASLs)
falling between 15 and 20 seconds.61 Sometimes prolonged shots were flaunted as
signs of showmanship or virtuosity. Citizen Kane (1941) and The Magnificent Amber-
sons (1942) called attention to their long takes, as did The Lady in the Lake (1947) with
its 24 shots, and Hitchcock's Rope (1948) with its 11. But in these latter two films,
and in most others of the 1940s, the long take achieved variety through fluid camera
movements. Citizen Kane had been criticized for its static single-shot sequences, and
most directors preferred to extend their takes by tracking and panning. But no, said
Scope's defenders; in obedience to this "theatrical" technology, the camera had to give
up the fluidity of the past few years.
If the camera was to sit still, a great deal of a scene's import would depend on
ensemble staging, and Clarke, like other Fox defenders, had recourse to the compari-
son with theatrical blocking. But his recommendations are characteristically silent
on exactly how to arrange the actors in the scene, and for good reason. In addition
CinemaScope 295
Figure 10.9 Although lining up characters Figure 10.10 A standard medium two shot
in a row perpendicular to the camera was the from The Mark ofZorro (1920).
most common staging strategy of the 1910s,
Assunta Spina (1915) exemplifies one sort of
depth that was also employed. The foreground
plane is fairly far from the camera, making
the action in the background quite distant.
For other examples, see Figures 1.1-1.2 and
9.1-9.5.
to restricting close-ups and camera movement, Scope initially induced a crisis in
Hollywood staging practices.
Hollywood's visual style has its roots in the silent era. In the years from 1906 to
1915 or so, filmmakers in various countries refined a "tableau" cinema based in long
takes. Usually the characters were arranged in a horizontal line across the frame, but
sometimes the blocking moved them diagonally into the distance. Either way, there
might well be intricate blocking to carry the drama across the tableau (Figure 10.9).
By 1920, this system had been transformed by a standardized approach to editing.
Cuts broke the scene into smaller bits and varied the camera angle. At the same time,
though, the horizontal array of players remained the dominant staging technique.
Although the actors might face each other, their bodies tend to be pivoted some-
what toward the viewer (Figure 10.10). The various close-ups, reverse angles, over-the
shoulder shots, and the like took the viewer around this lateral layout, and the char-
acters might be spread out further in the set, but even in long-shot framings the most
common layout remained fairly shallow (Figure 10.11). Depth staging became rarer
than it had been in the 1910s cinema.
Lining up the actors like clothes on a line is well suited to the building block of most
narratives, the dialogue exchange between characters. The flow of conversation is pre-
sented with clarity and point, showing faces and bodies so as to highlight expressions,
gestures, and bits of business that nuance the situation. OTS shots and singles of each
player stress particular lines or facial reactions. This schema prevailed throughout the
early sound era. Players are arrayed in the classic two shot (in the knees-up framing
known as the plan-americain, or in a medium shot, as in Figure 10.12). When more
than two characters are involved, the camera either shoots from somewhat farther
back or crowds the actors closer together (Figure 10.13). Sometimes the actors stand
296 Poetics of Cinema CinemaScope 297

Figure 10.11 Even in a long shot, characters
line up more or less in clothesline formation
(The Mark ofZorro).
Figure 10.12 The linear two shot persists as
a staple of sound cinema (Jezebel, 1938).
Figure 10.15 In the 1940s, recessional stag-
ing became more common, setting characters
into considerable depth (Caught, 1949).
Figure 10.16 Citizen Kane (1941) made
extreme depth staging central to its aesthetic;
here Mrs. Kane signs her son away to a guard-
ian while the boy plays outside unawares. This
shot, like several in the film, was achieved
through a rear-projected film of the boy seen
through the window.

Figure 10.13 Several characters grouped so Figure 10.14 Stacking rows: A crowded
that their bodies, facing us, fill out the frame scene in The Thin Man (1934) still obeys the
on a single plane (Little Women, 1933). clothesline principle, in layers.

Figure 10.17 Depth-oriented directors could

take,advantage of the nearly square propor-
in not one but two rows, one behind the other but still more or less 90 degrees to the tions of the Academy ratio to build up vertical
lens (Figure 10.14). These "clothesline" arrangements, spreading several players across compositions [The Little Foxes, 1941).

a perpendicular plane in profiled or fairly frontal views, became a basic technique for
dialogue scenes of 1930s cinema.
I'm not saying that 1930s cinema was excessively static or "theatrical." Some
directors exploited depth behind the main plane and explored the changing angles
afforded by camera movement. During the 1920s, Ernst Lubitsch and other directors
explored a more complex version of continuity, with the camera at the center of the
characters' dialogue exchange. 62 In other films, large-scale scenes did allow the cam-
era to penetrate the space more fully. In ballroom dances, sporting events, and courts
of law, the drama unfolds in several zones, and the camera tends to be positioned in
a space within those fields; we get a sense that the action is taking place all around
us. Passages of physical action likewise display great freedom of angle (high, low) and
depth. This is especially true of outdoor work, as we'd expect, because an exterior
set yields greater choice of camera position than an interior one, and natural light
permits greater depth of field. Even in big scenes and outdoor filming, however, the
clothesline schema tended to be the default staging.
Another set of options appears sporadically during the silent era and the 1930s,
but it becomes more prominent in the 1940s. Employing what art historian Heinrich
Wolfflin calls "recessive" composition, a scene could be staged along diagonals. 63
Recessive staging activates depth, placing one character notably closer to the cam-
era than the others (Figure 10.15). The depth can be relatively shallow or quite steep.
Sometimes the diagonal option gives us two or more distinct playing areas. We may
have independent actions taking place in both foreground and background (Fig-
ure 10.16). The layout may be lateral (foreground on left or right, and background on
the opposite side) or vertical (foreground at bottom or top; Figure 10.17).
This schema poses problems of visibility—if the foreground character is facing the
distant one, then she's turned from us and we can't clearly see her face—so some
compensations are called for. The most common fix is to let shot/reverse-shot cutting
favor first one character, then the other, creating "stretched" OTS shots. Another
 Poetics of Cinema CinemaScope
298 299

Figure 10.19 The "baroque" side of reces- Figure 10.20 Small and rather cheap Warner Figure 10.21 A tight composition from The
Figure 10.18 A shot from Gun Crazy (1949)
sional staging (The Killers, 1946). Such com- Bros, sets are well concealed by recessive Purple Heart (1944). The precise alignment of
exemplifies the extremes to which the deep-
positions can be found throughout world layouts of figures in The Maltese Falcon (1941). props and players, particularly the foreground
space staging of the 1940s could go.
cinema in earlier decades, but they became judge's head, is typical of post-Kane American
far more common during the 1940s.

compensating maneuver is to turn the foreground actor toward us, motivating this
as the character's refusal to face the other figure. The frontally positioned foreground
character became a common device in 1940s dramas, allowing us to see what the
background character can't. In either type of staging, the action can develop along the
diagonal, with characters moving toward or away from one another, perhaps in zigzag
paths as well. This schema is well suited to building suspense and heightening tension,
so it's not surprising that thrillers and psychological dramas are its natural home.
Does recessive staging signal a return to the tableau schemas of the 1910s? No,
because directors of the 1940s typically put the foreground figure much closer to the Figure 10.22 John Ford favored bold depth Figure 10.23 Films shot on color stock, which
compositions from the very beginning of his needed more light and didn't allow for great
camera than in the earlier era. In many shots the foreground figure is presented in
career. Here Wyatt Earp in the foreground depth of field, continued to favor the clothes-
looming close-up, and the background figures can be either distant or fairly close, is about to slide his pistol down the bar to line layout (Leave Her to Heaven, 1945).
packing the frame (Figure 10.18). In the recessive strategy, the amount of playing his brother under the nose of Doc Holliday
(My Darling Clementine, 1946).
space is greater than in the horizontal arrangement. Because the camera captures an
optical pyramid far deeper than it is wide, distance between characters can increase
with depth. This yields the "baroque" extremes of size and position that we some-
usually associated with big sets, having the camera hug the axis of action also permitted
times find in 1940s cinema (Figure 10.19).
As directors began experimenting with recessive staging in the 1930s and early sets to be more compact (Figure 10.20). In general, 1940s compositions became tighter, as
1940s, it became clear that planes so far from one another could not be kept easily in the urge to fill the frame created layouts that click neatly into place (Figures 10.21-10.22).
focus, especially if the foreground needed to be quite close. Most directors learned Because the 1.33 frame was firmly established as the standard, directors could count
to live with this, either letting one plane drift a little out of focus or keeping the on their clenched compositions being retained in most movie houses.
foreground fairly far from the camera. The "deep-focus" style heralded by Citizen For all its popularity, recessive staging remained a secondary option for 1940s
Kane (1941) provided technical solutions: lots of light, faster film, coated lenses, exact filmmakers. Most films continued to use variants of clothesline arrangements
diaphragm stops, and special effects trickery. Now one could have very deep shots (especially in color, which did not permit great depth of field; Figure 10.23). Many
and full focal range. Welles' flamboyant staging schemas would be toned down and directors adopted a moderate approach to depth, avoiding the most outre composi-
normalized throughout the 1940s and 1950s.64 tions and blending depth staging with more lateral layouts within a single shot. A
Once filmmakers started exploring diagonal staging schemes, they seemed to have film shot predominantly in the clothesline manner might include a few deeper com-
realized that standard establishing shots became less obligatory. A scene could start positions. Even directors who began in the 1920s and 1930s became accustomed to
right on a deep composition, then reveal the set as necessary. Although depth staging is diagonal staging options, if only as an occasional resource (Figure 10.24).
300 Poetics of Cinema CinemaScope
301

Figure 10.24 Alfred Hitchcock experimented Figure 10.25 The classic two shot remains a Figure 10.26 In Adam's Rib (1949), Cukor Figure 10.27 Ihe "intimate" two shot in
with recessional compositions after coming basic resource into the 1950s, as shown here also accedes to the new impulse: A wife trail- Demetrius and the Gladiators (1954). Compare
to America; this striking shot is from Lifeboat in George Cukor's The Marrying Kind (1952). ing her husband is caught in a sharply angled Figures 10.10, 10.12, and 10.25.
(1944). depth shot on location.

Such was the menu of staging schemes in force when Scope appeared in 1953, and the ? ,N> *• • •••'"!

ultra-wide format made mischief with several of them. By 1950, directors had grown
accustomed to having the recessive staging option available. But CinemaScope seemed
to take it away. No more deep-focus shots taken at vivid angles, with heads dotting
the frame high and low. Now the entire frame couldn't be grasped as a single forceful
totality. Criticizing widescreen processes, cinematographer Boris Kaufman asserted a
classic 1940s premise: "The space within the frame should be entirely used up in com-
position."65 It seems likely that Fox's staging recipe worried filmmakers because they
had mastered the 1940s recessive schemas, and now those options were banished.
Instead, Scope seemed to push staging practices back to the 1920s and 1930s—the
planar option of two figures facing one another, perpendicular to the camera. In shoot-
ing A Star Is Born, George Cukor complained, "Everything had to be played on a level
plane—if someone were too much upstage, they would be out of focus."66 Evidently
referring to Clarke's recommendations, Cukor's art director Gene Allen recalled,
Fox had given us this whole list of rules, like lining up your actors in a straight
row, because of perspective problems, focus problems, and all. Well, Cukor said,
"I don't know how the hell to direct people in a row. Nobody stands in rows."67
What's fascinating here is that Cukor and his peers knew very well how to direct actors
in rows. He started doing it in the 1930s and continued right up to the advent of Scope
(Figures 10.13 and 10.25). But he had also exploited recessive staging (Figure 10.26),
and Scope threatened to banish that tool from his kit.
Worse, filmmakers couldn't easily return to the planar layouts because now these
looked a little silly. Lining up two or more bodies in the 1.33 frame permitted, at the
very least, an unobtrusive encounter of two or more characters at close quarters. But
how do you compose the same encounter in Scope? Put them in the center of the frame,
and suddenly this traditional array looks awkward (Figure 10.27). There's acreage
stretching out on either side of the figures, violating Kaufman's rule of thumb about
the composition utilizing the entire frame. But if you move the couple apart, you're
SL* :i •
Figure 10.28 A first meeting, rendered in Figure 10.29 Clothesline staging with a
another two shot (Love Is a Many-Splendored vengeance in the soda-fountain epilogue to
Thing, 1955) that leaves large areas to be filled How to Marry a Millionaire (1953).
with props and decor.
creating a gulf in the center and turning an intimate encounter into a more detached
one (Figure 10.28). Add more characters, and you're likely to follow the line of least
resistance: an almost comical clothesline composition (Figure 10.29). The clumsiness
of such shots is implicit in Cukor's worry about directing strings of people. Having
taken away the deep-space schemas of the previous decade, Scope also made the tradi-
tional planar arrangements look embarrassingly artificial. "Nobody stands in rows."
Scope, then, seemed to limit camera movement and close-ups, reduce cutting rates,
ban deep focus, and expose as artificial one of the most basic staging tactics. The new
process seemed to have taken away virtually all of a director's visual resources. What
were filmmakers to do?
Taming a New Technology
Directors responded to the advice of Clarke and his colleagues in ways as various as
we might expect, given the cussedness of human nature. Some followed the guide-
lines, and some didn't. In the first couple of years particularly, many directors avoided
close-ups and kept the camera well back. Others accepted CinemaScope mumps (did
the audience notice, or care?) or somehow cured them. The climax of Joshua Logan's
Bus Stop (1956) includes surprisingly undistorted "choker" close-ups that look forward
 Poetics of Cinema CinemaScope 303
302

Figure 10.31 Bus Stop: more typical Figure 10.36 The lines of rowing sailors Figure 10.37 The Scope frame broken into
Figure 10.30 An unusually big close-up for
ensemble shot. create vectors that culminate in Paris and two symmetrical chunks by windows fram-
early Scope (Bus Stop, 1956).
Aeneas, lounging in the foreground (Helen of ing the characters in The Bridge on the River
Troy, 1956). Kwai (1957).

Figure 10.32 The bank robbery in Violent
Saturday (1955) makes ambitious use of edge
framing. As a bank officer dives under his
desk and a customer shrieks, one ol the thieves
stands poised at the right side o f t he frame.
Figure 10.33 The telltale eyeglasses in Com-
pulsion (1959) sit at the very bottom edge of
the shot. No director today would dare try
this, because film projection varies so much
from theater to theater.
Figure 10.38 Recessive anamorphic com- Figure 10.39 In Brigadoon (1954), the shots
positions could be built around tables and spread out the mythical Scottish town in a
other set elements (Battle Cry, 1955). fresco of dance and color.

Figure 10.34 As the officer in the center real-
izes that his rival has been promoted, we're
expected to notice the new star on the man's
shoulder, at extreme left (D-Day the Sixth of
June, 1956).
Figure 10.35 Aboard the airplane in The Figure 10.40 Brigadoon: Once the adman
High and the Mighty (1954), the characters protagonist is back in New York, however,
are framed by the seat on the left and the the compositions are cramped and closed-off,
stewardess's body on the right. suggesting stifling metropolitan life.

to today's monstrous faces (Figure 10.30). These have a powerful impact in a scene
that otherwise relies largely on distant shots and deep space (Figure 10.31). Likewise,
directors who worried about edge distortion placed their action in the central half or
three-fifths of the image. Now and then, though, key elements would be thrust to the
very side of the frame, to create dramatic tension or to induce the viewer to scan the
shot actively (Figures 10.32-10.33). Even Koster, once out from under The Robe, tried
his hand at edge framing (Figure 10.34).
As Delmer Daves' recollection of Zanuck's briefing of the Fox troops indicates,
directors were particularly worried about directing attention in the Scope frame.
"We have spent a lifetime," Hawks remarked, "learning how to compel the public
to concentrate on [a] single thing. Now we have something that works in exactly
the opposite way, and I don't like it very much."68 The most defensive reaction was
to deemphasize the empty stretches of the frame. Filmmakers filled the holes with
props or flanking figures, and blocked off chunks altogether (Figure 10.35). Fred
Zinnemann recalled spending most of his time "inventing large foreground pieces to
hide at least one-third of the screen."69 Other traditional devices might highlight an
item/Lines in the set could link or lead to characters (Figure 10.36). Actors could be
framed within corners, columns, and doorways, which broke the big screen into more
readable modules (Figure 10.37).70
Some directors set up recessive compositions despite Scope's depth of field problems.
Often sets create diagonals along which the players arrange themselves (Figure 10.38).
Because horizontals warped considerably in Scope (Figure 10.5, above), filmmakers
tended to shoot rectilinear solids from a 3/4 angle, which makes the distortions of
parallel lines less apparent and also creates a deeper space, though not all of it might
be used for dramatic purposes. 71 Minnelli's Brigadoon (1954) assigns thematic weight
to alternative stylistic schemas, using very frontal clothesline compositions for the
fantasy world of the Scottish village but presenting claustrophobic depth shots for the
modern Manhattan to which the hero returns (Figures 10.39-10.40).
 Poetics of Cinema CinemaScope
304 305
Cutting rates also indicate how differently filmmakers responded to Scope's pro-
claimed limitations.72 Early on, several directors accepted the challenge of the anamor-
phic long take. Three Scope films from the inaugural year of 1953 boast longish ASLs:
13.2 seconds for King of the Khyber Rifles, 15 for The Robe, and 21.2 for How to Marry
a Millionaire. This trend continued for a little while. Of the 68 Scope films I surveyed
from 1954 and 1955, about a third have ASLs falling between 12.0 and 19.9 seconds, and
Figure 10.41 The brilliant southwest sun- Figure 10.42 Elia Kazan cants the Scope
this is a higher proportion than we find in flat films of those years. Nine more films have light of Bus Stop permits striking depth frame throughout East of Eden (1955). Here
ASLs running over 20 seconds, a proportion not seen since the early silent years.73 compositions, such as this one showing the it creates a dynamic depth composition, with
Evidently, directors who had made the long take integral to their style found no rea- cowpoke protagonist tying on the scarf of the Cal on the porch while his father and Abra
woman he loves, while she watches, in frame wait at the party inside.
son to change. Otto Preminger had already become Hollywood's principal long-take center, from the stands.
director with Laura (1944), 21 seconds ASL; The Fan (1949), 21.8 seconds; and Fallen
Angel (1945), 33 seconds. George Cukor, Vincente Minnelli, Joseph Mankiewicz,
and Billy Wilder also favored fairly long takes during the pre-Scope era. So it's not
surprising that several Scope films by these directors boast lengthy ASLs, coming in
between 16 seconds (Cukor's A Star Is Born, 1954; Minnelli's Lust for Life, 1956) and
34 seconds (Minnelli's Brigadoon, 1954). Preminger's Carmen Jones (1954), running
35 seconds per shot, may well be the longest-take CinemaScope film ever made.
Figure 10.43 A low-key, low-angle framing Figure 10.44 The Tarnished Angels (1958): A
Significantly, in their Scope long takes, these directors often make use of extensive recalling film noir for Jubal (1956).
wide-angle composition reminiscent of 1940s
camera movements, regardless of Fox's warnings to the contrary. The lateral tracking style made possible by black-and-white Scope.
in the mess hall of Carmen Jones is only one instance of many.
Yet Scope didn't oblige all directors to give up rapid editing. Two titles released in
1953 are cut fairly fast: a 9.1-second ASL for Beneath the 12-Mile Reef and a startling between planes. These were easiest to accomplish outdoors, where brilliant sunlight
6.9 seconds for Knights of the Round Table. For 1954-1955, between 30 and 40% of the permitted even Bausch & Lomb lenses to achieve robust depth of field. Many of the
ASLs I examined run between 8 and 12 seconds. Most surprisingly, nearly a fifth of most remarkable shots in Scope can be found in Westerns, in adventure yarns, in
the films sampled for these years have ASLs shorter than 8 seconds. If quick cutting ancient world sagas shot in Italy or Spain, and in contemporary dramas set in the
on the big screen made viewers uncomfortable, nobody told Henry Hathaway (Prince blasting daylight of the Southwest (Figure 10.41). Even under studio illumination,
Valiant, 1954,6.6 seconds ASL), Robert Wise (Helen of Troy, 1956, 5.4 seconds), or the
though, some depth was achievable. In the canted framings in East of Eden (1955),
animators of Lady and the Tramp (1955, 4.5 seconds).74 shot both on location and in the studio, Elia Kazan seemed to be trying to become
The dynamic of recovery continued during Scope's life span, as directors absorbed
the Orson Welles of Scope (Figure 10.42). More discreetly, Delmer Daves relied on the
the format into more normal cutting rhythms. Really long takes become increasingly
deep-focus look in both exteriors and interiors for his Western Jubal (1956), built out
rare. For the 1956-1960 period, the center of gravity in my sample shifts, and two
of deep shots reminiscent of film noir (Figure 10.43).
thirds of the films' ASLs fall between 7 and 13 seconds. Fewer films average longer
Two other factors helped directors recover the 1940s depth aesthetic. One was
takes than 13 seconds, but more ASLs come in at less than 7 seconds.75 It seems that
Buddy Adler's decision to permit Scope films to be shot in black and white. Because
when Scope was introduced, long takes offered a line of least resistance, particularly
black-and-white film required much less light than color, cinematographers could
given all the other problems of filming with the system, but from the start any film-
maker from any studio who preferred to cut frequently could do so. The same options, stop down the lens diaphragm and get sharper images with better depth of field. With
incidentally, were available in most other widescreen systems.76 As the years passed, color, Scope filmmakers shot most close-ups with long, shallow-focus lenses, but
filmmakers working in both flat and anamorphic formats tended to accelerate their black and white allowed freer use of wide-angle lenses. The crisp, wide-angle imagery
editing pace. In the 1960s, double-digit shot lengths began to become almost extinct of Douglas Sirk's The Tarnished Angels (1958; Figure 10.44) would have been virtu-
77 ally impossible in color (as his Interlude, 1957, shows). The bivouac scenes of Edward
in all Hollywood movies.
As in the early sound era, artisans struggled to normalize the new technology, Dmytryk's The Young Lions (1958) present compositions as tensely jammed as any-
to throw off its constraints and restore earlier options. For many directors, this thing from the 1940s (Figure 10.45). Black-and-white Scope has a special following
entailed recovering the look they had come to prize in the 1940s: tightly composed among cinephiles, perhaps because images like these announce the triumph of
images, taken from high or low angles and yielding striking differences of scale aggressive style over the academic blandness promoted by Fox's spokesmen. 78
 Poetics of Cinema
306
Elvis gets Scope mumps (Love
Figure 10.45 A composition even more Figure 10.46
tightly articulated than Figure 10.21, in black- Me Tender).
and-white Scope (The Young Lions).
Figure 10.47 Elvis is cured, thanks to Pana- Figure 10.48 Panavision lenses permit a
close and foil-frame composition in Under-
vision optics (Jailhouse Rock, 1957).
water Warrior (1958).
Figure 10.49 Strangers When We Meet
(1960): Panavision renders good focus, in
close-up and color.
A second pressure toward normalization has already appeared in our story: the
emergence of Panavision. The new system was sold largely on its ability to provide
acceptable close-ups (Figures 10.46-10.47). Panavision's president Robert Gottschalk
had claimed that stars would soon refuse to appear in Scope films because the lens
made them look fat.79 In winning an Academy Award for Scientific or Technical
Achievement in 1958, the process was praised for its ability to "substantially reduce
photographic lateral distortion and thereby improve close-up quality and overall
definition." 80 A bonus, however, was the ability of Panavision lenses to handle recessive
staging without loss of focus. Early (and uncredited) Panavision films, mostly black-
and-white projects, display remarkable close-ups and depth of field (Figure 10.48).
Soon Panavision was offering a range of specially made lenses that could render depth
of field far more crisply, and when the process was used for color, the results were
impressive as well (Figure 10.49). By the early 1960s, the big-foreground wide-angle
look was attainable in widescreen, and Panavision was in the driver's seat.
In all, despite its peculiarities and constraints, Scope was absorbed into the norms
of classical continuity. Almost from the start, the new screen format was displaying the
CinemaScope 307
Figure 10.50 Guys and Dolls (1955): Rows of Figure 10.51 The "Glorious Technicolor"
figures fill out the frame and highlight the cen- sequence of Silk Stockings (1957) mocks the
tral information as Nathan Detroit challenges impossible ratio.
Sky Masterson to date Salvation Army Sarah.
cutting rates and some of the image schemas that had become familiar in the 1940s. The
goal—to make cinema in the 2.35 ratio as much like that in the 1.33 ratio as possible—
was being achieved. Even Charles G. Clarke gave in. His cinematography on Flaming
Star (1960) yields plenty of singles, close shots, and camera movements, and the average
shot length is a brisk 6.8 seconds. Clarke's tacit repudiation of the Fox aesthetic is one
sign that traditional techniques of cutting and framing had absorbed Scope.
But was this all there was to the CinemaScope revolution? Did it contribute noth-
ing of aesthetic value in itself? To answer this, I think we can profitably look more
closely at staging practices. Promoters of the Fox aesthetic were partly right: Moving
the actors around the frame was a crucial part of Scope aesthetics. But it was not to be
"theatrical" in exactly the sense that Koster, Negulesco, and company probably had
in mind.
Some Virtues of Clotheslines
Scope movies, of course, rely on clothesline staging. Shot after shot presents, at various
scales, a pair of characters facing each other on the same plane (Figures 10.27-10.28).
Bars, lunch counters, and other horizontal settings encourage directors to string
several characters across the frame. As in 1930s films, the schema also accommodates
horizontal layers of figures, as well as flanking figures to fill in the sides (Figure 10.50).
The perpendicular layout is the foolproof Scope default, the main source of our sense
that early Scope films are rather uninteresting, and the object of Zanuck's undying
love. "The greatest kick I get is when one person talks across the room to another
person and when both of them are in the scene and near enough to be seen without
getting a head closeup."81 He ordered his directors to place characters a good distance
apart, because the stereophonic sound was more pronounced that way.82 This practice
is parodied in the "Glorious Technicolor" number of Silk Stockings (Figure 10.51).
Once lateral staging supplies a baseline, the filmmaker can move in for standard
OTS framings, as Clarke's suggestions indicated. Heavy reliance on shot/reverse-shot
editing is a major source of brief average shot lengths in early Scope films. To com-
plete the package, framings presenting only one character are rare but not forbidden.
These singles avoid exact centering and leave an open area on the right or left, usually
to imply something offscreen that is the object of the character's look (Figure 10.52).
308 Poetics of Cinema CinemaScope 309

Figure 10.52 The off-center single shot, Figure 10.56 Love Is a Many-Spiendored Figure 10.57 . . . to be filled.
implying action taking place off frame (The Thing: A clothesline array leaves a vacancy on
Robe, 1953). the r i g h t . . .

Figure 10.53 Pat and Mike (1952): The cop

cut off at frame left restages the scene in
which Pat rescues Mike from the thugs.

Figure 10.58 The Man in the Gray Flannel Figure 10.59 A Star Is Born (1954): Televi-
Suit (1956): Scope breadth to dramatize a sion versus cinema, balanced at opposite ends
couple's fraying marriage. of the frame.

ters can be stationed at opposite edges of the image (Figure 10.58). Other filmmakers
use the horizontal sweep more delicately, as Jacques Rivette predicted:

Figure 10.54 Pat and Mike: As the reenact- Figure 10.55 . . . and he slips into a gap
ment starts, the changed positions reveal a between characters to put in his version of
waiter in the back tier . . . events.
Shot with a long lens, such a "single" risked losing focus, but at least an off-center face
was less likely to contract CinemaScope mumps.
Yet Scope didn't simply replicate the clothesline layout of previous decades; it
added something too. What Scope initiated wasn't horizontal staging as such, but
spacious horizontality. Consider a shot from Pat and Mike, released in 1952, a year
before CinemaScope was introduced. Here Cukor presents a fairly distant long take
by skewing the row of people slightly into depth, shooting from a slightly high angle,
and moving figures gracefully through apertures and a central zone of emphasis
(Figures 10.53-10.55). An equally packed shot is rare in early CinemaScope: The same
number of characters would be spread out more widely. The new format tended to
push people apart, forcing more air between them. In Love Is a Many-Splendored
Thing (1955), a composition showing four people flanking a priest (Figure 10.56) leaves
a fairly wide aperture available for a new character to enter (Figure 10.57). Compare
the tinier, more angular slot that the face of the excitable white-coated waiter enters
in Pat and Mike (Figure 10.55).
The greater distances between figures in turn became sources of expressive effect.
The most common example is emotional separation. Conflicting or estranged charac-
The director will learn how he can sometimes claim the whole surface of the
screen, mobilize it with his own enthusiasm, play a game that is both closed
and infinite—or how he can shift the poles of the story to their opposites, create
zones of silence, areas of immobility, the provoking hiatus, the skilful break.
Quickly wearying of chandeliers and vases brought into the edges of the image
for the "balance" of the close-ups, he will discover the beauty of the void, of free,
open spaces swept by the wind. 83
In A Star Is Born (1954), horizontality combines with edge framing to create shots
that oblige us to scan the full stretch of the image. When the fading star Norman
Maine talks with his producer at home, other guests are watching a film in the screen-
ing room. Cukor presents the two men fixed between the film image on the right
and the upstart medium of TV on the left (Figure 10.59). The shot is echoed later
when Norman's wife, Vicki Lester, receives her Oscar. A vast long shot shows her
stranded in the middle of the stage, but a big-screen TV image of her is pasted in at
the upper right (Figure 10.60). Suddenly her face starts to get larger, and we must shift
our eye leftward to detect the cause: a TV camera coasting slowly in from offscreen
(Figure 10.61). The effect doesn't feel forced because the shot remains plausibly
spacious; a 1940s film would have had to pack the frame more tightly, perhaps having
the TV camera nose into the frame from lower left and fill the foreground.
Rivette's precepts can be honored even in shots that aren't rigidly horizontal. In
Ronald Neame's The Man Who Never Was (1956), a low-angle, 3-minute take shows
a grieving father deliberating whether to let his dead son's corpse be used in a spy
310 Poetics of Cinema CinemaScope 311

Figure 10.60 A Star Is Born: Later, when Figure 10.61 A Star Is Born: The TV camera Figure 10.65 Abstract masses in Land of the Figure 10.66 Bracing for a hit, the German
Vicki Lester receives her award, her image on coasts in at frame left, recalling the earlier Pharaohs (1955), with the toiling slaves as submarine crew is fanned out like a poker
the television monitor swells. Why? scene's interplay of film and television. verticals balanced by striated horizontals. hand (The Enemy Below, 1957).

Figure 10.62 A father decides whether to Figure 10.63 Rebel Without a Cause (1955) Figure 10.67 The Scope frame as an abstract
give his son's remains to the war effort (The puts Judy in the center, framing Jim outside slice of space (Picnic, 1955).
Man Who Never Was). in the left window and revealing Plato and
the family maid on the far right.

Figure 10.64 Geometric elements of the
setting frame characters in medium shot (The
Badlanders, 1958).
mission rather than given a decent burial. The officer who has proposed the mission
has turned discreetly away (Figure 10.62). The angled depth recalls the 1940s, but in
the 1.33 era, the players would have to be jammed together, and the result would seem
airless and perhaps overwrought. The 2.35 proportions allow Neame to create "zones
of silence" that respect the solitude of each man while still letting us see the father's
agonizing choice play out over his face.
Although filmmakers in the 1.33 ratio have long used architecture to segregate
areas of the frame, Scope's width invited—demanded, some directors felt—a parti-
tioning of the visual field. This creates a strip of modules, and these can be juxtaposed
in breadth or depth, in order to isolate characters or to establish relationships. We
see this already in our Star Is Born scenes and many others I've invoked in this essay.
Nicholas Ray's frames-within-frames in the opening of Rebel Without a Cause (1955)
highlight the three main characters for us, and the partitioned setting connects them
before they even know each other (Figure 10.63). Closer views can be subjected to the
same rhythmic division and repetition (Figure 10.64).
The partitioning strategy tends to treat the screen not as a wraparound window on
a large chunk of reality but rather as a surface to be broken into rhythmic units. This
tendency can be heightened when the director emphasizes shapes, color contrasts,
and other pictorial features. The result comes close to an abstract configuration of
elements. Even the tiresome biblical spectacular can be assigned a majestic geometry
(Figure 10.65). The Enemy Below (1957) largely treats the interiors of its submarine
as a nest of rectangles cradling its crew, but just before impact the men's bodies are
fanned out like the fingers of a hand (Figure 10.66). Picnic, one of the most arrest-
ing 1955 Scope releases, has many points of interest, including daring close-ups and
flamboyant depth staging, but it's also noteworthy for its commitment to pictorial
abstraction. Joshua Logan and master cinematographer James Wong Howe provide
bold compositions outside and inside Midwestern grain elevators (Figure 10.67).
Such shots show that in a sense, Scope didn't expand the visual field; it cropped it.
"I never felt the screen was truly wider," Minnelli recalled. "It just tended to cut off the
top and bottom of the picture." 84 This tank-turret slit, by hiding so much, can yield
abstract imagery. Once the shot becomes an arbitrarily chopped-out strip of space, it
can be vividly decorative or expressive. In Kazan's long-lens portrait of rednecks plot-
ting against the Tennessee Valley Authority, with tattered posters balancing them in
the wide frame (Wild River, 1960), one can glimpse the sort of stylization that Godard
would accentuate further in Made in USA (1966) (Figures 10.68-10.69). In Bonjour
Tristesse (1958), one of the most painterly of Scope films, one scene begins with only
heads and shoulders, ends with only legs, and in the middle features an abstract
swoosh of blue to punctuate a moment of passion (Figures 10.70-10.73). The pure
wash of color underscores Cecile's erotic outburst, but once the umbrella is grounded
it becomes a prop again, masking off the couple's faces and forcing us to watch their
urgently moving hips and legs.
312 Poetics of Cinema CinemaScope 313

Figure 10.68 A planimetric composition for
a late Scope film (Wild River, 1960).
Figure 10.69 Once more, Jean-Luc Godard
takes anamorphic matters a step further, cre-
ating a Picasso-like collage of ragged strips of
space (Made in USA, 1966).
Figure 10.74 Midrange depth in Demetrius
and the Gladiators, with characters at differ-
ent distances from the camera.
Figure 10.75 The Robe: Demetrius, near
death, lies in Marcellus' family home, with
characters dotted around the frame. Ini-
tially the drama plays out on frame left, with
Marcellus and Diana at the window as he's
tormented by guilt.

Figure 10.70 Bonjour Tristesse (1958): Cecile
and her new boyfriend are talking about
Anne, who has started a romance with Cecile's
father.
Figure 10.71 Bonjour Tristesse: Carried
away by the idea of Anne's happiness, Cecile
flings herself into a kiss with Philippe.
Figure 10.76 The Robe: After a cut-in to
Marcellus and Diana, we return to the master
shot as John the Baptist comes in. The doctor
in the foreground swivels as Marcellus greets
John, clearing a space for him . . .
Figure 10.77 . . . which allows John to
stride to f r a m e center as other characters
rearrange themselves.

Figure 10.72 Bonjour Tristesse: The beach Figure 10.73 Bonjour Tristesse: They tumble
umbrella he's holding drops forward, to the floor, with the umbrella now masking
momentarily filling the frame with brilliant their faces and throwing all the emphasis on
blue before revealing the couple behind. their writhing legs.
Reinventing the Tableau
Such quasi-abstract images (and we have to imagine them projected on a screen over
60 feet wide) confirm the Cahiers critics' faith in the horizontal power of the image.
Rivette again: "Wouldn't great mise-en-scene, like great painting, be flat, hinting at
depth through slits rather than gaps?"85 Yet he and his colleagues probably went too
far in seeing the anamorphic format as a step beyond the baroque deep-focus of the
1940s. Just as CinemaScope forced directors to revise horizontal schemas that had
emerged in earlier years, so it obliged a rethinking of the 1940s depth schemas, gaps
and all. And it pushed some directors back to a mode of deep-space staging that had
preceded the development of classical continuity.
Scope's initial problems with distortion, focus, parallax, and depth-of-field pre-
vented filmmakers from achieving the big foregrounds and wire-sharp focus that they
had come to prize. The fact that early CinemaScope films were in color intensified
the difficulty, because Eastman stock was relatively slow and required a great deal
of light to get even moderate depth of field. For all these reasons, directors typically
brought the nearest figures no closer to the camera than 8 or 10 feet, and most shots
placed them quite a bit farther away. Yet the playing space was not as utterly flat as
clothesline staging might have made it seem. A cinematographer obeying Clarke's
recommended exposure (f/4.5) could have focused the standard 40mm and 50mm
Scope lenses at various distances, some of which would create playing areas between
10 and 30 feet. A playing area 20 feet deep allows considerable flexibility in staging. 86
Despite warnings about depth of field, many filmmakers freely checkerboarded their
figures in midrange layers (Figure 10.74).
Within this midrange playing space, a resourceful director could revive the
Hollywood tradition of graceful group dynamics within a general shot. This tech-
nique, virtually forgotten today, involves inconspicuously highlighting first one
player, then another. A character takes up a spot, then shifts to another place just
as a second character moves to fill the gap. Despite focus problems in The Robe, for
instance, Koster can sometimes move his players smoothly into and out of central
zones of attention (Figures 10.75-10.77). This choreography is made more felicitous
when characters cross each other's path, or rhythmically compensate for each other's
change of position (Figures 10.78-10.80). A much emptier set, the dusty hotel lobby
in Bad Day at Black Rock (1955), allows John Sturges to coordinate the ominous
movements of Reno Smith's men as they plan to put pressure on the mysterious
314 Poetics of Cinema CinemaScope 315

Figure 10.78 In Island in the Sun (1957),
the colonial family argues about their family
secrets, now exposed in the newspaper. The
brilliant daylight allows many planes of dis-
tant depth to be activated.
Figure 10.79 Island in the Sun: As they Figure 10.84 Somewhat like Jacques Tati's Figure 10.85 Kiss Them for Me: When new-
quarrel, characters advance, cross the paths restaurant scene in Play Time (1967; Fig- comers arrive, the clothesline layers become
of others, and turn from the camera . . . ure 7.15), the party in Kiss Them for Me (1957) slightly more recessive.
creates layers of distinct actions.

Figure 10.80 . . . before a servant, after qui-
etly tending to her duties in the far distance,
steps into a gap to announce the meal.
Figure 10.81 The hotel lobby in Bad Day
at Black Rock (1955) is used imaginatively
in each scene set there. Here the doctor tries
feebly to challenge Reno's authority.
Figure 10.86 Kiss Them for Me: And as the Figure 10.87 At the end of one shot of
group breaks into conversing pairs, their mild 20,000 Leagues Under the Sea (1954), all look
recession is broken by a woman towing a man off, awaiting the entrance of Captain Nemo.
through along a slightly opposing diagonal.

Figure 10.88 20,000 Leagues Under the Sea:

Figure 10.82 Bad Day at Black Rock: Soon
Hector comes into frame center to bully the
old m a n as Coley leaves the window and
moves to the middle ground.
Figure 10.83 Bad Day at Black Rock: Reno
goes outside as the men rearrange themselves.
Throughout the film, the lobby windows
create another zone of space for us to notice.
After a cut to Nemo, we follow him stepping
into the frame and halting in the prime spot,
with faces on different levels surrounding
him. Conseil (Peter Lorre) obligingly turns
from us to favor Nemo.

stranger, Macreedy, while the town doctor feebly protests. When a character speaks
a crucial line, he tends to come forward or mask off others, moving and occupying a
spot with the precision of chessplay (Figures 10.81-10.83). Sometimes the onlookers
swap places silently on the fringes of the action, resettling the composition in ways
at once subtle and transparent. These small adjustments may rebalance the frame, or
clear space for new characters to come into the shot.
Scope proved very amenable to cramped choreography too, and it becomes the source
of comedy in Kiss Them for Me (1957). In the party scenes, Stanley Donen uses several
strategies to shift attention from one line of action to another. He constantly breaks up
his clothesline arrays by having people intrude from the sides or the rear, exploiting
what depth of field he can get (Figure 10.84). Donen also pivots the clothesline array
slightly into depth to allow for other sorts of interruptions (Figures 10.85-10.86), or
just lets partygoers in the foreground pass between the camera and the principals
to block our view momentarily. Similar shallow-space maneuvers occur in Richard
Fleischer's 20,000 Leagues Under the Sea (1954). Fleischer claims to have taken to
Scope "like a duck to water," because it encouraged lengthier takes. 87 At 8.5 seconds,
the average shot length of this film is reasonably short, but some shots are precisely
choreographed long takes—in a submarine, no less. Fleischer lines up his characters,
but never in obtrusive clothesline arrays: Usually a slope or slant will skew the line of
figures, or a foreground body will close the composition, incidentally reinforcing the
cramped quarters of Captain Nemo's Nautilus (Figures 10.87-10.88).
Such choreography in fairly shallow space isn't the only way depth could be used in
Scope. If a director wanted a deeper playing space and reasonably sharp focus across
that, he was obliged to set the foreground plane fairly far from the camera. This tactic
makes the foreground element relatively small within the vast screen. The result is
316 Poetics of Cinema CinemaScope 317

Figure 10.89 Distant depth reminiscent of Figure 10.90 A distant silhouette framed Figure 10.94 Beneath the 12-Mile Reef (1953): Figure 10.95 Beneath the 12-Mile Reef: The
the 1910s in Demetrius and the Gladiators. for saliency (Ride Lonesome, 1959). As the policemen thread their way through dancing couple drifts further back, and the
Compare Figure 10.9. the bar, a young Greek boldly dances with his jilted fisherman steps to frame center, turned
rival's girlfriend. from us.

Figure 10.91 'Lite Great Locomotive Chase
(1956): A busy foreground and partial views
through apertures, another strategy charac-
teristic of 1910s cinema as well.
Figure 10.92 Fuller combines modular com-
position, edge framing, and distant depth in Figure 10.96 Beneath the 12-Mile Reef: Our Figure 10.97 Beneath the 12-Mile Reef: The
Forty Guns (1957). view of the couple is blocked first by a cop, couple rush outside in the distance, but when
then by the rival's brawny back, but then we the fisherman rushes to pursue them down
see them moving toward the faraway door- the left aisle, the boy's father halts him in the
way as he reaches for his knife. foreground.

Figure movement in the foreground can be designed to block and reveal faraway
niches of action. In Beneath the 12-Mile Reef (1953), sponge fishermen in their local
bar are about to fight, but the arrival of the police makes them fake comradeship.
Figure 10.93 The Bridge on the River Kwai:
Saito's order is passed down the chain of . To take advantage of it, a young Greek dances out the door with another man's girl-
command, farther and farther into the dis- friend, and the drama develops in striking depth (Figures 10.94-10.97). Our vision
tant buildings. has to shift from the foreground to a small, out-of-focus background region and
then back to the foreground again, in about three seconds—and across several feet
of screen space, in the original theatrical setting. Putting this sort of demand on
perhaps the most striking invention—or perhaps we should say rediscovery—facili-
tated by Scope: a kind of return to the 1910s, when filmmakers exploited the rich the viewer helps energize the experience of long-shot views. King Richard and the

possibilities of midrange foregrounds and remarkably remote background planes.
(See Figures 1.1-1.2.) For example, in Demetrius and the Gladiators (1954), the devious
Messalina comes into the (already quite distant) foreground while her husband and
Demetrius study her from the terrace fairly far back (Figure 10.89).
But if the foreground plane is set quite far back, how to highlight relevant infor-
mation? The partitional tactic comes in handy here. A break in a ruined wall in Ride
Lonesome (1959) encloses the very distant woman while two cowboys watch her yearn-
ingly (Figure 10.90). Some directors seek a much more cluttered foreground and a
middle ground with many apertures (Figure 10.91). Samuel Fuller's Forty Guns (1957)
presents a face-off from inside a gun shop, with each gunslinger framed in a different
window (Figure 10.92). During the British officers' briefing on how to build a proper
bridge over the Kwai, Colonel Saito calls for tea, and his order is relayed among staff
members visible in the buildings far behind them (Figure 10.93).
Crusaders (1954) shows that as in the 1910s, even a small slot between two players can
be activated for dramatic purposes (Figures 10.98-10.102). The technique of wedging
story points into the crevices of a dense visual field, all but forgotten today, became
reinvigorated in the Scope era.
These are fairly brief moments of foreground-background interplay, but suspense
can be built through a nagging suggestion about what's going on in the distance. In
Bad Day at Black Rock, Macreedy visits the cafe for a bowl of chili, and Coley is deter-
mined to pick a fight. As he harasses Macreedy in the foreground, we are uneasily
aware that Coley's boss Reno is hovering around behind—sometimes at the pinball
machine on the left, sometimes hidden by Coley, and sometimes watching warily
from the edge of the doorway (Figures 10.103-10.104). Our vigilance about what
happens in the rear is eventually rewarded when Macreedy's judo flips hurl Coley
through the same doorway (Figure 10.105). Today's director would put the camera at
318
Figure 10.98 King Richard and the Crusaders
(1954): After an attempt on the king's life, his
subordinates are gathered around his sickbed.
Sir Kenneth of Huntington studies the nearly
fatal arrow and looks up and offscreen.
Figure 10.100 King Richard and the Crusad-
ers: Meeting Sir Kenneth's gaze, the conspira-
tors leave, passing behind the crowd in the
middle of the frame . . .
Figure 10.102 King Richard and the Crusad-
ers: When they turn and leave, the camera
pans with Kenneth as he follows them out,
revealing new layers of men outside.
Figure 10.104 Bad Day at Black Rock: As
Coley provokes Macreedy, they advance to
the middle ground, Reno now warily closer
to the door's edge.
P o e t i c s of C i n e m a
Figure 10.99 King Richard and the Crusad-
ers: A cut to a new angle shows Sir Giles, whom
we know is behind the assassination attempt.
Figure 10.101 . . . before stopping in a gap to
confront Sir Kenneth's suspicions.
Figure 10.103 Bad Day at Black Rock: Coley
douses Macreedy's food with ketchup, with
Reno barely visible over his shoulder.
Figure 10.105 Bad Day at Black Rock: In
a new composition, Macreedy has thrown
Coley out the door and confronts Reno, now
retreating to the corner.
CinemaScope 319
Figure 10.106 A tight Scope two shot, but in Figure 10.107 Pete Kelly's Blues: Pete and Ivy
the center there is some out-of-focus move- turn, creating a very unusual shot for Holly-
ment in the doorway behind (Pete Kelly's wood anamorphic, one more reminiscent of
Blues, 1955). the opening of Godard's Vivre sa vie (1962).
Figure 10.108 Pete Kelly's Blues: Pete Figure 10.109 Pete Kelly's Blues: The camera
advances, with the composition still not creeps forward, and Ivy moves slightly aside
revealing why. to reveal George the cop.
Figure 10.110 Pete Kelly's Blues: George
questions Pete, who refuses to cooperate. The
shot has peeled away a layer of space in a sus-
penseful gesture.
exactly t h e o p p o s i t e p o i n t i n s p a c e — l e t t i n g Coley b e flung out f r o m t h e d o o r w a y i n t o
t h e audience's f a c e — b u t Sturges' a r r a n g e m e n t activates o u r a w a r e n e s s m o r e keenly,
f o r c i n g u s to a t t e n d to a s m a l l p a r c e l of t h e s c r e e n s u r f a c e .
A m o r e o v e r t i n s t a n c e of "spatial s u s p e n s e " is exploited b y Jack W e b b i n Pete Kelly's
Blues (1955). Pete a n d h i s g i r l f r i e n d Ivy are o n t h e club's balcony, w h e r e t h e y d i s c u s s
m a r r i a g e in a p r o l o n g e d , p r o f i l e d t w o shot. T h e y kiss, b u t t h e n t h e y s e p a r a t e a b r u p t l y
(Figure 10.106). W h y ? T h e r e is a d i m , o u t - o f - f o c u s figure s h i f t i n g b e h i n d t h e m . They
t u r n f r o m u s in a s t a r t l i n g l y G o d a r d i a n p l a n a r shot (Figure 10.107) as Pete says,
" W h a t ' d y o u get, a b l e a c h e r seat?" Pete w a l k s s t r a i g h t to t h e rear, a n d t h e c a m e r a
m o v e s f o r w a r d , t h r o w i n g Ivy g r a d u a l l y o u t of f o c u s (Figure 10.108). She steps aside
to reveal t h e t e r m i n a l l y disheveled cop G e o r g e , w h o has b e e n p r e s s u r i n g Pete t o give
details o n t h e s h o o t i n g of h i s y o u n g d r u m m e r (Figure 10.109). A s Pete d r a w s u p to
G e o r g e , t h e c o p tells Pete t h a t a n o t h e r of h i s f r i e n d s m a y have b e e n killed by t h e m o b
(Figure 10.110). The f o r w a r d t r a c k i n g s h o t a n d Ivy's sidestep flout t h e Fox aesthetic,
b u t W e b b treats t h e e n t i r e scene as a n exercise in overt, p u r e l y v i s u a l teasing. 8 8
320 Poetics of Cinema CinemaScope 321

Figure 10.111 River of No Return
Matt, the gun, and its holster.
(1954): Figure 10.112 River of No Return: Later, in
a scene with the gambler Harry, the gun is
gone, but we're unlikely to notice.
Figure 10.115 California sunshine allows Figure 10.116 House of Bamboo (1955): The
Nicholas Ray to produce one of the most climax on the slowly spinning rooftop globe.
exciting compositions in early CinemaScope,
as the gang member's switchblade rises omi-
nously up from the lower frame edge (Rebel
Without a Cause).

Figure 10.113 River of No Return: Only Figure 10.114 . . . does Preminger reveal
when Matt reaches for it out of habi t . . . that Harry has taken it in case Matt won't
cooperate.
At the opposite extreme, the tactic of distant depth can put something impor-
tant plainly in the background and not call our attention to it. The best instance of
this I know occurs in Preminger's River of No Return (1954). Early in the film, it's
established that Matt keeps his rifle in a holster near the cabin door (Figure 10.111).
Later, while relaxing after dinner, the gambler Harry offers to pay Matt to accom-
pany him through Indian territory (Figure 10.112). Matt refuses. He rises, Harry
leaves the shot, and Matt reaches mechanically for his rifle (Figure 10.113). Abruptly
the rifle protrudes into the frame from the right foreground, in the hands of Harry
(Figure 10.114). Preminger has played fair with the audience by keeping the empty
holster prominently centered during the whole scene. Those who notice the rifle is
missing will experience some suspense, whereas those who do not notice will be
startled by Harry's gesture.
Just as Scope horizontal choreography draws upon skills cultivated in the early
sound era, the use of distant depth brings a 1910s technique up to date, with fresh and
engaging results. Assimilation of a new technology has led to not only recovery but
also discovery—or, rather, rediscovery.
T h e End of Screen Ratios?
Filmmakers, then, had several staging strategies available. They could treat the wide
screen as more—a horizontal expansion of the standard ratio, demanding to be filled
up in ways that modified clothesline staging or depth composition. They could treat
the new format as less, a slice of the old frame that blew up details and created quasi-
abstract compositions. Or they could investigate depth in a tactful way, activating
the remote reaches of the shot so that the spectator had to be alert for slight changes.
All these payoffs couldn't have been foreseen in the "theatrical" agenda promoted by
Fox. Yet Scope permitted, and to some degree encouraged, them. By the end of the
1950s, these impulses coexisted with several more traditional schemas, often combin-
ing within a single film. Some scenes relied upon the default schemas—little depth
of field, clotheslines or shallow staging, partitioning of the frame, and standard OTS
shots. Others presented edge framing or surprisingly big close-ups. And any scene
might rely on long takes or rapid cutting. For those few years during which Cinema-
Scope was in the ascendant, it was adjusted to the demands of the classical style, but in
the hands of imaginative filmmakers, it also yielded uniquely valuable results.
Some historians have argued that cinema benefited from originating as a silent
medium; creators were forced to develop distinctly pictorial storytelling traditions.
Similarly, CinemaScope's initial drawbacks spurred filmmakers to work around
them or find creative alternatives. The sheer variety of stylistic choices available in
the first anamorphic era is exhilarating. Hathaway and Preminger, Minnelli and
Fleischer, Cukor and Jack Webb all used Scope in ingenious and powerful ways. As
often happens, energetic pulp proves more exhilarating than high-minded kitsch.
The ballyhooed productions (The Robe, How to Marry a Millionaire, The King and I,
Anastasia, and The Inn of the Sixth Happiness) usually look pachydermous, whereas
more modest genre efforts like Violent Saturday (1955) and The Enemy Below bristle
with pictorial intelligence. Apart from some brilliant "big pictures" in Scope—A Star
Is Born, Rebel Without a Cause (Figure 10.115), and Seven Brides for Seven Brothers
(1954)—the most intriguing explorations of the format are to be found in Westerns,
adventure movies, thrillers, war pictures, and melodramas. (One rule holds firm:
If a Scope film runs longer than 100 minutes, it's likely to be visually uninspiring.)
Samuel Fuller's efforts in workaday genres illustrate what could be done with nearly
all the items on the menu. His first Scope film, Hell and High Water (1954), relies
on straightforward lateral playing in zones of a submarine set. In House of Bamboo
(1955), Fuller imaginatively uses the gridded layout of Japanese rooms to segregate
figures in layers, and by cutting to various angles, Fuller turns the climax on a rotat-
ing globe into an angular play of curves and ellipsoid shapes (Figure 10.116). Once
Fuller moves to black-and-white Scope, eccentricities rule. Forty Guns (1957) gives us
322 Poetics of Cinema
Figure 10.117 A Leone-like extreme close- Figure 10.118 Griff steps out into the back
up in Forty Guns. yard of a mortuary, with a rifle visible in the
window above: an outlandishly low and deep
angle from Forty Guns.
Figure 10.119 China Gate (1957): The squad
gathers around a wounded soldier.
flamboyant close-ups of eyeballs and steep low-angle shots with overpowering depth
(Figures 10.117-10.118). China Gate (1957) includes simply staged long takes, rhyth-
mically cut jungle skirmishes, posterlike abstraction, and looming wide-angle shots
that plunge the camera into the center of the action (Figure 10.119). Fullers films
alone suggest that the Scope era may have been the last period of genuine stylistic
variety in Hollywood.
Panavision offered greater flexibility, but it also allowed all films, whether anamor-
phic or flat, to be stylistically similar. In another curious historical throwback, the
result was to revive much earlier norms. Once widescreen close-ups, especially singles
of stars, became feasible once more, what director could resist? Directors of the 1960s
began cutting faster and dwelling on big faces—both technical options characteristic
of late silent films. Sergio Leone and others recovered the one-point-per-shot style of
Lubitsch or Harold Lloyd. An assistant to Leone recalls that close-ups were problem-
atic in the Techniscope format: "When you wanted a close-up to bring the audience s
attention to a face, an entire landscape opened up behind you: an entire town could
fit in, so you could forget putting the attention on your characters!" 89 As a result,
Leone and his cinematographer decided to shoot his gunslingers in extreme close-ups
from chin to hat brim, and this framing became his signature (Figure 10.120). Today
comparable shots can be found in most Hollywood films, blockbuster or indie, flat or
Scope (Figure 10.121).
At first, splashing a close-up across the gigantic screen made some critics recoil.
Dwight MacDonald remarked that in Preminger's The Cardinal (1963), "even Romy
Schneider's face is distractingly ugly when it has to fill that wide screen, while [John]
CinemaScope 323
Figure 10.120 Lee Van Cleef, the m a n with Figure 10.121 A similarly cropped shot from
Techniscope eyes, in For a Few Dollars More Confidence (2002).
(1965). (But compare Figure 10.30, which
anticipates this framing.)
Figure 10.123 The iwo shot becomes a stan-
dard Scope composition; compare Figures
10.28 and 10.70.
Figure 10.122 A two shot from The Thin Man
(1934).
Huston's looks like a relief map of the Dakota badlands." 90 But complaints couldn't
stop the march of a new style that featured big heads (isolated in singles), unfettered
camera movement, and fast cutting. Although the clothesline layout remained the
default for much Scope filming, and indeed widescreen shooting in general, into the
1960s, directors soon began to abandon the two shot and rely on recessive layouts,
singles, and OTS framings. More intricate staging options began to wane, and what
I've called "intensified continuity" began its long rule of Hollywood screens. 91
The triumph of widescreen, in both 1.85 and anamorphic forms, also allows us
to see how norms of earlier decades were not so much overthrown as adjusted. The
information contained in the old Academy ratio was preserved and, we might say,
reedited in the 1960s. In many 1920s and 1930s shots, the Scope proportions seem
to lie uncannily nestled within 1.33 clothesline compositions. Take a long shot or
plan-americain from a 1930s film and mask it to the 2.35 ratio. The result is often
a recognizable Scope framing (Figures 10.122-10.123). By contrast, we can't easily
recrop 1940s recessive compositions; too much information is jammed into the top
and bottom of the frame (Figures 10.15-10.22, above). The vertical elements would
have to be moved down and across in the wider format (Figure 10.124). But once
you're working with a wide frame and an aesthetic of close views, you're likely to
turn traditional medium shots into close-ups and traditional close-ups into extreme
close-ups. "What pulled me into shooting close-ups," Steven Spielberg admits, "was
when I shifted to the widescreen format." 92
324 Poetics of Cinema CinemaScope 325

Figure 10.124 Scope turns a vertical space—a
ground floor and the stair landing above—into
ahorizontal one (The Violent Men, 1955). Com-
pare the Little Foxes staircase (Figure 10.17).
Figure 10.125 The long lens already Figure 10.126 Filling the anamorphic frame Figure 10.127 Getting to know her stu-
abstracts space, but in THX 1138 (1971), in Three Kings (1999). dents, the hired tutor Anna spreads her vast
George Lucas designs his anamorphic shots dress across the CinemaScope screen (The
to flatten and compartmentalize his futuris- King and 1,1956).
tic locale further.

As with every innovation, the fact that anamorphic imagery is no longer a problem
Anamorphic filming never went away, and it enjoyed a resurgence in the 1990s that has its benefits as well as its costs. Panavision opened new possibilities for all-over
continues unabated. 93 Today any movie can be comfortably shot or released in 2.40:1. composition in the anamorphic ratio, that maximal use of the screen format that
Indeed, any image can be repackaged in any ratio. Filmmakers working with Scope Boris Kaufman valued. Many directors have taken advantage of it, assuming that
could be fairly confident that their images would be shown more or less as they wished. their densely composed images would be displayed in full (Figure 10.126). And in
Many of our examples from the 1930s to the 1960s are so precisely composed that fairness, I have to say that most directors today face problems no less pressing than
careless projection would ruin them. But today's directors must frame loosely, know- the Scope format. How can one develop computer-driven spectacle? Or make full use
ing that many shapes and sizes will be carved out of their images (megaplex projec- of digital sound? Or endow children's fantasy, high school comedy, teenage horror,
tion at anything from 2.4 to 1.85, full-frame TV at 4:3, widescreen TV at 16:9, and the and comic book superheroics with freshness, beauty, and intelligence? Most of these
peephole displays on handheld devices like cell phones). The success of Super-35mm, weren't on the agenda for filmmakers of the 1950s.
which slices a variety of ratios out of a single square picture, is an acknowledgment For the student of film poetics, though, the Scope era can be seen as giving a cluster
that at some basic level, compositional precision is just less important. of classical staging options one final run-through. A system deplored for its tech-
The acreage afforded by Scope challenged directors to fill it up, and some found nical rigidity became, however briefly, a museum of quite varying achievements in
mise-en-scene. Perhaps that's an underlying reason that cinephiles born before 1950
thrilling ways to shift bodies around the screen space. By the end of the 1960s,
(like me) find Scope movies so intriguing. Their images invite us into realms where
however, most directors had no interest in articulating a scene through staging.
people have bodies and move in real time, and the shape and size of the screen encour-
Cutting and camera movement were enough, aided of course by close-ups of gripping
aged sheer graphic gamesmanship as well. Who among today's filmmakers would risk
performances. There emerged a generation of talented directors who loved movies,
the nuttiness of spreading Deborah Kerr's crinoline across 60 feet (Figure 10.127)?
who could spin engaging yarns and elicit memorable performances, and who had
Screen proportions may persist, but styles can change.
an eye for anamorphic abstraction, often aided by the long lens (Figure 10.125). But
they scarcely knew how to move actors around a set.94 To this extent, the triumph of
Panavision contributed to the defeat of ensemble staging.
Put it another way: Artists struggling with problems of craft can be spurred to
innovate, but the widescreen format is no longer sensed as a problem. Ratios now
offer no resistance. Yes, you still have to fill the wide image, but technology allows
you simply to post a head shot. The zone of facial expressivity—eyebrows, eyes,
and mouth—fits rather nicely into the horizontal slit. Could a filmmaker today
orchestrate several bodies moving across that expanse without looking awkward or
old-fashioned? Also, of course, widescreen films will be seen on TV, either cropped or
letterboxed, and a tangle of bodies doesn't command the increasingly small screens
that viewers are learning to live with. Oddly, the severe constraints of CinemaScope
pushed directors toward ingenious exploration, but the versatility of Panavision has
fostered a lockstep style.
 II.

Who Blinked First?

According to one tradition, if you're a scholar you make progress by learning more
and more about less and less, until you know everything about nothing. I'm happy
to report that this essay fits firmly into that tradition. If it does make some progress
toward understanding how films work, it does so by focusing on some fairly minute
matters. Blink and you might miss them.
Not that the general problem is trivial. Despite decades of discussion of The Gaze
and "visuality," it seems to me that we know very little about eye behavior in cinema.
How do film characters gaze or glance or peer or simply look at each other? What
patterns of looking can we find, and what functions can we assign them? How do
these patterns shape performance, and how might they accord with broader stylistic
strategies employed by filmmakers? How do we as viewers respond to these patterns?
Such issues are important, because eye behavior is central to understanding human
action, both onscreen and offscreen. As an effort toward answering these questions,
I want to consider some aspects of eye behavior in mainstream narrative films. Before
that, though, we need to consider how looking works in everyday situations.

T h e Tightrope
Although novels and poems portray eyes as fierce or dreamy, by themselves eyes
can express very little. As social signals, they normally function as part of the face.
Features, particularly the eyebrows and the mouth, work together with the eyes to
create what Paul Ekman has called a "facial action" system.1 Anger is prototypically
signaled less by the eyes than by the knitted brows, the tense mouth, and the set of the

327
328 Poet ics of Cinema
jaw, perhaps aided by a flushed complexion or a loud tone of voice. A Fantomas-style
cowl shows in a disquieting way how the eyes alone are rather uncommunicative.
Nonetheless, the eyes do have some locally significant features. The color of the iris
is distinctive, and the degree to which the eyes are closed is informative. (The droop-
ing eyelids of my students don't express quite the sensuality seen in portraits of Italian
Renaissance ladies.) The size of the pupil was presumably an important cue in our
evolutionary heritage, for a dilated pupil can be a sexual signal. 2 Particularly signifi-
cant is the direction of a person's gaze, a cue to which we (and perhaps other primates)
are highly sensitive. 3 When a person is looking off at something, this "deictic gaze"
triggers an interest from other parties, who tend to follow the direction of the look.
This behavior is apparent in babies' responses to their mother's glance. 4 The deictic
gaze, Noel Carroll points out, is a crucial cue in point-of-view editing. 5
Eye direction is not, however, a snapshot affair; our glance is often shifting,
and in quite patterned ways. A natural place to study longer-term eye behavior is
in conversations, both in real life and on film. To keep things simple at the start,
I assume a two-person dialogue.
Research in interpersonal communication suggests that in Western societies, talk
between two parties displays patterns of looking and looking away. These patterns are
regulated by turn taking, as the conversants switch the roles of speaker and listener.
Most commonly, the speaker looks away from the listener more frequently than the
listener looks away from the speaker. Perhaps surprisingly, the two parties seldom
share a look for very long. It appears that stretches of mutual gaze, with eyes locked,
are infrequent and brief. Michael Argyle found, with two people conversing, the
listener typically gazes at the speaker 75% of the time, the speaker gazes at the listener
40% of the time, and the two make eye contact 30% of the time. Argyle also found
that both people's eye directions changed often, with the typical one-sided glance
lasting only 3.0 seconds and the mutual gaze a mere 1.5 seconds. 6 Other researchers
indicate that eye contact tends to occur when partners switch speaker-listener roles.7
In sum, shared looks alternate constantly with "gaze avoidance" or other eye move-
ments, such as looking upward to recall something or glancing to the side to monitor
the environment.
What creates these patterns of interaction? The usual explanation is that the speaker
is expending more cognitive resources and needs to concentrate on formulating
speech, but she or he still must return at intervals to check the listener's uptake. The
listener, on the other hand, concentrates not only on what the speaker says but also
on other cues that carry meaning, such as the speaker's expression, hand gestures,
shrugs, and the like. So naturally the listener tends to pay more attention to the
stream of information. In addition, to look away too often might suggest boredom,
inattentiveness, or disagreement.
Imagine by contrast a situation displaying more prolonged staring between parties,
with sustained mutual eye contact. This is rare in ordinary life because, depending
on the context, the mutual stare typically signals either aggression or deep affinity.
We have, on the one hand, Travis Bickle's "Are you looking at me?" and, on the other,
the rapture of lovers lost in each other's eyes. Here's another reason why in ordinary
W h o Blinked First? 329
life people don't look at each other more often: Locking onto someone's eyes too
frequently can send a signal that could be interpreted as hostility or erotic interest. 8
. So much for everyday conversation. What do we find when we turn to film? A few
surprises, I think. Take a scene from L.A. Confidential (1997). (For reasons that will
become clear shortly, I've picked one in which the parties are in basic agreement,
displaying neither hostility nor affection.) Sergeant Edmond Exley has been sum-
moned by his superiors, who ask him to testify that policemen have beaten prisoners.
The officials want to make a public relations effort to clean up the LAPD image, and
they need officers willing to snitch on their colleagues. Exley immediately agrees, in
exchange for a promotion to lieutenant, and he offers suggestions on how they can
force another cop, Jack Vincennes, to testify as well.
The dialogue portion of the scene lasts about 2 minutes and 4 seconds. Exley is
standing at attention before a desk, with his superiors seated around it. The scene is
broken up by editing that alternates medium shots of Exley with group shots and indi-
vidual shots of his superiors. The officials take turns talking with him, occasionally
talking with each other, while he addresses himself to the police commissioner, the
most powerful man in the room. In the course of the scene, individuals look intently
at each other, either when they are speaking or when they are listening.
If we time the intervals in which any man listening is not looking at the speaker
and any man speaking is not looking at his addressee, they add up to very little—no
more than 10 seconds. And during many of these intervals, when one man is not
looking at the speaker or his own listener, he is exchanging glances with another
listener. For example, the officials glance at one another when they realize that Exley
has devised a plan for his benefit.
This sequence appears to invert the default case. The L.A. Confidential scene pres-
ents a world in which a speaker looks far more frequently and fixedly at a listener,
and the listener concentrates on the speaker even more intently, than in the normal
case. Why this result? After examining several scenes like this, I'd argue that the
standard cinematic case indeed alters the ordinary scenario. Movie characters rarely
look away from one another, and they often make mutual eye contact. Indeed, they
often seem to be staring into each other's eyes. Yet the stare doesn't necessarily signal
either hostility or love.
By contrast, a movie scene that presents something like the normal real-life case
risks sending the wrong signals. In a film conversation, when a character avoids look-
ing back at her or his partner, gaze avoidance takes on an expressive tint. A viewer
might construe it as evasiveness, furtiveness, lack of interest, or the like (the very
attributions made in real life when someone looks from a speaker too long or too
frequently). In films gaze avoidance, far from being a normal part of the rhythm
of conversational interaction, is rare and highly informative about the character's
psychological state.9
We can see this condition in an early scene in Chinatown (1974). Detective Jim
Gittes is visited by a woman claiming to be the wife of Hollis Mulwray, an official in the
Los Angeles Power and Light Commission. She sits at his associate's desk and explains
that she suspects that her husband is having an affair. She occasionally looks away from
330 Poetics of Cinema
Gittes, and he frequently looks away from her, glancing at his colleagues or frowning at
the floor. By my count, a total of 45 seconds of the 118-second scene consists of gazes in
which either the speaker or the listener doesn't look at the other partner.
If we examine these moments, however, we find that the deflected glance is psy-
chologically revealing. At times, Gittes shares a glance with his assistants, as did the
officials in the L.A. Confidential scene. More important, Mrs. Mulwray looks away
when she is flustered, as when she voices her suspicion that her husband is seeing
another woman. Responding to her eye behavior, Gittes lowers his eyes and looks to
the ceiling before returning her gaze. We know from the previous scene that he s a
cynic, and so we tend to read his exaggerated gravity as a sign that (a) he is pretend-
ing to be shocked by a man's peccadillo, and (b) he's not surprised that Mulwray has
strayed from such an unattractive wife. As the conversation goes on, it is clear that
Gittes is reluctant to take such a banal case, and this is expressed in fairly frequent
glances to the side and to the floor, as if he's searching for a way out. In later scenes,
though, once Gittes gets caught up in the investigation and starts to believe he is
unraveling a scandal, his gaze at others becomes much more unwavering.
Certainly the contrast between the real case and the filmic case is revealing, but
if we look a little further, the inversion isn't perfect. For one thing, aggression and
affinity—the feelings that promote prolonged looking on the part of the speaker in
normal life—are common bases of dramatic action in movies. So one could argue
that many scenes in fact conform to the rule that mutual looking depends on these
emotional circumstances. In fact, I had to search a bit to locate fairly neutral scenes
like the ones in L.A. Confidential and Chinatown, for most scenes I found had at least
the hint of mutual hostility or mutual attraction. This may suggest that these two
areas of emotion are at the emotional center of most scenes in mainstream movies,
whereas neutral encounters are fairly uncommon. At the very least, scenes of con-
frontation or enthrallment are far more frequent in fiction films than in life, so a
greater degree of shared looking is to be expected.
Secondly, I'd argue that the cinematic default isn't a true inversion of the normal case
because the prototypical cinematic conversation takes the characters' basic attitude to
be mutual attentiveness to the situation. The norm is that the speaker is paying strict
attention to the listener's response because (unlike most conversations in real life)
something of consequence hangs upon it. In effect, the eye behavior characteristic of
the listeners role in ordinary interaction is mapped onto the speakers role as well. The
fiction film presents a world in which speakers are constantly monitoring the effects
of their self-presentation on listeners, searching for the slightest reactions. It would
not be too great an exaggeration to say that one sign of fictional drama is people look-
10
ing intently at one another.
"What we do," says Michael Caine in his instructive tape on acting technique, "we
actors who are in the movie, is: We hang onto each other's eyes. That's the most impor-
tant thing." A more recent manual is just as explicit: "The eye-line is a tightrope that
keeps an actor aloft."11 Because the characters pay constant attention to one another,
we're encouraged to pay attention too. The drama, after all, is about them. A conversa-
tion on film omits the fluctuating eyelines we'd find in life in order to highlight the
"Who Blinked First? 331
ongoing mutuality of interest—that is, the dramatic issue. Correspondingly, when
an actor looks away, the act has far more dramatic import than it would have in life.
Gur acting manual goes on to remark, "Once eye contact is established between two
actors, the moment when it is broken becomes very significant Breaking eye
contact always makes a statement." 12
This reliance on mutual gaze to rivet us to the action nicely supports Ed Tan's
theory that the ground of our emotional engagement with films is the attitude of
interest. 13 But is there a way to detect mutual interest among the characters more
precisely? We might think of cases when the mutual interest isn't present. If a listener
is oblivious to what a speaker is saying, as in the case of the TV-watching husband or
a bored theater audience, the listener is usually shown looking away from the speaker.
Still, I think there's another way to chart mutual interest in conversation scenes, one
that brings out some unexplored aspects of acting technique as well.
T h e Strength of the Stare
In the 1970s I became fascinated with watching Judy Garland films, not just because
I found her a captivating performer but also because I noticed that she seemed almost
never to blink. At the time, I put this down to her having been fed pharmaceuticals
as a child star. Years later I began to notice that she wasn't the only nonblinker. Most
actors seldom blinked. The puzzle didn't exactly rocket to the top of my research
agenda, but it continued to intrigue me.
Only after reading a pop biography of Michael Caine did I get a hunch about the
process. Caine claims that as a youth, he read Pudovkin's treatise on film acting and
learned that he should never blink. 14 Caine then practiced staring without blinking
until he could do so for minutes on end. 15 In his tape on acting, produced many years
later, he explains why.
If I keep blinking, it weakens me. But if I'm talking to you and I don't blink
[stares at camera] and I keep on going and I don't blink [continues to stare at
camera], you start to listen to what I'm saying. And it makes me a very strong
person, as opposed to someone who is sitting there going [blinks several times],
which is someone who's completely flustered.
Thespian lore appears to hold that strength, menace, or some other intense quality
is best conveyed by the rocklike look.16 Anthony Hopkins maintains that in playing
Hannibal Lecter, he strove never to blink: "If you don't blink, you can keep the
audience mesmerized." 17 Likewise, Samuel L. Jackson credits his success at playing
disturbing roles to winning the no-blinking game as a child. "I have this habit of
being able to stare unblinkingly at you until you break."18
Now, playing a determined, menacing role might seem to call for unblinking eyes,
and there's no doubt that Caine, Hopkins, and Jackson excel in such parts. But I think
that the absence of blinking is far more widespread than these reports indicate. Judy
Garland isn't very threatening. Moreover, we have evidence that filmmakers want to
control blinking behavior whenever it occurs. According to a friend of mine, when
332 Poetics of Cinema W h o Blinked First? 333

he was directing a film, his editor felt he had to make a certain cut in order to elimi- seems to have become a convention of fictional narrative generally. No novel reports a
nate an actor's blink. 19
George Lucas, pointing out the postproduction advantages of character's every blink, but when the narration mentions one, it's important.
digital video, employs a telling example: "If someone blinks right where I'm making
"I don't understand," Meyer said, puzzled. "Did she look thirty, or did she ...?"
the cut and I can't make the cut because it doesn't work with the blink, I just get rid
of the blink." 20 It seems likely, then, that the suppression of blinking in films occurs "Well, how would I know how she looked, man?"
fairly often and fulfills purposes beyond the enhancement of Michael Caine's career.
Meyer blinked. "What do you mean?" he said.
In ordinary life blinks lubricate our eyes, and when relaxing or conversing, we blink
between 10 and 25 times per minute. A blink lasts about one third of a second. Interest- "She was wearing a mask," Bones said.
ingly, in conversation, playing the role of speaker tends to raise the blink rate, whereas
"A mask?" Meyer said, and blinked again. "At a wedding?"
playing the role of listener lowers it.21 Once more, in film, aspects of the listener's role
are transferred to the speaker, and the speaker becomes less of a blinker, just as his or "Oh," Bones said. "Yeah." He blinked, too. "Maybe I got something mixed up,
her gaze wanders much less. And Caine is right to worry about looking flustered. We huh?" he said. 28
blink faster when we are excited or in other states of arousal, such as feeling anxious,
Our L.A. Confidential policemen blink very little. They blink most while shifting
addressing a large crowd, or telling lies. One psychological researcher, Joseph Tecce,
the angle of their gaze, and occasionally they blink to register a reaction to what is
has specialized in studying U.S. presidential candidates' eyeblinks during televised
said; Exley, for example, keeps his face severely composed, but he blinks in response
debates, on the presumption that stress and anxiety increase the blink rate. In 1996,
to chastisement from his captain. My Chinatown example, where only Gittes' and
Robert Dole set a recent record by blinking an average of 147 times per minute. 22 Do
the purported Mrs. Mulwray's faces are discernible, consumes 118 seconds, about
observers pick up on such inadvertent signals? Another study found that people rated
the same length as the L.A. Confidential scene. Here I count 53 blinks, making the
a frequently blinking person to be more nervous and less intelligent than one who
average (about 13 blinks per person per minute) a little closer to the real-life norm.
blinks rarely.23
Again, though, the blinks tend to be dramatically meaningful. Mrs. Mulwray blinks
When do we not blink? It seems that absorption in a visual task creates longer
when she talks of her husband's infidelity; as with Exley, her facial expression appears
intervals between blinks. Several Japanese researchers have found that blink rates slow
unconcerned, but her blinking shows her to be agitated. Similarly, Gittes' efforts to
down when people are engrossed in television watching. 24 In movie dialogue scenes,
avoid taking the case are registered by several blinks that convey not only hesitation
the absence of blinking is a very direct way to convey each partner's attentiveness
and avoidance but also an elaborate effort to be polite. At one point, as Gittes strains
and mutual interest. If my L.A. Confidential scene were an everyday conversation, we
to charm Mrs. Mulwray, Jack Nicholson makes Gittes positively flutter his eyelashes.
should find around 160 blinks in total (20 blinks a minute x 2 minutes x 4 men). Yet
, Thus the demands of film acting build upon normal patterns of blinking but
I can find only 34 blinks shown onscreen. Each man seldom blinks when he is speak-
functionalize them: Actors strive to make this natural, necessary act a tool of their
ing, especially when he is the subject of a "single," or a shot framing only him. Actors
craft. One study has indicated that in ordinary life blinks occur with greater fre-
whom I've asked seldom report that they decide to avoid blinking, but they do invoke
quency at the start of an utterance or word, 29 but actors tend to blink when they want
the common actor's advice that credible performance involves watching and listening
to the other actors. When actors concentrate on what the other players are doing, fewer to punctuate an utterance, often after a meaningful phrase. Frequent blinking, as
we would expect, is a tool of expressive performance, with implications shaped by
blinks may become a by-product.
Yet blinking isn't outlawed altogether. A few blinks make our characters human. context. Just as a change of eye direction will not be read onscreen as the gaze drift
Even the "limited animation" of Japanese anime needs to present eye movement and characteristic of normal conversation, a series of blinks is likely to be taken not as
blinks. Chuck Jones points out that Bugs Bunny and Daffy Duck "blink when some- natural lubrication of the eye but rather as betraying a particular emotional state—all
25
one's talki ng. That's big stuff. To establish that the character is alive." Consequently, those variants of Caine's "weakness" we can call apprehensiveness, anxiety, remorse,
the eyes that never blink may be thought inhuman. Tecce commented that during fatigue, or sadness (blinking back tears). In The Guns of Navarone, when the David
the 1992 presidential debates, Ross Perot came across as having a "reptilian stare." 26 Niven character confronts a member of the demolition team with damning evidence
In shooting SlmOne (aka Simone, 2002), the director told the actress playing a that she is a spy, he holds the screen with unblinking force. The other team members
computer-generated performer not to blink; when she inevitably did, the blink was watch her warily, and their vigilance is expressed through a pronounced absence of
digitally removed. 27 blinks. By contrast, the suspected spy blinks to indicate that her facade is cracking.
An occasional blink humanizes, but the trick is to make it significant. In everyday Yet in the next scene, when Niven is waiting for the moment to launch the raid on the
life, a blink is often read as a signal of surprise, concern, or bafflement. This tendency guns, he blinks mightily. He's afraid.
334 Poetics of Cinema
A little observation of screen performances shows that there are blinking-related
tricks of the acting trade. For example, it's easier to keep from blinking if you're not
fixating on something (so all praise to Caine, Hopkins, and Jackson for managing an
unswerving stare). Film players have discovered that even slight changes of eye direc-
tion can help hold back a blink, as when two actors looking at one another seem to
search each other's faces. Actors also find ways to conceal their blinks. A performer
can sneak a blink by turning the head (the L.A. Confidential tactic) or by lowering the
eyes, as if in modesty or deep thought (one of Nicholson's tactics in the Chinatown
scene). Some actors squint, in the process making themselves look more adorable
(e.g., Renee Zellweger) or implacable (Charles Bronson).
Do various acting styles find alternatives to the patterns I've been pointing out?
Robert Bresson slows down the blink so that it becomes a dramatic event in itself; his
players lower their eyelids with such deliberation that they seem to be shutting down
an electrical circuit. Yet such variants seem quite rare. We might also expect cultural
variations in the eyeblink repertory, but my preliminary searches don't reveal any. In
films from various countries and periods, it seems that the actors avoid blinking, and
they watch each other as fixedly as Hollywood performers do. This uniformity seems
to occur despite a culture's rules about eye behavior. 30 Japanese etiquette discourages
people from looking fixedly at their conversational partner, but in films they do so
frequently, and they seldom blink. Ozu Yasujiro's films are remarkable repositories of
staring, nonblinking conversations. 31
Did actors in other cultures learn from U.S. films to restrain their blinking? Or
did they independently rework some transcultural norms of eye behavior? It would
be worth studying different films from different traditions and periods to plot the
ways in which actors conceal or manifest the simple act of blinking. Similarly, we
might consider how various shooting and staging techniques have made blinks more
or less salient. Today's insistence on singles, particularly close-ups, would seem to
demand actors who can hold back blinks. 32 But now you see why I began by saying
that I'm approaching the academic ideal of knowing more and more about less and
less. Fortunately, though, this exercise harbors some more general implications.
Streamlined Behavior
How may we best understand cinematic conventions? They are often built out of
ordinary-life behaviors, but not just any behaviors. The ones favored seem to put
people's social intelligence on display. One important function of art may well be
the opportunity it affords for us to test, refine, and expand our knowledge of why
others do what they do. To this end, for example, faces in films become of partic-
ular interest because they're informative on many levels—they provide informa-
tion about attention and interest, as well as mental and emotional states. From an
evolutionary standpoint, our interest in others' inner states can be seen as a problem
in "Machiavellian intelligence." We know we can fool others, so we're on our guard
against being fooled. It's important for us to detect deceivers, and we've evolved many
mechanisms to help us read minds. 33 So films present cowards or liars as blinking
W h o Blinked First? 335
characters who won't look others in the eye, thereby reinforcing and sometimes fine-
tuning our sensitivities.
. Filmmakers have forged conventions that piggyback on the most salient cues for
mental states that we encounter every day. The purpose is clear-cut. If you want to
tell stories on the screen, you'll normally seek to keep the viewer fastened on the
flow of information, especially in character encounters. As practical psychologists,
filmmakers know intuitively that the shared gaze and the absence of blinking are
two well-defined social signals for mutual attentiveness. They signal to an outside
party, the audience, that the characters are participating in a significant exchange of
story information. We sense that the situation is dramatic partly because characters'
eye behaviors indicate deep engagement. These signals of mutual engagement also
hold the viewer's attention from moment to moment, an important consideration in
a time-bound medium like film.
If cinema, like other artistic media, often models social intelligence, it doesn't
simply copy the relevant behaviors. Like all representations, it simplifies what is rep-
resented according to purpose and relevance. Accordingly, many devices of film style
rework social acts for clarity and expressive effect. The second essay in this volume
suggested that one function of shot/reverse-shot cutting is to accentuate the typical
patterns of conversational uptake and turn taking. Similarly, acting already stylizes
normal human interactions; it amplifies behavior for our quick understanding. The
actor's simplified enactment of psychological states is further amplified by framing,
lighting, color design, cutting, and other cinematic techniques. A close-up can make
eye direction unambiguous, and cutting can delete blinks.
Such considerations lead us, I think, to think of cinematic conventions from a
theoretical standpoint of moderate constructivism, something akin to what Torben
Grodal called "ecological conventionalism." 34 Conventions are constructed, yes; but
they're constructed out of preexisting regularities of human action. Some of those
regularities are social, and some aren't limited to a single time or place. Historically,
filmmakers have taken as material ordinary social behaviors, often of sorts that are
readable across many cultures. But the filmmakers have reworked those behaviors,
usually for the sake of greater clarity and force. Cinematic style often streamlines ordi-
nary human activity, smoothing the rough edges, and reweighting its features in order
to create representations that are densely informative and emotionally arousing.
If something like this is right, then gaze and blinking turn out not to be utterly
trivial things to study. The same goes for facial expressions, which are starting to get
the attention they deserve. 35 So let's move on to hands in cinema, and mouth move-
ments and even eyebrows—a realm in which, I suspect, John Wayne will prove to be
just as resourceful as Judy Garland.
 12.

Visual Style in Japanese

Cinema, 1925—1945

During the 1950s and 1960s, most critics believed that Japanese film's golden age
lay in the postwar period, the years of masterpieces by Kurosawa Akira, Mizoguchi
Kenji, Ichikawa Kon, and others. During the 1970s, the spotlight began to shift. As
more works became available, and as critics began to turn to the neglected early films
of Mizoguchi and Ozu Yasujiro, the 1920s and 1930s came to seem the most exciting
and innovative era.
It now seems clear that Japanese cinema of this period was one of the richest of
all filmmaking traditions we have known. It deserves to be as celebrated as Soviet
montage cinema and French poetic realism. Even its minor genres are endlessly
intriguing. I would argue that Japan's swordplay films from the mid-1920s to 1940
were not only the best action movies in the world at the time; they also merit the sort
of study we lavish on Sergei Eisenstein's work. That this period of Japanese cinema
isn't sufficiently appreciated (even in Japan itself) is due largely to the inaccessibility
of the films and the absence of any school or movement, let alone an avant-garde, that
can be easily labeled.
The ignoring of this cinema also owes something to snobbish reluctance to rec-
ognize the authentic energies of mass-market filmmaking. Japan's great tradition
emerged, unembarrassed, from a thoroughly mercantile studio system. Firms turned
out a total of several hundred films annually, perhaps as many as 800 in some years.
Even allowing for untrustworthy statistics, it's likely that Japan was about as prolific as

337
 Poetics of Cinema
338
America, which produced 400-650 titles a year from the late 1920s to the 1940s. This
large-scale production brought with it all the compromises that Western commenta-
tors bemoan about mass-market filmmaking. Directors had to make films in well-
recognized genres with stars under contract. They had to adapt trashy fiction, submit to
censorship, produce sequels and imitations (with films including Charlie Chaplin and
Harold Lloyd imitators), and remake local and imported hits (including King Kong).
Films promoted hit songs or popular magazines, they launched toys and clothing fads,
and they shamelessly plugged consumer brands through product placement.
The first lesson to be learned from our Japanese excursion, then, is that film art
isn't inevitably corrupted by large-scale commerce. The second is that for a poetics
of cinema, any tradition this powerful demands attention. The power is chiefly that
of quality, no doubt; filmmaker for filmmaker, I'd say Japan is matched only by
Hollywood cinema of the same years. But interwar Japanese films, even minor ones,
are seductive in a unique way. This cinema can be unashamedly pure in its emotional
expression, subtle in its atmospheres, and unpredictable in its storytelling. (The Bob
Hope-Bing Crosby Road comedies have nothing on the mind-bending antics of the
vaudevillian Enoken in a film like Heirokus Dream Story, 1943.) These films can blend
humor, melancholy, and a vivid sense of lived reality in ways that we associate with
Jean Renoir or the neorealists. At the same time they offer formal ingenuity of a sort
that we associate with the Western avant-garde. This is, in short, everything we want
cinema to be.
A single essay, even one resulting from 30 years' study, can hope only to suggest
some ways in which a poetics of cinema can track a central appeal of this cinema: its
diverse and creative uses of film style. There is much, much more to be said than I can
say here, so I try to concentrate on questions. In what respects does Japanese film style
in the period 1925-1940 differ from the dominant norms of Western, particularly
Hollywood, cinema? How might the influence of traditional culture of earlier centu-
ries explain such differences? And what impact did contemporary social conditions,
particularly the China and Pacific hostilities, have on the development of film style? 1 longer than it did in the West. In Japan, this meant freeing the film from the need for
A Classical Cinema?
Scarcely any films survive from the first 25 years of Japanese filmmaking.2 It seems
likely that during that period, most staged films were adaptations of scenes from
Kabuki plays, presented in self-consciously theatrical performances in painted
flat sets or outdoor locales. This genre of recorded performance, calling upon the
audience's existing knowledge of the traditional stories, persists to the end of the
1910s. The acting was likely very frontal, but we have some instances displaying
depth staging as well (Figures 12.1-12.2).3 Most surviving films make little use of
cutting within the scene. (During production, actors often simply froze in position
while the camera was reloaded.) 4
Like early Western cinema, Japanese films mixed media. Before World War I,
most films were parts of a larger performance. Films were taken into stage acts, song
recitals, and that blend of live theater and film called rensageki ("chain drama"). Most
Visual Style in Japanese Cinema, 1925—1945 339
Figure 12.1 Chushingura (1913-1917): The Figure 12.2 Chushingura: Other shots are
classic play staged planimetrically, with rendered along diagonals reminiscent of
pockets of space on the top and sides. European cinema.
often, films were shown with an accompanying commentator, the benshi, as in tradi-
tional vocal theater like Noh and bunraku (puppet theater).
Japanese filmmakers had access to Western models. European and American
films appeared from the moment that the moving picture was introduced. The
establishment of the Nikkatsu combine in 1912 and a new burst of American imports
two years later marked the beginning of awareness of international stylistic devel-
opments—chiefly, the proto-continuity style bred in North America and Western
Europe. And in the years just before the 1923 Kanto earthquake, reformers were
pressing for modernization. 5 One 1911 writer sounds a theme familiar to Western
intellectuals impatient with "theatrical" cinema. "Watching a Japanese film is like
watching a slow dirge. It willfully ignores the need to keep the screen in perpetual
motion. . . . All this is because the films are slavish recordings of stage plays. What
could be more boring?" 6
Films came to be more self-sufficient in their visual storytelling, but the change took
benshi commentary by adding expository titles and, in the early 1920s, dialogue titles.
The Tokyo quake accelerated the modernization of the Japanese cinema. Production
houses had to rebuild, and directors were prepared to rethink their storytelling
methods. The contemporary-life drama (gendai-geki) absorbed Western influences,
largely from Hollywood comedy and drama. In Kyoto, Shozo Makino launched a
series of swordplay films that used the rapid cutting and close framings characteristic
of American cinema. 7
By 1925 at the latest, the norms of staging and cutting in Japanese cinema were
very close to those of mainstream Western filmmaking.8 Murata's Minoru Souls on
the Road (Rojo no reikon; 1921), the principal example of the "modernizing" film,
looks a bit like an American film of the late 1910s. It contains over 800 shots and
utilizes virtually every device of Western editing technique, including crosscutting.
Winter Camellias (Kantsubaki; 1921) uses longer takes but still averages over a dozen
shots per scene. This feature, like the swordplay films Shibukawa Bangoro (1922),
Serpent (Orochi; 1925), and The Wandering Gambler (Horo zanmai; 1928), adheres
340 Poetics of Cinema Visual Style in Japanese Cinema, 1925—1945 341

Figure 12.3 In an early scene of Love and Figure 12.4 . . . a cut-in to Namiji (played Figure 12.7 Yoshi Castle (1928): A sword- Figure 12.8 Yoshi Castle.
Sacrifice (1922), the hero Teruo meets Namiji. by an oyama, a male actor impersonating a fight, filmed in axial shot/reverse shot.
The establishing shot is followed b y . . . woman)...

W *
I 1 «t

Figure 12.6 Love and Sacrifice: A later shot
Figure 12.5 . . . and then a reverse shot of
of Namiji in the conversation is even closer
Teruo.
than Figure 12.4, emphasizing her reaction.
Elsewhere in the film, dialogue scenes are pre-
sented in lengthy over-the-shoulder shots.
Figure 12.9 In the opening scene of A New
Family (1939), a wife looks up at her husband,
from right to left . . .
•
Figure 12.10 . . . and he returns her look,
from right to left. Hollywood practice would
consider this a disorienting cut; it crosses
the imaginary axis running between the two
characters.

to continuity principles. Winter Camellias also contains smooth matches on action,
diagonal compositions, and careful staging in depth through a doorway. Love and
Sacrifice (1922) looks a little old-fashioned because it features a female imperson-
ator (oyama) and relies on the benshi to furnish all dialogue, but it presents familiar
patterns of decoupage (Figures 12.3-12.6). By the time of Yoshi Castle (1928), we find
straightforward mastery of over-the-shoulder shot/reverse shots, matches on action,
and point-of-view cutting (Figures 12.7-12.8).
Thereafter, decoupage standards of this sort clearly remained in force. Dozens of
films, from Kaigari Ippei (1930) to Three Men From the North (1945), are virtually
indistinguishable in this respect from ordinary films from any European or North
American country. Indeed, many gendai-geki centering on urban romance or success
in show business are highly reminiscent of Hollywood program pictures.
In one respect, however, editing habits of the period seem to run counter to those of
the West. Directors violated the 180-degree axis of action somewhat more frequently
than did their counterparts in the United States. As a result, occasionally eyelines
or frame entrances and exits are inconsistent (Figures 12.9-12.10). What should we
make of these apparently deliberate violations of continuity?
First, these deviations are far outnumbered by completely conventional uses.
Many films obey Hollywood continuity completely. Any deviations usually occur
only in certain scenes, with the rest of the film obeying the 180-degree system.
Furthermore, mismatched shots typically occur in an unproblematic context. Often
the depicted space is already well established. In the example from A New Family
(1939; Figures 12.9-12.10), a long-shot framing has already indicated the placement of
the characters. Within the shots, the fact that the standing character is looking down-
ward and the seated one is looking up should imply that the characters are looking at
one another. Sometimes too, as in Hollywood, the dramatic situation may motivate
the breaking of the 180-degree line, as when the aberrant shot establishes an area (a
new room, the outdoors) before the character enters.
Noel Burch has argued that such nontraditional editing works to undermine the
spatial orientation that mainstream cinema provides. 9 But our orientation can't be
wrecked at a stroke, by a single technical choice. Hollywood's continuity system is
342 Poetics of Cinema Visual Style in Japanese Cinema, 1925—1945 343
plainly a redundant one. It relies on many mutually reinforcing stylistic cues, as well
as on the viewer's knowledge of real space and human behavior. So subtracting a cue
or two won't necessarily upset things. European filmmakers of the 1920s were a little
cavalier about shot/reverse shot too; it seems that American filmmakers cared more
than most about getting eyelines and reverse angles completely consistent. Put it this
way: Japanese and European filmmakers occasionally reduced Hollywood's degree of
stylistic redundancy in staging and cutting. The presence of other factors that over-
ride these little bumps would explain why most viewers, east or west, simply don't
notice these deviations from Hollywood continuity.
There is plenty of evidence, then, that Japanese films of the 1920s and 1930s adhere
Figure 12.11 Instead of cutting f r o m one Figure 12.12 . . . to a gambler attacking a
to baseline norms of classical staging and continuity cutting. This conclusion obliges
swordsman to a reverse angle of his ninja woman, seen from an angle that answers
us to question any claims that these films are sui generis, radically other to the antagonist, The Mysterious Edogawa Ranzan to the one of Figure 12.11. This is followed
Western mainstream. Yet employment of norms doesn't entail lockstep conformity. (1937) cuts . . . by...
Haydn, Mozart, and Beethoven adhere to shared norms of Viennese classicism
(indeed, they created them), but their compositions sound very different. John Ford,
Otto Preminger, and Anthony Mann adhered to the norms of classical Hollywood
filmmaking, but their works instantiate those norms in markedly different ways. A
group style includes not only a repertoire of recurrent devices but also functional
principles and a range of choices operating on different dimensions.
Despite the adherence to premises of Hollywood visual storytelling, Japanese
films present perhaps the widest range of directorial styles of any national cinema
of the period. The ingenious variations that these filmmakers developed on basic
premises of continuity still remain unsurpassed. In The Mysterious Edogawa Ranzan
(1937), for example, one scene intercuts shots of a swordfight with shots of a woman's Figure 12.13 ... a reverse angle of the woman, Figure 12.14 . . . does the director supply a
struggle to escape from a gambler. But the director works a startling variation on and only t h e n . . . shot of the ninja that replies to the first image
(Figure 12.11).
the shot/reverse-shot formula. He cuts from a tight angle on one participant to a
correctly angled answering shot, but showing a character in the other line of action
(Figures 12.11-12.14). Such passages, interrupting a film that otherwise obeys the
as in Hollywood's conception of the axis of action. Ozu can then place his camera at
protocols of continuity, show how novel effects can be built out of received devices.
various spots around the circle, but not just any old spots; he typically shifts the angle
What E. H. Gombrich calls stylistic schemas, the traditional patterns that form the
of view in multiples of 45 degrees. 11 An early example of his ambitions, from Tokyo
artist's repertory, serve as the indispensible starting point for innovation. 10
A more far-reaching instance is Ozu Yasujiro, who built an alternative system of Chorus (1931), presents the rudiments of a system that he would tweak and refine
decoupage out of the premises of Hollywood continuity. Ozu retains all the devices of across 30 years (Figures 12.15-12.18). The result is a unique creation, a world that
the American approach: establishing shot, analytical editing, reverse angles, eyeline invites the viewer to explore cinematic space as never before. Yet it remains completely
matches, matches on action, and enlargement of the scene from distant views to legible for viewers accustomed to classic continuity norms.
close-ups. He also grants the need for spatial redundancy, so character placement It's worth mentioning another recasting of a standard device. Even after the triumph
and movement are signaled through real-world cues like eyelines and our knowl- of continuity in the American manner, Japanese directors cultivated unusually distant
edge of how people sit and stand in certain situations. Ozu absorbs these devices and dense establishing shots. In Figure 12.19, Sadao Yamanaka arrays a remarkable
into a unique set of cinematic coordinates. By setting his camera consistently lower number of faces and bodies across and into the space, each element precisely posi-
than what it films, he brings the bottom half of the shot into greater prominence and tioned. Most directors don't play out entire scenes in such framings, preferring to cut
creates a striking graphic similarity across images. He intensifies this similarity by into and aroiind them as per analytical editing strategies; but such packed shots are
compositional devices, such as arranging figures so that they match in position and more salient than their parallel framings in Western cinema. The denouement of The
contour. Above all, he conceives of dramatic space as a circle rather than as a line, Groom Talks in His Sleep (1935), a charming marital comedy by Gosho Heinosuke,
344 Poetics of Cinema Visual Style in Japanese Cinema, 1925-1945 345

Figure 12.15 Tokyo Chorus (1931): Ozu
provides an establishing shot of the husband
dressing for work.
Figure 12.16 Tokyo Chorus: He follows this
with a 180-degree reverse, matched on his
movement of pulling on his pants. The flagrant
cheating of the wife's position is partly masked
by the graphic similarity of the shoji walls on
the right side of the two shots.
Figure 12.19 I n Yamanaka Sadao's Kochiyama Figure 12.20 A dense establishing shot in
Soshun (1936), as in Ozu's shots, a low height The Groom Talks in His Sleep (1934).
accentuates foreground elements like the jar
on the right and brings out layers of depth.

Figure 12.21 A slightly high angle show-

Figure 12.17 Tokyo Chorus: A closer view of
the husband . . .
ing samurai dotting the frame is a common
Figure 12.18 . . . is followed by a compara- image schema of the period (Crow of the
bly framed view of the wife, with her head in Moonlight Night, 1939).
almost the same spot of the frame as his. In
later films, Ozu's graphic matching of bodies
and locales would become eerily exact
Voices in the Dark

presents a doctor's visit to the newlyweds in over 50 shots, but the cutting is punctu-
ated by reestablishing shots filled with business. At one point, as the doctor tries to
work his cure on the sleep-talking husband, we see a group of family and friends set
at a considerable distance. They provide a string of variously angled faces, including
the worried mother on the left and the groom himself at center, far off, half-turned
from us, and nearly blocked by the others (Figure 12.20). Most Japanese films of the
period include at least a few such framings; one convention of the swordplay film is
the squad of kneeling warriors spread meticulously across the frame (Figure 12.21).
Household architecture can add to the sense of a very full image, as we'll see shortly,
and some directors will take this tactic quite far, asking us to scan a dense field and
pick out slight changes in the drama. The replete establishing shot is a good instance
of a Hollywood convention being creatively reworked by a tradition that senses that
an abundantly filled distant view is an arresting accomplishment.
Something similar can be said about the benshi. This chanter and declaimer, stand-
ing or sitting beside the screen, pouring out a stream of commentary, explanation,
conversations, and cries—in several voices—appears to be the most visible sign of the
radical differentness of Japanese cinema. The benshi s ties to traditional theater and
oral storytelling would seem to be incontrovertible evidence that the Japanese cinema
preserved feudal traditions unchanged into the 20th century. In Noel Burch's argu-
ment, the benshi serves as the central example of the anti-illusionistic basis of Japanese
film forms. By putting speech outside the narrative world, the benshi presented
a reading of the diegesis which was thereby designated as such and which thereby
ceased to function as diegesis and became what it had in fact never ceased to
be, a field of signs. The most "transparently" representational film, whether
Western or Japanese, could not be read as transparent by Japanese spectators,
because it was already being read as such before them, and had irrevocably lost
346 Poetics of Cinema
its pristine transparency. The benshi removed the narrative burden from the
images and eradicated even the possibility of the images producing a univalent,
homogeneous diegetic effect.12
According to Burch, the "presentational" side of Japanese film incarnated in the
benshi was never wholly lost over later decades.
But a comparative perspective complicates things. Early cinema in many countries
gave a role to offscreen speech. There were lecturers in American theaters into the
1910s, initially to accompany travelogues or Passion Play films but eventually to help
the audience understand increasingly complicated stories.13 Practices varied, but
some European and North American theaters employed lecturers until the end of the
teens.14 The benshi, although an extreme case, doesn't seem different in principle from
the sort of lecturers developed in Western countries.
More theoretically, several of Burch's assumptions can be questioned. Does a
physical separation of voice from story necessarily create an aesthetic dissociation? 15
Do spectators of a Western film experience an illusion?16 Does simply explaining a
story as it goes along reduce the diegetic effect? (This doesn't seem to occur when a
parent reads a story to a child and explains the action.) More to our purpose here,
other research, principally by Joseph L. Anderson, makes Burch's historical conclu-
sions less plausible.17
It seems certain that the benshi's original function was to clarify the film's action.
The benshi apparently answered the early need to explain foreign films' stories to
Japanese audiences. Imported silent films were never given Japanese intertitles because
the nation presented too small a market to make it worthwhile. If one really wanted
to weaken the diegetic effect and emphasize each shot as a collocation of signifiers,
surely the best way to do so would be to avoid using any commentary and show the
films silent, with all their plot obscurities intact. By contrast, if audiences wanted to
understand and get absorbed in untranslated foreign films, the benshi was virtually
the only solution available. The emergence of the benshi testifies to the audience's
appetite for narrative comprehension.
Just as important, the benshi s earliest application to domestically made films
centered not on narrative commentary but on dialogue, a function that Burch mini-
mizes. Until the 1920s, many Kabuki films used several benshi, lined up alongside the
screen to provide the voices for all the characters. In some cases, these reciters were
the very actors who performed in the film.18 Even in solo performance, which became
more common during the 1920s, the benshi s virtuosity consisted partly in mastering
kowairo, or vocal imitation, enabling the performer to assume the voice of a child, an
old woman, or a samurai.
The centrality of dialogue to the benshi s performance suggests a commitment to
what Burch would call illusionism. "Because of [the benshi]," observed Inagaki Hiroshi,
"we directors were aware that when you went to a movie theatre, voices came out of
the characters on the screen."19 And a 1915 account explains the benshi's function:
"They act [as] dialogists for the players in the picture and some of them make a dia-
logue so skilfully [it seems] as if the players in the picture were really speaking." 20
Visual Style in Japanese Cinema, 1925—1945 347
Another aspect of the performance calls into question Burch s idea that the benshi
blocked diegetic "transparency." In addition to explaining the action and portray-
ing the characters, the benshi intensified the expressivity of the film. A 1925 writer
noted that the benshi sought to "intoxicate the audience" and to "induce . . . mass emo-
tional response." 21 Through impersonation or commentary, the benshi could guide the
audience's emotional experience of the film: speeding up the patter for rapidly paced
sections, letting sudden cries split a silence, or prolonging sentimental scenes by a
lyrical delivery. Some benshi were famous for singing during scenes. Again, we must
ask, If the goal was "semantic dissociation" and the designation of the screen as simply
a play of signifiers, is this emotional rendition of the story's climaxes a reasonable path
to take? Rather, are not such efforts exactly what we would expect if the audience avidly
desired a heightened involvement in the twists and turns of the narrative?
Rather than going to the movies to have the narrative world "designated as such"
by a dissociated, demystifying commentary, Japanese spectators seem to have wanted
something else. The evidence suggests that they sought a rich multimedia experience
that blended narration, impersonation, expressive intensification of the situations,
and a self-conscious display of vocal and histrionic skill. If such a package of plea-
sures breaks the "diegetic effect," so does opera.
This is not to say that the benshi doesn't represent a distinctive aspect of Japanese
film. But the benshi's legacy is a complicated one. In providing a mixture of narra-
tive denotation and expressive heightening, all the while emphasizing virtuosity, the
benshi typifies many tendencies within Japanese visual style as well. The saturation of
the performance situation has a kinship with the films' tendency toward highly expres-
sive and decorative visual treatment. (See the next essay.) Moreover, the presence of
the benshi almost certainly did allow directors to be somewhat more elliptical in their
narration. Late silent films often have a freedom in their use of dialogue intertitles
quite rare in Western filmmaking. For instance, instead of the usual speaker-title-
speaker or speaker-title-listener patterns, the Japanese explored such variants as
listener-title (from offscreen speaker)-listener and scenic detail-title (from offscreen
speaker)-scenic detail. 22 The benshi $ mimicry could specify the speaker even if she
or he were not onscreen.
Exercising the Eye
The benshi's kowairo mimicry and expressive amplification all serve to point up
salient aspects of the image. We have other reasons to presume that Japanese
directors, like their counterparts abroad, sought to guide the viewer's attention. Most
of Figures 12.1 through 12.21 show a reliance on standardized centering, selective
focus, and the like. More complicated cases also exhibit the sorts of staging skills
cultivated in Hollywood and Europe in the 1910s. Fallen Blossoms (1938) displays
the wholly familiar tactic of providing cues of frontality and composition to direct
attention to dialogue (Figure 12.22). When the same principle is put to work in tighter
spaces, one can only admire the skill of the director. In A Mother Is Strong (1939),
Ryuku arrives home to find her mother ill. The discovery is handled in a densely
348 Poetics of Cinema Visual Style in Japanese Cinema, 1925—1945 349

ft "

•» < •

Figure 12.22 Fallen Blossoms (1938): The
heads of the geisha in the foreground are
turned from us or blocked, so as to con-
centrate attention on the centered, frontal
women in the background as they talk.
Figure 12.23 A Mother Is Strong (1939): A Figure 12.26 Heat of the Earth (1938): Figure 12.27 As the father is taken away
wide-angle lens shows Ryuku entering the Decorative use of aperture framing. through the gate in the far distance, the boys
house. ask their mother what is happening (Children
in the Wind, 1937).

Figure 12.24 A Mother Is Strong: The cam- Figure 12.25 A Mother Is Strong: The cam-
era pans left with her as she looks in and finds era pans still further left as she pauses in the
her mother ill. The doctor looks up. foreground, back to us. The doctor turns
away, giving compositional salience to the
sick mother, now centered and brightly lit.
packed series of planes, unfolding the salient elements through cues of frontality,
figure movement, camera movement, and disclosure (thanks to the sliding door, or
fusuma). (See Figures 12.23-12.25.)
As if to flaunt their virtuosity in concealing and revealing screen space,
directors make much use of aperture framing. We can find this tactic in American
and European cinema of the 1910s, but the Japanese push it very far. Most typically,
an establishing shot will tuck a character into one cell of that gridwork characteristic
of the Japanese house (Figure 12.26). Aperture framing may also press us to the limits
of visibility. Throughout Shimizu Hiroshi's Children in the Wind (1937), the two boys
are seen in deep-space shots spying on the adults. When their father is arrested, the
policeman takes him away in a distant slot formed by the opened gate (Figure 12.27).
Through concealment and disclosure, aperture framing, and selective focus,
directors often make what is important to notice hard to see. If we scan the frame
somewhat more in Japanese cinema than we might in Western films, it is partly because
Figure 12.28 First Steps Ashore (1932): Figure 12.29 First Steps Ashore: Later, the
Characters slotted into a decorative screens same divider yields a new array of figures.
portholes.
staging and lighting sometimes block access to the narratively important information.
When that is missing or insufficiently rich, the viewer must search for other cues. In
this process of ransacking the shot and straining to see, the spectator explores the
overall composition as a visual design as well as a transmitter of story information.
At the same time, the process of bumping up against relatively stingy zones of
narrative information can create a sharpened awareness of nuance. We become alert
to the slightest difference in shape or lighting or posture, the minutest movement.
Shimazu Yasujiro's First Steps Ashore (1932) provides a dazzling example. Two
conmen walk into a waterfront bar; they are filmed through a porthole-pocked
grillwork (Figure 12.28). Later in the scene, a slightly varied camera position creates
a new geometry as the two conmen put their heads together and size up the hero,
peering through various portholes in the course of the scene (Figure 12.29). We must
strain to pick out the characters and follow their action, all the while appreciating
how graphic modulations within the grid are created by shifts in character position.
Throughout the ensuing quarrel, Shimazu's pseudo-modern room divider continues
to slice up the figures.
350 Poetics of Cinema
Figure 12.30 Partial disclosure and decen- Figure 12.31 Lily of the Valley (1935).
tering in The Little Foxes (1941): At screen left,
Oscar is almost wholly screened by the curtain
as Birdie tells Xan about her family's faults.
Figure 12.32 Lily of the Valley. Figure 12.33 Lily of the Valley.
Blockages of story information aren't unknown in other national cinemas, but
few directors exercise the audience's eye as gracefully as a Japanese director can. In
The Little Foxes (1941), as Birdie tells Xan of her unhappy life with Oscar, he stands
behind them, concealed by a doorway curtain (Figure 12.30). The spectator waits
anxiously for his outburst. Chancy as this staging is by Hollywood standards, it looks
a bit contrived alongside the finesse displayed by Kawate Jiro in the boarding school
comedy Lily of the Valley (1935). At an assembly in the dining hall, our heroine becomes
ashamed of her part in a theft. As the headmistress asks the girls to show their purses,
Kawate cuts from a medium shot of the heroine hanging her head (Figure 12.31) to
a shot of the headmistress in another part of the room. In that shot (Figure 12.32),
the girl's bowed head is just barely visible, a dim shape between the mistress' right
shoulder and another girl (far better lit and more frontally positioned). Even more
daringly, Kawate then cuts to an extreme long shot of the dining hall (Figure 12.33),
with our attention still drifting to our girl—under the clock, in soft shadow, hers the
only head lowered in shame.
The demand that the viewer probe the shot for just barely noticeable elements
is explored to an unprecedented degree in Japanese film. Western filmmakers have
Visual Style in Japanese Cinema, 1925—1945 351
Figure 12.34 Record of My Love (1939). Figure 12.35 Record of My Love.
Figure 12.36 Record of My Love. Figure 12.37 Record of My Love.
long known how to stage scenes so that figures block, reveal, and highlight visual
information at crucial moments. They developed this skill set fairly early in the silent
era, and it was part of cinematic storytelling well into the 1950s. Japanese filmmakers
.proved adept at this, but they again revised the schema in ways that created a vibrant
game of vision, teasing us with partial views and offscreen action. The holes and
crannies created by settings in Heat of the Earth (Figure 12.26) and First Steps Ashore
(Figure 12.27) can also appear spontaneously when people shift their positions,
however slightly. In Record of My Love (1939), a paralyzed war veteran is visited by
an old friend. The friend's discreet sorrow, observed by the soldier's wife, is played
out in a long shot that creates crevices of visibility between the bodies. The scene
culminates when both friend and wife twist their heads slightly away from us in grief
(Figures 12.34-12.37).
The vigor, refinement, and nuance manifested in these variants of Western con-
tinuity principles impel us to ask whether the style can plausibly be traced to broad
cultural causes—for aren't these just the qualities prized in Japanese screen paint-
ing, e-makimono scrolls, woodblock prints, and the like? Don't we find the same
dynamic perspectives, veiled views, and decentered compositions? We do, but only
up to a point. It's untenable, I suggest, to posit a continuous tradition in which cinema
becomes the bearer of centuries-old models of representation. The situation is more
complicated, and more interesting.
352 Poetics of Cinema
Constructing and Reconstructing Japaneseness
When asked in 1955 why he made so many historical films for export, Nagata Masaichi
replied, "America was making action pictures. France had love stories, and Italy
realism. So I chose to approach the world market with the appeal of Japanese historical
subjects. Old Japan is more exotic than Westernized Japan is to Westerners." 23 The
idea that "Japaneseness" could be packaged was perhaps more evident in the post-
war era, when geisha and cherry blossoms seemed naively nostalgic in an aggressive
industrial economy. But many decades before, particular versions of Japanese tradi-
tion were being produced for both foreign and domestic consumption. And cinema
had a role to play in the process.
After Japan opened to the West in the late 19th century, it began to redefine its
own culture over and against the Europe and America it was coming to know. Meiji
modernization included more than principles of banking and education, military
organization and technological growth. Japan was also to mature through a selective
assimilation of Western culture. French-style painting came to prominence in art
schools. Tokyo's Conservatory of Music copied its curriculum from Europe, and the
teaching of classic Japanese music was not permitted there until 1936. Many writers
of the period went abroad, and most were influenced by English, French, and German
literature. At the level of daily life, the growth of Japanese capitalism created a new
urban lifestyle based around white-collar workers in government bureaucracies and
private firms. Professionals and office workers wore Western clothes at work. An
expanding economy and growing city centers gave Japanese more disposable income.
The taste for popular magazines, novels, films, and music on Western models emerged
in the 1910s, forming a middle-class mass culture (taishu bunka).
Genres of "traditional" culture declined, becoming devotee arts or vessels of
nostalgia. By 1900, the Noh and Kabuki were largely incomprehensible to the masses,
and shimpa, the melodramatic Westernized drama, would soon be vanquished by the
upstart medium of film. During the Meiji era, "traditional" Kabuki underwent a series
of reforms; the new theaters had Western-style auditoriums, with proscenium arches
and drop curtains. During the same period the Edo-style woodblock print died, and
the New Prints movement began to create numbered editions on the Western model,
aimed at foreign collectors. Noh chanting and flower arrangement became divorced
from the everyday course of life; they were pursued as hobbies or subjects for adult
education courses. As mass media expanded, the Japanese became estranged from
many of those practices that most identified their culture to the outside world.
The very word culture (bunka) came to be identified with the importation of Occi-
dental customs. Cinema played a key role in this process. Asakusa Park in Tokyo
became a pulsing center of popular entertainment, with the film theaters at the center
of a honeycomb of Western-style restaurants, milk parlors, and coffeehouses. 24 But if
cinema was to become a part of middle-class culture, then it would have to compete
with the modernized films of Hollywood. Kaeriyama Norimasa and other intellectuals
advocated a more up-to-date technique, and the Shochiku company was founded to
produce modern pictures that could compete with the American product. Japanese
Visual Style in Japanese Cinema, 1925—1945 353
who had worked in the American studios were brought home to make films. Through-
out the 1920s, Japanese film executives toured Hollywood and the European studios,
buying equipment, studying production methods, and watching the latest product.
Directors developed their craft by studying the Western movies that were flooding
into Japan's theaters. Partly for efficiency, partly to attract a more upscale audience,
and partly in hope of export, Japanese filmmaking began adopting American staging
and shooting techniques.
After the disastrous Kanto earthquake of September 1923, Tokyo was rebuilt, and
this process intensified the modernization of urban life. Now gas, water, and electricity
were more widely installed, and the streets were laid out for auto traffic. Because the
fires during the quake had burned out many of the urban poor, suburbs grew, and
the central city became even more dedicated to business and amusement. Japanese
artists fell under the influence of movements such as surrealism and expressionism,
while throngs embraced baseball, radio, phonograph records, American and Japanese
comic strips, and the "mass novels" serialized in daily newspapers.
The 1920s saw a new urban dandyism, a fascination with mass-produced commodi-
ties (canned food, candy, and fountain pens), and a carefree, mildly sexy view of life
known as ero-guro-nansensu: "erotic-grotesque nonsense." Observers at the time
discussed the culture as one dominated by a taste for Westernization and "modernism":
The ethos of sacrifice and self-improvement promulgated in the Meiji era became
replaced by consumer indulgence and fan subcultures. 25 Cinema was a central means
of disseminating and glamorizing the new urban lifestyle. The modern-life films teem
with flapper-like mogas ("modern girls") and mobos ("modern boys") sporting roydo
(glasses like those worn by Harold Lloyd).
During this period, Japanese directors completely mastered Western film style. The
conception of a Westernized modernity continued even during the early years of the
Depression era; the industry's first successful talkie, Madam and Wife (1931), gave a
starring role to Japanese jazz. The transition to sound evicted the benshi and increased
the rationalization of production. A new company, Toho, organized its management
along the lines of Hollywood's central-producer system. By the mid-1930s, Japanese
cinema was second only to Hollywood as a mass-production, vertically integrated
studio system.
But resistance to Westernization was growing. In the political sphere, efforts toward
party-centered democracy were collapsing; conservative factions demanded military
expansion, most notably into China. In the culture, there began a trend identified by
Japanese as the "return to Japan" (nihon kaiki), the sense among middle-aged writers
who had embraced Westernization that they had lost touch with indigenous tradi-
tions. Probably the most widely known statement at the period is Tanizaki Junichiro's
essay "In Praise of Shadows" (1933), which acknowledged that traditional culture
was collapsing in the face of modernization. Tanizaki urged writers to rediscover
indigenous sources of beauty. Similarly, Suzuki Daisetsu's lectures describing Zen
Buddhism as a central force in the national culture sought to redefine Japaneseness.
The government's 1937 edict Kokutai no hongi ("Cardinal Principles of the National
Entity of Japan") set out an official version of what it was to be Japanese: intuitive and
354 Poetics of Cinema
ascetic (as contrasted with the rational and analytical West), loyal to the group (rather
than individualistic), and opposed to social revolution and class warfare. 26
The artificiality of such exercises springs partly from their deliberate limiting of
what is to count as "inherently" or "purely" Japanese. Peter Dale has forcefully criti-
cized writings on Japanese culture for "their simple faith in the idea that Japans vast
and varied tradition may be summed up in one 'key word.'" 27 Japan's culture and art
have been crisscrossed by conflicting trends. Ian Buruma suggests that Japan has at
least two cultural traditions: a Shinto-inspired popular celebration of sensuality and
the grotesque, and a more aristocratic Buddhist aesthetic imported from China and
Korea and highlighting austere spirituality. 28
The very idea of Japanese culture was open to tactical redefinition. Many customs
widely believed to be centuries-old traditions were self-consciously introduced by
Meiji social engineering. The cultural and political elites drew from the samurai
ethos to construct an official ideology for a nation about to enter the world capitalist
system. 29 The new policies put in place seasonal work, seniority ranking, and lifelong
loyalty to the firm. Even veneration of the emperor, supposedly an eternal tradition,
was a Meiji invention; before 1868, most Japanese did not know they had an emperor.
During the 1920s and 1930s particularly, an image of Japan's distinctive traditions
was being created for domestic and foreign consumption. 30
Similarly, the 1930s "return to Japan" sought to redefine Japanese tradition over
and against the foreign influx of the 1910s and 1920s. Watching Kabuki under
electric lights, for instance, teaches Tanizaki the importance of shadows in his nat-
ion's unique tradition. He savors what he takes to be an inherently Japanese plea-
sure in textures half-discerned, somber emotions, the glimpse of a woman in the
dusk—now all scorched away by a loud, bright, cheerful urbanity. The nostalgia of
Tanizaki, Suzuki, and many others responded to popular culture's unauthorized,
helter-skelter, and potentially dangerous flirtations with Western fads. Citing recent
decades' failure to "digest things thoroughly," Kokutai no hongi complains that "since
the days of Meiji so many aspects of European and American culture, systems, and
learning have been imported, and that, too rapidly."31 The text echoes Meiji policy in
demanding not a complete cutoff from the West but an absorption of its most useful
inventions into Japan's national identity. This sense of unity, recast in still firmer
terms in the 1941 edict "The Way of Subjects," was to dictate the official policy of the
war years: "It is an urgent matter for Japan to realize the establishment of a structure
of national unanimity in politics, economy, culture, education, and all other realms
of national life."32
It isn't enough, then, to establish the influence of Western practices on twentieth-
century Japanese culture. We should also consider that when an artist draws on pre-
Meiji tradition, the activity is likely to be mediated by twentieth-century purposes
and concerns. For example, the exposure to the West set writers a problem. How to
contribute to world literature and remain, according to some conception, distinctively
Japanese? In parallel fashion, the Kanto earthquake gave filmmakers a chance to start
from scratch, to modernize Japanese film along Hollywood and European lines. But
Visual Style in Japanese Cinema, 1925—1945 355
at the same time, it posed a problem: How to avoid becoming simply a copy of the
West? How to infuse a modern cinema with a sense of Japaneseness?
Japanese identity was being redefined throughout the period during which the
nation's film was developing as industry and art. This makes it problematic to find
that cinema a straightforward transmitter of pre-Meiji traditions. If filmmakers at
the outset wanted to be "true to Japanese tradition," they could have shot their scenes
from a very high angle, filming in extreme long shot actions taking place in roofless
buildings enveloped in mist—adopting, that is, the visual principles of picture scrolls
and woodblock prints. No cinematic conditions prevented them. Instead, they filmed
their Kabuki and shimpa dramas in straight-on long shots that largely resemble those
of early cinema in the West. 33 Later, when directors had more technical resources,
they could have relinquished Western-style cutting altogether (much as their samurai
predecessors gave up guns) and filmed all their scenes in the exploded-rooftop man-
ner of e-makimono painting. Again, they did not.
What did the filmmakers do? They accepted, broadly and basically, Western con-
ventions of staging, cutting, and storytelling. In many instances, they then inflected
these by means of stylistic devices that, according to contemporary conceptions, might
make these films recognizably Japanese. That is, the norms of the West provided a
framework within which more distinctively "Japanese" elements could be situated.
In this respect, they behaved like the painters who tried to merge plein-air elements
with received traditions in design, like the novelists who inserted lyrical landscape
descriptions into intrigue-driven plots.
The institutional context favored this synthesis. As film studios grew, they adopted
rationalized Western production methods. They developed stars and publicity
campaigns, and they set up script departments that sent out staff to analyze the latest
U.S. product. The main genres emerged out of popular literature and theater and
contemporary fashion. In all this ferment, the films' visual conventions were shaped
by post-Meiji forces, particularly by Tokyo's mass culture. 34 By the time pictorial
strategies reminiscent of older traditions showed up in films, they weren't smoothly
transmitted across centuries. They operated mostly as knowing citations, marking
the product as distinctively "Japanese" while achieving particular aesthetic and
social ends.
Many of the striking stylistic figures I've highlighted can be seen as citing indig-
enous artistic traditions by embellishing the more "Western" stylistic protocols
governing the film. The occasional opacities of framing and the game of blockage
and disclosure can be read as instances of inpei (concealment or veiling), an aesthetic
value fulfilled in traditional painting by mists, screens, and steeply angular walls.
Aperture framing may self-consciously recall similar tactics in prints and paintings.
The fact that figures often kneel, sit on the floor, stand, or sit on a raised dais allows the
shot to accede to that traditional division of the picture format into low foreground,
higher middle ground, and highest distant point (Figure 12.38).35 And although the
exploded-roof convention is unknown in the films, certain shots of courtyards and
palaces from the monumental jidai-geki of the late 1930s seem deliberately calcu-
lated to summon up the great tradition of the narrative picture scrolls (Figure 12.39),
356 Poetics of Cinema
Figure 12.38 Massive and geometrical crowd Figure 12.39 Iemitsu and Hikozaemon
scenes characterize The Abe Clan (1938). (1941): A high angle with cherry blossoms.
answering to that "return to Japan" being called for by the government in the 1930s
and 1940s.
My point here differs from the common strategy of searching within Japa-
nese artistic traditions for analogues or sources of filmic elements. 36 As adequate
accounts of Japanese film style, such comparisons fall short. A film's style creates
a systematic context that swallows up discrete elements. To isolate generic features
of Japanese design—say, a fondness for diagonals or an interplay of empty and full
space—and then to find such features within a shot or two is to ignore the larger
dynamic of the film, its overall stylistic orientation, within which these elements
fulfill specific purposes. Such debts to traditional practices can often be seen as
momentary heightenings, self-conscious elaborations serving to add expressive or
decorative qualities. It's not just that Ozu occasionally employs wide-angle views
reminiscent of Hiroshige in certain shots (for we still have to find the principles
informing all those thousands of shots in which he doesn't). The point here is that
a filmmaker, working in the upstart medium of film, can self-consciously tap the
great tradition of woodblock prints in order to redefine Japanese urban life in the
contemporary era. 37 At the limit, such images may come to signify "Japaneseness"
itself, becoming a marker of cultural distinction. This possibility becomes available
once the image makers have become aware of other methods of making, as Japanese
directors certainly were by the early 1920s.
Stylistic Trends
Japanese film style, it seems to me, can best be understood as involving an assimi-
lation of continuity techniques seen in the West alongside an experimental impulse,
itself often mediated by a self-conscious sense of "Japaneseness." The more specific
stylistic trends of the period can show us these processes at work.
Surprisingly, the period film contributed as much to the modernization of style as
did the modern-li fe film. Although the reformers of the late 1910s and early 1920s had
Visual Style in Japanese Cinema, 1925—1945 357
Figure 12.40 Lieutenant Colonel Tachibana: Figure 12.41 . . . Tachibana striking down-
Tachibana slashes downward with his right ward and across with his left arm. The effect
arm. Cut to . . . is remarkably like the jump cuts of the Cos-
sack slashing with his saber at the end of
Potemkin's Odessa Steps sequence.
stressed applying Hollywood-based technique to tales from contemporary life, it was
the jidai-geki and particularly the swordplay film that emphasized shot/reverse-shot
decoupage, crosscutting, fast cutting in scenes of violent action, and other classical
strategies. To these were added violent thrusts of action to the camera, accelerated
motion (achieved through undercranking), and swift (often handheld) camera move-
ments, usually during fight scenes but sometimes during conversations or even across
scenes, as transitions. Judging by the few films available, Ito Daisuke and Makino
Masahiro, the two most prominent directors of the swordplay film, exploited these
tactics for both narrative and expressive ends. This flamboyant style continued into
the 1930s.
Lieutenant-Colonel Tachibana (Saegusa Genjiro, 1926) shows that this style could
exploit disjunctive cutting to a degree comparable to French impressionist and Soviet
montage cinema. An early battle scene, featuring explosion bursts, scurrying soldiers,
and jouncy framings made by mounting the camera between horses pulling a wagon,
includes a string of shots ranging between 10 and 15 frames in length. A kendo match
is even more jittery, with shots lasting a half second or less. The climactic battle,
complete with upside-down images, averages about half a second per shot as well,
with a culminating passage cutting together mismatched close-ups of Tachibana's
charge just before an explosion (Figures 12.40-12.41). Lieutenant-Colonel Tachibana
looks remarkably like a Soviet montage film such as Pudovkin's Mother, released in
Russia the same year.
The flamboyant style also displayed great ingenuity in staging frame entrances
and exits. How frenzied this process could become is shown by the lengthy sword-
fight in The Twenty-Six Japanese Martyrs (Les vingt-six martyrs japonais; 1931).38
The shots are framed tightly, and swordsmen plunge in and out with high-speed
abandon; the director proves particularly adroit in using the bottom frame edge
(Figures 12.42-12.46). Films like Tachibana and Twenty-Six Martyrs boldly push

358
Figure 12.42 The Twenty-Six Japanese Mar-
tyrs (1931): A shot begins with a swordsman
in the middle ground.
Figure 12.44 Th I ly-Six Japanese Mar-
tyrs: Anotherfijlire ~ es into the shot from
the left middle ground and passes through to
leave on frame right. (He's the blurred figure
on the right side of this illustration.)
Figure 12.46 The Twenty-Six Japanese Mar-
tyrs: After dodging a blow, the hero pivots
and thrusts his attacker away just as another
fighter leaps into the fray from off left.
Poetics of Cinema
Figure 12.43 The Twenty-Six Japanese Mar-
tyrs-. The hero pops into the foreground from
the lower frame edge.
Figure 12.45 The Twenty-Six Japanese Mar-
tyrs: The hero steps further into the shot and
raises his sword just as another fighter rushes
into the shot from off right, very close to the
camera.
Visual Style in Japanese Cinema, 1925—1945 359
classical technique to expressive limits, and they remain exhilarating to watch
today. If they have a contemporary counterpart, it would probably be Hong Kong's
kinesthetic policiers, sword dramas, and martial arts films (which themselves were
influenced by 1960s Japanese successors to the classic chambara). Ito and Makino
would probably relish the way Ringo Lam supplies a bullet's point of view as it
ricochets around a nightclub in Full Contact (1993).
The force of the flamboyant jidai-geki was felt by the new generation of directors
emerging in the 1920s, and alternative stylistic resources began to be explored. Yoda
Yoshikata, screenwriter for Mizoguchi, recalled that at the time two trends emerged—
one putting the emphasis on shots and their combination, and the other on the scene
as a unit. 39
After the Kanto earthquake, the Shochiku studio encouraged directors to adapt
American continuity editing for use in shomin-geki, dramas and comedies about
contemporary lower- and middle-class families. Ordinary people's efforts to get by
were examined in what producer Kido Shiro called "a warm-hearted, approving
way."40 Shochiku's studio in the Tokyo suburb of Kamata fostered a cutting-centered
option I'll call piecemeal decoupage. In this method, modeled on Chaplin's Woman
of Paris (1923) and Lubitsch's The Marriage Circle (1924), every scene is broken into
a great many shots. Through titles and close views of faces, body parts, and details of
setting, the audience can follow the emotional flow of the drama. Sometimes, as in
Shimizu Hiroshi's Eternal Heart (1929), the film might provide very few establish-
ing shots and rely extensively on medium shots and close-ups of individuals. Ozu,
Shimizu, Gosho, Naruse Mikio, Shimazu Yasujiro, and Nomura Hiromasa were the
most famous exponents of the Kamata style. Piecemeal decoupage proved adapt-
able throughout the 1930s, continuing in such Shochiku vehicles as A New Family
(Figures 12.9-12.10) and Land of Cherry Blossoms (1941; Figures 12.47-12.49), as well
as in other studios' products. When Naruse left Shochiku for Toho, he continued to
employ it.
The piecemeal method could become a vehicle for expressive and decorative elabo-
ration as well. Naruse's Street Without End (1934), a melodrama about a poor woman
marrying into a snobbish family, pushes the principle of incessant scene breakdown
to an extreme: The average shot length is less than 4 seconds, and there is an aston-
ishing variety of setups. The climax is stunning, both dramatically and stylistically.
After fighting for respect in a household dominated by a domineering mother and
sister, Sugiko visits her husband in the hospital. "You're a good man," she tells him,
"but you're weak." Sugiko then charges her in-laws with selfish pride. Sugiko walks
out on them as the husband falters. This deeply untraditional scene, played out in
5 and a half minutes, 84 shots, and 18 dialogue titles, works virtuoso variations on the
characters' positions (Figures 12.50-12.51).
Sugiko goes out to the hospital corridor. There she learns that her husband might
live. Abruptly, Naruse presents three shots based on the same pattern. There is a blank
frame, into which Sugiko steps (Figures 12.52-12.53). She stops. Cut to a new blank
frame, into which she steps, moving in a different direction (Figures 12.54-12.55).
360 Poetics of Cinema
Visual Style in Japanese Cinema, 1925—1945
361

Figure 12.47 In Land of Cherry Blossoms
(1941), one scene begins with a close-up
of paper balloons on Girls' Day, a national
holiday.
Figure 12.48 Land of Cherry Blossoms: Cut
Figure 12.50 Shifting arrangements of fig-
to a basket of knitting, the ball of yarn echo - Figure 12.51 Street Without End.
ures in depth in the hospital climax of Street
ing the shape of the balloons. Without End (1934).

Ijir
Figure 12.52 The abruptly cut moment in
Figure 12.49 Land of Cherry Blossoms: The the hotel corridor in Street Without End. Figure 12.53 Street Without End.
scene's action starts with a cut to the two
female protagonists talking about married
life, with the ball of yarn anchoring them in
relation to the previous shots.

She stops. Cut to a new blank frame. She steps in, from yet a different direction, and
stops (Figures 12.56-12.57). The three shots maintain narrational uncertainty. Is she
returning to the sickbed in a final gesture of love? Is she moving away in a definitive
act of rejection? Is she hesitating, stepping this way and that? The momentary disori-
entation provided by the empty frames and the abrupt, accentual close-ups reduce
i• • K ' n
her movement to a precise rhythmic geometry, but the passage remains spatially and Figure 12.54 Street Without End.
Figure 12.55 Street Without End.
psychologically ambivalent.
Along with the flamboyant style of chambara action films and the finely broken-
down montage of the Kamata-inspired films, there was another trend, which I'll Night (1939; average shot length 23 seconds), there are very few close-ups, and estab-
call the pictorialist approach. Here the director relies on significantly longer takes, lishing long shots are held for many seconds. This tendency probably coincides with
more distant framings, less cutting in to details, and more leisurely scanning of the that Nikkatsu-based custom pointed out by Yoda, a tendency to take the scene as the
film's basic unit.
image. Morning Mist in Edo (1935), for instance, plays several of its dialogue scenes in
The pictorialist approach emphasizes the individual shot as a rich visual
lengthy takes. (The average shot lasts 14 seconds.) During many scenes, a conversation
design, summoning light, texture, and geometric shape to create stable, graceful
is handled in a few fairly distant framings (Figure 12.58). In Crow of the Moonlight
compositions. The shot is often divided up into severe vectors, either from a high
362 Poetics of Cinema
Figure 12.56 Street Without End. Figure 12.57 Street Without End.
Figure 12.59 Hie solemn landscapes of the
Figure 12.58 In Morning Mist in Edo (1935),
family confrontations are often punctuated monumental jidai-geki: The shogun strolls
by distant, static framings. thoughtfully along his garden in Iemitsu and
Hikozaemon.
angle to create diagonals or from a straight-on position that generates horizontal
bands. The geometrical variations of these can become strikingly decorative.
As these examples suggest, the pictorialist style achieves full range in the period
film. Here vassals and retainers kneel in rows in castle halls, long corridors provide
sharp diagonals that set off the curves of the human figure, and gardens and court-
yards yield picturesque vistas (Figure 12.59). Out of the somber stillness evoked
by the pictorialist style, the monumental jidai-geki of the late 1930s build images
of majestic sorrow and fortitude. Crane shots from one poised long-shot compo-
sition to another heighten the serene gravity of feudal tradition, as in Iemitsu and
Hikozaemon (Makino Masahiro, 1941). The Abe Clan (1938) becomes a procession
of splendid compositions, imbuing the story of a samurai family's lost dignity with
an epic grandeur. (See Figures 12.38-12.39.) By this point, the monumental style was
well-suited for conveying government policies seeking to redefine martial virtues. 41
Of all efforts in this vein, Mizoguchi Kenji's Genroku Chushingura (1941-1942) is
by common consent the most hieratic. The most expensive Japanese film made to
that time, it plays the classic tale of honorable samurai retribution on a colossal scale
Visual Style in Japanese Cinema, 1925—1945 363
Figure 12.60 The attack on Lord Kira that Figure 12.61 A monumental high angle
opens Genroku Chushingura (1941-1942) for Lord Asano's ritual suicide in Genroku
employs vast and deep sets, with action
unfolding diagonally.
(Figures 12.60-12.61). It unfolds gravely and unhurriedly, in extraordinarily long
takes and solemn movements of characters and camera.
There is some evidence that these stylistic trends came to dominance in roughly the
order I've considered them, with the flamboyant method emerging in the mid-1920s,
piecemeal decoupage becoming widespread at the end of the decade, and pictorialism
emerging in the mid-1930s. There is also evidence that as the 1930s went on, direc-
tors staged action more slowly and let takes run longer. All the 1925-1933 films in my
sample displayed an average shot length of less than 12 seconds. During the transition
to sound in 1934-1937, average shot lengths spread across a wider range, with 30%
of the means running 12 seconds or longer. In the years 1938-1945, over half of the
84 films I examined had average shot lengths of 12 seconds or more. (Only six of these
films, be it noted, were by Mizoguchi.) 42
Blending and Refining
The flamboyant style, the piecemeal method, and the pictorialist style all constitute
variants of standard continuity, that pervasive style on which ex hypothesi the world's
mass-market cinemas are based. The more distinctive qualities of each style, shaped
by local demands of genre and institutional policy, stand out against that fundamental
stylistic commitment. For this reason, the three approaches are seldom seen in a pure
state. The flamboyant style and the pictorialist method rarely dominate an entire film.
A swordfight film might well utilize Shochiku-style scene dissection during conversa-
tion sequences, then present combat scenes in much more florid ways. Most commonly,
a picturesque long shot may appear within a scene of traditional analytical cutting.
Naruse, Yamanaka, Gosho, and other directors reveal a comparable eclecticism
across their careers, exploring different stylistic approaches within the flexible
boundaries of Japanese classicism. Perhaps the most striking example is that young
director who began his career near the end of the war, Kurosawa Akira. His wartime
364 Poetics of Cinema Visual Style in Japanese Cinema, 1925—1945 365

Figure 12.62 The Most Beautiful (1944): As
the girls working in the lens factory argue
about whether one is strong enough to go
on, Kurosawa presents the characters' varied
reactions in a skillfully composed full shot.
Figure 12.63 The back-to-the-camera strat-
egy recurs throughout The Most Beautiful,
as when the protagonist Watanabe is turned
from us on the extreme left and a girl on frame
right excitedly points out planes overhead.
Figure 12.64 Sanshiro Sugata (1943): San- Figure 12.65 Sanshiro Sugata: After Monma
shiro's flying adversary, seen in a burst of lands, a window screen floats down in slow
overlapping cuts. motion.

efforts offer veritable anthologies of the editing and staging experiments of the
previous two decades.
Kurosawa's judo film Sanshiro Sugata II (1945), fully predicated on classical continu-
ity, also presents bodies flung to the camera, rapidly cut fight scenes embellished by
swooping crane shots, and swift axial cuts. (Using abrupt cut-ins or cut-backs along
a line became a Kurosawa trademark across his career.) The Most Beautiful (1944), a
patriotic film about girls working in a lens factory, exploits not only dynamic jump
Figure 12.66 Sanshiro Sugata: Elliptical Figure 12.67 . . . to old Murai already land-
cuts but also subtly modulated ensemble shots, with heads strung across densely cutting suggests power: From Sanshiro's leg ing very far off.
packed frames and distant, laconic staging that makes the emotional climaxes almost trip...

tranquil (Figures 12.62-12.63).

Kurosawa's first film, Sanshiro Sugata (1943), offers an archetypal narrative struc-
ture for hundreds of Asian martial arts films. Members of rival schools stage matches
to determine the supremacy of one style. The old master Yano defeats the pupils of
Monma, and Yano's student Sanshiro soon defeats Monma himself. The decadent,
Westernized Higaki beats Yano's other pupils; Sanshiro bests Higaki's sensei Murai
before trouncing Higaki himself in the film's climactic battle. The plot of Sanshiro is
thus already combinatorial.
Sanshiro is something of a resume film, displaying the young director's versatility
by treating each combat quite differently. At the dockside Yano tosses the thuggish
jujitsu students into the water, and Kurosawa handles this in long shots and slow,
symmetrical camera movements. By contrast, at a public match Sanshiro defeats
Monma in a flurry of feet, grappling hands, and airborne bodies. Kurosawa borrows
from the jidai-geki tradition and accelerates the editing of the three shots (27 frames,
18 frames, 5 frames) showing Monma flying through the air (Figure 12.64). This
rhythmic burst abruptly slackens: We don't see Monma land; instead, the camera
pans gravely around the audience before settling on his body; then a broken window
floats down in ghostly slow motion (Figure 12.65). A later match is handled even more
elliptically. We see Sanshiro's feet bracing after a throw (Figure 12.66); cut to a shot of
his opponent, old Murai, instantly hitting the mat very far away (Figure 12.67). This
last strategy not only expresses superhuman strength but also makes Sanshiro's art
seem to be impossible for the camera to catch.
This implication is intensified in the final combat. As the wind rages in the field,
again and again the camera pans tall grass to catch the sinister Higaki already falling
in defeat. Sanshiro's winning moves go almost completely unseen. A stylistic progres-
sion across Sanshiro Sugata steadily minimizes the adroitness of judo and empha-
sizes the struggle for purity leading up to the decisive moment. 43 By plotting out a
continuum running from the physical to the spiritual side of judo, Kurosawa builds a
significant pattern out of technical options made available by reigning norms.
Kurosawa turned eclecticism into a formally rich pluralism. Other directors took
a narrower path. They created distinctive styles out of recognizable mainstream
schemas but treated these schemas with unusual rigor. I've already indicated how
Ozu's distinctive style is grounded in classical continuity. Now we can see that his
system constitutes one long exploration of the minute possibilities of piecemeal
decoupage. Having set himself rigorous constraints, virtually a set of private rules for
staging and cutting, he could then apply them to make any scene fresh. For instance,
 Visual Style in Japanese Cinema, 1925-1945 367

Figure 12.68 Tokyo Chorus-. Salarymen check
their bonus envelopes by retreating to the
toilet; Ozu uses the swinging doors to block
our vision.
Figure 12.69 Tokyo Chorus: A typical 180-
degree cut soon creates a purely graphic play
of the doors, as another employee starts to
enter.
Figure 12.70 In Passing Fancy (1933), Ozu
uses distant depth and aperture framings in
his characteristic spatial style. When Kihachi
comes into Otome's cafe, he's shown bowing
and pivoting to go to a table.
Figure 12.71 Passing Fancy: Ozu's typical
180-degree cut follows him to the rear as he
sits with Jiro, barely visible in the distance.
(Compare Figure 12.19.)

his 180-degree cuts, usually applied to people, could also transform empty spaces,
activating purely graphic qualities of the decor and even creating whimsical gags
(Figures 12.68-12.69). He draws on many local norms, but the pressure of his stylistic
system makes them wholly his own (Figures 12.70-12.74).44
Shimizu Hiroshi found his approach somewhat later. Starting at Shochiku in
1924 at the age of 21, he had by the time of his death in 1966 directed over 150 films.
(In 1927 and 1929, he averaged one a month.) Shimizu was skillful at urban
melodrama, but he developed an interest in filming in rural Japan. Shochiku let him
go on the road with a small crew and come up with stories on the way. The results,
films like Mr. Thank You (1936) and Star Athlete (1937), are full of lyrical good humor Figure 12.72 Passing Fancy: He's blocked Figure 12.73 Passing Fancy: She walks down
and gentle pathos. Shimizu's staging in interiors relies on piecemeal decoupage and when Harue, now hired as a waitress, enters the corridor to present the food to the men.
to show Otome the trays.
the depth and aperture effects characteristic of the period. Once he gets outdoors,
though, he becomes maniacally insistent on moving his characters abreast of one
another. He then treats their common direction as an axis along which he will shoot.
The camera thus either faces the action or frames it straight from behind.
Consequently Shimizu frequently films his characters in incessant tracking shots,
moving only forward or backward. All the devices of conventional analytical edit-
ing, then, must be applied to characters walking, running, or riding directly to or
from the camera. In Star Athlete, a sequence of college boys out marching on military
maneuvers consumes nearly 40 consecutive traveling shots (Figure 12.75), complete
with point-of-view inserts of people whom the troop passes on the road. Mr. Thank
You allows Shimizu to vary his approach by filming a long bus journey, but again he
Figure 12.74 Passing Fancy: Ozu's low-
shoots rigorously down the center aisle, limiting analytical editing to a few rectilinear
45
camera version of an over-the-shoulder shot
camera setups (Figure 12.76). presents her giving Kihachi his food. He's
Mizoguchi Kenji also drew upon current tendencies to forge a rigorous and dis- pleased to find her there.
tinctive style. He began his career working for Nikkatsu, and his few surviving films
from the 1920s and 1930s suggest that he gave the pictorialist method considerable
prestige. Taki no shiraito (1933), The Downfall ofOsen (aka Osen of the Paper Cranes;
1935), The Poppy (1935), and Oyuki the Virgin (1935) are largely tied to Western
 Poetics of Cinema Visual Style in Japanese Cinema, 1925—1945 369
368

r
Figure 12.76 Mr. Thank You: Shimizu sub- Figure 12.77 In The Straits of Love and Hate
Figure 12.75 Star Athlete (1937): The endless Figure 12.78 The Straits of Love and Hate:
jects continuity cutting to his habit of film- (1937), Mizoguchi resolutely shoots most
march of the students as the camera tracks After some performers have cleared the fore-
ing the passengers only from the back or the scenes from a fair distance, while plunging
backward. ground and middle ground, the actor charged
front of the bus. them into semidarkness and employing very
with watching Ofumi's baby goes to the rear
distant depth. We must concentrate intently
to watch the play, dimly visible through a
on the drama. Here we're backstage, with
back doorway. Kenkichi strides in and seizes
actors making up both in the darkened fore- her child.
decoupage principles, employing many cuts and extensive shot/reverse-shot passages. ground and further to the rear.
But sporadically in these films and systematically afterward, Mizoguchi seized upon
the long take as a way to deepen the audience's concentration upon the image. During
the pre-1936 period he occasionally presents a shot that impedes our sight of charac-
ters, and particularly of faces, but he does so without recourse to the somewhat
recherche devices used by other directors. A scene sometimes arranges the figures so
that camera distance, posture, lighting, and architectural features such as walls and
doorways cooperate to create distant, opaque compositions.
His long takes allow him to make the "illegible" shot far more structurally promi-
nent, and to develop the composition gradually, creating a fluctuating pattern of
visibility. Decentering, distant framing, and patches of darkness create this effect
throughout Sisters ofGion (1936). A climactic scene in Naniwa Elegy (1936) presents Figure 12.79 The Straits of Love and Hate:
Figure 12.80 The Straits of Love and Hate:
Ayako's shamefaced confession to her boyfriend Fujino as a series of retreats from Mizoguchi holds on the empty frame as the
She comes through and to frame center, dis-
play goes on in the distance, with O f u m i
the camera, so that much of the scene is played with both figures in long shot, turned covering the baby is gone. The entire shot has
barely visible as she bows.
lasted nearly three and a half minutes.
from the viewer. 46 This "dorsality" strategy, an extension of the principle of dramatic
discretion we find in films like Record of My Love (Figures 12.34-12.37), reappears
throughout Mizoguchi's career. His films of this period also pursue the idea that
instead of bringing characters closer to the camera as the scene's drama intensifies, Tiie Pacific War: Toning Down Technique
one could make them back off. In combination with stately camera movements and We might expect that the distinctiveness of Japanese film style would reach a height
unexpected frame entrances and exits (Figures 12.77-12.80), his long takes could turn
during the "dark valley," the years of 1937-1945 during which war spread through
establishing shots into the building blocks of the drama. In his hands, the replete long
China and the Pacific. Certainly many ultranationalist writers urged that the wartime
shot characteristic of Japanese cinema became prolonged and more nuanced, ripen-
cinema convey the nation's essence. Here is one:
ing through small changes in particles of visual information.
In sum, when Japanese films of the 1930s diverge stylistically from Western films,
It is unthinkable that we should ever allow our splendid Japaneseness to fall into
we can most fruitfully understand them as transforming the schemas of the continuity
ruin. On the contrary, this love for things Japanese is destined to well up within
tradition in three primary directions. All of these may rechannel the flow of narrative
information, heighten expressivity, or add a decorative element to the scene. Individual [the creative artists among us], taking on marvelous new forms. Seen from this
directors developed their own variations on the dominant trends, creating styles that vantage point, the way forward for Japanese cinema is the bringing to real life of
stood out within their national cinemas and, eventually, within international cinema. that unique beauty which is native to Japan. 47
370 Poetics of Cinema
By the end of the 1930s, the "national policy" (kokusaku) film celebrating Japaneseness
had appeared in several guises. What I've been calling the monumental jidai-geki
was known as the rekishi eiga ("historical film"), in which samurai and clan dramas
48
became stately displays of feudal virtues. Contemporary-life films portrayed farms
and villages (Horse, 1941), intrigue in China (China Night, 1940), domestic problems
on the home front (Record of My Love; Tale of Guidance, 1941), and stories of combat,
ranging from the "humanistic" war films of the late 1930s to the more spectacular
battlefront sagas.49 The dominant strain in all these genres was "spirit-ism"
(seishinshugi), a portrayal of those prized Japanese ideals of patience, endurance,
courage, loyalty, and dedication to nation and emperor.
Many of these films develop the Japanese film tradition in significant directions.
The "humanistic" combat films of the China War, for instance, show soldiering as
grinding routine occasionally broken by the ferocity of battle. The most famous,
Tasaka Tomotaka's Five Scouts (1938), remains notable for its episodic narrative
line and its portrayal of quiet male camaraderie. 50 Mud and Soldiers (1939) recalls
All Quiet on the Western Front in its stress on the sheer physical exhaustion of war,
the hours of slogging through marsh and mud. In addition, these films work less
through the large-scale narrative arc of Hollywood war pictures than through an
assembly of vignettes. In this respect, they recall the most famous China War novel,
Hino Ashihei's Barley and Soldiers (1938), and other reportage produced by fiction
writers sent abroad to cover the front. 51
The films' celebration of quiet stoicism served an important propaganda purpose,
embodying a sense of the spiritual discipline necessary to win what would be a long
war. Still, the combat films don't present war as wholly a dreary affair. All contain
exciting battle scenes, accentuated by dynamic camera movements and rapid editing.
Five Scouts starts with vigorous combat, laying its credits over breathless tracking
shots and brief images of men firing, as if the attack were so ferociously unexpected
that the film could not spare a moment for orthodox opening titles. In the same
film, suspenseful traveling shots follow the scouts into a marsh as they search for the
enemy. The China War films also draw upon conventions of war films elsewhere: the
blood-soaked flag, the heroic charge, the moment when the soldier studies the photo-
graph of wife and child, and the inevitable passing of a cigarette from man to man. As
happens so often in the combat genre, fierce fighting is juxtaposed with lyrical detail.
When, during one charge in Mud and Soldiers, a soldier is hit and sinks to the ground,
the camera holds on the tall, gently waving grass that enfolds him.
Critics have noted that these films rarely present or even mention the enemy. Often
a battle is filmed simply from the Japanese side, showing no opposing soldiers. When
a battle is won, the scene fades out before the Japanese are shown taking the territory
and seizing prisoners. Even in the Pacific War films, the enemy is rarely portrayed
directly. In Victory Song of the Orient (1942), Manila is a decadent city, full of American
shops and advertising. The night before the attack on Pearl Harbor, sailors in The War
at Sea from Hawaii to Malaya (1942) listen pensively to dance music broadcast from
a Hawaii nightclub, where drunken U.S. sailors shout, "I'm wide awake and ready to
go!" In rare cases, an enemy is shown briefly; in Story of Tank-Commander Nishizumi
Visual Style in Japanese Cinema, 1925—1945 371
(1940), Japanese soldiers help a Chinese woman give birth, only to have her sneak off
after killing her baby.
John Dower suggests that the absence of enemy forces follows from the peculiarly
inward thrust of wartime ideology. By and large, propaganda presented the hostilities
as a means to regain a lost moral and racial purity. After decades of Westernized self-
indulgence, Japan's subjects would have to expunge foreign influences, live austerely,
and prepare to die for the emperor. 52 This is the period during which a soldier could
jot a pacifist haiku on a homebound postcard:
When I close my eyes,
I can see valleys
Flowing blood red.
The censor who intercepted the card demanded only that "blood red" be changed to
"pure blood." 53 Dower argues that the films emphasized not victory over the enemy
but the struggle to forge a pristine self. The enemy was not central because the war
was not to be presented as about conquest or the clash of competing political interests.
The war was about Japan's effort to recover its hereditary spiritual nobility, bearing
hardship with dutiful patience. 54
In addition, the absence of the enemy in these films may be a response to the Home
Ministry's 1937 constraints on what could be represented on the screen. One pre-
cept urged that directors not "exaggerate the cruelties of war with overly realistic
depictions." 55 Because offending films could be cut or banned, directors may have
steered to the side of safety by keeping carnage offscreen. A similar strategy governs
the more spectacular Pacific War films: Bombers and antiaircraft guns dealt death at a
distance, so long shots involving miniatures or stock footage avoided "overly realistic"
images of suffering.
, Whatever the narrative stratagems employed for propaganda purposes, can we
maintain that the 1940s visual style continues the experimentation of the earlier
years? On the whole, no—at least not on the basis of the dozens of works I've viewed.
Most films from 1941 to 1945, historical or contemporary, set on the battlefield or
home front, seem to be relatively orthodox stylistically. For the most part, what seems
to have been "innately Japanese" for purposes of national policy were the stories,
themes, and values depicted, not the mode of representation itself.
One might expect that the war films would draw upon the flamboyant technique
pioneered in the chambara, but they do not. The jump cuts of Lieutenant-Colonel
Tachibana and the pulsating combat of Twenty-Six Japanese Martyrs may have seemed
too frantic and undignified for the war picture as it was conceived under the new
order. Shots grew lengthier and more decorous during the years 1941-1945, and these
longer takes don't employ much intricate staging. We can still find some of the most
typical devices, such as rather distant establishing shots, but these seem to me less
prominent. Although some directors adhered to a strategy of piecemeal decoupage,
stylistic play of the sort seen in Street Without End and the work of Shimizu and Ozu
became infrequent. Most directors seem to have timidly staged their dialogue scenes
372 Poetics of Cinema
in standard ways. Mizoguchi's surviving 1942-1945 works are less daring than his
immediately previous work, and Ozu's sole film of the Pacific War years, There Was
a Father (1942), is one of his most stylistically conservative. Kurosawa's films would
seem to be among the few wartime efforts that preserved some of the experimental
verve of the 1920s and 1930s.
Ironically, a period of political exceptionalism and chauvinism appears to have
brought Japanese film style closer to that of Hollywood. How to explain this? It's worth
remembering that the rich experimentation of the 1920s and 1930s arose not from
isolation but from a fruitful commingling of international film styles with contem-
porary Japanese impulses toward highly expressive, decorative, and dynamic visual
presentation. With national mobilization after 1937, the international decoupage
norms underlying Japanese filmmaking came to the surface. Not surprisingly, the
propaganda needs of the war period worked against directors' flagrant and playful
manipulation of style. It seems likely that by the end of the 1930s, the norms of
continuity were no longer felt as foreign borrowings but simply as the basis of any
competently made movie.
During the 1920s and 1930s, Japanese filmmakers created one of the world's
most formally ambitious film traditions. For purposes of expressivity, decoration,
and narrational complexity, they revised the schemas of continuity cinema. In the
process, the directors created distinctive variants—the flamboyant chambara style,
the pictorialist approach, and piecemeal decoupage. Each depended upon and recast
principles of cinematic style current in both the West and East. Further, the revisions
were inflected by Japanese conventions in adjacent arts. Yet those conventions were
largely absorbed through second-degree conceptions of Japaneseness. And those con-
ceptions were in turn redefined within contemporary popular culture and in relation
to the specific needs of cinematic storytelling.
Finally, the reversion to a strongly classical style in films of the Pacific War period
suggests that the previous decade's explorations were indeed recognized as somewhat
eccentric. For state agencies and studio executives, the stabilization of the style may
have seemed necessary for the transmission of a clear propaganda message. We might
conjecture that Japan's postwar golden age, and the new burst of experimentation
that followed in the New Wave, heralded returns to the adventurous explorations that
graced the 1920s and 1930s.
Afterword
This essay, developing some points I made in Ozu and the Poetics of Cinema, was
originally written in 1994. I've revised it considerably, but to take full account of all
the new work on Japanese cinema of the period would make it unwieldy. In any case,
I think that my fundamental claims still hold good. Some have been reaffirmed by
other commentators. Donald Richie, for instance, agrees that it's useful to conceive
three major stylistic traditions at work in the period, and he treats popular culture
as a more proximate influence on the films than Edo culture or other more distant
traditions. 56 Likewise, the research of Komatsu Hiroshi confirms the interplay
Visual Style in Japanese Cinema, 1925—1945 373
of Japanese filmmaking with that of the West. By 1927, Komatsu notes, "Japanese
cinema was recognized to be at the same stage as European and American cinema.
The problem of imitation and originality lies there. To achieve an innovative character,
Japanese cinema had to be confronted with foreign cinema." 57
In the wake of studies by Carol Gluck, John Dower, and others cited in the essay,
many scholars have reevaluated cultural change in the period. Most historians now
seem to agree that a surprising number of social practices considered deeply tradi-
tional were in fact recently devised in top-down fashion. An exemplary collection
is Stephen Vlastos, ed., Mirror of Modernity: Invented Traditions of Modern Japan
(Berkeley: University of California Press, 1998). See also Leslie Pincus, Authenticating
Culture in Imperial Japan: Kuki Shuzo and the Rise of a National Aesthetic (Berkeley:
University of California Press, 1996); Sheldon Garon, Molding Japanese Minds: The
State in Everyday Life (Princeton, N.J.: Princeton University Press, 1997); Sharon A.
Minichiello, ed., Japan's Competing Modernities: Issues in Culture and Democracy
1900-1930 (Honolulu: University of Hawai'i Press, 1998); Ian Buruma's Inventing
Japan: 1853-1964 (New York: Modern Library, 2003); and Barbara Molony and
Kathleen Uno, eds., Gendering Modern Japanese History (Cambridge, Mass.: Harvard
University Asian Center, 2005). In The Idea of Japan: Western Images, Western Myths
(Chicago: Ivan. R. Dee, 1996), Ian Littlewood examines how Japan has been conceived
by Western historians, artists, and thinkers.
A fascinating series of essays explores the process by which Western painting tech-
niques were absorbed by the artists and institutions of Japan, and the process reveals
a constant reworking of native traditions. See Ellen P. Conant, ed., Challenging Past
and Present: The Metamorphosis of Nineteenth-Century Japanese Art (Honolulu: Uni-
versity of Hawai'i Press, 2006).
Although I've often criticized scholars' use of modernity theory as an all-purpose
explanation for things cinematic, my Ozu book and this essay did invoke this concept
as a way to understand some subjects and themes informing the films. The approach has
been developed with regard to intellectual history by Harry Harootunian in Overcome
by Modernity: History, Culture, and Community in Interwar Japan (Princeton, N.J.:
Princeton University Press, 2000), and in History's Disquiet: Modernity, Cultural
Practice, and the Question of Everyday Life (New York: Columbia University Press,
2000). Harootunian's references to cinema are, however, minimal and general. Isolde
Standish's A New History of Japanese Cinema (New York: Continuum, 2005), is more
detailed. The book applies modernity theory in what seems to me a fairly reflectionist
way, and its discussions of film style are not particularly precise, but it offers a wealth of
concrete information about studio business practices and government policies.
English-language readers now have access to a detailed account of the benshi,
thanks to Jeffery A. Dym. 58 Dym makes it clear that the benshi sought to involve the
spectator imaginatively in the film's action and to add an expressive overlay to the
visuals. He quotes one benshi saying that the effort to impersonate characters vocally
(.kowairo setsumei) "must appear as if the force is emanating from the screen."59 Dym
also shows the extraordinary control that benshi exercised over filmmaking and film
exhibition, insisting on long takes (they were easier to synchronize) and sometimes
374 Poetics of Cinema

even reediting films tofittheir accompaniment. In addition to Dym's study, see the
informative The Benshi: Japanese Silent Film Narrators, edited by Friends of Silent

\l
Films Association (Tokyo: Urban Connections, 2001).
The essential book on politics and filmmaking during the "dark valley" is Peter B.
High, The Imperial Screen: Japanese Film Culture in the Fifteen Years' War, 1931-1945
(Madison: University of Wisconsin Press, 2003). This incorporates some of the
material mentioned in endnotes 42 and 45 of this chapter. See as well Barak Kushner,
The Thought War: Japanese Imperial Propaganda (Honolulu: University of Hawai'i
Press, 2006). John Dower followed his magnificent War Without Mercy, cited in my A Cinema of Flourishes
essay, with a study of the occupation years, Embracing Defeat: Japan in the Wake of
World War II (New York: Norton, 2000).
Several directors I've mentioned have received study since my essay, and the Decorative Style in 1920s and
authors have sometimes contributed to our understanding of form and style. Two
invaluable anthologies are Shigehiko Hasumi and Sadao Yamane, eds., Mikio Naruse
(San Sebastian, Puerto Rico: Festival Internacional del cine, 1998); and Li Cheuk-to, 1930s Japanese Film
ed., Shimizu Hiroshi (Hong Kong: Hong Kong International Film Festival, 2004). In
The Cinema of Gosho Heinosuke: Laughter Through Tears (Bloomington: Indiana
University Press, 2005), Arthur Nolletti Jr. offers detailed examinations of Gosho's
surviving films from the period. My discussion of Mizoguchi in chapter 3 of Figures
Traced in Light follows up some points broached in this essay and the next. See also
Mark Le Fanu, Mizoguchi and Japan (London: British Film Institute, 2005). Other
technical options are analyzed in Peter Rist, "Camera Movement in Japanese Silent
Films," Asian Cinema 14, no. 2 (Fall-Winter 2003): 197-205. Friends of Silent Film
Association, eds., Recalling the Treasures of Japanese Cinema (Tokyo: Urban Connec-
tions, 2003), is an entertaining survey of lost films from the era.

In one scene of Tomu Uchida's Police (1933), the plainclothes protagonist leaves his
suspect sitting at a cafe table. Uchida shows the protagonist coming toward us through
a curtained doorway (Figure 13.1). He walks out frame right. The camera now holds
on the narrow slit of the curtain (Figure 13.2). Abruptly the policeman's hand shoots
into the frame and lifts the curtain away a bit, revealing his prey still sitting at the
table (Figure 13.3).
This shot occurs in a context that is wholly comprehensible within the norms of
ordinary Western filmmaking. The film's narrative line, about a cop who discovers
that his boyhood friend is the crook whom he must bring to justice, poses no
difficulties to viewers accustomed to 1930s Warner Bros, films. Stylistically, Police
depends mostly upon Hollywood conventions of staging, framing, and editing. In
these respects, the film can stand as an emblem of mainstream Japanese cinema of the
1920s and 1930s. Most of the time, most Japanese films of this era behave like most
American and European films of the same period. This is, on the whole, a "classical"
narrative cinema.
Yet this shot seems to deviate from what we usually see. For one thing, it's oddly
gratuitous. Why use the curtain to hide and then reveal the villain? Why not simply

375
 Poetics of Cinema
376
Figure 13.1 Police (Uchida Tomu, 1933). Figure 13.2 Police.
Figure 13.3 Police.
show the hero peering through the curtain's slit? Why let him act as a sort of impre-
sario of our vision? Moreover, the shot flaunts its virtuosity: The rack focus from the
curtain to the distant villain is exactly timed to the hero's gesture of brushing open
the curtain. This is, to put it plainly, a flashy shot. In this respect, too, the shot is
emblematic, for in this classical cinema, some moments stand out by virtue of their
ostentatious technique.
At first it might seem that Japanese directors have simply taken over a wider range
of stylistic devices than Western cinema of the period has accustomed us to. Japanese
cinema is eclectic in its free inclusion of unusual techniques, such as fast-motion,
handheld shots, and whip pans. But such devices aren't unthinkable in Hollywood;
they're simply motivated in other ways. Fast motion can be justified for comedy, or to
speed up a dangerous stunt. A handheld shot can be justified as the view of a newsreel
cameraman. Whip pans can link scenes, as in The Trial ofVivienne Ware (1932). A play
with vision somewhat like that in Police might be found in a scene of murder or seduc-
tion. True, such outre devices are more common in Japanese films; but they are more
common because the functions that they fulfill are more prominent there than in most
Hollywood and European cinema. One reason to study the Japanese cinema of the
1930s is that it pushes us toward reflections on the ways in which we may describe and
analyze the functions of film style.
A Cinema of Flourishes 377
Some Functions of Style
Critics and theorists tend to assign three principal roles to film style. First, style is
assumed to channel the flow of story information, as when the choice of a framing
allows us to grasp pertinent story action. This denotative function is usually con-
sidered to be so obvious as to require no discussion, but in narrative films it is the
principal, most systematic, and most pervasive function of style, and we still haven't
sufficiently explored it.
A second role often ascribed to style is that of thematic meaning. A critic may claim
that a shot that shows a man looking into a mirror implies that he is narcissistic,
or that a shot that shows him turning from a mirror implies that he lacks self-
knowledge. The thematic function of style has been one that critics have proven very
keen to explore; it has become a central object of critical interpretation. 1
Another function of style, which film critics have fairly often noticed, is the
expressive one. Here some perceptual qualities of the stylistic device signal a feeling-
ful quality, related either to the action depicted or to a character's inner state. Japanese
films of this period, for example, often use style to convey kinesthetic qualities of
rapid physical action. In another scene of Police, a football scrimmage is filmed in a
bumpy, handheld tracking shot. An Actor's Revenge (1935) employs very fast cutting
in a scene in which the hero hurls a knife into the villain's wrist. Extreme examples
of the expressive function are to be found in the chambara, or swordfight film. Here a
remarkable range of conventions increases the sense of frenetic combat: fast motion,
rapid editing, jump cuts, handheld shots, whip pans, and overlapping shot changes.
Few films from any national cinema mobilize so many expressive techniques so
energetically, and they give the lie to notions that all Japanese cinema is tranquil and
serene. (In the period we're examining, it was surely the world's most violent cinema,
and it probably remains so.) Frenzied sword duels and yakuza gunfights in films of
recent decades carry on this tradition. Films in other genres use style expressively, too,
as when a tearful wife's calling after her husband is presented in a dialogue title that
goes out of focus (Respect for the Emperor, 1928). Granted, such effects can quickly
become cliches, but the point is that the devices were repeatable because the reigning
style reserved a canonized function for them.
Clearly, the flashy qualities of Japanese films of the late 1920s are partly traceable
to their reliance on expressive functions of film style. There is, however, one more
stylistic function that deserves attention—not least because it is most frequently over-
looked. I shall call this the ornamental or decorative function of style. Here style takes
narrative denotation or an expressive quality as an occasion for exhibiting perceptual
qualities or patterns.
Decorative conceptions of style have gotten a bad reputation in recent centuries.
Many theorists have argued that all artworks, or at least all good ones, should be
organic wholes. Style should therefore function to convey meaning and to work
harmoniously with other components. The organicist would grant that style may
function denotatively, thematically, and expressively, but would reject the possibility,
or at least the desirability, of its functioning as decoration.
378 Poetics of Cinema
Yet this position overlooks one of art's most historically important functions: the
arresting and engagement of our perceptual activity. Poetry is at least as much sound
as sense. Painting aims to arouse and alter our visual habits. Music sharpens our
discrimination of acoustic texture and structure. Art characteristically invites us
to go all the way down to the minutest features of the medium, and the decorative
function of style fulfills this function by highlighting those features in a particularly
vivid way. Initially, we might say, the decorative function of style gives a perceptual
salience to features of the medium. When a stylistic device seems not to function
denotatively, thematically, or expressively, it stands out in itself, as a device utilizing
concrete materials and processes.
Beyond this atomistic saliency, there lies the possibility that devices may be
patterned. The vividly marked device can be prolonged, developed, or varied. How can
this be sustained without relying on denotation, expressivity, or thematic meaning?
E. H. Gombrich has proposed the general hypothesis that humans possess both a
sense of meaning and a sense of order. Although he doesn't explain the distinction
in detail, I take it that our sense of order is what enables us to scan our environment,
to grasp stable patterns and regularly recurring events. Abstract art embodies this
search for order in a pure state. On the other hand, the search for meaning includes,
for Gombrich, all attempts to interpret our environment, ranging from object
recognition to complex attributions of symbolic significance. 2 At moments, Gombrich
treats the search for order and the search for meaning as working in tandem, particu-
larly for evolutionary ends. At other times, when discussing decoration, Gombrich
treats them as conflicting impulses.
Gombrich's hypothesis is strengthened by research and theorizing in cognitive
science. Our perceptual systems are adapted to providing us with a highly reliable
representation of a world of three-dimensional objects, events, and states of affairs.
All the sensory processes that yield that output are necessarily hidden from us; we
register just the results. Yet the processes are pretuned to assume unity, regularity,
symmetry, constancy, and the like—in short, order. Those assumptions may, of course,
be overridden by the nature and structure of the stimulus input; if so, they must then
be revised. But these assumptions, probably evolved in response to environmental
regularities, remain necessary points of departure for any perceptual activity.
Film style can swerve from fulfilling purely narrative functions and can solicit
attention in its own right. For my purposes here, decorative functions of style emerge
when Gombrich's "search for order" outruns the meanings that can plausibly be attrib-
uted to the particular devices and patterns we encounter. At a larger level, Gombrich s
wealth of examples shows that we need not accept limited definitions of style if our
purpose is to analyze the functions of style. If we think of a film's style as a system of
technical choices instantiated in the total form of the work, itself grasped in its relation
to pertinent and proximate stylistic norms, then we need not adopt either an organic
or an ornamental definition of style a priori. We need only say that in some cases, style
may work "organically" to convey meaning or expressive qualities, and that in other
cases, it may seem "applied," or laid over other components or structures.
A. Cinema of Flourishes 379
The metaphors of superposition in the last sentence should remind us that deco-
ration usually implies two kinds of processes. We decorate when we attach some-
thing to a free-standing support—a wall, a page, the human body. In cinema, the
analogy would make the film's narrative and its other stylistic functions serve as the
"support" for decorative uses of style. Yet there's a second way of posing the problem
too. Decoration can also be seen as a process of embellishment, the elaboration of
a simpler or plainer version. In a musical composition, ornamentation can be con-
ceived as the decoration of a simpler structure that has already been heard or that is
absent but implicit. Some musical passages seem to be heard as elaborations of more
skeletal forms that are implied by the overall organization of the piece. 3 Similarly,
in the visual and verbal arts, ornamentation takes an implicit, more or less recog-
nizable representation as its basis. What the ornament may adorn is not only some
physical support but also the result of our literal understanding. In taking an image as
distorted or stylized, we must, however tentatively, assign some referential meaning
to it. The intertwined curlicues on a wallpaper design are, after initial processing,
revealed to us as flowers' tendrils; then, by positing an earlier, minimally realistic
schema for them we can perceive them as unrealistically stretched. 4 Constructivist
research in perceptual and cognitive psychology has established the priority of object
recognition and story comprehension schemas in our experience of artworks, and it
seems likely that these manifestations of the "sense of meaning" are points of depar-
ture for recognizing embellishment. 5
This second option would seem to be a fruitful way of conceiving of decorative
style in cinema, because it brings into play tacit norms and conventions. It seems
likely that decorative impulses build upon the more stable responses supplied by
familiarity with comparatively simple elements and functions. At least in some
cases, the viewer could see style as imposing purely ornamental patterns of technique
upon quickly recognized narrative and expressive qualities. 6 The surplus in stylistic
treatment is construed as decorative elaboration of more conventional forms. Thus,
in our opening shot from Police, the tacit norm for staging and shooting the scene
would run something like this. The policeman comes toward the camera, closes the
curtain, and turns as he opens the curtain and peers into the cafe. His back blocks
our view. Cut to a subjective point-of-view shot that shows the suspect sitting at the
table, perhaps framed by masking that simulates the edges of the curtains. Uchida's
refusal of this more straightforward alternative prompts the spectator to notice the
device as such; as critics, we can attribute this to the shot's status as an elaboration of
the simpler, more conventional representation.
In sum, we can hypothesize that a stylistic device or pattern is operating decora-
tively when it appears to exceed its denotative, thematic, or expressive function. The
stylistic device or pattern can plausibly be treated as an elaboration of a simpler, more
norm-bound device or pattern that would suffice to achieve these other functions.
It seems likely that such conditions obtain in popular cinema around the world.
Hollywood filmmaking often presents intricate montage sequences, self-conscious
compositions or camera movements, technical flamboyance (as in Welles or Ophuls),
and certain fillips encouraged by genres like the comedy, the musical, or the thriller.
 Poetics of Cinema A Cinema of Flourishes 381
380

In what follows, I want to show that the Japanese films of the 1920s and 1930s seem
striking to us chiefly because they make style more often and more vividly ornamental
than most Western films. Without permanently displacing narrative structure from
its central place—without, that is, ceasing to remain "classical"—the films give greater
prominence to the decorative functions of style.

Games ^Vith Vision

The greater degree of decorative tendencies in Japanese film style is apparent in some
fairly common conventions, such as flashy transitions. Like its Hollywood counterpart,
Figure 13.4 The Twenty-Six Japanese Martyrs Figure 13.5 Little Man, Do Your Best (aka
Japanese classicism needs linking material between scenes, and in both traditions the (1936). Flunky, Work Hard!; Naruse Mikio, 1931).
film's narration is likely to become more overt during these shifts. Still, the Japanese
filmmaker is more likely to take the transition as a pretext for stylistic embroidery.
For instance, many filmmakers commonly utilize what we could call the dissolve-in-
place. Here, a time gap is conveyed by keeping the framing and locale constant while
dissolving to an action taking place there later. The dissolve-in-place can function
decoratively because it exceeds simple narrative denotation; the dissolve itself would
suffice to convey the passage of time. Moreover, the device usually doesn't cue the-
matic meanings or expressive qualities. Instead, by holding the composition constant
and letting the dissolve display all the similarities, the transition calls the spectator's
attention to the image as a graphic configuration. Like a rhyme in a narrative poem,
the dissolve-in-place imposes an abstract pattern on the telling of the story. Figure 13.6 Dr. Kinuyo (1937). Figure 13. I h. kinuyo.
This device recurs in so many Japanese films of the period that we may treat it as a
decorative cliche. Other conventional transitions include cutaway shots to landscapes
or to seasonally significant details. Sometimes transitions are charged with an expres- expressive moment with ornamental force, as Naruse does in Little Man, Do Your Best
sive or commentative role, but frequently they function simply as poncifs, set pieces to (aka Flunky, Work Hard!; 1931): The father who learns of his son's accident is treated
prolong the change of scene and to engage our interest in isolated compositions or a in an orgy of optical-printing effects, including diagonal wipes that slice up the man's
suite of connected shots. Naruse Mikio's Apart From You (1933) and Nightly Dreams face (Figure 13.5). Even the old standby, the shot/reverse-shot conversation scene, can
(1933) link scenes by having a character walk into the camera so as to block the lens, come in for decorative treatment. Take the banter between Saburi Shin and Tanaka
and then beginning the next scene with a character walking away from the camera. Kinuyo in Dr. Kinuyo (1937). Here a shot/reverse-shot passage cants each composition
Like the transition between scenes, the intertitle is a narrational staple that can
sharply (Figures 13.6-13.7). No need to posit expressivity or thematic commentary
launch decorative treatment. In Japanese films of the 1920s and early 1930s, intertitles
("Their world is out of kilter"). The canted shots decoratively dynamize a static scene,
may zoom out at the spectator or pull rapidly into focus. In the early 1930s, titles,
and incidentally camouflage the fact that we have moved from location shooting to
either dialogue or expository, are often set against live-action backgrounds. In Apart
the studio. To the objection that such a flagrant device cannot be gratuitous, that it
From You, a young woman speaks to a young man at the seashore, and her words are
must mean, I reply that stylistic decoration may function to awaken our perception,
superimposed on the crashing waves. A film's final intertitle, being a conventional
sharpen our attention, and make the scene more vivid. Japanese cinema of the period
spot for overt narration, is a prime candidate for such embellishment. The comic
teems with exactly such arbitrary choices, but they're gratuitous only by standards of
chambara Iwami Jutaro (1937) closes with a shot of the colossal swordsman looking
the organic theory of style.
down at a beach and laughing; the camera swivels down to reveal "The End" written
The decorative tendency becomes even more explicit when Japanese filmmakers
in the sand.
Almost any cliched or normalized element—generic, stylistic, or narrational— hide or teasingly block the main action. They have long enjoyed inserting geometrical
can function as the basis of ornamental novelty. A montage sequence can utilize patterns between characters and viewer (Figures 13.8-13.9), giving our attention a
extreme low angles and wide-angle lenses, as in The Twenty-Six Japanese Martyrs workout. In Shimazu Yasujiro's First Steps Ashore (1932), for instance, a bar scene is
(1930; Figure 13.4). When a subjective sequence is needed, the director can imbue the filmed through a grillwork that imposes a grid over the figures (Figures 12.28-12.29);
382 Poetics of Cinema A Cinema of Flourishes
383

Figure 13.8 Cuckoo (Ikeda Yoshinobu, 1922). Fig re 13.9 Apart From You (Naruse Mikio, Figure 13.10 Apart From You: As her son
waits by the sickbed of the woman he loves, Figure 13.11 Apart From You: She walks
the mother stands by the window. The cord of abruptly out of the shot, brushing the
cord...
the window blind bisects the shot.
we must strain to see the action, and we simultaneously appreciate the modular dif-
ferences that shifts in character position can create. Other compositions make use
of the doorways and sliding panels of the Japanese house to create squarish frames
within frames that hide and reveal figures on various planes (Figures 12.26-12.27).
Even the most slender visual element can serve as a decorative distraction, as in one
scene of Apart From You (1933; Figures 13.10-13.12). Perhaps Josef von Sternberg's
films became popular among Japanese filmmakers because they used composition,
lighting, movement, and optical transitions such as dissolves to thwart easy recogni-
tion of the action and to insist upon the shot as a total design. Here, it would seem, is
one premise underlying our shot from Police: the embellishment of a scene through a
game of vision that arouses, sharpens, and rewards the spectator's attention. Figure 13.12 . . . which continues to swing Figure 13.13 Fallen Blossoms (1938).
in the foreground after she has departed. It's
The game appears in another decorative convention of Japanese classicism: the use both a realistic detail that renders the scene
of selective focus. As a general rule, Japanese filmmakers use a great deal of depth in convincing and an abstract line that swings
staging, often with considerable depth of focus. However, 1930s directors also had a across the frame metronomically.

great tolerance for putting significant objects and characters out of focus. Sometimes
the entire shot will be diffuse to various degrees on all planes, as in the opening of
The Scarlet Bat (1931). Or one plane will be in focus while a distant one—often an impor-
tant one—is not. The effect is also evident in another cliche of the period, the rack-focus
shot (as, for instance, in Gosho Heinosuke's LAmour [1933]). Such processes create bold
geometric shapes and pulverize objects and figures into granular patches of light.
So far, all of my examples have been fairly isolated ones, in which a single shot or
pair of shots makes one technical device palpable. But the decorative function of style
calls attention to the patterning of devices as well. Editing is the principal means of
creating such patterns. Indeed, the Japanese film of any period may use editing to
impose an abstract structure on a scene. Figure 13.14 Fallen Blossoms. Figure 13.15 Fallen Blossoms.
This is evident when a conventional scene gets elaborated at the level of style. For
example, Japanese cinema of the 1930s frequently shows a character in a static pose, and
from three angles (Figures 13.13-13.15). As a swordsman in Faithful Servant Naosuke
the style will enliven the moment by what we can call "intensification" cuts: brief shots
(1939) stands poised with his blade raised, four shots rapidly bring us closer to him.
from disparate angles or distances, usually progressively closer to the figure. In Fallen
This decorative arrangement of shots is analogous to what Viktor Shklovsky
Blossoms (1938), as Akira sits thinking about her lover, we get three quick shots of her
referred to as the "geometricization" of linguistic devices in verse. 7 Through
384 Poetics of Cinema
Figure 13.16 Nightly Dreams (1933). Figure 13.17 Nightly Dreams.
short-range repetition, any device may become geometricized. Japanese directors
are fond of cutting together short, sharp camera movements in simple patterns. In
Kinugasa Teinosuke's An Actor's Revenge, as a tough actress mocks the protagonist,
we get three brief shots, each one a quick pan from her to him. A similar triple
pattern of camera movements occurs in Police and Lily of the Valley (1935). Gosho
Heinosuke uses an A-B-A pat terning in Burden of Life (1935): A scene of a mother
and father's conversation is repeatedly interrupted by cutaway track-ins to the son
in bed. An action may be framed by symmetrical camera movements, as is the Bat's
interrogation of an old man in The Scarlet Bat.
Once such tactics have become conventional, however, the filmmaker in the grip of
the decorative impulse must go beyond them. A device that has become straightfor-
ward and predictable loses its ornamental force. There is thus a tendency for directors
to search out less codified devices that can achieve saliency. In Iwami Jutaro, a scene
begins with a close-up of a small hanging sign that reads, "Put out fires," which is
then yanked up out of the frame to reveal a woman in her steaming bath. Titles can
also be manipulated in fresh ways. Three Beauties (1935) uses a dialogue title that flips
over to reveal another title underneath. In Mito Komon Part II (1932), a shot of the
hero eavesdropping is invaded by a superimposed title reporting what the villains are
saying; later in the same film, the camera pans rapidly from a speaking swordsman
to . . . a dialogue title.
If the stylistic norms of a period encourage ornamentation, directors will not only
search for idiosyncratic devices but also prolong and emphasize commonplaces in
the hope of restoring their force. Naruse Mikio's work of the period offers several
striking examples of this process. Naruse will stress the interplay between sharp and
unfocused planes, as in the climactic scene of Nightly Dreams. The spectator is urged
not only to notice the switch of foreground and background elements from shot to
shot but also to appreciate the way a figure jumps from clear outline to indistinctness
(Figures 13.16-13.17). Naruse will also employ marked accentual cutting, as in the
deathbed scene in Street Without End (1934; see Figures 12.52-12.57). He will make
insistent use of symmetrical camera movements too. In Nightly Dreams, Apart From
You, and Wife, Be Like a Rose! (1935), a rapid track-in to a character in close-up will
A Cinema of Flourishes 385
Figure 13.18 An Actor's (Kinugasa Figure 13.19 An Actor's Revenge.
Teinosuke, 1935).
be followed by several more such track-ins or will alternate with equally paced track-
backs, so that what starts as an expressive heightening of the action becomes, via
immediate repetition, a purely rhythmic pattern—like the zoom-ins and -outs during
fight scenes in Hong Kong martial arts movies of the 1970s.
Further evidence of this wide-ranging search to refresh decorative commonplaces
may be the tendency toward stylistic flourishes that we find in these films. In decora-
tive art, the term flourish retains two of its original senses: ostentatious embellishment
and showy movement. The florid stroke of penmanship is the trace of the gesture that
produced it. Gombrich treats the flourish as "the expression of the joyful exuberance
of the craftsman who displayed both his skill and his inventiveness." 8 The flourish is
thus the epitome of the perceptual and ludic possibilities of ornament:
The flourish is easily understood as a playful product, a paradigm of the relation
between sign and design. Even where it enters into a symbiosis with the sign,
serving as a means of emphasis or enhancement, it never quite surrenders its
freedom from the constraints of signification. 9
Japanese films of this period, I suggest, are full of flourishes. When, in An Actor's
Revenge, guards who are searching for a thief move aside to reveal the thief far above
their heads, leaping from roof to roof (Figures 13.18-13.19); when, in any number of
chambara swordfights, a victim is hurled out at the camera, blocking our view, before
falling aside to reveal the hero in another spot, fighting a new adversary; when, in
Mito Komon Part II, a fight inside a room reaches a climax and a swordsman spins
around and slashes his way through a wall, straight out at us (Figures 13.20-13.21);
when, in Little Man, Do Your Best, a montage sequence mixes negative images, blank
frames, and fireworks into a swirl of very short shots—when, in sum, the decorative
impulse of this filmmaking tradition encourages the filmmaker to display his assured
mastery over the medium, we are in the presence of that "pleasure in control" that
Gombrich attributes to the flourish and that he considers central to decorative art.
In all the instances I've mentioned, the flourish prolongs the game with vision
that we've already observed in the abstract patterning of shot composition. A sword-
 Poetics of Cinema A Cinema of Flourishes 387
386

Figure 13.21 Mito Komon Part II. Figure 13.26 Komatsu Riyuzo Part II: After
the camera has panned left to follow the
fight, the hero edges back to the front of the
wagon, his back to us. His position conceals
his enemy.
Figure 13.27 Komatsu Riyuzo Part II: His
enemy makes the fatal thrust and stands
revealed as the hero falls back into the wagon,

Figure 13.23 Komatsu Riyuzo Part II: Cut to

Figure 13.22 Komatsu Riyuzo Part II (1932).
a low-angle view with the wagon in the fore-
One of the heroes is shown near a wagon as
ground. As one man charges, the hero dives
his enemies approach him.
down out of sight. Figure 13.28 Komatsu Riyuzo Part II: The Figure 13.29 Komatsu Riyuzo Part II: The
wagon bed flips up one last time, now regis- camera tilts down to show the hero's sword
tering the hero's collapse. dropping as he dies.

in our face, its aggressive rhythm amplifying the action in an indirect but vivid way.
Like the climax of Sanshiro Sugata, which conceals the combat behind tall grass
(p. 365), this shot ingeniously embellishes a conventional scene. Naruse's mid-1930s
films exploit the same intermittent masking of sight by foreground objects that shear
across the frame, erasing or revealing the main action. The stylistic flourish thus
becomes one mark of self-conscious narration. 10 And given the likelihood of an
Figure 13.25 Komatsu Riyuzo Part II: The
Figure 13.24 Komatsu Riyuzo Part II: His artistic arms race, filmmakers will strive to outdo each other in finding ingenious
wagon bed suddenly drops to reveal one of
action tips the wagon u p into the extreme ways to fulfill this function.
the men pursuing the hero off left.
foreground, blocking our view of the fight.
We can, then, best consider the flourish as an unlikely decorative choice that
displays a degree of technical virtuosity and that pushes toward a narrational self-
consciousness. Our opening shot from Police succeeds on all these counts. The more
fight in Komatsu Riyuzo Part II (1932) is handled in a lengthy single shot that keeps a
straightforward technical option is avoided in favor of an unusual one. The shot
wagon bed in the foreground, seesawing up and down, alternately blocking and free-
requires a split-second control over timing and focus. And the hand that peels back
ing our vision (Figures 13.22-13.29). By holding on one camera position and letting
the curtain acknowledges that this narration addresses itself to the viewer. The shot is
the wagon impede our view, the director sacrifices a clear rendition of the combat.
made for us, and it declares as much. We have perhaps come full circle: Now style, in
But in exchange he supplies the dynamic thrust of the wagon heaving up and down
388 Poetics of Cinema
the furthest reaches of decorative elaboration, affects narrative denotation, governing
rather directly how we construct and construe the simplest story action.
The narrational self-consciousness of the flourish makes it ideal for comic scenes,
but it's significant that most of my examples here come from dramas. In any genre,
the decorative impulse yields a gamelike approach to art, an acceptance of norms as
occasions for engaging display. Treating art ludically also asks the spectator to be a
sort of connoisseur—knowing the conventions, demanding to be dazzled, applaud-
ing the artist's virtuosity. In this sense, even the most serious scene can furnish the
pretext for a playful handling of style.
Even at its most flamboyant, however, the flourish needs stable stylistic norms
to set it off. The discontinuously cut swordfight relies on spatial continuity in the
scenes around it. Opaque long shots alternate with nearer, more legible framings.
Out-of-focus planes will be crisp in adjacent shots. The grid laid across the action
(Figure 13.8-13.9, above) will keep the figures within the central zones of the shot,
whereas the curtained slit of Police squeezes the salient information into the vertical
axis of the frame (Figures 13.2-13.3, above). Not only does the ornament revise an
implied, simpler device, but it also requires a subdued context to set it off. Again,
decoration presupposes denotation.
A Tradition of Ornamentation
Considering film style's ornamental function will affect the ways in which we do film
theory, criticism, and history. By making distinctions among various functions of
film style, a poetics of cinema can help us identify crucial features of the medium
and its culturally varying conventions. In particular, the decorative function of
style points to certain aspects of part-whole relations that theorists have neglected.
For instance, the Police passage that I cited at the outset stands out partly because
there is nothing else like it in the film. Neither the curtain device, nor the strategy
of intermittently blocked vision, nor the quality of self-conscious narration achieved
through implausible character gesture recurs or develops across the film. We have
only a momentary flourish. The same holds true, at least as far as I can tell, for the
other sequences I have mentioned in this essay. Yet if we pursue the sense of order
across an entire film's stylistic patterning, we may find large-scale systems that are
not easily subsumed to the sense of meaning. Some narrative films exhibit what I have
elsewhere called parametric narration.
Decorative passages tend to detach stylistic patterning from the films ongoing
representation of a fabula, the inferred causal-chronological chain of the action. 11 Like
decorative flourishes, parametric systems are motivated artistically. They call attention
to the materials and forms of the medium. But unlike a flourish, parametric narration
is organized across the whole film according to distinct and developing principles. 12
When the ranking of perceptual order over meaning is pursued consistently, accord-
ing to an additive logic of variation, we have parametric narration. When the ranking
is more occasional or eclectic, we have a narration that's occasionally embellished by
isolated moments of aesthetic motivation.
A Cinema of Flourishes 389
The concept of a decorative use of style can also assist us in doing criticism. When
we analyze an individual film, we no longer need to make every salient stylistic event
thematically or expressively significant. The foregoing pages should illustrate how the
critic might discuss decorative passages. But we want also to talk about filmmakers
whose works bear a more complicated relation to the norms of their peers and period.
Let me sketch how this might go by taking as an instance the work of one of the most
significant filmmakers of the era.
In my book Ozu and the Poetics of Cinema, I argue that Ozu Yasujiro's 1930s films
can be characterized by three simultaneous tendencies. First, there is an adherence to
classical filmmaking precepts, as seen in both Hollywood and Japan. Ozu's films rely
upon basic units of narrative composition (scene, summary, and transition), and they
draw upon broad norms of stylistic construction (e.g., eyeline matching, matches on
action, and centered framing). They also presume that the spectator will understand
narrative cinema with the aid of certain pragmatic assumptions about real-world
events, such as the likelihood that people talking to each other normally face one
another. Besides this classical tendency, Ozu also displays a recognition of decorative
elaboration as conceived by his colleagues. For instance, Ozu's contemporaries mined
the narrationally self-conscious transition and the cutaway to items of decor or land-
scape, and Ozu was obviously sensitive to such possibilities. But here enters the third
tendency. Ozu goes beyond his peers in organizing the decorative possibilities of film
style into a unique parametric system. He restricts his devices to a smaller set than do
other directors; he treats that set as possessing a range of minute, sometimes barely
noticeable differences; and he spreads that range of differences across the film as a
whole, developing them permutationally in complex interaction with the syuzhet, the
film's presentation of the unfolding narrative action. One can study at some length
how Ozu's parametric narration obeys its own principles of staging, composition, and
cutting, coaxing the spectator into a playful but rigorous experience. 13 It would be
hard to call the dynamically developing nuances of Ozu's style "flourishes" in my
sense here. He has, as it were, planed his style down to a level where it is constantly,
systematically, subtly decorative.
If we're justified in speaking of Japan's version of classical filmic storytelling, we
must also search for the specific historical factors that have allowed decorative aspects
of style to gain more prominence than in Western, and particularly Hollywood,
filmmaking. Further historical inquiry would, I think, reveal that these decorative
emphases are encouraged by many Japanese aesthetic traditions.
Japanese poetics and painting have tended to treat art as a matter of technical skill.
In many pre-Meiji arts, the cliche holds an honored place: It is a fixed counter in a
formal game, to be deployed in new patterns. Konishi Jin'ichi calls this the tradi-
tion of "inorganic" unity, as exemplified by the patterning of images in renga (linked)
verse or by the kata, the decomposed movements of the Noh play.14 It would be
misleading, therefore, to say that Japanese artistic traditions invariably scorn novelty
or originality. Art becomes a game of pattern making within fixed rules that stipulate
how the medium is to be handled, and an artist wins praise by submitting the rules to
390 Poetics of Cinema A. Cinema of Flourishes 391

fresh and virtuoso treatment. For the same reason, the decorative impulse in Japanese
music and visual arts would be a plausible forerunner of this attitude toward cinema.
In the previous essay, though, I argued that the cinematic style of the period has
more complicated sources. First, there are several distinct Japanese aesthetic tradi-
tions, undergirded by significantly varying assumptions about the nature and roles of
art. (Indeed, the tradition associated with the Heian period, 794-1155, is at least partly
a construct of later centuries.) Second, films from Japan of this period are largely
comprehensible according to Western protocols of viewing. Directors' embrace of
Western-style continuity can't be attributed to age-old traditions. In general, it seems
very likely that in many arts, older formal practices have been mediated through
Figure 13.30 Morning for the Osone Family Figure 13.31 Morning for the Osone Family:
post-Meiji modernizations (not least the popular urban culture of this century's (Kinoshita Keisuke, 1946): After her uncle As her mother enters, the camera arcs right-
first decades). By the time such features show up in films, they are not unreflectingly has forced her to break off her engagement, ward to pick her up.
assimilated but are knowingly cited to achieve particular formal ends. 15 Yuko is shown reading a letter from her
fiance, Akira.
To explain the ornamental function of style in the Japanese filmmaking tradi-
tion, I submit, it's not enough to involve distant traditions. The analyst should also
recognize that the introduction of cinema participates in broader twentieth-century
trends toward the absorption of Western cultural practices. Recall the adaptation of a
compromise Westernized scale for the nation's music, and in Japanese visual arts the
mixing of European graphic conventions with more traditional ones, from Hiroshige's
use of wide-angle perspective to 1920s avant-gardes. Most important, perhaps, is the
Meiji-era encounter with the long and comparatively unified narrative forms of the
West. Historians of Japanese literature generally agree that lyric poetry, not drama or
epic, lay at the center of pre-Meiji poetics. One consequence was an emphasis on short
forms and on episodic construction in longer ones.16 Renga verse was composed by
creating local connections between relatively independent stanzas. Again, because of Figure 13.32 Morning for the Osone Family.
Yuko's hand emerges to show her the letter, in
the centrality of the lyric, prose narrative tended to grow out of glosses on poems or to which Akira declares he won't make her wait
become a foil for verse outbursts. It's hardly surprising, then, that the flood of foreign while he's at the front. As so often happens,
translations in the nineteenth and early twentieth centuries offered Japanese writers the decorative approach treats the drama as
an occasion for pattern making while muting
new compositional options, but also some difficulties. One of the most frequent results
the expression of strong emotion (here, hid-
was the synthesis of quasi-European plot structure with "poetic" interludes, as in the ing Yuko).
shi-shosetsu ("I-novel") or in the works of Shimazaki Toson and Tanizaki Junichiro. 17
The possibilities of synthesis were posed even more acutely for filmmakers. It's clear
that by 1925, the norms of Western filmmaking had shaped the narrative construc-
tion of Japanese films.18 Tanizaki's script for Amateur Club (1920) and the Lubitsch- Most Beautiful (1944) exemplifies the decorative tendency. And certainly films like
inspired Woman Who Touched the Leg (1926) of Yutaka Abe are only the most Morning for the Osone Family (1946) and Between War and Peace (1947), along with
illustrious instances of how "modernizing" the cinema meant adapting Hollywood the postwar work of Ozu, Mizoguchi Kenji, and other directors, continue to embellish
conventions. In accepting Hollywood norms, Japanese filmmakers largely relegated canonical elements of Japanese classicism (Figures 13.30-13.32).
decorative treatments to inserted or hyperstylized moments—themselves marked as In the 1950s, one could argue that the arrival of widescreen cinema made the deco-
"Japanese" touches. Filmmakers could consciously draw on the conventions of haiku rative impulse, if anything, more pronounced. The Robe played Japan in December
or kodan to punctuate a straightforward plot with the sort of lyrical interludes already 1953, followed by many more titles, but local filmmakers had to wait several years
common in artistic fiction. 19 to try anamorphic widescreen themselves; the first domestic anamorphic release
Once in place, the ornamental impulse in cinema remained fairly strong after didn't appear until 1957.20 TohoScope, DaieiScope, and other formats sprang up, and
the 1930s. Although a certain stylistic severity can be seen in films of the Pacific soon most productions were anamorphic. Clearly all of the options I've sketched in
War period, the wartime work of Kurosawa Akira in Sanshiro Sugata (1943) or The Chapter 10 appealed to Japanese directors. Many sought to give widescreen films
392 Poetics of Cinema A Cinema of Flourishes 393

Figure 13.33 The final, gun fight in Suzuki
Seijun's Underworld Beauty (.1958) displays
a typically abstract use of the anamorphic
format.
Japanese filmmaking of the era before 1950 is one of the richest cinematic legacies we
possess. The variety, quality, and sheer emotional power of the films are paralleled, in
my experience, only by those of the Hollywood studio era. My probes into the area surely
need more nuancing and revision. But I think that we have good reason to work under
the assumption that most Japanese cinema of the 1920s and 1930s is neither a radical
signifying practice nor a direct product of ancient Japanese traditions. It is a classical
Figure 13.34 Kato Tai's persistent reliance
cinema, though perhaps the most vivid and vivacious classicism we've yet discovered.
on low-angle framings and wide-angle lenses
helped him create the crowded establishing
shots characteristic of Japanese cinema (Red
Peony Gambler: Flower Cards Match, 1969).
Compare Figure 12.73.

1 1 A A
V :
«n

Figure 13.35 A long-lens establishing shot J n

from Sanjuro (1962), complete with grid.
Compare Figures 12.21, 12.28-12.29, 13.8,
and 13.9.
Figure 13.36 Dangan Runner (Tanaka
Hiroyuki, aka Sabu, 1996): The camera arcs
around theyakuza in the car, but whereas most:
directors would reframe to show both passen-
ger and driver fully, here we watch only one of
the driver's eyes as he talks with his partner.

deep-focus compositions and assertive close-ups in the Hollywood manner. They

were also, as we might expect, drawn to schemas yielding distant depth and abstract
compositions. The splayed geometry of the 2.35:1 frame called forth compositions
more daring and somewhat more unmotivated denotatively and expressively than
what was typical in American CinemaScope efforts (Figures 13.33-13.34).
The decorative results arose partly because the Japanese anamorphic processes had
a significantly different optical basis than the U.S. ones. Filmmakers placed the ana-
morphosing lens behind, not in front of, the prime lens, which allowed more choices
about what lenses could be used. Decisive as well was the Japanese tendency to choose
as a prime lens a very powerful zoom. 21 Because the zoom could provide extremes of
focal length, directors could create striking wide-angle and telephoto compositions.
Kurosawa is the prime instance. He relied more than most directors on long lenses
before he took up the anamorphic format in The Hidden Fortress (1958), and in the
films that followed he adapted their flattening effect to the geometrical conventions of
Japanese shot composition. The recurring gag in Sanjuro (1962) is that the plot features
many brothers not because of dramatic demands but because the ensemble can neatly
fill the wide screen (Figure 13.35). Today, even though anamorphic ratios are less com-
mon, one can still find hints of the 1930s decorative aesthetic in the widescreen format,
particularly in shots that play hide-and-seek with characters' faces (Figure 13.36).
'

H.

Aesthetics in Action

Kung-Fu, Gunplay, and

Cinematic Expression

Late in the 1980s, scarcely before I knew it, many of the Hong Kong action films
I enjoyed, along with a great many more completely unknown to me, became the
object of admiration among festival programmers, baby boomers, and graduate
students living on ramen noodles. By the mid-1990s, the genre had become a fixture
of American youth culture. Our campus video store stocked hundreds of Hong Kong
tapes, and one of my students who clerked there told me, "John Woo is God."
In trying to discuss such movies, I may well label myself as another gwailo seeing
only the most obviously exotic and outlandish side of the territory's cinema. So let
me mention that I deeply enjoy Hong Kong comedy and drama as well. Still, there's
no denying that the action genre is what first attracted me to Hong Kong film. My
fondness goes back to my own twentysomething days, specifically to a double bill of
Fist of Fury and Five Fingers of Death/King Boxer in the Majestic Theatre, Madison,
Wisconsin, in 1974. But although I'm a loyal fan, my knowledge is a long way less than
encyclopedic. There are people who can talk for hours about the niceties of Centipede
Horror (1982) and Mad Mad Ghost (1992), but I'm not one of them. Worse, I know
no Chinese. What I do hope to offer is a view of some artistic strategies of Hong
Kong cinema that largely go undiscussed, even by those Chinese critics whose work
is available in languages I read. This essay, then, should be taken as an interested

395
396 Poetics of Cinema
outsider's attempt to broach the possibility of stylistic analysis and history of Hong
Kong cinema, taking the most outre (translation: over-the-top) forms of that cinema
as a point of departure.
This purpose sets my project apart from the celebratory appreciations provided by
the connoisseurs and the fanboys of my culture. In 1996 a major American publisher
issued Sex and Zen and a Bullet in the Head, one of the first in a string of charming,
flippant guides to what these writers consider the most headbanging, let-it-rip film-
making on the planet. "Gone are the flying pigtails and contrived fist-thuds of your
father's favorite chopsockies," the blurb on the back tells us. S&Z&BH is an entertain-
ing read, and the peppy plot synopses play up the films as seedy, sexy, weird, bloody,
and nutty. But the authors' introduction warns that "film school polemics," dosed
with "pointy-headed, white-wine-and-baked-brie philosophizing," cannot adequately
describe the "scalding propulsion" of these movies. 1
These barbs strike me like a flurry of ninja throwing stars. I'm old enough to be
the father of the young fans, and I've long loved kung fu movies. Worse, I'm a film
studies professor. I don't drink white wine, and I get baked brie only at faculty parties,
but I do spend time trying to figure out what makes movies work. So, in my pointy-
headed way I want to ask other questions. Why do Hong Kong action movies trigger
such unbridled passions? How are they put together? How do they exploit the film
medium? What is the craft behind them? After you walk out of the best Hong Kong
action movies, you are charged up. You feel that you can do anything. How can mere
movies create such feelings?
This essay explores some answers to these questions. Our itinerary will take us
through contemporary Hollywood, with side trips to some technicalities of film
style and some detours into the writings of that earlier fan of Asian action, Sergei
Eisenstein. I offer these ideas tentatively, as hypotheses to be refined or discarded. But
I'll try to show that if we want to understand these movies as well as enjoy them, it
turns out that film school polemics can actually help.2
Hollywood Action: The 1980s and After
During the 1980s, the action picture—the policier or crime thriller, the adventure
film, the war picture, the science fiction saga, the movie of chase and combat—
became one of the dominant Hollywood genres. Often such films won places onto the
top-10 grossers list, and the home video boom assured an almost endless stream of
less prestigious product. Action pictures have become a central Hollywood genre at
all scales of production.
The action movies of the last 15 years built upon stylistic strategies and tactics that
had flourished in the decades before. The mold was created, I think, in the period
from Goldfinger (1964) and succeeding James Bond titles, Bonnie and Clyde (1967),
Bullitt (1968), The Wild Bunch (1969), Butch Cassidy and the Sundance Kid (1969), and
The French Connection (1971) to Jaws (1975), The Wind and the Lion (1975), and the
somewhat overlooked Thief (1981). By the time the action revival was firmly in place,
directors could draw upon a batch of conventional techniques.
Aesthetics in Action 397
Consider an example from Lethal Weapon (1987) by Richard Donner, a highly
successful action director (rumored to be an aficionado of Hong Kong film). In the
climactic chase of the film, Mel Gibson has escaped from Gary Busey's torture chamber
and now pursues him through the night streets of Tos Angeles. Busey seizes a passing
car, firing his automatic rifle into the crowds, and Gibson sprints after him through
traffic. From an overpass, Gibson shoots Busey's car. The engine bursts into flames,
and Busey smashes into a telephone pole. As Gibson approaches, Busey commandeers
another car just as Gibson is knocked flat by a passing cab.
Donner handles this situation in ways typical of Hollywood action sequences. He
aims at a maximal sense of movement, with action flowing through every image;
often no movement will be completed within a shot. The cutting is rapid, creating
an average shot length of two and a half seconds. Few would make an action picture
in long takes, of course, but it's worth noting that the cutting may sometimes be too
rapid. Donner, like most of his Hollywood peers, prefers to cut after a movement has
begun but before it's completed, even when the second shot shows an entirely different
action. For instance, when Gibson leaves Danny Glover behind at a lamppost, Glover
starts out of the shot but Donner does not give him what editors call a "clean" exit;
we are on to the next shot while Glover is in the middle of his action. Moreover, two
thirds of Donner's shots contain some camera movement. And Gibson is virtually
never in repose. Except when he stops to fire at Busey's getaway car, he races down
streets and across lanes of traffic, even leaping up onto cars parked by the curb.
This scene seems to me characteristic of the Hollywood approach in trying to
produce an overwhelming but sketchy impression of physical activity. There is,
for instance, no effort to dramatize the fact that Gibson's run through traffic is
dangerous. Although Gibson races through lots of streets, for instance, he never
comes close to any moving vehicle. Donner's long lenses, in lessening distances
between Gibson and the surrounding traffic, merely suggest that he might eventu-
ally be in danger (Figure 14.1). There is a moment early on when it seems that a car
might strike him from behind, but Donner cuts away to Busey, and when we return
to Gibson later, the potentially threatening car has disappeared. At the scene's
climax Gibson is struck by a car, but now Donner has recourse to the common
technique that Pudovkin called "constructive editing." 3 No overall view establishes
that Gibson is at risk. Instead we get 10 quick shots of slowing taxicab wheels, the
driver's startled face, and Gibson (actually a stunt man) already rolling around on
the hood of the cab (Figure 14.2). The moment when the car actually hits the Gibson
character is omitted.
Hong Kong fans rightly object to such bogus action, but we can generalize from
their annoyance. The sequence gives us the idea; we understand that the car strikes
Gibson. But the collision isn't directly delineated, and so the scene has no wallop. In
substituting an editing stratagem for a more direct presentation, this moment exempli-
fies a broader Hollywood strategy. Put generally, the actor's performance is minimized
and other cinematic techniques compensate for that. The rapid cutting, the constant
camera movement, and dramatic music and sound effects must labor to generate an
excitement that is not primed by the concrete event taking place before the lens.
398 Poetics of Cinema
Figure 14.1 Lethal Weapon (1987): the chase Figure 14.2 Lethal Weapon: Mel Gibson
through traffic. rolling away from us on the hood of the taxi
that felled him, at a safe distance from us.
Stunts and special effects likewise deemphasize the actor's performance. Donner
gives pride of place to the sparks struck by bullets and to a string of auto stunts. Busey
skids around corners, drives the wrong way on a one-way street, and steers a flaming
car into a telephone pole. Donner renders these actions in medium-long shots, long
shots, and even extreme long shots throughout the sequence. This assures him that
the action is broadly covered, but it sets the action, however spectacular, at a consider-
able distance from us. Donner offers us few close-ups of characters, giving Busey and
Gibson one quick reaction shot apiece; most of the close-ups show pieces of the car
shattered or pockmarked. Moreover, the line of movement is constantly broken by
impediments (parked cars that Gibson must hop over) and countermovements (traffic
passing to and fro).
So the sequence is physically unspecific in two respects. First, the mechanics of
the individual actions—a burst of gunfire, a tricky maneuver during a pursuit—are
sacrificed to a broad sense that something dynamic is going on. Second, instead of
conveying a specific expressive quality—Gibson's risky but tenacious pursuit, Busey's
desperation in the face of impending capture—the busyness of the style creates a
relatively undifferentiated visceral arousal. We understand that Gibson is in danger,
but we don't have any reason to sense it. The result is a diffuse feeling of excitement.
Admittedly, this loose style can be considered somewhat realistic, in that such a
chase in real-life traffic would create graceless, helter-skelter movement. And even
the interruptive cutting and occasional jerky camera movement can be justified as
plausibly reflecting the messiness of extreme action in the world we know. But there
is a cost as well. Michael Bay, director of The Rock (1996), explains how he shot inserts
for the chase through San Francisco:
I film actors driving in these scenes from a dolly a few feet in front of a station-
ary car. I do whip pans and whip zooms and violently shake the camera, trying
to make the whole screen rumble. Used in snippets, it looks as if the actors are
driving ferociously. 4
The director seeks to suggest the ferocity of the action chiefly through camerawork.
But when only the screen rumbles, we carry away just an impression of a violent chase.
Aesthetics in Action 399
And the actor has almost nothing to do, because the cinematography is carrying the
emotional jolt of the movements.
During the 1980s, there emerged two broad approaches to creating this loose sense
of incessant movement. The nervous, blurry approach typified by Lethal Weapon can
be found as well in the Beverly Hills Cop series, Die Hard 2 (1990), and Heat (1995).
The increasing jerkiness of this style can be traced in the car chases in the three Lethal
Weapon films, reaching a new high in The Rock's chase, with its 1.2-second average
shot length and the spasmodic camerawork of which Bay is so proud. Here the aim is
to create the impression that everything is in hyperkinetic motion. People and vehicles
bob and weave, the handheld camera wobbles through each shot, and cuts fracture
every instant of action into a barrage of brief, sometimes barely legible images.
Admittedly, a more poised and stable approach also emerged during the 1980s.
It relies on wide-angle compositions, somewhat more legible vectors of movement,
and more complex staging and shot design, with much rack focusing and foreground-
background interplay. This style is in evidence in the Indiana Jones series, as well as
in Aliens (1986), the original Die Hard (1988), The Fugitive (1993), and Speed (1994).
In both approaches, however, directors remain committed to the belief that an action
scene must move, and move incessantly. All other things being equal, continual
activity is the goal. This overall strategy has influenced action pictures around the
world, including those made in Hong Kong. But there are significant differences, and
these are, I submit, one source of the intuitive sense that something original is going
on in the Hong Kong action cinema.
The Mechanics of Movement
In Law Man's cop thriller Hearty Response (1986), police officer Chow Yun-Fat breaks
into a tattoo parlor. There he finds that Joey Wang has been tattooed and raped by her
former boyfriend. The tattooist leaps out the window and Chow gives chase, followed
quickly by Joey, who has grabbed a gun. As Chow searches the streets, the tattooist
tries to run him down by smashing a car through a fence. But Joey catches up with her
assailant and fires, wounding him just before she is struck by a car. Still, she runs on,
chasing the rapist into a nightclub. When Chow discovers the two, a gunfight ensues.
Chow blocks a bullet meant for Joey, and she then advances on the tattooist and,
sobbing with nervous rage, fires round after round into his body.
This climactic sequence is virtually identical in length to the Lethal Weapon chase,
but it offers instructive stylistic contrasts. A serious Hong Kong fan will note imme-
diately Law Man's much more vivid handling of the car accident. After wounding her
rapist with her first bullet, Joey starts to run across the street toward him. Cut to a
shot taken from the front seat of a car bearing down on her (Figure 14.3). The image
creates an immediate and concrete sense of danger that Donner's shots of tires and
a driver's expression can't summon up: We see that she and the car will inevitably
collide. Cut to a reverse angle, the car approaching the camera and going into a skid
as she continues across the street, now moving left to right. The car sideswipes her and
knocks her out of the frame (Figure 14.4). In a still closer view, but from the opposite
 Poet ics of Cinema
400
Figure 14.3 Hearty Response (1986): The car Figure 14.4 Hearty Response-.lhe car strikes
bears down on Joey Wang, who is running her, hurling her toward the camera.
into its path.
Figure 14.5 Hearty Response: The wide angle
exaggerates the traffic pinning Joey in place.
Compare Figure 14.1.
angle again, she hits the sidewalk and rolls bumpily rightward across the pavement.
Cut to a tight, brief shot of the car skidding and smashing into another car. The last
shot shows her still rolling, now leftward, before she slaps down her hands to stop
herself. She pauses briefly to gather strength before rising to pursue her quarry.
Of course, the Hearty Response action was faked in the sense that a stunt man (in a
wig and miniskirt) took the impact and the fall. Nonetheless, the kinetic impact of
the action is much stronger than in Lethal Weapon. This is because we more directly
register the danger of the situation and the sickening thud of the car's blow. The
controlled execution of the act was made central to the mise-en-scene, and other film
techniques were devoted to presenting that vividly. Each shot has the schematic clarity
of a cartoon panel, and the lighting renders everything in bold relief. No telephoto
lens creates a safe illusion of risk (Figure 14.5); no elliptical editing covers imprecise
gestures. The car crash is very modest by Hollywood standards, and it takes up only
one shot, as if it's inserted simply to show us the consequence of that element of the
situation. (Also, the production probably couldn't afford to cover the crash with two
cameras.) In five shots adding up to 6 seconds, a range of film techniques has put the
emphasis squarely on the sheer physical and emotional impact of the action.
Naturally, we must remember that U.S. censorship was probably more squeamish
about showing a bociy ricocheting off a car than the Hong Kong censor was. And
certainly the low budget of Hearty Response forces Law Man to create his thrills
cheaply, with small-scale stunts and people risking injury. But these factors reveal
only that Law, like other directors, turned opportunities and constraints in his
Aesthetics in Action 401
filmmaking situation to aesthetic advantage. Moreover, we might expect a low-budget
production to have even more recourse to shortcuts like Pudovkin-style construc-
tive editing. Here the big budget picture, eager to protect its star and its audience's
sensibilities, embraces the cheaper and more roundabout solution, whereas the
exploitation picture gives us the event more forthrightly.
Much the same could be said of the tattooist's attack on Chow Yun-Fat, a 9-second
barrage of trim medium-long shots and close-ups in which Chow dodges a car and
then saves a boy from being run down. As in the collision with Joey Wang, Chow's
split-second avoidance of the car relies on artifice, but not the sort that, as Andre
Bazin puts it, turns an actual event into something imaginary. 5 Someone, if not Chow
the actor, very nearly gets smashed.
My point is not that Hong Kong films employ death-defying stunts; that's not news.
What's important is that the stunts are staged and shot and cut for crisp readability.
In the auto's lunges at Chow, the rapidity of our uptake starts from the smoothness
and cogency of the physical action that is shown. The entire Hearty Response sequence
is even more quickly cut than the passage from Lethal Weapon, with an average shot
length of about 1.5 seconds; but it doesn't feel as jerky. This is because the shots tend
to be readable at a glance. Close, simply composed, each image displays only one or
two trajectories of movement. And significantly, Law Man uses half as many moving
camera shots as Donner does, thereby making it all the more important for the action
in front of the lens to engage our attention.
The lesson of this comparison is quite general. If Hollywood movies sketch a perva-
sive but often inexact sense of physical action, the Hong Kong norm aims to maximize
the action's legibility. From the 1960s swordplay films (wuxia pian) and 1970s kung
fu movies to the cop movies and revitalized wuxia pian of the 1980s, this filmmaking
tradition has put the graceful body at the center of its mise-en-scene. In order to follow
the plot, one must be constantly apprised of the actor's behavior, down to minute
changes of posture, stance, or regard. Like Soviet films of the 1920s, which took their
inspiration from the precise gymnastics of popular theater, Hong Kong cinema has
emphasized the concreteness and clarity of each gesture. Doubtless, traditions of
martial arts and Beijing Opera—cultural factors quite different from those governing
Hollywood style—have been central to this aesthetic, but here I want to concentrate
on how the filmmakers' cinematic choices create distinctive action sequences.
How does a director achieve gestural clarity on film? Several important tactics are
on view in your father's prototypical kung fu film of the 1970s and early 1980s.6 Most
obviously, the director must provide an unobstructed view of the action. The bare stretch
of earth that provides the arena for so many kung fu duels, though often the sign of a
skimpy budget, has the virtue of making salient every instant of the fight. By contrast to
Michael Bay's efforts to make his screen rumble, Hong Kong cinematography seeks not
to hide the action. The baseline is the long shot or medium-long shot, with closer views
serving to enlarge details or offer breathing space between stretches of combat.
At times, of course, difficult or impossible physical feats may be faked by means of
the Kuleshov effect, as when cutaways cover tremendous leaps or dextrous juggling.
But again the presentation of these feats will be diagrammatically clear. To take a
 Poetics of Cinema
402
Figure 14.10 The Killer.
typical instance: In Chor Yuen's 1972 film The Killer (aka Sacred Knives of Vengeance),
some thugs invade the hero's room. He is concealed in the rafters watching them. A
low-angle shot shows him leaping down (Figure 14.6), and as he passes out of the
lower frame edge, he drops down into a long shot of the group of men (Figure 14.7).
Some space has been reserved in the center of the frame for him, and as soon as he
hits the floor he springs up (thanks, presumably, to a mini-trampoline) and plunges
head first through a window, already visible in the rear of the set (Figure 14.8). Cut
to a new angle of him diving through the window (Figure 14.9), somersaulting over
the porch, and landing in the courtyard, ready to face more gang members, all conve-
niently spread out to create a vacant spot for him to occupy (Figure 14.10). The staging
cooperates with the cutting to make each leap and landing maximally evident.
Similarly, key factors of manual combat—the demeanor of each fighter, and espe-
cially the exact distance between them—must be depicted unambiguously. Cinemato-
graphy enters not as a substitute for the physical feat but as an enhancer of it, as when
zooms show a detail of the fight or when slow motion allows us to examine an action
that would otherwise pass too quickly.
Aesthetics in Action 403
The martial artist's performance can achieve clarity through the focused precision
of a gesture, as well as through the effort to invest the entire body's energy in each
gesture. Eisenstein can help us understand the latter tactic a bit better. He believed
that every movement involves the entire body, and so in theater and film one must
"sell" each action by exaggerating the body's role in forming it. Eisenstein imagines
how one might add bodily expressivity to a line of dialogue like "But there are two,"
with the actor showing two upraised fingers on the final word.
How much the pex*suasiveness of the phrase itself will be strengthened, the
expressiveness of the intonation, if on the first words, you make a recoil move-
ment with the body while raising the elbow, and then with an energetic move-
ment you throw the torso and the hand with the extended fingers forward.
Furthermore, the braking of the wrist will be so strongly directed that the wrist
will vibrate (like a metronome). 7
This sort of stylized clarity will be quite different from the more subdued per-
formance style characteristic of contemporary Hollywood acting, which tends to
emphasize the face rather than the whole body. When called upon to create gestures,
today's American actor is unlikely to produce something as exaggerated and cleanly
defined as Eisenstein's ideal. At the end of our Lethal Weapon scene, when Gibson
gets up after the taxicab accident, he starts to spar with the driver before dropping his
guard and half-sweeping his hands up and down in a mixture of dismissiveness and
angry frustration. The gestures are tentative and incomplete. In Eisenstein's sense,
they aren't articulated for impact.
By contrast, in the Hong Kong action sequence actors leap, twist, and scramble
with an energetic explicitness that reveals a dynamic of forces at work on the entire
body. No one just falls down in a Hong Kong sequence. In Hearty Response Joey Wong
hits the ground sidewise, in a vigorous roll that testifies to the pitch and impetus of
the launching force. Sometimes an actor lands with a splat, arms and legs splayed
wide. Or the actor may land on his neck or spine, creating the very picture of an
awkward, agonizing fall. By exploiting the actor's entire body, the Hong Kong action
film can present the specifics of each action with diagrammatic clarity.
This tendency is aided by the actor's ability to build the performance out of fast,
crisp gestures. The Hong Kong performer has recourse to something like Eisenstein's
idea of "recoil," because the actor's key movements, however rapid, are often separated
by noticeable points of stasis. We might describe this as the pause-burst-pause
pattern. A punch lands, and there is a pause; it misses, and the extended arm is held
poised, if only for a fraction of a second. The hero leaps and lands, resting in place
briefly. The tattooist in Hearty Response glances back to see Joey gaining on him;
instead of glancing over his shoulder, as he might in a Hollywood film, he runs fran-
tically, stops, swivels his head, registers Joey's pursuit, swivels his head again, and
then pounds downstairs into the nightclub. Throughout Hong Kong action scenes,
a pause often enframes each instant of movement, giving it a discrete, vivid identity.
The result is a kind of physics of combat and pursuit: Out of quiescence rises a short,
sharp action, which ceases as energy is switched off and stored for the next action. A
404 Poetics of Cinema Aesthetics in Action
405
parallel strategy rules the overall scene of fight or pursuit: Moments of near-stillness
alternate with bursts of smooth, rapid-fire activity.
Somewhat like meter in music, the pause-burst-pause pattern creates a regular
and recognizable pulse. Any one piece of combat can realize this accentual pattern
in different rhythms and tempos. Some action choreographers sustain the pauses by
dwelling on blocks and parries; this can achieve remarkable comic effects in such
films as Yuen Woo Ping's Snake in the Eagle's Shadow (1978) and Dance of the Drunken
Mantis (1979). Sammo Hung and Jackie Chan seek a faster pace, and so they allow Figure 14.11 John Woo's The Killer (1989): Figure 14.12 The Bodyguard From Beijing
only a split-second of pause between the lightning punches and kicks. Both artists pause-burst-pause in intimate surroundings. (1994): Li at rest.
sometimes enhance their presto tempo with fast-motion cinematography.
A lovely example of a more measured rhythm occurs early in Legendary Weapons
of China (Lau Kar-leung, 1982). In an inn, two guests, each bent on assassinating a
swordsman, unknowingly converge in a storeroom above his room. Their attempts
to stab through the floorboards to his bed are played out in rhyming thrusts and
dodges, each conveyed in a shot with its own pause-burst-pause rhythm. When the
swordsman eludes them, the two assassins turn on one another and fight. In one shot
the young man flings a dart at the woman; in the reverse shot she dodges away and
Figure 14.13 The Bodyguard From Beijing: Figure 14.14 The Bodyguard From Beijing:
the dart plinks into a grain sack beside her. (Beat.) A closer view of her shows her The first frame of the next shot; Li's head is Li's head snaps erect into the frame.
staring at the dart. (Beat.) Then she flings a claw-headed weapon. Reverse angle: He already starting to pop up at the frame edge.
ducks, and it clatters against a rake. A new angle shows the rake in the foreground,
out of focus. (Beat.) Then the rake falls abruptly backward as she yanks on the cable
attached to the weapon. In the next shot, the rake continues to fall and strikes the
young man. (Beat.) Now the young man lies in the foreground under the rake while
the young woman crouches in the background. (Beat.) She glances down through the
floorboards to see if anyone below has heard. (Beat.) Throughout, the scene manifests
a perceptible pulse in the pause-burst-pause pattern.
For a parallel instance from apolicier, we need look no further than Woo's The Killer
Figure 14.15 The Bodyguard From Beijing. Figure 14.16 The Bodyguard From Beijing.
(1989). The famous gun-to-gun confrontation between Chow Yun-Fat and Danny Lee
depends on briskly alternating the swift movements of each one's pistol hand with
long periods of immobility (Figure 14.11). Less obvious is the way in which the crisp
rhythm of the final shootout of Bodyguard From Beijing (1994) is laid down at the very
start of the movie. Jet Li is being addressed by his superior officer. In one shot, Li is
sitting calmly (pause; Figure 14.12), but at the cut his head is already popping up across
the lower frame edge (burst; Figure 14.13) to stop smartly before his superior (pause;
Figure 14.14). The climactic fight offers more vivid instances. The villain's gang has
invaded the heroine's living room, and Li must fight them in the dark. At the begin- Figure 14.17 The Bodyguard From Beijing. Figure 14.18 The Bodyguard From Beijing.
ning of one passage, Li twists and rolls into view before his face rises abruptly into
close-up (burst), immediately freezing warily (pause; Figure 14.15). Looking down, he
sees a sneaker poking out from behind a pillar (Figure 14.16). (Beat.) Li's pistol snaps
abruptly into the shot and fires (Figure 14.17). (Beat.) In a reverse angle, we see Li
from a low angle as the opponent starts to heave painfully down into the foreground
(Figure 14.18). Li's pistol snaps up and fires again (Figure 14.19), and his victim is
hurled out of the frame (Figure 14.20). The metronomic fall and rise of the gun arm
suggest a robotic efficiency appropriate to the killer-automaton Li is portraying.
406 Poetics of Cinema
The sequence also demonstrates how a director can squeeze the pause-burst-pause
pattern into glances and small gestures within confined spaces.
Very likely the martial arts tradition, with its repertoire of forms and combinations,
cultivated a belief that combat involved a balance between poised stillness and swift
attack or defense. The technique owes something as well to the Beijing Opera tradi-
tion of Hang hsiang ("bright appearance"), which presents a frozen pose assumed for
an instant after an acrobatic feat. (Significantly, Asian theater was also an important
source of Eisenstein's theory of expressive movement: He noted, for example, Kabuki's
reliance on "transitionless acting," by which he meant sudden switches from one pos-
ture to another.) 8 Possibly, directors also borrowed from the long pauses and outbursts
of violence found in Japanese swordplay movies and Sergio Leone's Westerns.
The pause-burst-pause pattern is far from an invariable rule, but it seems to occur
so rarely in Hollywood action films that we can treat it as an important mark of Hong
Kong's difference. In Lethal Weapon the continual, often lumbering movements of the
actors do not isolate discrete gestures on this moment-by-moment basis. Similarly,
all the storyboarders in the world cannot give Indiana Jones and the Temple of Doom
(1984) a visual pulse on celluloid. The opening fight scene in the Shanghai nightclub
is at once over- and underwrought because its compositions and cutting impose a
rhythm on haphazard runs and dives, whereas the film's protracted climax, showing
Jones in a mine tunnel grappling with a thuggee on a conveyor belt, presents a vague,
achingly slow tussle. In Hollywood, too often when a fight scene is fast it's not clear,
and when it's clear it looks laborious.
So when fans praise a Woo or Ringo Lam sequence for its "continuous action," I'm
suggesting a slight correction. It's Hollywood that relies on nonstop bustle, and the
sequence often suffers from it. One source of a Hong Kong sequence's power is a clarity
born of discontinuity, of lightning switches between quick, precise gestures and punc-
tuating poses. This staccato performance tactic gives a fistfight or gun battle or martial
arts bout a visual snap almost completely lacking in the American action picture.
Toward an Ecstatic Cinema
There are surely many other tactics that promote the Hong Kong film's gestural clarity,
but the ones I've itemized may start to characterize its action aesthetic. To press further,
we can usefully remind ourselves that the vividness of Hong Kong action depends on
more than visual intelligibility. There is as well the expressive amplification of action.
At bottom, this represents a willingness to go beyond Western norms of restraint and
plausibility—to surpass that appeal to realism that makes the typical Western action
scene comparatively diffuse in its stylistic organization and emotional appeal. Like
a good caricature, the Hong Kong action sequence selects and amplifies to achieve
a precise effect. The Hong Kong action sequence arrests us not because it mimics
normal behavior but because it felicitously exaggerates the most emotion-arousing
features of pursuit or combat. The stylization of Eisenstein's "But there are two" line
not only clarifies the gesture and the meaning, but also provides expressive thrust
and provokes a distinct response in the audience. A Hong Kong punch or tumble or
Aesthetics in Action 407
agonized death may seem cartoonish to viewers who do not recognize that in this
tradition, every movement gains its proper impact from being "sold" in Eisenstein's
sense. Expressive amplification is one major way in which the Hong Kong film goes
over the top.
I call this tendency expressive amplification because it goes beyond the desire
of Hollywood directors to convey merely an impression of the action. Instead,
the filmmaker tries to endow that action with vivid physical or emotional quali-
ties. The goal is, we might say, to characterize each chase or fight distinctly. A given
sequence will vividly exemplify power or elegance or vengeful fury or awkwardness
or indefatigability, or some combination of such qualities. In our Hearty Response
sequence, Joey Wang's performance is defined by her frantic, quavering despera-
tion mixed with a heedless impulsion for vengeance. This informs every action—
her firing at the tattooist, scrambling up from the pavement, crumpling under his
blows in the nightclub, and finally advancing, pistol trembling, on her trapped
attacker. In addition, a scene can articulate contrasting expressive qualities, build-
ing emotion by their alternation. In Woo's Killer, the gangsters who have invaded
Chow Yun-Fat's apartment die in grotesque poses of slow-motion agony, whereas in
his shots he cruises serenely through the frame.
A Hollywood action scene may occasionally display such expressive qualities, but
they tend not to be as sharply delineated on a moment-by-moment basis. In our Lethal
Weapon case, we know that Mel Gibson is furious, but although he pounds along
the pavement urgently, nothing else in his demeanor evokes the rage his character
is presumably feeling. And Donner supplies no close-ups comparable to those that
specify the emotional flow in Hearty Response.
Nor does much emphasis fall upon the sheer physical effort Gibson puts forth;
unlike Hong Kong characters, he never gets so winded he must collapse in tears. A
performer like Jackie Chan creates an extended suite of postures and facial gym-
nastics just in order to exhibit the strain of executing a stunt. This can be played for
laughs, as Jackie frequently does when, after fiercely delivering a blow, he half-turns
to hide his smarting hand. Similarly, in Fong Sai-Yuk (1993), after a barrage of blows
to the villain, Fong (played by Jet Li) steps back, scrunches up his face, and flaps his
arms frantically, creating a vivid image of stinging pain. For such reasons, the expres-
sive characterization of the action may well mix in recognition of the physical harm it
involves. The hair-raising stunts of Jackie Chan, Sammo Hung, Yuen Biao, and many
other adepts modify the action's dominant expressive quality—say, resourcefulness
or tenacity—by reminding us how easily the hero might be hurt.
I call this strategy expressive amplification because the emotional qualities aren't
presented laconically; the filmic handling magnifies them. Thus the actor may exag-
gerate the gesture to underlie its expressive quality, as Jet Li does in the Fong Sai-Yuk
instance. Or the expressive quality may be amplified through ensemble playing, the
exchange and rhyming of gestures among the performers. This was of course a staple of
the 1970s martial arts film. Later in the decade, the comic kung fu movies turned each
combat episode into a distinctive, almost dancelike symmetry of action and reaction.
The synchronized running, leaping, and diving of Waise Lee and his partner in the
408 Poetics of Cinema
incendiary climax of The Big Heat (1988) add up virtually to a circus act. By American
standards, many Hong Kong films (of all genres) look broadly played, perhaps seem-
ing closer to Warner Bros, cartoons than post-Method Hollywood. But I suggest that
this is part of a distinct aesthetic, in which expressive amplification is central to the
performance of actor and ensemble.
I've proposed that the U.S. action films of the 1960s and 1970s provided an array of
techniques that were exploited by directors of the 1980s. The martial arts cinema of the
same decades seems likewise to have bequeathed Hong Kong directors a rich reposi-
tory of tactics for expressive amplification. Take, for instance, the much-maligned
zoom shots. The rapid zoom itself often manifests the pause-burst-pause pattern
found at the level of the performance, as pose-strike-pose can be underscored by
static framing-fast zoom-static framing. Moreover, the zoom often intensifies the
expressive qualities of the bodies' postures and movements. Although 1980s action
films tend to avoid the zoom, the technique illustrates a broader tendency of the
Hong Kong tradition to take a commonplace stylistic device and exploit it to make
the action larger than life.
Editing also offers a rich array of tactics for amplifying the action, and here the
1980s action picture owes an even larger debt to the martial arts program picture.
Editing can create a smooth continuity of action, tracing out the geometry of the
body's trajectory (Figures 14.6-14.10). Cuts can mark the pause-burst-pause pattern
in a variety of ways: The rapid action may start at the end of one shot, continue across
one or more shots, and halt in a final shot. Or the pause-burst-pause pattern may be
confined to a single shot, and a series of shots may play the actions off against one
another. Right before the moment I've already discussed in Legendary Weapons of
China, the assassins hop abruptly apart, each one assigned a separate shot and a path
that mirrors that of the other.
Cutting may also strengthen the expressive force of running, jumping, punch-
ing, or kicking. For instance, the editing can underline the gesture by repeating a
portion of the movement at the cut. This tactic runs wild in the Hollywood action
picture, often in order to fake stunts like Mel Gibson's roll across the taxicab hood.
Overlapping editing finds a richer application when it serves, as usual in Hong Kong,
to clarify key gestures by distending the time they take onscreen. True to their pro-
gram, however, Hong Kong directors often exploit cutting that unabashedly repeats
entire actions. The most famous examples are Jackie Chan's instant-replay stunts,
such as his fall from the clock tower in Project A (1983), his slide down the post in the
climactic mall fight in Police Story (1985), and his leap from a motorcycle to a pallet
dangling from a crane in Armor of God II (1991). Another example is Conan Lee's
appalling drop from a lamppost in Tiger on the Beat II (1989). One way to make an
action transparently evident is simply to run it again, but this means of clarification is
not available to the more realistically committed Hollywood filmmaker.
Repetitive cutting to bring out a movement's trajectory is similar to the expan-
sion of an action by slow motion, and of course this has long been a favored resource
of 1970s and 1980s filmmaking. For the Hong Kong action film, slow motion can
intensify the fury or effort or danger of a blow while also stressing its grace. By the
Aesthetics in Action 409
same token, fast-motion cinematography can emphasize the precision and timing
of the kicks, punches, and parries. All the 1980s experiments with variable-speed
movement, ranging from Tsui Hark's and John Woo's efforts to intercut shots display-
ing different rates of slow motion to the stuttering pause-burst-pause printing one
finds in Wong Kar-Wai's work, probably owe a debt to the tradition of amplifying an
action's emotional overtones by playing with speed of motion.
Like the zoom, the soundtrack of the martial arts movie has had its share of bad
press, and not just among writers of hip guidebooks like Sex and Zen and a Bullet
in the Head. Westerners are often put off by the impossibly loud, absolutely uniform
whacks, smacks, whooshes, and air-smiting thuds to be heard in film after film. Yet
these can be understood as conventional signals that a blow has landed or missed. The
sounds, that is, clarify the action on a moment-by-moment basis. (I'm especially fond
of the rustle of fabric or invisible wings, which announces that someone has taken
flight.) Just as important, these implausible noises give the action an extra expressive
force, like the crash of cymbals during combat scenes in Beijing Opera.
The martial arts scene, then, can draw on a wide range of cinematic techniques
that underscore and magnify the action. But this profusion presents a new problem:
for if every blow or block is bracketed by pauses; if every kick or punch emits the same
thud; if variable-speed photography and printing constantly clarify and accentuate
a gesture—then how can there be any variation of emphasis? How can the decisive
punch stand out from all the hundreds of other blows that might look and sound
exactly the same? The solution that filmmakers devised was to create a multiaccentual
system that could underscore certain pieces of action on various channels. The
performance, filmed in long shot or broken into closer views, along with a normalized
level of sound effects (i.e., loud), yields a baseline. Then certain techniques can be
layered over that base at a particularly intense moment. In the kung f u film, a key
action can be presented as a stupendous piece of acrobatics, but also treated in super-
slow motion, accompanied by a piercing cry from the attacker and a throbbing, high-
reverberation thwack on the soundtrack.
Urban gunplay films follow this precept straightforwardly. In Hard-Boiled (1992),
during the massive gun battle in the hospital, Tony Leung accidentally shoots a fellow
policeman. How can John Woo distinguish this crucial moment from the dozens of
other gunblasts that have preceded it? As Leung hears an elevator open offscreen right,
he whirls and fires, and the multiaccentual system comes into play. In an instant, the
shot slips into step-printed slow motion, the camera pans right and moves in on the
victim, and the natural sound drops out and is replaced by a spectral tinkling and
roiling synthesizer music. The camera swings back to Leung, who halts abruptly. The
camera holds on him as Chow steps to him from behind and pushes him into the
elevator. Woo refuses one obvious accentual cue, a cut to a closer view of the victim,
evidently because that would disrupt the very long traveling shot that has followed
Leung and Chow through a maze of corridors and will cling to them in the eleva-
tor and through the firefight on the next floor. And step printing by itself would not
distinguish the moment from other stretches of slow motion in the shot. Woo had
to laminate several received techniques together to intensify this pitiful accident. By
 Poetics of Cinema
410
combining cues for expressive qualities, the filmmaker can underscore an action dou-
bly or trebly.
I don't mean to suggest that clarifying and amplifying physical movement are only
matters of conveying important story information to the viewer. These processes do
that, but they also aim at a piercing arousal of the spectator's senses and emotions.
Clearly many Hong Kong filmmakers aim, as Yuen Woo Ping puts it, to make the
viewer "feel the blow."9 Not only must the action be legible and expressively amplified;
it must be communicated, as energy is communicated from one body to another.
It must be stamped on the spectator's senses. And this is what Hollywood fights, falls,
and car crashes so often fail to do. We watch them, sometimes with keen interest
and anticipation, but we seldom feel them because all the resources of performance,
filming, and editing are not focused upon transmitting the sensory and emotional
expenditure of energy that propels the scene from instant to instant.
All this is, once more, far closer to the tradition of Sergei Eisenstein than of Raoul
Walsh or Steven Spielberg. Throughout his career Eisenstein, always a man of the
theater, emphasized that expressive movement was the core of cinematic mise-en-
scene. First the filmmaker had to discover concrete actions that an actor could be
trained to execute in simplified and stylized form. Then the filmmaker had to devise
ways of framing (mise en cadre) and editing (montage) that would sharpen and further
dynamize the expressive movements. And at key moments these techniques could
cooperate, double, and intensify one another in a massive assault on the spectator's
senses. If the effort was successful, the force of the movement and its onscreen pre-
sentation would stir in the viewer's body a palpable echo of the actor's gesture. "It is
precisely expressive movement," Eisenstein writes,
built on an organically correct foundation, that is solely capable of evoking this
emotion in the spectator, who in turn reflexively repeats in weakened form the
entire system of the actor's movements; as a result of the produced movements,
the spectator's incipient muscular tensions are released in the desired emotion. 10
Eisenstein believed that affective qualities were the result of agitation in the
nervous system, so expressive movement on the screen could yield a uniquely exhila-
rating experience, at once physical and emotional. He dreamed of an ecstatic cinema,
one that literally carried spectators away, tearing them "out of stasis" and into a realm
of rapt, electric apprehension of sheerly pictorial and auditory momentum. 1 1 How
successful his own films were in producing this effect is a matter of dispute, but it may
be that, just as the unabashedly popular filmmaking of Mack Sennett and Douglas
Fairbanks and William S. Hart launched a tradition of what Eisenstein considered a
"cinema of attractions," in Hong Kong's grassroots cinema that impulse continued
to flourish.
It's not the whole story, of course. Part of our pleasure is created by an apprecia-
tion of innovation—imaginative variations on formulas, unexpected uses of familiar
schemas. Consider the device of one fighter tossing the gun to his buddy. Woo uses
the gesture to suggest the rituals of courtly honor when the opponents of A Better
Tomorrow 2 (1987) exchange pistols. In Pom Pom and Hot Hot (1992), however, it
Aesthetics in Action
411
Figure 14.21 The Bodyguard From Beijing. Figure 14.22 The Bodyguard I rom l^ijing.
becomes a manic gag; two cops caught in a firefight must share the same automatic
and flip it back and forth, always just in time to dispatch an attacker.
And certainly part of our pleasure comes from the film's display of skill for its own
sake. In any art, we enjoy the virtuoso's mastery of craft, the way she or he sets up
difficult problems and solves them with easy brilliance. Later in our specimen sequence
from Bodyguard From Beijing, Jet Li is fighting the main villain around, behind, and
on top of a kitchen counter. Earlier, director Corey Yuen Kwai has conveyed the
pause-burst-pause visual pulse by coordinating cutting with performance, but now
he manifests it within a single flashy shot. To prevent his adversary from reaching
water, Li cracks a towel like a whip, snapping the swiveling faucet sharply from side to
side before our eyes (Figures 14.21-14.22).
Still, these more knowing pleasures build on the sheer kinetic transport we're
offered by doings that are energetic, functional, and laden with infectious emotion.
These actions are the opposite of brute force. They rivet our attention because they
are clearer, more expressive and expansive than ordinary running or jumping or arm
swinging. Hollywood offers an all-purpose blanket excitement. Directors try to turn
graceless stars into action heroes by having them run through a loud, vague bustle.
From the 1960s to the 1990s, Hong Kong action genres at their best offer balletic
performers carrying out concrete, intelligible physical tasks. Each gesture is honed to
a fine edge and triggers specific expressive qualities, standing out by virtue of a visual
design stripped of clutter and amplified by particular resources of film technique.
This conclusion needs more refining, but it affords some initial insights into the
distinctive artistic accomplishments of these movies. Other questions remain to be
asked, too. Aren't there distinct stylistic trends within the domain I've sketched? What
more specific lines of influence can we trace out—not just within Chinese traditions
and between Hong Kong and Hollywood, but also between Hong Kong and Japan?
(The Zatoichi and Lone Wolf and Cub films, for instance, achieve a more sober, atten-
uated, and grotesque expressivity, and this resurfaces, it seems to me, in the films of
Kitano Takeshi.) And how much theoretical merit is there in Eisenstein's account of
the spectator who repeats, in less forceful form, the muscular rhythms displayed on
the screen? Approaching Hong Kong action movies as an ecstatic cinema may lead us
toward understanding why so many of these movies infect even film professors, heavy
with middle age and polemics if not baked brie, with the delusion that they can vault
calmly over the cars parked outside the movie theater.
 15.
Richness Through

Imperfection

King Hu and the Glimpse

As one of Chinese cinema's finest directors, King Hu (Hu Jinquan) rewards study
from many angles: his treatment of historical periods, his thematic preoccupations,
his stratagems of plotting, and his handling of character point of view.1 In what follows
I I hope to clarify an aspect of his work over which he labored patiently—visual style.
Working on Come Drink With Me (1966) he realized that "if the plots are simple,
the stylistic delivery will be even richer." 2 Visual design was very important to him;
he drew every shot in advance and supplied the cast and crew with photocopies. 3
Declaring himself ignorant of the martial arts ("Kung fu, Shaolin tales—I don't under-
stand anything about that"), 4 he derived his films' combat techniques from Beijing
Opera and compared his fight scenes to dances. He lavished attention on his set
pieces, spending 25 days shooting the confrontation in the bamboo forest in A Touch
of Zen (1970).5 Whatever his other preoccupations—Zen, China's history—Hu is an
unabashed aesthete, and an aesthetic approach to his style seems natural.
Just as naturally, the most inviting objects of stylistic inquiry are his magnificent
fight scenes. True, Hu might discourage critics from concentrating wholly on them.
Legend of the Mountain (1979) and All the King's Men (1983) pointedly avoid the fren-
zied fights that made his reputation. He told one interviewer that The Fate of Lee Khan
(1973) was initially conceived as a theater-like piece centering wholly on interweaving
intrigues, but happily for us the distributors and exhibitors insisted on combats. 6

413
 Poetics of Cinema
Figure 15.1 Anamorphic deep focus in The
Fate of Lee Khan (1973).
Predictably, Hu's dramatic and expository scenes aren't completely without pictorial
interest. There are picturesquely decentered widescreen framings, elegant tracking
shots, and an occasional striking depth image (as in the famous shot of Lee Khan
pondering in the foreground, while Wang Shih-Cheng is interrogated in the back-
ground, seen through a doorway; Figure 15.1). In all, his films display an effortless,
polished professionalism that seldom becomes academic.
Undeniably, though, King Hu's continuing renown derives largely from his superb
treatment of physical action in his Hong Kong and Taiwanese films of the 1960s and
1970s. Whatever his other virtues, he will be remembered as a director of running,
jumping, and ferocious combat. How do his action scenes deliver the stylistic richness
he sought? We can start to answer this question by examining some of the traditions
that he embraced and transformed. 7 in a full long shot. A medium-shot framing allows us to concentrate on the essen-
Inherited Norms
There are many ways to stage and shoot a martial arts combat. Some say that the
most felicitous method is to present the fight in fairly distant shots and in longish
takes. That way, the argument runs, the totality of the action is preserved. The pro-
tracted long shot is common in Japanese jidai-geki (the Zatoichi films offer notable
instances), and can be found in some kung fu films, like Ng See-Yuen's Secret Rivals
(1976) and Jackie Chan's Young Master (1980). But much Asian action cinema relies
heavily on two other stylistic norms.
One of these norms can be easily visualized. Instead of fighting in one spot, the
hero or heroine advances toward a string of attackers. Framed in medium shot, the
warrior moves more or less laterally, the camera tracking alongside and revealing a
string of opponents one by one. Each luckless enemy moves into the frame, fights,
and is slain. Sometimes the fight is staged so that the action shifts diagonally toward
the camera or into depth, and sometimes the protagonist will break with the shot's
overall trajectory, step back, and counter an attack from the rear. I'm not sure when
this sort of sustained, action-packed tracking shot came into favor, but examples can
be found in Japanese films of the 1930s, and certainly by the 1960s it was common
in jidai-geki. A vivid example is the grueling swordfight in a rice paddy that climaxes
Uchida Tomu's Musashi Miyamoto Part 4 (1964). Here Musashi takes on dozens of
men seriatim in a remarkably lengthy moving shot.
Richness Through Imperfection 415
Figure 15.2 Near the climax of Dragon Gate Figure 15.3 Dragon Gate Inn: The camera
Inn (aka Dragon Inn-, 1967), the swords- tracks leftward as she swivels to drive back
woman Zhu Hui wades into an attack from two men.
Cao's minions
Figure 15.4 Dragon Gate Inn: She dispatches Figure 15.5 . . . before falling away to reveal
one in the foreground, who freezes momen- the decisive blow that Zhu deals his partner.
tarily . . .
The one-by-one tracking shot shows the advantages of not staging a fight scene
tial action—the encounter of the hero with each adversary—and to see the thrusts
and parries clearly. Moreover, because the scene depends on a series of attackers, the
closer framing allows each one to shift into prominence at the proper moment; our
attention is not distracted by idle opponents awaiting their turn at the hero. Indeed,
the relatively tight framing of the medium shot frees the filmmaker from specifying
how many attackers there are, or where they are lurking.
From the start of his career, King Hu's combat scenes seem strongly influenced by
Japanese techniques. He uses the convention of mortally wounded victims freezing
in place before toppling over, as well as judicious cuts back from a sword stroke to
an extreme long shot that shows several victims falling at once. In a similar way, Hu
borrows the one-by-one tracking shot. The fight scenes in Come Drink With Me use
long traveling takes into which fighters burst unpredictably. The one-by-one tracking
shot is central to Dragon Gate Inn (1967) (Figures 15.2-15.5), A Touch of Zen, The Fate
of Lee Khan (with abrupt attacks from the foreground), and The Valiant Ones (1975),
which features a brilliant shot of Wu's wife (played by Xu Feng) moving rightward,
zigzagging into the foreground, and ending in a close view at the extreme right frame
edge. Raining in the Mountain (1979) largely avoids the device, but its principle of
dynamic displacement within medium-shot framings governs the games of grab-the-
scroll that shift characters around abruptly and demand sudden thrusts to and from
the camera.
A second stylistic norm that King Hu exploits reaches farther back into film history
than the one-by-one tracking shot. It's that form of "constructive editing" pioneered
416 Poetics of Cinema
by American filmmakers of the 1910s and recast by Soviet filmmakers of the 1920s.
An action is broken up into bits, each one rendered in a separate shot, and the viewer
constructs a unified action out of the details.
Constructive editing, as this technique is usually called, is ideal for combat scenes.
In the Shanghai film Heroine White Rose (1929), we get the following passage:
Long shot: White Rose fires an arrow.
Medium shot: Her enemy clutches at the arrow in his chest.
Long shot: He falls to the ground.
Constructive editing proved particularly useful in suggesting the vaulting leaps
executed by powerful swordsmen. In Zhang Che's One-Armed Swordsman (1967),
Fong Kang (Jimmy Wang Yu) leaps up in medium shot. Cut to a low-angle shot show-
ing him soaring into the air and starting to descend. A third shot shows him landing.
In such ways, the wuxia pian of the 1960s induced the audience to imagine that the
hero has mastered all the skills of martial chivalry. Again, this stylistic norm violates
the premise that the hero's feat should be executed in a single take; cutting allows the
filmmaker to suggest what could have been shown only with costly and perhaps uncon-
vincing special effects.
King Hu mastered this constructive cutting. In Dragon Gate Inn, the hero Xiao
Zhaozi tosses a bowl of noodles to a bully at a neighboring table. This action is shown
in four easy-to-stage shots:
Medium shot: Xiao tosses the bowl to the left.
Extreme long shot: The inn dining room; the bowl flies leftward across the room
as the bully leaps back.
Close-up: The bowl spins into place on the table and stops.
Medium shot: The bully reels back, startled.
Throughout his career, King Hu would use constructive editing to present similarly
amazing feats of discipline and skill.
If King Hu had relied upon only the one-by-one tracking shot and constructive
editing, he would have remained fundamentally like his Asian action movie con-
temporaries—more elegant and meticulous than most, but not essentially different.
Instead, like Kurosawa Akira in Yojimbo and Sanjuro, like Kato Tai in Love for a
Mother (1962) and Stormy Days of the Sanada Clan (1963), and like Misume Kenji
in the Lone Wolf and Cub series (1972-1974), Hu recast inherited norms in original
ways. In the process, he created a distinctive, highly engaging approach to presenting
physical action.
A Problem and Some Solutions
Hu's combat scenes were obliged to present near fantastic feats of martial prowess.
The chivalric knights in the wuxia pian had to display preternatural speed and agility.
They had to strike a blow that sends their opponent reeling back many paces. They
had to avoid an attack by somersaulting over their enemy. They had to vault into trees,
Richness Through Imperfection 417
spin, and come diving down at their opponents. Any such acts can be shown quite
clearly if the director is able to spend time and money on wires and special effects.
But apart from the expense, special effects inevitably risk looking fake. The problem
wasn't entirely surmounted by wirework, which often makes combatants look like
they are hauled and swung around. Alternatively, gravity-defying feats could be
implied through constructive cutting, but this technique gives us the event in isolated
bits rather than as an integral whole.
It seems to me that King Hu wanted to avoid making martial feats look artificial.
He was proud of not employing trick photography. 8 Instead of trying to put the feats
on the same plane as ordinary sequences, as special effects and orthodox construc-
tive cutting tend to do, he sought a stylization that set these extraordinary feats apart
from mundane reality.9 Not that he wanted to glorify the warriors as superhuman.
The powers that they display aren't supernatural: They spring from the mastery of chi,
or essential energy. King Hu's problem was to dignify and beautify these feats without
tipping them into implausibility and sheer fantasy. How, we might ask, can cinema
express the otherworldly grace and strength of these supremely disciplined but still
mortal fighters?
The solution he found was to stress certain qualities of these feats—their abrupt-
ness, their speed, and their mystery. And he chose to do so by treating these feats as
only partly visible. Normally, in the hands of Hu or others, the one-by-one tracking
shot and constructive cutting make the action perfectly lucid. But Hu also frequently
stages, shoots, and cuts his action so that it becomes too quick, too distant, or too
sidelong for us to register fully. Significantly, Hu's films of the 1960s and 1970s almost
never employ slow motion to dissect his fighters' technique. (His most famous use of
slow motion, showing the floating monks of A Touch of Zen, isn't a deceleration of
normal action; the monks drift toward the confrontation at a serenely unreal pace.)
Instead, Hu makes his action faster than the eye—and even, it seems, the camera—
can follow. Often we are allowed only a trace of the warriors' amazing feats. We don't
see the action so much as glimpse it.
Of course, the action can't be completely indeterminate. The fight cannot play out
wholly "offstage" or in pitch darkness. The filmmaker must make the action clear
enough to be comprehensible and then add "imperfections" that make it partially
indiscernible. Consider a specific example from A Touch of Zen.
The scholar Gu Shengzhai has spent the night in the ruined fortress with the myste-
rious woman Yang Huizhen. In the morning the sinister stranger arrives, and soon he
is sparring with Yang. They run outside, and the stranger flees with Yang in pursuit.
Gu comes to the window in time to see the stranger in midair, leaping over the wall
in a very brief shot (13 frames). Here our limited view could be said to be motivated
by Gu's act of looking (although what we see isn't from his optical point of view). Yet
as the scene goes on, the action becomes indiscernible in ways that can't be explained
by Gu's limited perspective.
The stranger suddenly drops from the sky and lands in a patch of tall weeds. He
starts to run. Yang lunges out to follow him, but instead of showing her vault over
the wall in constructive-editing fashion, King Hu cuts away to another shot of the
 Poetics of Cinema
418
Figure 15.6 A Touch of Zen (1970). Figure 15.7 A Touch of Zen.
Figure 15.10 A Touch of Zen. Figure 15.11 A Touch of Zen.
stranger fleeing. Cut back to Yang, who has already vaulted the wall and is landing on
a rooftop. The editing omitted the stranger's launch but shows his leap and landing;
now the editing shows us Yang's launch and landing, but not her leap. As she springs
up again, the camera whip-pans left, blurring the image. The next shot shows the
stranger fleeing, and Yang hurtling down and slashing at him. In the next shot, she
lands, and he pops in from offscreen. They begin to struggle, with most of their move-
ments hidden by waving weeds (Figure 15.6).
So it goes in the rest of the sequence. Shots showing pieces of action are inter-
rupted by cutaways to other actions; by the time we return, the first action is already
well advanced. Key moments of the fight are blocked by weeds or rendered vague by
distance—as when the fight on the distant wall turns the figures into tiny somersaulting
silhouettes (Figure 15.7). Whip pans try in vain to follow the fighters or their weapons
hurled through space. Either fighter may pop into the foreground, especially because
the very brief shots have rendered their position uncertain. Or an attack may take place
offscreen. The stranger runs furiously rightward, thrusting out his hand suddenly
(Figures 15.8-15.9). Cut to a symmetrical shot of Yang running leftward and pausing
at a doorway, her hair tumbling down her face (Figures 15.10-15.11). Her adversary has
somehow, behind the weeds, during the shot change, slashed away her scarf.
One could argue that these technical choices simply hide artifice. The walls and
long shots and tall weeds all conceal the trampolines. The actors and their doubles
execute the stunts piecemeal—shot of a leap, shot of flying, shot of landing—and so
Richness Through Imperfection 419
naturally the cutting is choppy. Yet other directors, facing the problems of presenting
this sort of action, did not resort to King Hu's solutions. They filmed leaps clearly,
centering them and letting low angles keep the trampolines below the frame line.
By contrast, Hu used long shots that allow the actors to vault in and out of blocking
material in the foreground. Instead of 1-2-3 constructive editing like Zhang Che's
in One-Armed Swordsman, Hu gives us only phase 1 or 2 or 3—launch or leap or
landing—or only two of them. And by making the shots extraordinarily brief, he goes
beyond his contemporaries; blink, and you miss the stunt.
Working within the same production constraints as other directors, King Hu
created willed imperfections in the presentation of the action, and one result was to
make many maneuvers seem too fast or too powerful for the eye to follow. The indirect
handling has the effect of making you unsure you've really seen what you thought you
saw. Did Cao Shaoqin leap over trees at the rapid-fire climax of Dragon Gate Inn7.
Did Yang really keep herself aloft by ricocheting among tree trunks in A Touch of
Zen? Whereas fantasy swordplay films since the late 1980s dwell on their aerobatics-
recall the end of Swordsman II (1992), when several characters fly and float for what
seem to be minutes—King Hu teases us with mere glimpses of the action. The recent
wuxia pian techniques give us time to savor the outrageousness of the stunts, but Hu's
glimpses tantalize rather than fill our appetites for action.
Consider, for instance, how features of the setting often obscure the combat. Fights
and flights can be partially concealed by fog or mist (the climax of Raining in the
Mountain), boulders (the closing of The Valiant Ones), darkness (the ambush on the
fortress in A Touch of Zen), and, especially, furniture and foliage. King Hu is adept
at hiding his trampolines behind inn tables and foreground bushes, but this artifice
becomes less evident when the entire frame is cluttered up, as when tall weeds masks
essential moments of the Touch of Zen duel between Yang and the stranger. Likewise,
many of Hu's forest glade fights fill the frame with far more tree trunks and branches
than would be necessary to camouflage the trampolines. Whereas the backlot and
studio-interior forests of many Shaw Brothers martial arts vehicles are designed to
make the action maximally readable (e.g., the first fight in Zhang Che's The New
One-Armed Swordsman [1970]), Hu's exterior scenes, shot on location, motivate a
naturalistic density of detail.
King Hu is not alone in using bits of setting to hide his stunts; one thinks of the
way in which a foreground roof edge in The Jade Bow (1966) conceals one phase of a
series of leaps (which also involve in-camera reverse motion). But whereas other direc-
tors' blocking foregrounds favor clarity, a single center of attention, and full visibility,
Hu pushes toward a more dense, multicentered, and strategically opaque composi-
tion. Very often, a stupendous leap will be glimpsed through a patch of light leaking
through shadowy branches. Foreground elements can be used in more flamboyant
ways as well. In Dragon Gate Inn, when Mao (played by Han Yingjie) flees from Zhu
in the skirmish in front of the inn, he ducks behind a wall and reemerges instantly far
away, with superhuman speed, thanks to a hidden double (Figures 15.12-15.13).
Sometimes Hu lets the audience enjoy the humor of blocked views, as in the Tom-
and-Jerry evasions in the short film Anger (1970). When Ren and Jiao are on opposite
420 Poetics of Cinema
Figure 15.12 Dragon Gate Inn. Figure 15.13 Dragon Gate Inn.
11
Figiiu I •>.I I litiinihg in ilu Mountain. Figure 15.15 Raining in the Mountain.
Figure 15.16 Raining in the Mountain. Figure 15.17 Raining in the Mountain.
sides of the door, neither realizes his opponent is actually his ally. Hu's composition,
which lets Jiao's silhouetted head peek out in the distance, makes a joke of Ren's limited
perspective. Raining in the Mountain turns the monastery from a solid, symmetrical
mass into a booby-trapped maze of walls and jutting rooftops that obscure and then
reveal the characters. The transformation is initiated in one witty sequence. White
Fox (Xu Feng) and the boy Jin are furtively exploring the monastery, dodging passing
monks. The camera travels with them as they sneak from an austere, rectilinear room
(Figure 15.14) to an intricate diagonal space defined by a plunging roof edge and a
distant walkway. They rush away from us, down to the walkway (Figure 15.15), and
suddenly vault off it, over the stone wall and down into the gap between the two build-
ings. Without pause, they then race into the distance (Figure 15.16). Why their sudden
retreat? As their figures dwindle, a monk in blue appears at the upper center of the
frame, barely visible in a doorway of the terrace they have just left (Figure 15.17).
Hu makes story action even more opaque through shrewd tactics of framing. Surely
no director has been more in love with the oddly decentered action shot. His soaring
combatants often scrape the top of the image, whereas his divers sink along the frame
edge, as Peony does when she plummets from the balcony in The Fate of Lee Khan. The
final fight of Dragon Gate Inn expresses Cao's power through a weirdly vacant shot:
He leaps down (Figure 15.18) and goes out of the frame (Figure 15.19) before spring-
ing back in to somersault over his adversary (Figure 15.20). It's as if he moved too
suddenly and swiftly for the camera to frame him properly. At the climax of The Fate
Richness Through Imperfection 421
Figure 15.18 Dragon Gate Inn: Cao leaps Figure 15.19 ... and goes out of the frame
down...
Figure 15.20 . bdoic springing back in to Figure 15.21 m Gate Inn.
somersault o\ei his adu-i s.n
of Lee Khan, male and female fighters are paralleled in their attack method not only by
symmetrical cutting patterns but also by comparable edge framing.
Such moments remind us that the frame excludes as well as includes, and King
Hu is masterful in holding action outside the frame edge until it can burst in with
maximum impact. In the opaque combat between Yang and the stranger in A Touch
of Zen, the fighters swerve unexpectedly in and out of the shots, and Yang's scarf is
slashed by an offscreen stroke. When combined with editing, this sort of exit-entry
pattern can be powerfully abrupt. In Dragon Gate Inn, the villain Tsao swoops down
in a predatory leap (Figure 15.21). Cut to Xiao recoiling as the sword blade slashes in
at him (Figure 15.22). He deflects it and dodges out of the frame (Figure 15.23). Tsao
lands behind him and rises into the frame, poised and ready (Figure 15.24). In an
instant the fighters have traded places, although the cuts reveal images that cannot
hold them for very long.
 Poetics of Cinema Richness Through Imperfection 423
422

Figure 15.25 The Fate of Lee Khan. Figure 15.26 The Fate of Lee Khan. Figure 15.32 The Valiant Ones (1975): how Figure 15.33 The Valiant Ones: Cut, using
to make a whirlwind swordsman. The long a match on action, to a tight shot in which a
shot shows Bai Ling, the actor playing Wu, double, his back to us, has been substituted.
quite clearly. The moment that this false Wu dodges out of
the frame . . .

Figure 15.27 'I he Fate of Lee Khan. Figure 15.28 The Fate of Lee Khan.

Figure 15.34 ... the "original" Wu reappears

in the rear of the shot.

style of swordplay. But as the fight intensifies, Hu cuts to closer views. This prepares
Figure 15.29 The Fate of Lee Khan. Figure 15.30 The Fate of Lee Khan. for an astonishing effect. Wu rises into the right foreground and slides to the left
frame edge, back still to us as the Chinese slashes at him. But Wu ducks out of the
foreground, and immediately, in the very next frame, Wu springs into the left back-
ground, now behind the Chinese, who is stupefied by his sudden disappearance
(Figures 15.32-15.34). The rest of the fight continues to display Wu's ability to trans-
port himself instantly between foreground and background by ducking out of the
frame. In the next trial, against a Japanese swordsman, Wu (with Hu's help) employs
Figure 15.31 The Fate of Lee Khan. the same tactic. His whirlwind technique will be on still splashier display in the final
fight with the Japanese pirate leader, Hakatatsu, and at another point Wu's wife,
Ruoshi, gets a chance to demonstrate it too (Figures 15.35-15.38).
Even though working in the anamorphic ratio of 2.35:1, King Hu boldly blocks off
Of course, these scenes depend on artifice—an actor's double who, back to us, can
key portions of the action, pushing combat maneuvers so close to the camera that the
slip in and out of the foreground—but it is a brilliantly cinematic way to present a
fighters' tactics, by masking and revealing bits of action, give a distinct pulse to the
swordsman who can "attack from all sides." The solution is more fluid and integral
fight. The colorfully dressed serving girls in The Fate of Lee Khan prove to be of excel-
lent use in such scenes. When a hapless thief tries to escape, the female innkeeper flips than constructive editing, because it does not break the action into separate shots.

a plank through the air to bar his exit. He's caught in a hail of waitresses, who are at
once abstract streaks of color and very tangible fighters, before they hold him frozen
in a formidable low angle—the final beat of the sequence (Figures 15.25-15.31).
The Valiant Ones offers a striking use of framing to suggest heroic prowess.
Wu Jiyuan is the "whirlwind" swordsman, and King Hu lets him demonstrate his
technique in a powerful, purely cinematic way—again, one that depends on with-
holding a full view of the combat situation. Wu and his wife are in the pirates' lair
and undergoing a series of martial tests. A stout Chinese challenges him, and they
begin to fight. At first King Hu uses long shots to establish Wu's bobbing, ducking
Hu's innovation also exploits the principle of the glimpse. The medium-shot framing,
as in the one-by-one traveling shot, evidently shows us everything relevant, but Wu's
whirlwind tactic seems simply to defeat the camera. The framing cannot keep up with
him, and we, like Wu's opponents, can only gape at his agility.
Cuts and Slashes
It is above all editing that creates those "imperfections" that make Hu's action scenes
so distinctive. At the same time, a study of his editing principles reveals that the
424 Poetics of Cinema
Figure 15.35 The Valiant Ones: The dou- Figure 15.36 . . . and pops up in the fore-
ble act creates a whirlwind swordswoman. ground before he can turn.
Ruoshi in her turban dodges a pirate . . .
Figure 15.37 The Valiant Ones: The Ruoshi Figure 15.38 . . . by Ruoshi herself, who has
figure in the foreground vanishes, and the evidently moved too fast for the camera to
pirate is struck from behind . . . follow.
aesthetic of the glimpse succeeds within a context of clear, cogent presentation. It is
the dynamic between stability and momentary indiscernibility that yields Hu's most
original effects.
He was from the start a cutting-based director: Sons of the Good Earth (1964),
though it employs some ensemble long takes, still ends up with an average shot length
of 6.3 seconds. His action films are cut significantly faster. In Come Drink With Me,
Dragon Gate Inn, the "Anger" episode of Four Moods, and Painted Skin (1993), the
average shot runs between 5.4 and 6 seconds. In A Touch of Zen, The Fate of Lee Khan,
and Raining in the Mountain, it's around 4.5 seconds. As he got older, Hu evidently
speeded things up: The Valiant Ones and his short film The Wheel of Life (1983) are
the fastest paced of his works, each with an average shot length of around 3.3 seconds.
Hu's counterpart Zhang Che wasn't shy about cutting quickly, but his films don't
approach Hu's very low averages.
Even Hu's most static dialogue scenes rely on restless editing. He sometimes breaks
up scenes according to optical point-of-view structures reminiscent of Hitchcock;
scene by scene, A Touch of Zen and Raining in the Mountain are cut on the basis of
shifting viewpoints. Hu creates a nervous rhythm for such simple actions as allies
assembling at an outpost (The Valiant Ones) or rivals greeting each other in rapidly
intercut close-ups (Raining in the Mountain). He sometimes interrupts a longish take
with a peremptory shot; at the very beginning of Come Drink With Me, a shot of
a horseman dismounting is broken by a one-second shot of the villain advancing
before we return to the horseman striding off. The editing can formalize even transi-
tional moments, as when in quick succession each conspirator in The Fate of Lee Khan
rushes rightward through a curtained doorway, every shot framed and timed to create
a symmetry among them.
Richness Through Imperfection 425
Although cuts like these sometimes jolt, most of Hu's dramatic scenes strictly
obey conventions of continuity cutting—eyeline matching, shot/reverse shots,
smooth matches on action, and all the rest. When Hu resorts to constructive edit-
ing, he sometimes takes what we might call the "American" option: Scenes like Xiao
pitching the bowl of noodles across Dragon Gate Inn succeed thanks to precise
continuity. Hu's combat scenes, however, seem closer to the tradition of the Soviet
montage filmmakers, who made constructive editing more dynamic by accentuating
discontinuities. Like other Hong Kong filmmakers, Hu revived and refined Russian
montage principles for the demands of an action-based popular cinema. 10 It's as if
his orthodox, smoothly cut passages provide a ground or foil for the innovative,
decidedly disjunctive action sequences.
In such passages, Hu's already brisk editing tempo intensifies. The average shot in
the final reel of Lee Khan runs 2.4 seconds, and the climactic fight on the beach in
The Valiant Ones averages 1.6 seconds per shot (the same as the opening firefight in
Sam Peckinpah's The Wild Bunch). In Part I of The Wheel of Life, the opening attack
of the rebel archers on the corrupt government leader moves even more quickly; each
shot averages almost exactly one second. Hu told interviewers that he experimented
with shots of eight frames or less, and his fight scenes have become known for their
bursts of brief images. 11
Still, comparably short shots can be found in films of Hu's contemporaries. Zhang
Xinyan and Fu Qi's Jade Bow (1966) contains much rapid cutting, including a shot
only four frames long, and the opening of Tu Quangqi's First Sword (1967) features
a very swiftly cut leap. For this reason, just calling King Hu a rapid cutter takes us
only so far. What qualities of his quick cutting set him apart? Very often during the
action sequences, fast editing works with the visual blockages in the setting and the
tactics of oblique framing in order to suggest the awesome and mysterious power of
the fighters.
Virtually by definition, a brief shot offers only a glimpse of the action. Our sense
that the fortress skirmish between Yang and the stranger in A Touch of Zen develops
faster than we can follow stems partly from the fact that some shots last only about
half a second. At one point in The Fate of Lee Khan, when the serving girls turn
cartwheels over a table, the shot takes only 14 frames. But we should distinguish
between "legible" brief shots and more "illegible" ones. In Raining in the Mountain,
a dynamic sequence presents monks beating temple drums from several angles and
distances, and it preserves the same information from shot to shot. The passage builds
up excitement in a conventional way. Less orthodox is a rhythmically "geometricized"
passage in Come Drink With Me. During the attack on Chang's palanquin, Golden
Swallow faces off against her three principal opponents (Figures 15.39-15.43). The
shots are virtually identical in length (each about half a second), and each displays
a simple, strongly parallel graphic design. This construction recalls the arithmetical
cutting of 1920s silent film as practiced by Sergei Eisenstein, V. I. Pudovkin, and some
French experimenters. Not for the only time, a Hong Kong director spontaneously
rediscovers the lesson that shots with a simple, strong graphic thrust can be power-
fully cumulative if given a rigorous editing tempo.
426 Poetics of Cinema
Figure 15.39 Come Drink With Me (1966): Figure 15.40 Come Drink With Me: One
Golden Swallow hurries rightward (10 frames). opponent runs leftward toward her (10 frames).
Figure 15.41 Come Drink With Me: Another Figure 15.42 Come Drink With Me: Her third
opponent runs leftward toward her (10 frames). opponent races diagonally toward her (10 frames).
Figure 15.43 Come Drink With Me: All of the
fighters clash (12 frames).
Like most directors who employ fast cutting, King Hu usually takes care to ensure
that information is displayed and repeated clearly so that the cutting doesn't hamper
our understanding. The Come Drink With Me sequence just mentioned offers a good
example. But Hu is willing to tolerate an unusual degree of illegibility in certain
passages for the sake of rendering the action only partly discernible and hence more
miraculous. For example, some of his brief shots are whip pans that blur the scene. In
the tumultuous finale of Dragon Gate Inn, the whip pans disturb the conventional con-
structive-editing pattern: Characters launch, leap, and land, but disconcerting pans
make the action harder to capture.
The usual rule of thumb is that long shots must be left onscreen for a longer time
than closer views, because the viewer needs more time to scan a long shot. But one of
the most distinctive qualities of King Hu's editing style is his very brief long shots. What
other director would accompany the voice-over exposition that introduces The Valiant
Ones with 30 shots of documents and landscapes flashing by in about 70 seconds? In
Come Drink With Me, Golden Swallow leaps up to a window in a 25-frame long shot;
cut to an 18-frame long shot showing her passing through the window and landing.
During the celebrated battle that closes the first part of A Touch of Zen, there occurs
a string of long shots of Yang Huizhen soaring, spinning, and bouncing through the
Richness Through Imperfection 427
Figure 15.44 Come Drink With Me. Figure 15.45 Come Drink With Me.
bamboo forest, each shot lasting between 8 and 16 frames. Such brief long shots deny
us the leisure to browse over the image, ensuring that we catch only part of a dense,
often busy scene.12
Hu tends to combine indiscernibility devices like the whip pan and the long shot
with others I've already mentioned. Opaque settings and less than fully informative
compositions will cooperate with the editing in making the action seem miraculously
fleeting. The parallel diving attacks that Wang Shih-Cheng and Peach visit upon
Lee Khan and Wan-Erh are at once decentered long shots and very short shots. Cao's
leap and bounce in Dragon Gate Inn (Figures 15.18-15.20, above) are made obscure
not only by edge framing but also by virtue of being in a long shot that's quite brief.
Directors who rely on fast cutting often expand screen time. This can be accom-
plished by overlapping shots, that is, repeating the same bit of action; Eisenstein's
films provide some famous examples, such as the raising of the bridges in October
(1928). Alternatively, the director can use cutaways to stretch out time, as Peckinpah
and John Woo do. We see a fighter hit and start to fall; cut away to another part of the
battle; cut back to the first fighter, continuing to fall; cut away to another skirmish; cut
back to the first fighter, still falling; and so on. (Howard Hawks criticized Peckinpah,
complaining, "I can kill and bury ten guys in the time it takes him to kill one.")13
But just as King Hu seldom uses slow motion to draw out his fighters' feats, he rarely
resorts to time-stretching cuts. Instead, in accordance with his insistence on the
glimpse, he favors elliptical editing.
Flagrant ellipses in fast-cut passages characterize Hu's style from the start of his
career, as a beautiful moment in Come Drink With Me shows. When Golden Swallow
attacks the corrupt master Tiao Ching-Tang, we see him smack her backward in a close
shot, followed by a shot of her falling away, now several feet off (Figures 15.44-15.45).
The cut from medium shot to long shot has skipped over the time it took her to reel
backward. In Raining in the Mountain, Hu creates weird matches on action by showing
White Fox and Jin leaping up in one spot and landing in quite a different part of the
monastery. At the climax of The Fate of Lee Khan, Cao Yukun (played by Roy Chiao)
turns; we see the princess on the balcony (a 7-frame shot). In the next shot she is already
leaping downward toward him (16 frames), but she's not shown landing. In the next
shot of him, her sword is already in the frame, jabbing at his hand (8 frames).
Typically, then, King Hu shaves moments off the most flamboyant action. Some-
times he can use ellipses for grimly humorous effect. In The Valiant Ones, Wu and
his young pupil return from a lesson to see, in extreme long shot, pirates flail and fall
428 Poetics of Cinema Richness Through Imperfection 429

dead as Wu's wife settles back down to her embroidery. The aesthetic of the glimpse
could hardly go farther; trim any more from the scene, and we would have no fight
at all.
Hu is seldom this laconic. Most often, elliptical cutting mimics the blinding speed
of the protagonists' movements. Instead of floating or rocketing endlessly, as do the
heroes of Swordsman II, King Hu's warriors are usually up and down before you realize
it. When they stay airborne for an extended period, it is usually because they find
clever ways to cheat gravity for a few more instants. That is, instead of using cutting
or slow motion or wirework to extend fleeting actions, Hu prolongs a flight simply
by adding more fleeting actions—a foot striking a tree trunk, or the flash of a body
twisting and spinning to keep aloft a split-second longer. Surprising as it seems, the
aerobatics in the bamboo grove clash in A Touch of Zen aren't sustained by stretching
out time. When the cuts aren't temporally continuous, they are elliptical, showing us
Figure 15.48 A Touch of Zen Figure 15.49 A Touch of Zen.
only fragments of a startlingly wide range of airborne maneuvers.
The interplay of clear presentation and patchy indiscernibility allows Hu to keep
some stylistic elements stable and treat others elliptically. Take the canonical way of across the cut, he often pilots his players through an erratic trajectory: "For the combat
handling high jumps, seen in our One-Armed Swordsman instance: a shot of the hero
scenes, I always shoot a lot of material, some without regard for continuity, in order
starting to leap, followed by a low-angle shot of the leap itself, concluded by a shot
to be able to choose later."15 A character soaring leftward may suddenly plunge down
showing the fighter landing. Hu sharpens this schema in two ways. He often trims a
rightward or somersault clockwise. The discontinuities accentuate the fact that this
bit off each phase, as in Hsiao's counterattack on Cao at the end of Dragon Gate Inn. At
film will provide mere glimpses of the action.
other moments, as we've seen, Hu simply deletes one or two phases. So in the balcony
Yet in Hu's hands, principles of orthodox cutting can yield abstract pattern-
scene of Lee Khan and the final sequence of Raining in the Mountain, he can capture
making. During the forest combat near the end of The Valiant Ones, Xu Lian delivers
women warriors in midleap, without showing launching or landing. He can also, as in
ten thunderous kicks to the chest of Yu Dayou. Nothing in the passage violates any
the inn scenes of Come Drink With Me, delete the flying shot but preserve the depar-
rules of continuity editing, but it remains richly experimental. Hu gives the sequence
ture and landing. In the testing sequence of The Valiant Ones, the second Japanese
its aching force by adjusting the ellipses. A few simple movements—Xu running, leap-
swordsman jumps toward Wu's back, but Hu elides the moment of touchdown; at
ing, and kicking Yu, and Yu falling back and recovering—are shown again and again,
the cut, Wu is already turned and whacking the Japanese aside. In such passages, the
but each kick is cut differently. One of Xu's attacks is rendered in four shots, another
three-stage schema tacitly guides our comprehension, but Hu's condensed treatment
in three, another in five, and another in two. On each cycle a phase may be pro-
suggests action too quick for the orthodox pattern to catch.
longed, accelerated, dwelt on, or skipped over. Whereas kung fu films routinely repeat
Hu measures degrees and kinds of ellipses very subtly. During the martial test of
a crucial action in an instant-replay shot, King Hu gives us, in a mere 45 seconds and
The Valiant Ones, a Japanese swordsman leaps down from a second-story room, and
the movement is handled in straightforward, continuous duration. In reply Wu Jiyuan 32 shots, a staggering ten variations on how constructive editing can create ellipses.
leaps toward him, and this jump is treated in highly elliptical fashion, as if already At the same time, the style is fully expressive: The rhythmic repetitions capture the
marking him as superior to his foe. But when the renegade Chinese master Xu Lian pitiless efficiency of Xu's onslaughts.
hurries to confront Wu, his rush is presented in dazzling ellipses, announcing him as
Wu's equal. A Touch of Zen conveys the force of Xu Xianchun through a bold jump Glimpses of Marvels
cut carrying him from the background (Figure 15.46) to a confrontation with the
monk Hui Yan (Figure 15.47). Hui counters in an unnervingly powerful way, and the The one-by-one tracking shot and the canons of standard editing stabilize King
director obliges with a fresh presentational option: Hui smites Hsu in the forehead Hu's scenes, but the opacities of setting, the play with the bounding frame, the
(Figure 15.48) and knocks him back several feet—no ellipsis now—to where he had overinformative long shots, the disorienting whip pans, and the elliptical cutting
started (Figure 15.49). create a teasing degree of indiscernibility. In trying to understand this dynamic, I've
In sequences like these, King Hu moves toward a kind of pure cinema in which the pulled these devices apart and spread them out. But of course they're often combined
pattern and pacing of the images command attention as much as the story action. 14 in particular sequences, as we have seen from the start in A Touch of Zen s weedy
Although his contemporaries carefully preserve screen direction and match movement duel between Yang and the stranger. A typical King Hu combat will use one-by-one
430 Poetics of Cinema
Figure 15.50 Escort Over Tiger Hills (1969).
tracking shots and orthodox constructive cutting to establish or reestablish the
fighters in the location, whereas the tactics of imperfection will be used to build the
excitement and express abruptness and power. Hu's "richness of stylistic delivery
comes from a dynamic of stability and instability.
I've also had to ignore how each film distinctively develops a cluster of techniques.
For example, in A Touch of Zen the duel in the fortress ruins seems to mark a phase in
the film's overall progress toward a full, awesome revelation of the warriors powers.
In The Fate of Lee Khan, early fights are shown with great clarity, using long takes and
depth, whereas later combats become markedly more elliptical until the final burst of
kinetic abstraction. Raining in the Mountain moves from a leisurely opening to games of
hide-and-seek in the monastery before climaxing in an orgy of wildly elliptical cutting
as concubines rain down from the rocks. A complete analysis of King Hu's technique
would have to take into account such stylistic progressions within each film.
Nonetheless, we've gone far enough to appraise his originality. From today s
vantage point, it seems apparent that he had virtually no disciples. Quite quickly, his
innovations seem to have been registered and reworked. Escort Over Tiger Hills (1969),
a remarkable Cathay picture by Wang Xinglei, cranks up Hu's aesthetic of disorien-
tation (Figure 15.50). The future, however, belonged to a style that made construc-
tive editing ever more crisp, legible, and expressive. Lau Kar-Leung, Yuen Woo Ping,
Sammo Hung, Jackie Chan, Ching Siu-Tung, Yuen Kuei, Tsui Hark, and many other
creators ingeniously refined and sharpened the norms laid down in the 1960s and
1970s. These were norms that King Hu partly assimilated and partly rejected, the
better to experiment with an aesthetic of the glimpse. The mainstream style has given
us many beautiful and stirring films, but Hu's eccentric explorations evoke something
that other directors' works seldom arouse: a sense that extraordinary physical achieve-
ment, if caught through precisely adjusted imperfections, becomes marvelous.
Notes
Introduction
1. The whole text is now available at "Trans: A Visual Culture Conference at the Univer-
sity of Wisconsin-Madison," http://www.visualculture.wisc.edu/Conference/call.htm.
2. Donna Haraway, "The Promises of Monsters: A Regenerative Politics for Inappro-
priate/d Others," in Cultural Studies, ed. Lawrence Grossberg, Cary Nelson, and
Paula Treichler (New York: Routledge, 1992), 297.
3. Examples are Sarah Blaffer Hrdy, The Woman That Never Evolved (Cambridge,
Mass.: Harvard University Press, 1981); Sarah Blaffer Hrdy, Mother Nature:
Maternal Instincts and How They Shape the Human Species (New York: Ballantine,
1999); Alison Jolly, Lucy's Legacy: Sex and Intelligence in Human Evolution
(Cambridge, Mass.: Harvard University Press, 1999); Anne Campbell, A Mind of
Her Own: The Evolutionary Psychology of Women (New York: Oxford University
Press, 2002), especially 1-33; and Griet Vandermassen, Who's Afraid of Charles
Darwin? Debating Feminism and Evolutionary Theory (Lanham, MD: Rowman &
Littlefield, 2005).
4. Examples of work in this tradition are Ingo Rentschler, Barbara Herzberger, and
David Epstein, eds., Beauty and the Brain: Biological Aspects of Aesthetics (Basel,
Switzerland: Birkhauser Verlag, 1988); Joseph Carroll, Evolution and Literary
Theory (Columbia: University of Missouri Press, 1995); Joseph Carroll, Literary
Darwinism: Evolution, Human Nature, and Literature (New York: Routledge, 2004);
Ellen Dissayanake, Homo Aestheticus: Where Art Comes From and Why (Seattle:
University of Washington Press, 1995); Ellen Dissayanake, Art and Intimacy: How
the Arts Began (Seattle: University of Washington Press, 2000); Robert Storey,
Mimesis and the Human Animal: On the Biogenetic Foundations of Literary Rep-
resentation (Evanston: Northwestern University Press, 1996); and Brett Cooke and
Frederick Turner, eds., Biopoetics: Evolutionary Explorations in the Arts (Lexington,
Ky.: Paragon House, 1999). For overviews, see Joseph Carroll, "Adaptationist Literary
Study: An Emerging Research Program," Style 36, no. 2 (Winter 2002): 596-617;
and Brian Boyd, "Literature and Evolution: A Bio-Cultural Approach," Philosophy
431
432 Notes
and Literature 29 (2005): 1-23. See also Boyd's "Reduction or Expansion? Evolution
Meets Literature" (keynote address at the International Society for the Empirical
Studies of Literature Conference in Edmonton, Alberta, August 7, 2004).
5. Jonathan Gottschall and David Sloan Wilson, "Introduction: Literature—a Last
Frontier in Human Evolution Studies," in The Literary Animal: Evolution and the
Nature of Narrative, ed. Jonathan Gottschall and David Sloan Wilson (Evanston:
Northwestern University Press, 2005), xiv. See also, in the same volume, Dylan
Evans, "From Lacan to Darwin," in which a Lacanian recalls his search for an
academic venue in which the master's ideas could be examined critically (44-45).
6. Laurie Goodstein, "Teaching of Creationism Is Endorsed in New Survey," New York
Times, August 31,2005, sec. A, 9. See also "Evolution and Schools: Intelligent Design
Rears Its Head," The Economist, July 30, 2005, 30-31.
Americans' staggering ignorance on the subject of evolution might have one
benefit if it wakes up humanists to the stakes in the game. The assault of intelligent
design (ID) proponents on school boards might encourage academics to lose enthu-
siasm for notions of "situated knowledges." For who is to say that the beliefs about
the origin of life held by fundamentalist Christians in small towns are less valid
than creation stories told by Navajos? Don't both need to be respected as authentic
expressions of local culture? Perhaps academics look more charitably on supersti-
tions when they're entertained by more exotic cultures. The prospect of ID insinuat-
ing itself into suburban biology courses might kindle a new recognition that beliefs,
no matter how strongly clung to, aren't tantamount to knowledge.
7. For a swift and entertaining rebuttal to relativistic claims, see Ophelia Benson and
Jeremy Stangroom, Why Truth Matters (New York: Continuum, 2006), 45-64.
8. Most humanists' rejection of science's claims to truth appears to stem from an
undergraduate reading of Thomas Kuhn's 1962 book, The Structure of Scientific
Revolutions (Chicago: University of Chicago Press, 1962), and little since. There is
a vast tradition in the philosophy of science of arguing with Kuhn; recent critiques
can be found in Philip Kitcher, The Advancement of Science: Science Without
Legend, Objectivity Without Illusions (New York: Oxford University Press, 1993);
Lewis Wolper, The Unnatural Nature of Science (Cambridge, Mass.: Harvard
University Press, 1993); Roger G. Newton, The Truth of Science: Physical Theories
and Reality (Cambridge, Mass.: Harvard University Press, 1997); Peter Carruthers,
Stephen Stich, and Michael Siegal, eds., The Cognitive Basis of Science (Cambridge:
Cambridge University Press, 2002); and Susan Haack, Defending Science—Within
Reason: Between Scientism and Cynicism (Amherst, N.Y.: Prometheus, 2003).
9. See my "Film and the Historical Return (March 2005)," http://www.davidbordwell.net/
essays/return.php.
10. For a complete bibliography, see http://www.davidbordwell.net/cv.php.
Chapter 1
1. Rene Wellek and Austin Warren, Theory of Literature (New York: Harcourt, Brace,
1949); and Stanley Edgar Hyman, The Armed Vision: A Study in the Methods of
Modern Literary Criticism (New York: Knopf, 1948).
Notes 433
2. See, for example, Stanley Edgar Hyman, The Critical Performance (New York:
Vintage, 1956), Sheldon P. Zittner's Practices of Modern Literary Scholarship
(Glenview, 111.: Scott, Foresman, 1966), and Mary Ann Caws, ed., Textual Analysis:
Some Readers Reading (New York: Modern Language Association, 1986).
3. Bill Nichols, ed., Movies and Methods: An Anthology, vol. 1 (Berkeley: University of
California Press, 1976); and Bill Nichols, ed., Movies and Methods: An Anthology,
vol. 2 (Berkeley: University of California Press, 1985). TV studies followed the trend,
as in Robert C. Allen, ed., Channels of Discourse: Television and Contemporary
Criticism (Chapel Hill: University of North Carolina Press, 1987).
4. For more on the practical reasoning characteristic of film interpretation, see my
Making Meaning: Inference and Rhetoric in the Interpretation of Cinema (Cambridge,
Mass.: Harvard University Press, 1989).
5. Igor Stravinksy, Poetics of Music in the Form of Six Lessons (New York: Vintage,
1956); Alexander Tzonis and Liane Lefaivre, Classical Architecture: The Poetics of
Order (Cambridge, Mass.: MIT Press, 1987); Herbert Eagle, ed. and trans., Russian
Formalist Film Theory (Ann Arbor: Michigan Slavic Publications, 1981); and
Richard Taylor, ed., The Poetics of the Cinema (Oxford: Russian Poetics in Transla-
tion Publications, 1982). In recent years, poetics has become a minor term in cultural
studies, as witness books like Adam Krims, Rap Music and the Poetics of Identity,
New Perspectives in Music History and Criticism no. 5 (Cambridge: Cambridge
University Press, 2000); and William Echard, Neil Young and the Poetics of Energy
(Bloomington: Indiana University Press, 2005).
For surveys of the tradition of poetics in literature, see R. S. Crane, "Critical
and Historical Principles of Literary History," in his The Idea of the Humanities
and Other Essays Critical and Historical, vol. 2 (Chicago: University of Chicago
Press, 1967), 45-156; Earl Miner, Comparative Poetics: An Intercultural Essay on
Theories of Literature (Princeton, N.J.: Princeton University Press, 1990); and
Lubomir Dolezel, Occidental Poetics: Tradition and Progress (Lincoln: University
of Nebraska Press, 1990).
6. Gerard Genette, Narrative Discourse: An Essay in Method, trans. Jane E. Lewin
(Ithaca, N.Y.: Cornell University Press, 1980).
7. Samuel Johnson, excerpt from Rasselas, in Criticism: The Major Texts, ed. Walter
Jackson Bate (New York: Harcourt Brace, 1952), 207.
8. Samuel Taylor Coleridge, "On the Principles of Genial Criticism Concerning the
Fine Arts," in ibid., 364.
9. W. H. Auden, "Reading," in his The Dyer's Hand and Other Essays (New York:
Random House, 1962), 8.
10. W. H. Auden, "The Guilty Vicarage," in ibid., 150.
11. Andre Bazin, "The Evolution of the Language of Cinema," in his What Is Cinema?
vol. 1, trans, and ed. Hugh Gray (Berkeley: University of California Press, 1967),
23-40.
12. See my On the History of Film Style (Cambridge, Mass.: Harvard University Press,
1997), 68-75.
13. David Bordwell, Janet Staiger, and Kristin Thompson, The Classical Hollywood
Cinema: Film Style and Mode of Production to 1960 (New York: Columbia University
Press, 1985), 3-11, 70-84, 87-112, 243-61. Noel Carroll argues at a more abstract
level for such a model in "Film History and Film Theory: An Outline for an Institu-
tional Theory of Film," in his Theorizing the Moving Image (Cambridge: Cambridge
University Press, 1996), 375-91.
 Notes
434
14. On this point, see Richard Evans, In Defence of History (London: Granta, 2000),
73-74.
15. See Rick Altman, ed., Genre: The Musical (London: Routledge and Kegan Paul, 1981);
Rick Altman, The American Film Musical (Bloomington: Indiana University Press,
1987); and Donald Crafton, "Animation Iconography: The Hand of the Artist,"
Quarterly Review of Film Studies 4, no. 3 (Fall 1979): 409-28. Broader theories of
film genre in the spirit of a cinepoetics can be found in Rick Altman, Film/Genre
(London: British Film Institute, 1999); and Steve Neale, Genre and Hollywood
(London: Routledge, 2000).
16. See my Ozu and the Poetics of Cinema (London: British Film Institute, 1988),
45-50.
17. This was pointed out by Kristin Thompson in her essay, "Closure Within a Dream?
Point of View in Laura," in her Breaking the Glass Armor: Neoformalist Film Criti-
cism (Princeton, N.J.: Princeton University Press, 1988), 162-94.
18. See Alexander Zholkovsky, Themes and Texts: Towards a Poetics of Expressiveness
(Ithaca, N.Y.: Cornell University Press, 1984); and Yuri Shcheglov and Alexander
Zholkovsky, Poetics of Expressiveness: A Theory and Applications (Amsterdam:
Benjamins, 1987).
19. Yuri Shcheglov, "A Generative Approach to Thematics: Poetics of Expressiveness
and Modern Criticism," in The Return of Thematic Criticism, ed. Werner Sollors
(Cambridge, Mass.: Harvard University Press, 1993), 73-74.
20. See Michael Riffaterre, Text Production, trans. Terese Lyons (New York: Columbia
University Press, 1983), 75-89. See also Michael Riffaterre, Semiotics of Poetry
(Bloomington: Indiana University Press, 1978).
21. See David Bordwell, The Cinema of Eisenstein (New York: Routledge, 2005),
chs. 3-5.
22. Noel Carroll, "Film, Rhetoric, and Ideology," in his Theorizing the Moving Image,
280-85.
23. See Seymour Chatman, Story and Discourse: Narrative Structure in Fiction and Film
(Ithaca, N.Y.: Cornell University Press, 1978); and Edward R. Branigan, Point of View
in the Cinema: A Theory of Narration and Subjectivity in Classical Film (The Hague:
Mouton, 1984).
24. For discussion of these nonnarrative constructive principles, see David Bordwell and
Kristin Thompson, Film Art: An Introduction, 8th ed. (NewYork: McGraw-Hill, 2006),
ch. 10.
25. P. Adams Sitney, Visionary Film: The American Avant-Garde 1943-2000, 3rd ed.
(New York: Oxford University Press, 2002); in a similar vein is Paul Arthur, Line of
Sight: American Avant-Garde Film Since 1965 (Minneapolis: University ofMinnesota
Press, 2005).
26. Among many examples, see Graham Bruce, Bernard Herrmann: Film, Music, and
Narrative (Ann Arbor, Mich.: UMI Research Press, 1985); Kathryn Kalinak, Settling
the Score: Music and the Classical Hollywood Film (Madison: University of Wiscon-
sin Press, 1992); Jeff Smith, The Sounds of Commerce: Marketing Popular Film Music
(New York: Columbia University Press, 1998); James Buhler, Caryl Flinn, and David
Neumeyer, eds., Music and Cinema (Hanover, Conn.: Wesleyan University Press,
2000; and Pamela Robertson Wojcik and Arthur Knight, eds., Soundtrack Available:
Essays on Film and Popular Music (Durham, N.C.: Duke University Press, 2001).
Notes 435
Detailed insights into scoring practices are provided in Fred Karlin and Rayburn
Wright, On the Track: A Guide to Contemporary Film Scoring, 2nd ed. (New York:
Routledge, 2004).
27. For survey accounts, see Michael Podro, The Critical Historians of Art (New Haven,
Conn.: Yale University Press, 1982); and Michael Ann Holly, Panofsky and the
Foundations of Art History (Ithaca, N.Y.: Cornell University Press, 1984).
28. See, for comprehensive surveys, Victor Erlich, Russian Formalism: History, Doctrine,
3rd ed. (The Hague: Mouton, 1969); and F. W. Galan, Historic Structures: The Prague
School Project 1928-1946 (Austin: University of Texas Press, 1985).
29. Noel Carroll discusses the virtues of a question-centered approach to theorizing in
the last section of his Mystifying Movies: Fads and Fallacies of Contemporary Film
Theory (New York: Columbia University Press, 1988), 226-34.
30. Viktor Shklovsky, "In Defence of the Sociological Method," Russian Poetics in
Translation 4 (1977): 94.
31. Stephen Jay Gould, "Cardboard Darwinism," New York Review of Books, September
25, 1986, 47.
32. Quoted in Hans Christian von Baeyer, Information: The New Language of Science
(London: Orion, 2003), 30.
33. C. Wright Mills, The Sociological Imagination (London: Oxford University Press,
1959), 128.
34. See my essay, "Contemporary Film Studies and the Vicissitudes of Grand Theory,"
in Post-Theory: Reconstructing Film Studies, ed. David Bordwell and Noel Carroll
(Madison: University of Wisconsin Press, 1996), 26-30.
35. Boris M. Ejxenbaum, "O. Henry and the Theory of the Short Story," in Readings
in Russian Poetics: Formalist and Structuralist Views, ed. Ladislav Matejka and
Krystyna Pomorska (Cambridge, Mass.: MIT Press, 1971), 269.
36. On the surrealists' "irrational enlargement" of a moment in a movie, see the
Surrealist Group, "Data Toward the Irrational Enlargement of a Film: The Shanghai
Gesture, in The Shadow and Its Shadow, 3rd ed., ed. Paul Hammond (San Francisco:
City Lights, 2000), 121-29. A modern proponent is Robert B. Ray in The Avant-Garde
Finds Andy Hardy (Cambridge, Mass.: Harvard University Press, 2002). See also
Christian Keathley, Cinephilia and History, or the Wind in the Trees (Bloomington:
Indiana University Press, 2005).
37. See my Ozu and the Poetics of Cinema, chs. 5 and 6; and ch. 3 of my Figures Traced
in Light: On Cinematic Staging (Berkeley: University of California Press, 2004).
38. For a review of these positions, see Ben Brewster and Lea Jacobs, Theatre to Cinema:
Stage Pictorialism and the Early Feature Film (Oxford: Oxford University Press,
1997), 3-17.
39. For a general discussion, see my On the History of Film Style, 158-98; and ch. 2 of my
Figures Traced in Light.
40. Charles Rosen, The Classical Style: Haydn, Mozart, Beethoven, expanded ed. (New
York: Norton, 1997), 57-69.
41. On this practice, see my The Way Hollywood Tells It: Story and Style in Modern
Movies (Berkeley: University of California Press, 2006), 117-89.
42. See Bordwell, Staiger, and Thompson, The Classical Hollywood Cinema-, David
Bordwell, Narration in the Fiction Film (Madison: University of Wisconsin Press,
1985), pt. 3; and Thompson, Breaking the Glass Armor.
43. See Bordwell, Staiger, and Thompson, The Classical Hollywood Cinema, 341-52.
44. Thompson, Breaking the Glass Armor, chs. 3-4, 8-12.
436 Notes
45. Kristin Thompson and David Bordwell, "Space and Narrative in the Films of Ozu,"
Screen 17, no. 2 (Summer 1976): 55-64,66-70; Thompson, Breaking the Glass Armor,
ch. 12; and Bordwell, Ozu and the Poetics of Cinema, chs. 5-7.
46. For a survey of filmmakers' conceptions of cinematic expression, see Jacques
Aumont, Les theories des cineastes (Paris: Nathan, 2002).
47. For examples, see Bordwell, Staiger, and Thompson, The Classical Hollywood
Cinema; Bordwell, Ozu and the Poetics of Cinema, 17-50; Lutz Bacher, Max Ophuls
in the Hollywood Studios (New Brunswick, N.J.: Rutgers University Press, 1996);
and Emmanuel Grimaud, Bollywood Film Studio ou comment les films se font a
Bombay (Paris: CNRS, 2003). John Caldwell has been in the forefront of exploring
craft practices in modern media. See John Caldwell, Production Culture: Industrial
Reflexivity and Critical Practice in Film/Television (Durham, N.C.: Duke University
Press, forthcoming); John Caldwell, "Convergence Television: Aggregating Form
and Repurposing Content in the Culture of Conglomeration," in Television After TV,
ed. Lynn Spigel and Jan Olsson (Durham, N.C.: Duke University Press, 2004), 41-74;
and John Caldwell, "Industrial Geography Lessons: Socio-Professional Rituals and
the Borderlands of Production Culture," in Media/Space: Place, Scale, and Culture
in a Media Age, ed. Nick Couldry and Anna McCarthy (London: Routledge, 2004),
163-89.
48. See Bordwell, Staiger, and Thompson, Classical Hollywood Cinema, 125-27, 137-39.
49. John Madden, quoted in "One to One: Directors John Madden and Jonny Campbell,"
Screen International, May 27, 2005, 22.
50. Alissa Quart, "Networked," Film Comment 41, no. 4 (July-August 2005): 48-51.
51. For example, Martin Walker interprets Forrest Gump as reflecting the Clinton
administration in "Making Saccharine Taste Sour," Sight and Sound 10, no. 10
(October 1994): 16-17. A more recent instance is A. O. Scott, "Reading From Left to
Right," New York Times, September 25, 2005, sec. 2,1, 35.
52. To take a common instance: Group anxieties are often evoked as explanatory forces,
especially by scholars who hold that urban modernity altered film and its audi-
ences. For a skeptical critique, see Alan Hunt, "Anxiety and Social Explanation:
Some Anxieties About Anxiety," Journal of Social History 32, no. 3 (Spring 1999):
509-28.
53. I propose a more detailed account of what follows in Ozu and the Poetics of Cinema,
,ch. 8.
54. Virginia Yates, "'Rope' Sets a Precedent," American Cinematographer 29, no. 7
(July 1948): 230.
55. I believe that V. F. Perkins was the first critic to point this out. See his essay "Rope,"
Movie, no. 7 (1963): 35.
56. Although the last two shots are comparatively brief, they still add up to more than
10 minutes, and with the final credits the footage runs a bit longer. So the final pro-
jection reel would also have been a 2,000-foot one in most theatrical situations.
The reel-break list for the film is reprinted in Dan Auiler, Hitchcock's Secret
Notebooks: An Authorised and Illustrated Look Inside the Creative Mind of Alfred
Hitchcock (London: Bloomsbury, 1999), 483. The memo's layout of the reels corres-
ponds to mine except that the opening shot isn't included. For some reason, the
memo's marginal notations about running times are quite inaccurate. Particularly
striking is the note stating that the last shot is planned to run for 2 minutes; but in
the finished film, it runs 340 seconds, nearly 6 minutes. Perhaps the estimates were
Notes 437
made early in production; the total running time listed on the memo, 95 minutes,
may reflect the notion circulated before the film's release that the picture would run
90 minutes or so.
57. Besides shortening the shots, Hitchcock claimed that he accelerated the actors' per-
formances in the last two reels. See his remarks in R. M. Therond and J-C. Tacchella,
"Hitchcock se confie," L'Ecran frangais, no. 187 (January 25,1949): 3-4.
58. We might consider the possibility that two of the cuts might correspond to act
breaks in Patrick Hamilton's original three-act play. But none does. Interestingly, the
anonymous novelization of the film contains 12 chapters; whereas some of the film's
cuts coincide fairly closely with chapter ends, others don't. See Alfred Hitchcock's
Rope (New York: Dell, 1948), 14, 35, 50, 69, 85, 99, 114, 130, 139, 144.
59. This cut is also the most disruptive, because positions aren't matched accurately.
When the camera retreats from Brandon's back, Rupert is now much closer to the
right doorframe than he was before the cut.
60. Alfred Hitchcock, "Direction" (1937), in Hitchcock on Hitchcock: Selected Writings and
Interviews, ed. Sidney Gottlieb (Berkeley: University of California Press, 1995), 255.
61. The average shot lengths for his 1940s features are as follows: Rebecca (1940),
9.1 seconds; Foreign Correspondent (1940), 6.8 seconds; Mr. and Mrs. Smith (1941),
7.6 seconds; Suspicion (1941), 8.7 seconds; Saboteur (1942), 6.7 seconds; Shadow
of a Doubt (1943), 8.4 seconds; Lifeboat (1944), 6.4 seconds; Spellbound (1945),
9.1 seconds; Notorious (1946), 6.8 seconds; The Paradine Case (1947), 7.3 seconds;
Rope (1948), 7.3 minutes; and Under Capricorn (1949), 43.9 seconds. With Stage
Fright (1950, 8.7 seconds) and subsequent features, Hitchcock resumed cutting in
his pre-Rope range.
62. See Leonard J. Leff, Hitchcock and Selznick: The Rich and Strange Collaboration of
Alfred Hitchcock and David O. Selznick in Hollywood (New York: Weidenfeld and
Nicolson, 1987), 78-79.
63. Although Macbeth was released in October 1948, after Rope, it was shot in June
and July 1947, half a year before Hitchcock's project went into production. Welles
claimed that the initial footage of Macbeth included several long takes, "never
shorter than five minutes and often right up to a full reel in length. I think about
five reels were like that—in other words, without cuts." See Orson Welles and Peter
Bogdanovich, This Is Orson Welles (New York: HarperCollins, 1992), 212; see also
Jonathan Rosenbaum's comments in the same volume, 509-10.
64. Herb A. Lightman, "The Fluid Camera," American Cinematographer 27, no. 3
(March 1946): 103.
65. Bart Sheridan, "Three and a Half Minute Take . . . , " American Cinematographer
29, no. 9 (September 1948): 304. In the finished film, this take was interrupted by
three inserted singles; what remains of it are two shots running 57 and 48 seconds.
Many of Paradine's most elaborate takes didn't make it into the final film intact
because Selznick deleted them or demanded a breakdown into shorter shots. See
Patrick McGilligan, Alfred Hitchcock: A Life in Darkness and Light (New York:
HarperCollins, 2003), 394-95.
66. Quoted in Sheridan, "Three and a Half Minute Take . . . , " 305.
67. Screenwriter Arthur Laurents, quoted in aRope Unleashed," documentary supple-
ment to Rope, Warner Bros. DVD release 20671, 14:12-14:22.
68. Quoted in Francois Truffaut, Hitchcock (New York: Simon & Schuster, 1967), 130.
69. See Leff, Hitchcock and Selznick, 239.
 Notes
438
70. Production stills for these scenes reveal a jungle of cameras, lighting rigs, micro-
phones, and mixing equipment. See ibid., 253; and Bill Krohn, Hitchcock at Work
(London: Phaidon, 2000), 104.
71. Yates, "'Rope' Sets a Precedent," 231.
72. Quoted in ibid., 246.
73. A trade reviewer commented that an ordinary audience would find the "soaring"
camera movements distracting, and the review in Life noted that Hitchcock was "not
wholly successful" in making them unobtrusive. See "Rope," Variety, September 1,
1948,14; and "Movie of the Week: Rope," Life, July 26,1948,60. V. F. Perkins, in Film
as Film (Harmondsworth, UK: Penguin, 1972), finds that one shot's virtuoso timing
nicely expresses Brandon's self-assured precision (88-89). Hitchcock uses the phrase
"roving camera" in "My Most Exciting Picture" (1948), in Gottlieb, Hitchcock on
Hitchcock, 275.
74. Jack Cardiff explains the grueling process of shooting in "The Problems of Lighting
and Photographing 'Under Capricorn,"' American Cinematographer 30, no. 10
(October 1949): 358-59, 382. For a thoughtful analysis of the film's long takes, and
the ways they differ from those in Rope, see John Belton, "Alfred Hitchcock's Under
Capricorn: Montage Entranced by Mise-en-Scene," Quarterly Review ofFilm Studies
6, no. 4 (Fall 1981): 366-83.
75. Andre Bazin, "Theatre and Cinema," in his What Is Cinema? 1:76-124.
76. See "William Wyler," in Conversations With the Great Moviemakers of Hollywood's
Golden Age at the American Film Institute, ed. George Stevens Jr. (New York: Knopf,
2006), 215.
77. Quoted in McGilligan, Alfred Hitchcock, 469.
78. Action in these zones occupies 94% of Dial M s running time. There are also a few
shots of the street outside and the husband's club, from which he calls his wife
during the murder attempt.
79. Roy Armes, Patterns of Realism: A Study of Italian Neo-Realist Cinema (Cranbury,
N.J.: A. S. Barnes, 1971), 171.
80. For a recent discussion of this process, see Zenon W. Pylyshyn, Seeing and Visual-
izing: It's Not What You Think (Cambridge, Mass.: MIT Press, 2006), especially chs.
2 and 3.
81. An entertaining introduction to fast intuitive judgments is Malcolm Gladwell, Blink:
The Power of Thinking Without Thinking (New York: Little, Brown, 2005). More
formal accounts can be found in David G. Myers, Intuition: Its Power and Perils
(New Haven, Conn.: Yale University Press, 2002); Timothy D. Wilson, Strangers to
Ourselves: Discovering the Adaptive Unconscious (Cambridge, Mass.: Harvard Univer-
sity Press, 2003); and Ran R. Hassin, James S. Uleman, and John A. Bargh, The New
Unconscious (New York: Oxford University Press, 2005). See also Nalini Ambady and
Robert Rosenthal, "Thin Slices of Expressive Behavior as Predictors of Interpersonal
Consequences: A Meta-Analysis," Psychological Bulletin 111, no. 2 (1992): 256-74.
82. For an overview of mirror neurons, see V. S. Ramachandran, "Mirror Neurons
and Imitation Learning as the Driving Force Behind 'the Great Leap Forward' in
Human Evolution (June 1, 2000)," http://www.edge.org/3rd _culture/ramachandran/
ramachandran_pl.html.
83. In the long run, the discovery of mirror neurons is likely to refine the cognitive
perspective. Instead of treating thought as higher or central processes that rework
distant sensory stimulations, we may discover that some garden-variety thought,
including mental representation, takes place at the level of cells and cell networks.
Notes
439
84. What follows is a model I have developed in my teaching work since the mid -1980s. For
a more detailed account congruent with it, see Per Persson, Understanding Cinema:
A Psychological Theory of Moving Images (Cambridge: Cambridge University Press,
2003), 26-34. A pioneering study in this vein is Joseph Anderson, The Reality of
Illusion: An Ecological Approach to Cognitive Film Theory (Carbondale: University
of Southern Illinois Press, 1996).
85. See for example Colwyn Trevarthen, "Communication and Cooperation in Early
Infancy: A Description of Primary Intersubjectivity," in Before Speech: The Begin-
ning of Human Communication, ed. Margaret Bullowa (Cambridge: Cambridge
University Press, 1979), 321-47.
86. Paul Messaris, Visual Literacy: Image, Mind, and Reality (Boulder, Colo.: Westview,
1994), chs. 3 and 4.
87. S. Mineka and M. Cook, "Mechanisms Involved in the Observational Conditioning
of Fear "Journal of Experimental Psychology: General 122 (1993): 23-38.
88. E. H. Gombrich, "Image and Code: Scope and Limits of Conventionalism in Pictorial
Representation," in his The Image and the Eye: Further Studies in the Psychology of
Pictorial Representation (Ithaca, N.Y.: Cornell University Press, 1982), 289.1 develop
this point in relation to cinema in my Figures Traced in Light, 258-60.
89. Janet Staiger, Perverse Spectators: The Practices of Film Reception (New York:
New York University Press, 2000), 37.
90. See my Making Meaning, 224-48.
91. See Patrick Colm Hogan, The Mind and Its Stories: Narrative Universals and Human
Emotion (Cambridge: Cambridge University Press, 2003), 101-9.
92. Murray Smith, Engaging Characters: Fiction, Emotion, and the Cinema (Oxford:
Oxford University Press, 1995); Ed Tan, Emotion and the Structure of Narrative
Film: Film as an Emotion Machine (Teaneck, N.J.: Erlbaum, 1996); Torben Grodal,
Moving Pictures: A New Theory of Film Genres, Feelings, and Cognition (Oxford:
Oxford University Press, 1999); Carl Plantinga and Greg Smith, eds., Passionate
Views: Film, Cognition and Emotion (Baltimore: Johns Hopkins University Press,
1999); and Greg M. Smith, Film Structure and the Emotion System (Cambridge:
Cambridge University Press, 2003).
93. Robert H. Frank, Passions Within Reason: The Strategic Role of the Emotions
(New York: Norton, 1988).
94. An eloquent review of this research can be found in Ellen Dissayanake, Art and
Intimacy: How the Arts Began (Seattle: University of Washington Press, 2000), 26-42.
95. See Paul E. Griffiths, What Emotions Really Are: The Problem of Psychological
Categories (Chicago: University of Chicago Press, 1997), 122-25.
96. Hillary Anger Elfenbein and Nalini Ambady, "When Familiarity Breeds Accuracy:
Cultural Exposure and Facial Emotion Recognition," Journal of Personality and
Social Psychology 85, no. 2 (2003): 276-90.
97. On sympathy for protagonists, see Michael Hauge, Writing Screenplays That Sell
(New York: Harper, 1988), 44-49; and on hope and fear, see David Howard, How
to Build a Great Screenplay (New York: St. Martin's, 2006), 52-56. Patrick Keating
makes a strong case for Howard's position in "Emotional Curves and Linear Narra-
tives," The Velvet Light Trap no. 58 (Fall 2006): 4-15.
98. Smith, Film Structure and the Emotion System, chs. 1-4.
99. Ben Singer, "A Taxonomy of Pathos" (paper delivered at a convention of the Center
for the Cognitive Study of the Moving Image, Calvin College, Grand Rapids, Mich.,
July 22, 2004).
440 Notes
100. Hogan, The Mind and Its Stories, 86-101.
101. For collections that exemplify the range of cinematic cognitivism, see "Cinema et
cognition," special issue of Cinemas 12, no. 2 (Winter 2002); and Joseph D. Anderson
and Barbara Fisher Anderson, eds., Moving Image Theory: Ecological Consider-
ations (Carbondale: Southern Illinois University Press, 2005). See also Laurent
Jullier, Cinema et Cognition (Paris: L'Harmattan, 2002). The best introduction to
the "cognitive turn" in the humanities is Patrick Colm Hogan, Cognitive Science,
Literature, and the Arts: A Guide for Humanists (New York: Routledge, 2003). On
literature, see also Mark Turner, The Literary Mind: The Origins of Thought and Lan-
guage (New York: Oxford University Press, 1998); Peter Stockwell, Cognitive Poetics:
An Introduction (New York: Routledge, 2002); and Joanna Gavins and Gerard Steen,
Cognitive Poetics in Practice (New York: Routledge, 2003). Many essays on cognitive
approaches to the arts appear in the journal Poetics Today. See in particular the
special issue "The Cognitive Turn? A Debate on Interdisciplinarity," Poetics Today
24, no. 2 (Summer 2003). For a wide-ranging critique of the limited conceptions of
art undergirding many cognitive explorations, see Meir Sternberg's two-part essay,
"Universals of Narrative and Their Cognitivist Fortunes," Poetics Today 24, no. 2
(Summer 2003): 297-395; and Poetics Today 24, no. 3 (Fall 2003): 517-638.
102. See my Ozu and the Poetics of Cinema, 348-54, for fuller discussion.
103. Some scholars elsewhere are taking up a poetics-based framework. See for example
Michel Chion, Technique et creation au cinema: Le livre des images et des sons
(Paris: ESEC, 2002); Michel Chion, Un art sonore, le cinema: Histoire, esthetique,
poetique (Paris: Cahiers du cinema, 2003); Timothy White, "Historical Poetics,
Malaysian Cinema, and the Japanese Occupation," Kinema (Fall 1996), http://www.
kinema.uwaterloo.ca/white962.htm; Warren Buckland, Directed by Steven Spielberg:
Poetics of the Contemporary Hollywood Blockbuster (New York: Continuum, 2006);
Michael Z. Newman, "From Beats to Arcs: Toward a Poetics of TelevisionNarrative,"
The Velvet Light Trap no. 58 (Fall 2006): 16-28.
Chapter 2
This essay has benefited from comments made by audience members in various ven-
ues: the University of Wisconsin-Milwaukee Center for Twentieth Century Studies
in 1991, the 1992 Convention ofthe American Society for Aesthetics, the 1993 Society
for Cinema Studies convention, and the Annenberg School of Communication in
1994.1 also thank Ben Brewster, Noel Carroll, Lea Jacobs, and Kristin Thompson for
their criticisms.
1. We have, however, made progress. This essay has benefited particularly from points
made in several essays collected in Mette Hjort, ed., Rules and Conventions: Litera-
ture, Philosophy, Social Theory (Baltimore: Johns Hopkins University Press, 1992).
2. On the development of the shot/reverse shot in American cinema, see Kristin
Thompson, "The Continuity System," in David Bordwell, Janet Staiger, and Kristin
Thompson, The Classical Hollywood Cinema: Film Style and Mode of Production to
1960, (New York: Columbia University Press, 1985), 208-10.
3. V. I. Pudovkin, Film Technique (1926; reprint, New York: Evergreen, 1970), 70.
4. Barry Salt, Film Style and Technology: History and Analysis (London: Starword,
1983), 164.
Notes 441
5. Karel Reisz and Gavin Millar, The Technique of Film Editing, 2nd ed. (London: Focal
Press, 1968), 215.
6. Ibid.
7. Someone might object that a more sophisticated semiology of cinema would classify
automobile turn signals as Peircean "indexical" signs rather than arbitrary symbols
a la Saussure. After all, to signal a turn is rather like pointing to the direction in
which I wish to go. To analyze this objection fully would take us afield, but let me
quickly record my doubts.
First, I regard the appropriation of the Peircean triad of index-icon-symbol as
suspect because this is only one of the several "trichotomies" that Charles Sanders
Peirce sets forth. He proposes three "trichotomies" of signs in his manuscript
"Nomenclature and Divisions of Triadic Relations, as Far as They Are Determined"
(c. 1903). Three years later, he proposed ten trichotomies, accounting for 66 types of
signs! See Charles Hartsthorne and Paul Weiss, eds., Elements of Logic, vol. 2 of their
Collected Papers of Charles Sanders Peirce (Cambridge, Mass.: Harvard University
Press, 1932), 134-73. To my knowledge, no film theorists have confronted the dizzy-
ing array of sign relations set out within Peirce's overall system, even though this
would seem a prerequisite for any serious assessment of his theory's usefulness for
film study.
Furthermore, we should be chary of appropriating one piece of a conceptual
system; taking that piece out of its context can easily lead to misunderstanding or
inconsistency. For example, the Peircean trichotomy of symbols comports ill with
the hodgepodge semiotics practiced by most film theorists. Peirce remarks that the
Symbol actually denotes "a conventional sign, or one depending upon habit (acquired
or inborn)" (Elements, 167; italics mine). This claim poses problems for those film
semioticians who think of the Peircean Symbol as wholly conventional or arbitrary
("The third category of sign, the symbol, corresponds to Saussure's arbitrary sign.
. . . [The symbol] is conventional"; Peter Wollen, Signs and Meaning in the Cinema
[Bloomington: Indiana University Press, 1972], 123).
By one of Peirce's definitions, an index "refer[s] to an Object... by virtue of being
really affected by that object" (Elements, 143). Among his examples are weather-
vanes and photographs. But my turn signal does not refer to the "Object" right turn
by being affected by that Object, the way that the wind affects the weather vane or
the pattern of ambient light reflected from an object affects the chemically sensi-
tized emulsion of a strip of film.
Finally, let me just register a skepticism about the clarity and precision of the
positions taken in Peirce's writings. Film theorists (and semioticians in general)
write with breezy confidence about his conception of signs, as if his texts (largely
unpublished) were consistent, rigorously argued, and (above all) clear. You will not
find in the film literature a recognition of the puzzles arising from a passage like
this: "A rap on the door is an index. Anything which focuses the attention is an
index. Anything which startles us is an index, in so far as it marks the junction
between two portions of experience" (ibid., 161).
8. For an introduction to this line of argument, see Steven Pinker, The Language
Instinct: How the Mind Creates Language (New York: Morrow, 1994).
9. Some examples are discussed by Arthur Danto in "Description and the Phenom-
enology of Perception," in Visual Theory: Painting and Interpretation, ed. Norman
Bryson, Michael Ann Holly, and Keith Moxey (New York: HarperCollins, 1991),
209-11.
442 Notes
10. For an exploration of the empirical possibilities, see Donald E. Brown, Human
Universals (Philadelphia: Temple University Press, 1991).
11. Robin Horton, "Tradition and Modernity Revisited," in Rationality and Relativism,
ed. Martin Hollis and Steven Lukes (Cambridge, Mass.: MIT Press, 1982), 228.
More recently, Richard Dawkins has proposed the idea that the human mind has
evolved tofit"Middle World," "where the objects that mattered to our survival were
neither very large nor very small." See Richard Dawkins, The God Delusion (Boston:
Houghton Mifflin, 2006), 367.
12. See Roy Harris, Reading Saussure (La Salle, 111.: Open Court, 1987), 64-69.
13. This sense is close to that posited by David Lewis in Convention: A Philosophical Study
(Cambridge, Mass.: 1969). See also Paisley Livingston, Literature and Rationality:
Ideas of Agency in Theory and Fiction (Cambridge: Cambridge University Press,
1991), chs. 1 and 2.
14. The idea that conventions are designed for utility in action echoes Noel Carroll's
argument that many "arbitrary conventions" are in fact cultural inventions aimed at
achieving specific goals. See his Mystifying Movies: Fads and Fallacies in Contempo-
rary Film Theory (New York: Columbia University Press, 1988), 42-144.
15. E. H. Gombrich, "Image and Code: Scope and Limits of Conventionalism in
Pictorial Representation," in his The Image and the Eye (Oxford: Phaidon, 1982), 283.
16. E. H. Gombrich, "Illusion and Art," in Illusion in Nature and Art, ed. R. L. Gregory
and E. H. Gombrich (New York: Scribner's, 1973), 199-213.
17. See Julian Hochberg, "Representation of Motion and Space in Video and Cinematic
Displays," in Sensory Processes and Perception, vol. 1 of Handbook of Perception and
Human Performance, ed. Kenneth R. Boff, Lloyd Kaufman, and James P. Thomas
(New York: Wiley, 1986), 31-40.
18. See E. H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Repre-
sentation, 11th ed. (Princeton, N.J.: Princeton University Press, 2000), 240.
19. Margaret A. Hagen, Varieties of Realism: Geometries of Representational Art
(Cambridge: Cambridge University Press, 1986).
20. See Julian Hochberg, "The Representation of Things and People," in Art, Perception,
and Reality, ed. E. H. Gombrich, Julian Hochberg, and Max Black (Baltimore: Johns
Hopkins University Press, 1972), 67-73.
21. Gombrich, Art and Illusion, 86.
22. One theorist has proposed that "artificial perspective" as usually understood is
just such a package of devices. John Hyman suggests that the distinct techniques
of overlapping, foreshortening, and perspective diminution find "harmonious inte-
gration into a unified system" known as artificial perspective. See John Hyman,
"Perspective," in A Companion to Aesthetics, ed. David Cooper (Oxford: Blackwell,
1992), 324-27.
23. Noel Carroll, "Toward a Theory of Point-of-View Editing: Communication,
Emotion, and the Movies," Poetics Today 14, no. 1 (Spring 1993): 127-31.
24. Quoted in Jennifer Hillner, "J. J. Abrams, Spymaster," Wired, May 2006, 160.
25. See Vicki Bruce, Tim Valentine, and Alan Baddeley, "The Basis of the 3/4 View
Advantage in Face Recognition," Applied Cognitive Psychology 1 (1987): 109-10; and
Robert H. Logie, Alan D. Baddeley, and Muriel M. Woodhead, "Face Recognition,
Pose, and Ecological Validity," Applied Cognitive Psychology 1 (1987): 53-69.
Notes 443
26. Empirical research on these problems suggests that the conventions of shot change
and eyeline matching are quickly grasped at a young age and by members of widely
different cultures. For a lucid discussion, see Paul Messaris, Visual Literacy: Image,
Mind, and Reality (Boulder, Colo.: Westview, 1994), 74-86.
27. For a discussion of this principle, see David Bordwell and Kristin Thompson, Film
Art: An Introduction, 8th ed. (New York: McGraw-Hill, 2006), 231-251.
28. Gombrich, Art and Illusion, 146-91.
29. See Scott Atran, Cognitive Foundations of Natural History: Towards an Anthropology
of Science (Cambridge: Cambridge University Press, 1990).
30. Brown, Human Universals, 135.
31. Matt Ridley, Nature via Nurture: Genes, Experience, and What Makes Us Human
(New York: HarperCollins, 2003), 92-93. See also Peter J. Richerson and Robert
Boyd, Not by Genes Alone: How Culture Transformed Human Evolution (Chicago:
University of Chicago Press, 2005).
32. David Bordwell, "A Case for Cognitivism," Iris, no. 9 (Spring 1989), 11-40, http://
www.davidbordwell.net/articles.
33. When I wrote this essay, I was only slightly aware of the intense debate about human
universals rippling through the social sciences, chiefly in the emerging field of evolu-
tionary psychology. The seminal piece is John Tooby and Leda Cosmides, "The Psy-
chological Foundations of Culture," in The Adapted Mind: Evolutionary Psychology
and the Generation of Culture, ed. Jerome H. Barkow, Leda Cosmides, and John
Tooby (New York: Oxford University Press, 1992), 19-136. In the last decade the
field has exploded, with several introductions available, notably David M. Buss,
Evolutionary Psychology: The New Science of Mind (Needham Heights, MA: Allyn
& Bacon, 1999); Christopher Badcock, Evolutionary Psychology: A Critical Introduc-
tion (Cambridge: Polity, 2002); and Louise Barrett, Robin Dunbar, and John Lycett,
Human Evolutionary Psychology (Princeton, N.J.: Princeton University Press, 2002).
See also Charles Crawford and Dennis L. Krebs, eds., Handbook of Evolutionary
Psychology: Ideas, Issues, and Applications (Mahwah, N.J.: Erlbaum, 1998) and,
most entertainingly, Steven Pinker, The Blank Slate: The Modern Denial of Human
Nature (New York: Viking, 2002). See also John Tooby and Leda Cosmides, "The
Innate Versus the Manifest: How Universal Does Universal Have to Be?" Behav-
ioral and Brain Sciences 12 (1989): 36-37. For a critique of the research program, see
David J. Buller, Adapting Minds: Evolutionary Psychology and the Persistent Quest
for Human Nature (Cambridge, Mass.: MIT Press, 2005). Buller nonetheless grants
considerable force to evolution-based explanations if properly constrained, while
pressing a conception of contingent universals that is close to mine here.
34. Peter Singer, A Darwinian Left: Politics, Evolution and Cooperation (New Haven,
Conn.: Yale University Press, 2000), 34-38.
35. Slavoj Zizek, The Fright of Real Tears: Krzystof Kieslowski Between Theory and Post-
Theory (London: British Film Institute, 2001), ch. 1; David Bordwell, Figures Traced
in Light (Berkeley: University of California Press, 2005), 260-65; David Bordwell,
"Slavoj Zizelc: Say Anything" (April 2005), http://www.davidbordwell.net/essays/
zizek.php; and Miriam Bratu Hansen, "The Mass Production of the Senses: Classi-
cal Cinema as Vernacular Modernism," in Reinventing Film Studies, ed. Christine
Gledhill and Linda Williams (London: Arnold, 2000), 332-50. See my reply to
Hansen in Bordwell, Figures Trace in Light, 275-76.
36. Hansen, "The Mass Production of the Senses," 339.
444 Notes
37. David Bordwell, "La Nouvelle Mission die Feuillade; or, What Was Mise-en-Scene?"
The Velvet Light Trap, no. 37 (1996): 23.
38. Paul Pietroski and Stephen Crain, "Innate Ideas," in The Cambridge Companion
to Chomsky, ed. James McGilvray (Cambridge: Cambridge University Press, 2005),
166-67.
39. Hansen, "The Mass Production of the Senses," 341.
40. David Bordwell, "Foreword," in Moving Image Theory: Ecological Considerations, ed.
Joseph D. Anderson and Barbara Fisher Anderson (Carbondale: Southern Illinois
University Press, 2005), ix-xii.
41. See James J. Gibson, The Perception of the Visual World (1950; reprint, Westport,
Conn.: Greenwood, 1974); James J. Gibson, The Senses Considered as Perceptual
Systems (Boston: Houghton Mifflin, 1966); and James J. Gibson, The Ecological
Approach to Visual Perception (Boston: Houghton Mifflin, 1979). See also Claire
F. Michaels and Claudia Carello, Direct Perception (Englewood Cliffs, N.J: Prentice
Hall, 1981); and David Marr, Vision (San Francisco: Freeman, 1982). Gibson's life is
engagingly recounted by Edward S. Reed in James J Gibson and the Psychology of
Perception (New Haven, Conn.: Yale University Press, 1988).
42. On pictorial conventions, see Noel Carroll, "The Power of Movies," in his Theorizing
the Moving Image (Cambridge: Cambridge University Press, 1996), 80-81.
43. See my "Contemporary Film Studies and the Vicissitudes of Grand Theory," in Post-
Theory: Reconstructing Film Studies, ed. David Bordwell and Noel Carroll (Madison:
University of Wisconsin Press, 1996), 1-36.
44. Torben Grodal, "Film Lighting and Mood," in Moving Image Theory, ed. Anderson and
Anderson, in ibid., 152-163; Dolf Zillmann, "Cinematic Creation of Emotion," in ibid.,
164-179; Dirk Eitzen, "Documentary's Peculiar Appeals," in ibid., 183-199; Ed. S. Tan,
"Three Views of Facial Expression and Its Understanding in the Cinema," in ibid., 107-127;
John M. Kennedy and Dan L. Chiappe, in ibid., "Metaphors in Movies," 228-242.
45. William Evans, "Reality Programming: Evolutionary Models of Film and Television
Viewership," in ibid., 200-214; Claudia Carello, Jeffrey B. Wagman, and Michael
T. Turvey, "Acoustic Specification of Object properties," in ibid., 79-104; Sheena
Rogers, "Through Alice's Glass: The Creation and Perception of Other Worlds in
Movies, Pictures, and Virtual Reality," in ibid., 217-242.
46. See Gerd Gigerenzer, Adaptive Thinking: Rationality in the Real World (New York:
Oxford University Press, 2000).
47. Robert E. Shaw and William M. Mace, "The Value of Oriented Geometry for Eco-
logical Psychology and Moving Image Art," Moving Image Theory, ed. Anderson
and Anderson, in ibid., 28-47; James E. Cutting, "Perceiving Scenes in Film and in
the World," in ibid., 9-27.
48. See James J. Gibson, ed., Motion Picture Testing and Research: Army Air Forces
Aviation Psychology Program Research Report no. 7 (Washington, D.C.: Government
Printing Office, 1947), 181-212, 219-30. Reed explains the importance of these
experiments in James J. Gibson and the Psychology of Perception, 114-80.
Chapter 3
1. The example comes from Caroline Furness James, String Figures and How to Make
Them: A Study of Cat's Cradle in Many Lands (1906; reprint, New York: Dover, 1962),
16-20.
Notes 445
2. For a discussion, see Howard Morphy, Aboriginal Art (London: Phaidon, 1998),
111-21.
3. See Timothy D. Wilson, Strangers to Ourselves (Cambridge, Mass.: Harvard Univer-
sity Press, 2002), 167-73.
4. August William Schlegel, A Course of Lectures on Dramatic Art and Literature,
trans. John Black (London: Bohn, 1846), 362.
5. Michael appears in about 49% of the running time, and Vito in just under 28% of it.
These totals include scenes in which they appear together, but there are surprisingly
few of those. If we simply split the difference and assign half of their two-handed
scenes to one side or the other, we come up with about 48% for Michael and 27% for
Vito. One man or the other is present for 75% of the film's duration.
6. Murray Smith, Engaging Characters: Fiction, Emotion, and the Cinema (Oxford:
Oxford University Press, 1995); Ed Tan, Emotion and the Structure of Narrative
Film: Film as an Emotion Machine (Teaneck, N.J.: Erlbaum, 1996); Torben Grodal,
Moving Pictures: A New Theory of Film Genres, Feelings, and Cognition (Oxford:
Oxford University Press, 1999); Carl Plantinga and Greg Smith, eds., Passionate
Views: Film, Cognition and Emotion (Baltimore: Johns Hopkins University Press,
1999); and Greg M. Smith, Film Structure and the Emotion System (Cambridge:
Cambridge University Press, 2003).
7. See Stephen Halliwell, The Poetics ofAristotle: Translation and Commentary (Chapel
Hill: University of North Carolina Press, 1987), 39. Because the plot is also an
abstraction, an event-structure that can be manifested in different media, Aristotle
further distinguishes between the plot and the play, the drama text that presents it.
See Malcolm Heath, Unity in Greek Poetics (Oxford: Clarendon, 1989), 40.
8. Quoted in Veronique de Turenne, "Shopgirl," Variety, December 18, 2005, http://
www.variety.com/ac2006_article.
9. See Meir Sternberg, Expositional Modes and Temporal Ordering in Fiction
(Baltimore: Johns Hopkins University Press, 1978), 45-55; and Meir Sternberg,
"Telling in Time (II): Chronology, Teleology, Narrativity," Poetics Today 13, no. 3
(Fall 1992): 524-38.
10. Sternberg, Expositional Modes, 129-58.
11. See Smith, Engaging Characters, chs. 3-6; and Murray Smith, "Gangsters, Cannibals,
Aesthetes, or Apparently Perverse Allegiances," in Plantinga and Smith, Passionate
Views, 217-38.
12. David Bordwell, Narration in the Fiction Film (Madison: University of Wisconsin
Press, 1985). A portion of this argument is presented in David Bordwell and Kristin
Thompson, Film Art: An Introduction, 8th ed. (New York: McGraw-Hill, 2006), ch. 3.
13. Meir Sternberg, The Poetics of Biblical Narrative: Ideological Literature and the
Drama of Reading (Bloomington: Indiana University Press, 1985), 172-73.
14. Jane Smiley, 13 Ways of Looking at the Novel (New York: Knopf, 2005), 232-33.
15. Seymour Chatman, Story and Discourse: Narrative Structure in Fiction and Film
(Ithaca, N.Y.: Cornell University Press, 1978), 93-95.
16. The classic work is Georges Polti, The Thirty-Six Dramatic Situations, trans. Lucille
Ray (Eng. orig., 1906; reprint, Boston: The Writer, 1988). For a recent effort, see
Christopher Booker, The Seven Basic Plots: Why We Tell Stories (New York:
Continuum, 2004).
17. E. M. Forster, Aspects of the Novel (New York: Harcourt, Brace, 1927), 153.
446 Notes
18. On Elizabethan parallel plotting, see Richard Levin, The Multiple Plot in English
Renaissance Drama (Chicago: University of Chicago Press, 1971); and Thomas G.
Pavel, The Poetics of Plot: The Case of English Renaissance Drama (Minneapolis:
University of Minnesota Press, 1985). Carol Clover exposes stranding in the Nordic
sagas in The Medieval Saga (Ithaca, N.Y.: Cornell University Press, 1982).
19. Urban Gad, Filmen: Dens Midler og Mai (Copenhagen: Gyldendal, 1919), 247-51.
See also Ben Brewster, "Traffic in Souls: An Experiment in Feature-Length Narrative
Construction," Cinema Journal 31, 1 (Fall 1991): 50n5.
20. See Semyon Timoshenko, Chto dolzhen znat' kino-rezhisser [What a Film Director
Must Know] (Leningrad: Teakinopechat, 1929), 21-23.
21. Anonymous, "Reel Length From the Exchange and Projection Viewpoints:
Discussion," International Projectionist 7, no. 1 (June 1934): 13-15, 21-23.
22. See my Planet Hong Kong: Popular Cinema and the Art ofEntertainment (Cambridge,
Mass.: Harvard University Press, 2000), 180-82, 275, 283.
23. I review this tradition in my The Way Hollywood Tells It: Story and Style in Modern
Movies (Berkeley: University of California Press, 2006), 27-42.
24. See for example, Paul Joseph Gulino, Screenwriting: The Sequence Approach: The
Hidden Structure of Successful Screenplays (New York: Continuum, 2004), 1-19.
25. See Kristin Thompson, Storytelling in the New Hollywood: Analyzing Classical
Narrative Technique (Cambridge, Mass.: Harvard University Press, 1999); and
Kristin Thompson, Storytelling in Film and Television (Cambridge, Mass.: Harvard
University Press, 2003).
26. Quoted in Karl Iglesias, The 101 Habits of Highly Successful Screenwriters: Insider
Secrets From Hollywood's Top Writers (Avon, Mass.: Adams, 2001), 52.
27. He finds two-act and five-act patterns as well. See Francis Vanoye, Scenarios modeles,
modeles de scenarios (Paris: Nathan, 1991), 89-98.
28. Michel Chion, Ecrire un scenario (Paris: Cahiers du cinema, 1985), 144.
29. See for example the manual by Christian Biegalski, Scenarios: Modes d'emploi (Paris:
Dixit, 2003), 164-96.
30. I analyze Memento's four-part layout in my The Way Hollywood Tells It, 78-80. For
an analysis in relation to three-act structure, see Simon Kent, "Fragmented Narra-
tive: Non-Linear Storytelling," Scriptwriter, no. 20 Qanuary 2005): 25-29.
31. See for example Chatman, Story and Discourse, chs. 2 and 3; and Mieke Bal,
Narratology: Introduction to the Theory of Narrative, trans. Christine van Boheemen
(Toronto: University of Toronto Press, 1985), ch. 1.
32. Bordwell, Narration in the Fiction Film, ch. 3.
33. There is some evidence that conceiving of art as mimesis is basic to all art-making
cultures. See Ming Dong Gu, "Is Mimetic Theory in Literature and Art Universal?"
Poetics Today 26, no. 3 (Fall 2005): 459-99.
34. Marie-Laure Ryan, Possible Worlds, Artificial Intelligence, and Narrative Theory
(Bloomington: Indiana University Press, 1991), 51.
35. Ibid., 52.
36. Chatman, Story and Discourse, 117.
37. Gustav Freytag, Freytag's Technique of the Drama, trans. Elias J. MacEwain (1863;
reprint, New York: Benjamin Blom, 1968), 249.
38. I discuss the person schema's role in characterization in more detail in my Making
Meaning (Cambridge, Mass.: Harvard University Press, 1989), 151-57. That
discussion owed a good deal to Roy D'Andrade, "A Folk Model of the Mind," in
Notes 447
Cultural Models in Language and Thought, ed. Dorothy Holland and Naomi Quinn
(Cambridge: Cambridge University Press, 1987), 112-48. Murray Smith has explored
my proposal more extensively in Engaging Characters, 20-35.
39. See Alison Gopnick, Andrew N. Meltzoff, and Patricia K. Kuhl, The Scientist in
the Crib: Minds, Brains, and How Children Learn (New York: Morrow, 1999), for a
useful introduction to the research.
40. Marc Vernet, "Le personnage du film," Iris, no. 7 (1986): 108-11. See also Bal,
Narratology, 85-89.
41. For a clear discussion, see David G. Myers, Intuition: Its Power and Perils (New
Haven, Conn.: Yale University Press, 2002), 110-15.
42. See Richard Gerrig and David Allbritton, "The Construction of Literary Character:
A View From Cognitive Psychology," Style 24, no. 3 (Fall 1990): 381-83.
43. Michael Newman, "Characterization in American Independent Cinema" (Ph.D.
diss., University of Wisconsin-Madison, 2005), 136-50.
44. These heuristics are drawn from discussions in Daniel Kahneman, Paul Slovic, and
Amos Tversky, eds., Judgment Under Uncertainty: Heuristics and Biases (Cambridge:
Cambridge University Press, 1982); and Richard Nisbett and Lee Ross, Human
Inference: Strategies and Shortcomings of Social Judgment (Englewood Cliffs, N.J.:
Prentice Hall, 1985). See also Myers, Intuition, 115-24; and John Allen Paulos,
Once Upon a Number: The Hidden Mathematical Logic of Stories (New York: Basic,
1998), 46-56. For applications to narrative, see Richard G. Gerrig, Experiencing
Narrative Worlds: On the Psychological Activities of Reading (New Haven, Conn.:
Yale University Press, 1993), ch. 2.
45. Bobbi S. Low, Why Sex Matters: A Darwinian Look at Human Behavior (Princeton,
N.J.: Princeton University Press, 2000), 177-78.
46. See Vicki Bruce and Andy Young, In the Eye of the Beholder: The Science of Face
Perception (Oxford: Oxford University Press, 1998), 247-55.
47. The dominant position, proposing that there are some universal emotions and
that human faces express these in similar ways across cultures, is put forth in Paul
Ekman, ed., Darwin and Facial Expression (San Diego, Calif.: Academic Press, 1973);
Paul Ekman and Wallace V. Friesen, Unmasking the Face (Englewood Cliffs, N.J.:
Prentice Hall, 1975); Paul Ekman and Erika Rosenberg, eds., What the Face Reveals:
Basic and Applied Studies of Spontaneous Expression Using the Facial Action Coding
System (FACS) (New York: Oxford University Press, 1997); and a popularization,
Paul Ekman, Emotions Revealed: Recognizing Faces and Feelings to Improve Commu-
nication and Emotional Life (New York: Holt, 2003). The view that facial expressions
evolved not to express emotional states but to send social signals is argued forcefully
in Alan J. Fridlund, Human Facial Expression: An Evolutionary View (San Diego,
Calif.: Academic Press, 1994); and in James A. Russell and Jose Miguel Fernandez-
Dols, eds., The Psychology of Facial Expression (Cambridge: Cambridge University
Press, 1997). For an excellent review of this debate, see Ed S. Tan, "Three Views of
Facial Expression and Its Understanding in the Cinema," in Moving Image Theory:
Ecological Considerations, ed. Joseph D. Anderson and Barbara Fisher Anderson
(Carbondale: University of Southern Illinois Press, 2005), 107-28.
48. This follows from what Paul Grice calls the Cooperative Principle as a conversa-
tional maxim, as well as the "supermaxim" "Do not say what you believe to be false."
See Paul Grice, Studies in the Way of Words (Cambridge, Mass.: Harvard University
448 Notes
Press, 1989), 27-40. For a helpful exposition of Grice's conversational implicature
theory, see Geoffrey N. Leech, Principles of Pragmatics (London: Longmans, 1983),
ch. 4.
49. Tan, "Three Views of Facial Expression," 111-13.
50. Carl Plantinga, "The Scene of Empathy and the Human Face on Film," in Plantinga
and Smith, Passionate Views, 239-55. See also Smith, Engaging Characters, 95-105.
51. Ferdinand Brunetiere, "The Law of the Drama," in Barrett H. Clark, ed., European
Theories of the Drama (New York: Crown, 1961), 407.
52. This isn't a claim that can be considered "biological determinism," because a great
many narratives lack such goal-driven action patterns. If they are a universal, they're
a contingent one.
53. I argue this in more detail in my Ozu and the Poetics of Cinema (Princeton, N.J.:
Princeton University Press, 1988), 318-19.
54. Interestingly, the Claire Dolan entry on the Internet Movie Database (http://www.
imdb.com) "classicizes" the film: "A high-priced call girl, shocked by her mother s
death, decides to get out of the business and have a baby. The steps that she takes
to free herself from her pimp and find a father for the baby are the central story of
this movie."
55. In Scenarios modeles, 52-55, Francis Vanoye provides a brief but intriguing discus-
sion of "modern" models of character, including the "problematic" character in crisis,
the opaque character as seen in Duras or Bresson, and the noncharacter, the mask or
marionette, as in Last Year at Marienbad (1961) or some films of Peter Greenaway.
56. Nicholas Kazan, quoted in Jurgen Wolff and Kerry Cox, Top Secrets: Screenwriting
(Los Angeles: Lone Eagle, 1993), 134.
57. Lajos Egri, The Art of Dramatic Writing (1942; reprint, New York: Simon & Schuster,
1960), 61.
58. Ibid., 60-61.
59. See Bordwell, The Way Hollywood Tells It, 51-71.
60. Bordwell, Narration in the Fiction Film, 62.
61. Andre Gaudreault and Francois Jost, Le recit cinematographique (Paris: Nathan,
1990), 24-27; Albert Laffay, Logique du cinema: Creation et spectacle (Paris: Masson,
1964), 62-85; and Tom Gunning, D. W. Griffith and the Origins of American
Narrative Film: The Early Years at Biograph (Urbana: University of Illinois Press,
1991), 13-25.
62. Meir Sternberg has argued that even within quoted passages, we can detect the
presence of the external narrator. See his "Proteus in Quotation-Land: Mimesis and
the Forms of Reported Discourse," Poetics Today 24, no. 3 (Fall 2003): 107-56.
63. Quoted in Rudy Behlmer, ed., Memo From Darryl F. Zanuck: The Golden Years at
Twentieth Century Fox (New York: Grove, 1995), 190.
64. Bordwell, Narration in the Fiction Film, 61.
65. There may be an important point about literary conventions here. It seems that
insofar as a literary narrative relies on a communication situation, the character-
narrator is presumed to have one-point access to information. That is, she can tell
only what she knows (or supposes, or guesses, or imagines, or divines). As in real
life, all information we have comes filtered through that character's range of knowl-
edge. It does seem, however, that recounting situations in narratives in other media
don't follow the same convention. A flashback in a play or a comic book would seem
to have the same flexibility that we find in cinema.
66. Bordwell, Narration in the Fiction Film, 62.
Notes 449
67. Chatman, Story and Discourse, 151.
68. Seymour Chatman, Coming to Terms: The Rhetoric of Narrative in Fiction and Film
(Ithaca, N.Y.: Cornell University Press, 1990), 115-16.
69. Ibid., 127.
70. Ibid., 115.
71. Ibid., 134.
72. Ibid., 134.
73. Ibid., 136.
74. Chatman, Story and Discourse, 253-61.
75. Marie-Laure Ryan, "Introduction," in Narrative Across Media: The Languages of
Storytelling, ed. Marie-Laure Ryan (Lincoln: University of Nebraska Press, 2004), 15.
76. Chatman, Coming to Terms, 133.
77. See for example Gregory Currie, "Unreliability Refigured: Narrative in Literature
and Film," The Journal of Aesthetics and Art Criticism 53, no. 1 (Winter 1995):
19-23; George M. Wilson, "Le Grand Imagier Steps Out: The Primitive Basis of Film
Narration," Philosophical Topics 25, no. 1 (Spring 1997): 295-318; Paisley Livingston,
"Narrative," in The Routledge Companion to Aesthetics, ed. Berys Gaut and Dominic
Mclver Lopes (London: Routledge, 2001), 278-81; and Andrew Kania, "Against the
Ubiquity of Fictional Narrators," The Journal of Aesthetics and Art Criticism 63, no.
1 (Winter 2005): 47-54. For a critique of the communication model and an argu-
ment that even third-person literary narration may not proceed from a narrator, see
John Morreall, "The Myth of the Omniscient Narrator," The Journal of Aesthetics
and Art Criticism 52, no. 4 (Fall 1994): 429-35. I'm grateful to Paisley Livingston
for several helpful discussions on this topic and for sharing some of these sources
with me.
78. Ryan, "Introduction," 15.
79. See, for example, Gunning, D, W. Griffith, 24; Chatman, Coming to Terms, 127-30;
and J. Lothe, Narrative in Fiction and Film: An Introduction (Oxford: Oxford
University Press, 2000), 28.
80. Chatman, Coming to Terms, 133.
81. Ibid., 74.
82. Ibid., 75, 82, 86.
83. Ibid., 81.
84. See Wayne Booth, The Rhetoric of Fiction (Chicago: University of Chicago Press,
1961), 70-77.
85. Chatman, Coming to Terms, 75.
86. Chatman, Coming to Terms, 91.
Chapter 4
1. Christian Metz, "Problems of Denotation in the Fiction Film," in his Film Language,
trans. Michael Taylor (New York: Oxford University Press, 1974), 145.
2. Apart from the many critiques of poststructuralism in literature and philosophy,
see David Bordwell and Noel Carroll, eds., Post-Theory: Reconstructing Film Studies
(Madison: University of Wisconsin Press, 1996). Interestingly, this remains the
only anthology to mount such a critique within film studies, as Carroll's Mystify-
ing Movies: Fads and Fallacies in Contemporary Film Theory (New York: Columbia
 Notes
450
University Press, 1985) is the only monograph to take this position. Film scholars
have been remarkably reluctant to criticize the foundations of this paradigm, prefer-
ring to quietly switch over to a rival framework, that of cultural studies.
3. For a discussion, see my essay "Contemporary Film Studies and the Vicissitudes
of Grand Theory," in Bordwell and Carroll, Post-Theory, 3-36. An example of the
return to 1960s Metzian themes is Roger Odin's Cinema etproduction de sens (Paris:
Colin, 1991).
4. See the collection of Jerome Bruner's papers, Beyond the Information Given: Studies
in the Psychology of Knowing (New York: Norton, 1973).
5. I outline the emerging paradigm in my "A Case for Cognitivism." Iris, no. 9 (Spring
1989): 11-40. See also my "A Case for Cognitivism: Further Reflections," Iris, no. 11
(Summer 1990): 107-12.
6. On "gist," see Carol Fleisher Feldman, Jerome Brunei", Bobbi Renderer, and Sally
Spitzer, "Narrative Comprehension," in Narrative Thought and Narrative Language,
ed. Bruce K. Britton and Anthony D. Pellegrini (Hillsdale, N.J.: Erlbaum, 1990),
1-78.
7. The classic source of the concept of schema is F. C. Bartlett, Remembering: A Study
in Experimental and Social Psychology (Cambridge: Cambridge University Press),
1932. Useful orientations to later conceptions of schemas can be found in Reid
Hastie, "Schematic Principles in Human Memory," in Social Cognition: The Ontario
Symposium, vol. 1, ed. E. Tony Higgins, C. Peter Herman, and Mark P. Zanna
(Hillsdale, N.J.: Erlbaum, 1981), 39-88; and Ronald W. Casson, "Schemas in Cogni-
tive Anthropology," Annual Review of Anthropology 12 (1983): 429-62.
8. For information on the film, as well as an edition of the shooting script, see Albert
J. LaValley, Mildred Pierce (Madison: University of Wisconsin Press, 1980). Produc-
tion background can be found in Rudy Behlmer, Inside Warner Bros. (1935-1951)
(New York: Viking, 1985), 254-64.
9. Meir Sternberg, Expositional Modes and Temporal Ordering in Fiction (Baltimore:
Johns Hopkins University Press, 1978), 45-55. See also Sternberg's trio of essays:
"Telling in Time (I): Chronology and Narrative Theory," Poetics Today 11, no. 4
(Winter 1990): 901-48; "Telling in Time (II): Chronology, Teleology, Narrativity,"
Poetics Today 13, no. 3 (Fall 1992): 463-541; and "Telling in Time (III): Chronol-
ogy, Estrangement, and Stories of Literary History," Poetics Today 27, no. 1 (Spring
2006): 125-235.
10. Joyce Nelson, "Mildred Pierce Reconsidered," Film Reader 2 (1977): 67.
11. Parker Tyler, Magic and Myth of the Movies (New York: Henry Holt, 1947), 214-15.
12. Jean Mandler, Stories, Scripts, and Scenes: Aspects of Schema Theory (Hillsdale, N.J.:
Erlbaum, 1984), 22.
Chapter 5
13. Roland Barthes, S/Z, trans. Richard Howard (New York: Hill and Wang, 1974), 67,191.
14. A rich set of reflections along these lines can be found in the essays collected in
Michael Carrithers, Steven Collins, and Steven Lukes, eds., The Category of the
Person: Anthropology, Philosophy, History (Cambridge: Cambridge University Press,
1985). Concrete work supplementing this line of thinking is exemplified by Roy G.
D'Andrade, "Character Terms and Cultural Models," in Directions in Cognitive
Anthropology, ed. Janet W. D. Dougherty (Urbana: University of Illinois Press, 1985),
321-43; and Roy G. D'Andrade, "A Folk Model of the Mind," in Cultural Models
in Language and Thought, ed. Dorothy Holland and Naomi Quinn (Cambridge:
Notes
451
Cambridge University Press, 1987), 112-48. With respect to cinema, see Murray
Smith, Engaging Characters: Fiction, Emotion, and the Cinema (Cambridge:
Cambridge University Press, 1995), 21.
15. "The existence of an office," writes an anthropologist, "logically entails a distinction
between the powers and responsibilities pertaining to it and their exercise by dif-
ferent incumbents. Hence some concept of the individual as distinct from the office
is established" (J. S. La Fontaine, "Person and Individual: Some Anthropological
Reflections," in Carrithers, Collins, and Lukes, The Category of the Person, 138).
16. Barthes, S/Z, 11.
17. The odd thing is that in shot 2 of the first scene, a figure can be glimpsed ducking
out of sight in the passenger seat. This is a good example of what is not perceivable
under normal protocols of viewing.
18. It's been suggested to me that the filmmakers were simply sloppy; although they
intended both versions of Monte's death to be identical, they were unable to dupli-
cate the details of performance exactly. Even if that were so, when confronted with
two versions of Zachary Scott's delivery of the line "Mildred," the filmmakers put
the version most likely to throw suspicion on the heroine in the opening and used
the nonincriminating version in the revelatory flashback. The published version
of the screenplay doesn't call for the reenactment of Monte's final words, so evidently
the replay of the murder was devised in the course of shooting and postproduction.
See LaValley, Mildred Pierce, 233.
19. Pam Cook, "Duplicity in Mildred Pierce," in Women in Film Noir, ed. E. Ann Kaplan
(London: British Film Institute, 1978), 78.
20. Nelson, "Mildred Pierce Reconsidered," 65.
21. Tyler, Magic and Myth, 214.
22. Noel Carroll, The Philosophy ofHorror; or Paradoxes ofthe Heart (New York: Methuen,
1990), 59-96; Murray Smith, Engaging Characters, 81-86; Ed Tan, Emotion and the
Structure ofNarrative Film: Film as Emotion Machine (Teaneck, N.J.: Erlbaum, 1996);
Carl R. Plantinga and Greg R. Smith, eds., Passionate Views: Film, Cognition and
Emotion (Baltimore: Johns Hopkins University Press, 1999); and Greg R. Smith, Film
Structure and the Emotion System (Cambridge: Cambridge University Press, 2003).
See also Joseph Anderson, The Reality of Illusion: An Ecological Approach to Cogni-
tive Film Theory (Carbondale: University of Illinois Press, 1997).
23. David Bordwell, Making Meaning: Inference and Rhetoric in the Interpretation of
Cinema (Cambridge, Mass.: Harvard University Press, 1989).
24. See David Bordwell, Narration in the Fiction Film (Madison: University of
Wisconsin Press, 1985) chs. 10-12.
1. More radical avant-garde movements, such as Soviet montage filmmaking, sur-
realism, and cinema pur, seem to have been relatively without effect upon the art
cinema's style. I suspect that those experimental styles that did not fundamentally
challenge narrative coherence were the most assimilable to the postwar art cinema.
2. See Thomas Guback, The International Motion Picture Industry (Bloomington:
Indiana University Press, 1969), passim.
3. See, for example, Philip Rosen, "Difference and Displacement in Seventh Heaven,"
Screen 18, no. 2 (Summer 1977): 89-104.
452 Notes
4. This point is taken up in Christian Metz, "The Modern Cinema and Narrativity,"
in his Film Language, trans. Michael Taylor (New York: Oxford University Press,
1974), 185-227.
5. Arthur Knight compares the Hollywood film to a commodity and the foreign film
to an artwork: "Art is not manufactured by committees. Art comes from an indi-
vidual who has something that he must express.... This is the reason why we hear
so often that foreign films are 'more artistic' than our own. There is in them the
urgency of individual expression, an independence of vision, the coherence of a
single-minded statement." Quoted in Michael F. Mayer, Foreign Films on American
Screens (New York: Arco, 1965), vii.
6. "The strategy was to talk about Hawks, Preminger, etc. as artists like Bunuel and
Resnais" [Jim Hillier, "The Return to Movie," Movie, no. 20 (Spring 1975): 17].
I do not mean to imply that auteur criticism did not at times distinguish between
the classical narrative and the art cinema. A book like V. F. Perkins' Film as Film
(Baltimore: Penguin, 1978) insists not only upon authorial presence but also
upon the causal motivation and the stylistic economy characteristic of the classi-
cal cinema. Thus, Perkins finds the labored directorial touches of Antonioni and
Bergman insufficiently motivated by story action. Nevertheless, Perkins' interpreta-
tion of the jeep sequence in Carmen Jones in terms of characters' confinement and
liberation (80-82) is a good example of how Hollywood cutting and camera place-
ment can be invested with symbolic traces of the author.
7. See, for instance, Mark Nash, "Vampyr and the Fantastic," Screen 17, no. 3 (Autumn,
1976): 29-67; and Paul Willemen, "The Fugitive Subject," in Raoul Walsh, ed. Phil
Hardy (London: Edinburgh Film Festival, 1974), 63-89.
8. Norman Holland, "The Puzzling Movies: Three Analyses and a Guess at Their
Appeal," Journal of Social Issues 20, no. 1 (January 1964): 71-96.
9. See Steve Neal, "New Hollywood Cinema," Screen 17, no. 2 (Summer 1976):
117-33; and Paul Willemen, "Notes on Subjectivity: On Reading Edward Branigan's
'Subjectivity Under Siege,"' Screen 19, no. 1 (Spring 1978): 59-64. See also Robin
Wood, "Smart-Ass and Cutie Pie: Notes Toward an Evaluation of Altman," Movie,
no. 21 (Autumn 1975): 1-17.
10. See David Bordwell, The Films of Carl Theodor Dreyer (Berkeley: University of Cali-
fornia Press, 1981).
11. Kristin Thompson and David Bordwell, Film History: An Introduction, 2nd ed.
(New York: McGraw-Hill, 2003). On art-cinema traditions, see chs. 4-6, 8, 16-20,
23, 25-26, and 28.
12. Steve Neale made an early contribution to this line of thinking with "Art Cinema as
Institution," Screen 22, no. 1 (1981): 11-39. For an overview of state support of the
European cinema, see Anne Jackel, European Film Industries (London: British Film
Institute, 2003).
13. See Tino Balio, United Artists: The Company That Changed the Film Industry (Madi-
son: University of Wisconsin Press, 1987), chs. 7 and 9; and Peter Lev, The Euro-
American Cinema (Austin: University of Texas Press, 1993).
14. Vinca Wiedemann, quoted in Jacob Wendt Jensen, "Northern Lights," Screen
International, May 5, 2006, 16.
15. See Film Festival Today magazine and its website, http://www.filmfestivaltoday.com.
Notes 453
16. On film festivals' role in international film culture, see Thompson and Bordwell, Film
History, 716-18. A more extensive account is provided in Thomas Elsaesser, "Film
Festival Networks: The New Topographies of Cinema in Europe," in his European
Cinema: Face to Face With Hollywood (Amsterdam: Amsterdam University Press,
2005), 82-107.
17. The Tehran-based magazine Film International has kept track of festival entries
and awards. See for example the charts in Mohammad Atebbai, "Iranian Films
and the International Scene in 1997," Film International, no. 19 (1998): 17-20; and
Mohammad Atebbai, "Iranian Films in the International Scene in 1998," Film
International, no. 23 (1999): 10-14.
18. See my Planet Hong Kong: Popular Cinema and the Art ofEntertainment (Cambridge,
Mass.: Harvard University Press, 2000), 87-89.
19. For a thorough account of Wong's debts to prestigious literature and film, see
Stephen Teo, Wong Kar-wai (London: British Film Institute, 2005). I discuss Wong's
experimental impulses in my Planet Hong Kong, 266-89.
20. On Hou's context and development, see Emilie Yueh-yu Yeh and Darrell Davis,
Taiwan Film Directors: A Treasure Island (New York: Columbia University Press,
2005), chs. 1,2, and 4. Hou's stylistics and industrial context are considered in chapter
5 of my Figures Traced in Light (Berkeley: University of California Press, 2005).
21. Quoted in Stig Bjorkman, "Preface," in Lars von Trier, Breaking the Waves (London:
Faber and Faber, 1996), 8. For detailed discussions of Dogme and von Trier, see Mette
Hjort and Scott Mackenzie, eds., Purity and Provocation: Dogme 95 (London: British
Film Institute, 2003); and Mette Hjort, Small Nation, Global Cinema: The New Danish
Cinema (Minneapolis: University of Minnesota Press, 2005). I discuss modern Danish
film as an accessible art cinema in my "A Strong Sense of Narrative Desire: A Decade
of Danish Film," Film (Copenhagen), no. 34 (Spring 2004): 24-27, http://www.dfi.dk.
22. I discuss Hartley's adaptation of some European staging principles in "Up Close
and Impersonal: Hal Hartley and the Persistence of Tradition" (June 2005), http://
www.16-9.dk/2005-06/sidell-inenlish.htm, or by a link from my website http://
www.davidbordwell.net. It was first published as "Nah dran und unpersonlich:
Hal Hartley und die Beharrlichkeit der Tradition," in Die Spur durch den Spiegel:
Der Film in der Kultur der Moderne, ed. Malte Hagener, Johann Schmidt, and
Michael Wedel (Berlin: Bertz Verlag, 2004), 410-21.
23. See Geoff King, American Independent Cinema (Bloomington: Indiana Univer-
sity Press, 2005). See also David Bordwell, Janet Staiger, and Kristin Thompson,
The Classical Hollywood Cinema: Film Style and Mode of Production to 1960 (New
York: Columbia University Press, 1985), 372-77; and J. J. Murphy, Me and You and
Memento and Fargo (New York: Continuum, 2007).
24. See Andras Balint Kovacs' comprehensive study, Modern European Art Cinema From
the 1950s to the 1970s, Cinema and Modernity Series (Chicago: University of Chicago
Press, 2007). See also John Orr, Cinema and Modernity (London: Polity, 1994).
25. On the planimetric image, see my On the History of Film Style, 261-64; my Figures
Traced in Light, 167-68,173-76,232-33; and my essay "Modelle der Rauminszenierung
im zeitgenossishen europaischen Kino," in Zeit, Schnitt, Raum, ed. Andreas Rost
(Munich: Verlag der Autoren, 1997), 17-42.
26. For a study of Hou's staging, see ch. 5 of my Figures Traced in Light.
27. See David Bordwell, Narration in the Fiction Film (Madison: University of Wisconsin
Press, 1985), 213-28.
454 Notes
28. I discuss its use in Toto le heros (1989) in my "Toto le Moderne: Narration dans le
cinema Europeen d'apres 1970," La Revue beige du cinema, nos. 36-37 (April 1994):
33-39.
29. Varda discusses these and other matters in the excellent bonus material to be found
on the French DVD release of the film (Cine Tamaris disc no. CT2). See especially
the chapter "Musique et Travellings."
Chapter 6
After the first publication of this essay, I discovered Gerald Prince's lively essay, "The
Disnarrated" in Style 22, no. 1 (Spring 1988): 1-8. In studying those small-scale
moments in literary narrative that play with what might have been, Prince's paper
intersects with mine in intriguing ways.
1. Jorge Luis Borges, "The Garden of Forking Paths," in his Collected Fictions, trans.
Andrew Hurley (New York: Penguin, 1999), 125.
2. Ibid., 127.
3. Arguments for parallel worlds arise within several scientific domains—cosmol-
ogy, quantum physics, mathematics, and logic. For an overview, see Max Tegmark,
"Parallel Universes," Scientific American 288, no. 5 (May 2003): 41-51. The older
standard source is Fred Alan Wolf, Parallel Universes: The Search for Other Worlds
(New York: Simon 8t Schuster, 1988). See also Michael Shermer, Science Friction:
Where the Known Meets the Unknown (New York: Holt, 2005), 40-42.
4. Gary Saul Morson, Narrative and Freedom: The Shadows of Time (New Haven,
Conn.: Yale University Press, 1994), 232.
5. The classic studies on heuristics and rationality are collected in Daniel Kahneman,
Paul Slovic, and Amos Tversky, eds., Judgment Under Uncertainty: Heuristics and
Biases (Cambridge: Cambridge University Press, 1982). See also Richard Nisbett
and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment
(Englewood Cliffs, N.J.: Prentice Hall, 1980). A more recent account is Gerd
Gigerenzer, Adaptive Thinking: Rationality in the Real World (New York: Oxford
University Press, 2000).
6. Quoted in Mark Prendergast, The Ambient Century: From Mahler to Moby—the
Evolution of Sound in the Electronic Age (London: Bloomsbury, 2003), 118.
7. It seems that the Internet, which doesn't offer the predetermined trajectory that a film
does, is far more hospitable to widely branching narrative futures. Katherine Hayles
illustrates this point in "Reconfiguring Narrative in Electronic Environments" (paper
presented at the Narrative at the Outer Limits conference, University of California,
Santa Barbara, May 4,2001). The book format offers its own intricacies, as the Choose
Your Own Adventure series indicates. For a map of the possibilities in that series, see
Gregory Lord, "Analysis Essay," http://www.ethblue.com/cyoa/essay.html.
8. Quoted in Danusia Stok, ed., Kieslowski on Kieslowski (London: Faber, 1993), 113.
9. Bryce DeWitt, quoted in Tim Folger, "Quantum Shmantum," Discover, September
2001,42.
10. A clear example can be found in the Back to the Future trilogy. In the first film, the
switchpoint is established as the moment when Marty, having traveled back to 1955,
pushes his future father out of the path of the car driven by Lorraine's father. As a
result, George McFly doesn't win Lorraine's pity, they don't go to the prom together,
they don't kiss and fall in love and marry . . . and Marty doesn't get born. Having
Notes 455
disturbed the past, Marty must restage the kiss under new circumstances, along the
way allowing his father to become more courageous and self-confident. But Back to
the Future II, which presents an alternative future for Marty and his family, shifts
the switchpoint to an earlier moment on the day of the prom, when the villain Biff
receives an almanac from the future that will allow him to win any sports bet he
lays down. This earlier moment becomes the crucial fork for that second film in the
series—another aspect of the plot that Doc obligingly diagrams for us on a black-
board in his lab. See also p. 186.
11. Meir Sternberg, "Telling in Time (I): Chronology and Narrative Theory," Poetics
Today 11, no. 4 (Winter 1990): 901-48; Meir Sternberg, "Telling in Time (II):
Chronology, Teleology, Narrativity," Poetics Today 13, no. 3 (Fall 1992): 463-541;
and Meir Sternberg, "Telling in Time (III): Chronology, Estrangement, and Stories
of Literary History," Poetics Today 27, no. 1 (Spring 2006): 125-235.
12. Alain Masson, "Necessite et variations," in Krzysztof Kieslowski, ed. Vincent Amiel
(Paris: Jean-Michel Place, 1997), 57.
13. Tom Tykwer, Cours, Lola, cours, trans. Marie Ollivier (Paris: Fleuve noir, 1999), 134.
14. Perhaps Ayckbourn's work is a more general inspiration as well. His play How
the Other Half Loves (1969) presents two locales separated in space and time
"superimposed" on the same set. Here one table serves as two tables, one sofa as two
sofas, and the like, while two sets of actors play out the different scenes simultane-
ously. This virtuosic premise seems to anticipate Sliding Doors' scene showing the
two Helens in the same bar.
15. The Malcolm episode is "Bowling" (airdate April 1,2001). Thanks to Jonathan Frome
for pointing out this episode to me.
16. See Robert Coover, Pricksongs and Descants: Fictions (New York: Grove, 1969),
150-67. On the role of forking-path plots in contemporary literature, see Brian
McHale, Postmodernist Fiction (London: Methuen, 1987), 106-11.
17. There was an African American production, also called Eyes of Youth (1920), which
may have been a film record of a performance of the 1917 play. A later version is The
Love ofSunya (1927).
18. The opening monologue of Slacker (1991), delivered by the director himself, intro-
duces the notion of parallel worlds, along with the butterfly effect, chaos theory,
degrees of separation, and other pop science motifs. It's as if Richard Linklater is
opening up a box of formal devices for the independent cinema to explore in the
next two decades.
19. Michio Kaku, Hyperspace: A Scientific Odyssey Through Parallel Universes, Time
Warps, and the 10th Dimension (New York: Anchor, 1995), 254-64.
20. "Random Quest" is included in John Wyndham, Cover Her Face and Others
(Harmondsworth, UK: Penguin, 1961), 131-73.
21. For more on the phenomenon of the replay movie and its relation to narrative
experimentation and DVD consumption, see my The Way Hollywood Tells It:
Story and Style in Modern Movies (Berkeley: University of California Press, 2006),
89-94.
22. On the patterns of redundancy in Groundhog Day, see Kristin Thompson, Story-
telling in the New Hollywood: Analyzing Classical Narrative Technique (Cambridge,
Mass.: Harvard University Press, 1999), 131-54.
456 Notes
Chapter 7
1. Italo Calvino, "Two Interviews on Science and Literature," in his The Uses of Litera-
ture, trans. Patrick Creagh (New York: Harcourt Brace Jovanovich, 1986), 34.
2. See Evan Smith, "Thread Structure: Rewriting the Hollywood Formula," Journal of
Film and Video 51, nos. 3-4 (Fall-Winter 1999-2000): 88-89.
3. See my The Way Flollywood Tells It: Story and Style in Modern Movies (Berkeley:
University of California Press, 2006), 97-102.
4. Alexander Mackendrick, On Film-Making, ed. Paul Cronin (London: Faber and
Faber, 2004), 18.
5. In such cases, one or both partners must modify the goals if the romantic union is
to be consummated. In His Girl Friday, Hildy has to give up Bruce and surrender
to Walter. On dual protagonists generally, see Kristin Thompson, Storytelling in the
New Hollywood (Cambridge, Mass.: Harvard University Press, 1999), 44-49.
6. George Eliot, Middlemarch, ed, Rosemary Ashton (Harmondsworth, England:
Penguin, 1994), 141.
7. Margaret Anne Doody, The True Story of the Novel (New Brunswick, N.J.: Rutgers
University Press, 1996), 153-54.
8. See Richard Levin, The Multiple Plot in English Renaissance Drama (Chicago:
University of Chicago Press, 1975).
9. On the rise of network stories on episodic television, see Robert }. Thompson, Tele-
vision's Second Golden Age: From Hill Street Blues to ER (Syracuse, N.Y.: Syracuse
University Press, 1997), 70; and Glen Creeker, Serial Television: Big Drama on the
Small Screen (London: British Film Insitute, 2004), 1-11. This trend has continued
to the present, as shown in Jeff Sconce, "What If? Charting Television's New Textual
Boundaries," in Television After TV: Essays on a Medium in Transition (Durham,
N.C.: Duke University Press, 2004), 93-112; and Jason Mittell, "Narrative Complex-
ity in Contemporary American Television," The Velvet Light Trap, no. 58 (Fall 2006):
16-28.
10. Nellie Andreeva, "Lions Gate Gets 'Crash' Course in TV," Hollywood Reporter, June
28-July 4, 2005, 1, 58.
11. Francisco Ibanez, 13, rue del Percebe (Barcelona: Super Humor, 2005).
12. Daniel Clowes, Eightball, no. 22 (October 2001).
13. R. Alberich, J. Miro-Julia, and F. Rossello, "Marvel Universe Looks Almost Like a Real
Social Network" Condensed Matter (February 2002), available at <http://xxx.lanl.gov>
as <arXiv:cond-mat/0202174vl [cond-mat.dis-un], 11 Feb2002>.
14. See Robin Dunbar, Grooming, Gossip and the Evolution of Language (London: Faber
and Faber, 1996), 192.
15. Even, apparently, fifth- and sixth-order mind-reading can be managed, according
to Robin Dunbar's experiments reported in The Human Story: A New History of
Mankind's Evolution (London: Faber and Faber, 2004), 45-52. Beyond the sixth
level, however, people seem to fail.
16. Ibid., 52.
17. Sarah Blaffer Hrdy, Mother Nature: Maternal Instincts and How They Shaped the
Human Species (New York: Ballantine, 1999), 143-44.
18. William Empson, Some Versions of Pastoral (New York: New Directions, 1960), 25, 30.
19. Someone should examine non-Hollywood cinema for early instances. In Urlaub auf
Ehrenwort (Leave on Word of Honor, 1938), a commander gives his men leave for
a day on their way to the front, and the film follows each one's experiences. It was
Notes 457
adapted into a Japanese film, The Last Visit Home (1945). On the latter, see Peter B.
High, The Imperial Screen: Japanese Film Culture in the Fifteen Years' War 1931-1945
(Madison: University of Wisconsin Press, 2003), 298.
20. See David Bordwell, Planet HK: Popular Cinema and the Art of Entertainment
(Cambridge, Mass.: Harvard University Press, 2000), 282-89.
21. For more on experimental tendencies in U.S. independent cinema, see Jeffrey
Sconce, "Irony, Nihilism, and the American 'Smart' Film," Screen 4, no. 3 (2003):
349-69; and my The Way Hollywood Tells It, 72-103.
22. The study of networks is a paradigm case of interdisciplinary scientific work. For
reviews, see Manfred Kochen, ed., The Small World: A Volume of Recent Research
Advances Commemorating Ithiel de Sola Pool, Stanley Milgram, and Theodore
Newcomb (Norwood, N.J.: Ablex, 1988); Albert-Laszlo Barabasi, Linked: The New
Science of Networks (New York: Perseus, 2002); Mark Buchanan, Nexus: Small Worlds
and the Groundbreaking Science of Networks (New York: Norton, 2002); Duncan J.
Watts, Six Degrees: The Science of a Connected Age (New York: Norton, 2003); and
Steven Strogatz, Sync: The Emerging Science of Spontaneous Order (New York: Theia,
2003). The phenomenon is put into the context of the history of science in Philip
Ball, Critical Mass: How One Thing Leads to Another (New York: Farrar, Straus &
Giroux, 2004). See also Judith Kleinfeld, "The Small World Problem," Society 39,
no. 2 (January-February 2002): 61-6; and Thomas Blass, The Man Who Shocked the
World: The Life and Legacy of Stanley Milgram (New York: Perseus, 2005), 284-89.
23. Not all movies invoking this image are network narratives, of course; The Butterfly
Effect (2004), for example, isn't.
24. Meir Sternberg, Expositional Modes and Temporal Ordering in Fiction (Baltimore:
Johns Hopkins University Press, 1978), chs. 2-3.
25. Quoted in Stephen Farber, "A Half-Dozen Ways to Watch the Same Movie," New
York Times, November 13, 2005, 26.
26. Anonymous, "A Crash Course With Paul Haggis," Premiere, December 2005-
January 2006, 14. This is an advertisement presented as an interview.
27. For a complete list, see Monika Mauser, KrzysztofKieslowski (London: Pocket Essen-
tials, 2000), 38-65. For Kieslowski's comments, see Danusia Stok, ed., Kieslowski on
Kieslowski (London: Faber and Faber, 1995), 154-55. Interestingly, the German pro-
ducers of The Decalogue asked for an 11th episode in which all the characters would
converge—a convention of network narratives—but he felt incapable of doing this
See Vincent Amiel, ed., KrzysztofKieslowski (Paris: Jean-Michel Place, 1997), 99.
28. See Linda J. Cowgill, Secrets of Screenplay Structure (Los Angeles: Lone Eagle,
1999), 128.
29. Quoted in John Gribbin, Deep Simplicity: Chaos, Complexity and the Emergence of
Life (New York: Penguin, 2004), 49.
30. Lelouche remarks that a traffic tie-up was an ideal way to assemble all the characters
in a single scene, especially because grave consequences flow from such a trivial
convergence. "I truly believe that the flapping of the butterfly's wings can trigger a
cascade of events." See Yves Alion and Jean Olle-Laprune, Claude Lelouche: Mode
d'emploi (Paris: Calmann-Levy, 2005), 241.
31. Arguably, this does happen in Yuva. This film evidently models its opening traffic
violence on Amores Perros, but it doesn't seed its first flashback lead-in with enig-
matic scenes from its second and third tales. After the first flashback brings us back
 Notes
458
to the opening action, we know quite fully what has led the hitman and his victim to
the bridge. As a result we must rely on the intrinsic interest of the other two charac-
ters' backstories to carry us through the rest of the film's first 100 minutes.
32. Quoted in Nancy Hendrickson, "Happy Endings," Creative Screen-writing 12, no. 4
(July-August 2005): 30.
33. On parallelism in contemporary American films, see Murray Smith, "Parallel Lines,"
in American Independent Cinema, ed. Jim Hillier (London: British Film Institute,
2000), 155-61.
34. Judith S. Kleinfeld, "The Small World Problem," 65-66.
35. I discuss some of these in The Way Hollywood Tells It, 78-82, 89-94.
36. Despite all this maneuvering, the plot structure of 21 Grams still cunningly obeys
Hollywood act structure. The first half hour establishes most of the preconditions
for the action, the middle hour presents most of the scenes of the accident and its
immediate aftermath, and the final 20 minutes concentrate on Paul's and Chris's
pursuit of vengeance.
37. Quoted in Farber, "A Half-Dozen Ways," 18.
38. Christine's scenes alone or with secondary characters occupy nearly 19% of the
running time, Richard's occupy over 20%, and their scenes together amount to
24.3%, making a total of 63.4%.
39. For arguments along these lines, see my The Way Hollywood Tells It, 72-103.
40. On three-part and four-part script structures, see Thompson, Storytelling in the
New Hollywood, ch. 1; and my The Way Hollywood Tells It, 27-42.
41. Richard Curtis, Love Actually, DVD commentary, ch. 5, 36:02.1 discuss the network
structure of Love Actually in my The Way Hollywood Tells It, 100-2.
42. See Steven Johnson, Everything Bad Is Good for You (New York: Riverhead, 2005),
67-77, 108-15.
43. Quoted in Farber, "A Half-Dozen Ways," 26.
44. A lesser-known version of the film contains a prologue introducing the gigolo whom
we see at the end of the usual version, while a voice-over, presumably Yuddy's, talks
about seeing "him" upon getting back from the Philippines. In this version, then,
Yuddy is from the start an absent node that pulls together several strangers. This
prologue also introduces, in a fragmentary fashion, the policeman who conceives a
love for one of Yuddy's women, Su Lizen.
45. For a good discussion, see Richard Neupert, A History of the French New Wave
Cinema (Madison: University of Wisconsin Press, 2002), 275.
46. Quoted in Kevin Conroy Scott, Screenwriters' Masterclass: Screenwriters Talk About
Their Greatest Movies (New York: Newmarket, 2006), 101-2.
47. Quoted in Tom Wicker, "'Nashville': Dark Perceptions in a Country-Music
Comedy," New York Times, June 15,1975, D17.
48. Quoted in Alex Lewin, "It Happened in Nashville," Premiere 13, no. 11 (July
2000): 88-102.
49. Quoted in Lewin, "It Happened in Nashville."
50. Quoted in Jon Stuart, The Nashville Chronicles: The Making of Robert Altman s
Masterpiece (New York: Simon & Schuster, 2000), 133.
51. Quoted in Paul Gardner, "Altman Surveys 'Nashville' and Sees 'Instant' America,"
New York Times, June 13,1975, 26.
52. Robert Altman, Nashville, DVD commentary, 2:45.
53. Stuart, The Nashville Chronicles, 192. Most of the ensemble scenes were recorded
with 8-track equipment, itself unusual in filmmaking at the period.
Notes 459
54. Patrick McGilligan, Robert Altman: Jumping Off the Cliff (New York: St. Martin's,
1989), 398.
55. Stuart, The Nashville Chronicles, 210.
56. The prologue's first story is taken verbatim from Charles Fort, Wild Talents (1932;
reprint, New York: Cosimo, 2004). See Charles Fort, The Complete Books of Charles
Fort (New York: Dover, 1974), 848.
57. See Kristin Thompson, Breaking the Glass Armor: Neoformalist Film Analysis
(Princeton, N.J.: Princeton University Press, 1988), chs. 3 and 9.
58. Quoted in Serge Toubiana, "Le gai savoir: Entretien avec Otar Iosseliani," Cahiers
du cinema, no. 368 (February 1985): 5.
59. Quoted in Antony Fiant, (Et) le cinema d'Otar Iosseliani (fut) (Paris: L'Age d'homme,
2002), 140.
60. Quoted in Toubiana, "Le gai savoir," 8.
61. Quoted in ibid., 6.
62. Quoted in ibid., 8.
63. Quoted in ibid., 6, 9. See also Michel Ciment, "Entretien avec Otar Iosseliani,"
Positif, no. 287 Qanuary 1985): 14.
64. "The fragment allows one to explore tiny incidents and flashes [eclats]." Quoted in
Jean-Francois Rauger, "'Avant de tourner mes plans, je les reve,"' Le monde, July 3,
1999; reprinted in Prix de I'age d'or / Cinedecouvertes 1999 (Brussels: Royal Film
Museum of Belgium, July 1999), 67.
65. Serge Kaganski and Frederic Bonnaud, "Achille, Hamlet 8c Don Quichotte," Les
Inrockuptibles, June 2, 1999; reprinted in Prix de I'Age d'or, 61.
66. "AFM Buzz," Screen International, October 28, 2005, 23.
67. Filmmakers, we're told, are now "taking on subjects that can't be contained by
traditional linear storytelling" (Farber, "A Half-Dozen Ways," 18).
68. C. E„ "Crash," Premiere, November 2005, 112.
69. Todd McCarthy, "Happy Endings" Variety, January 24-30, 2005, 44.
70. Leslie Felperin, "Festival," Variety, July 25-31, 2005, 37.
71. Quoted in Mike Goodridge, "Babel Comes Into View," Screen International, May 27,
2005, 7.
72. I've seen Babel since this was written. For comments, see my "Lessons From Babel
(December 15, 2006)," http://www.davidbordwell.net/blog/?p=147.
73. Quoted in Missy Schwartz, "The Deal Report," Entertainment Weekly, March 18,
2005, 13.
74. Tim Kirkman, quoted in Farber, "A Half-Dozen Ways," 26.
75. "World Film Production/Distribution," Screen Digest, June 2005, 174-75.
Chapter 8
1. See the special issue "American Directors," Film Culture; no. 28 (Spring 1963) and
Pauline Kael, I Lost It at the Movies (Boston: Atlantic-Little Brown, 1965).
2. Philip Lopate evokes this era well in "Anticipation of La Notte," in his Totally,
Tenderly, Tragically: Essays and Criticism From a Lifelong Love Affair With the Movies
(New York: Anchor, 1998), 9-15.
3. Andrew Sarris, The American Cinema: Directors and Directions 1929-1968 (1968;
rpt., Chicago: University of Chicago Press, 1985).
460 Notes
4. Consider, for example, Peter Wollen's comparison of Ford and Hawks in Signs and
Meaning in the Cinema (Bloomington: Indiana University Press, 1972), 78-102.
5. David Bordwell, Making Meaning: Inference and Rhetoric in the Interpretation of
Cinema (Cambridge, Mass.: Harvard University Press, 1989), 48-50.
6. Andrew Sarris, "Afterword: The Auteur Theory Revisited," in his The American
Cinema, 276.
7. Rudolf Arnheim, Film trans. L. M. Sieveking and Ian F. D. Morrow (London: Faber
and Faber, 1933); and Raymond Spottiswoode, A Grammar of the Film: An Analysis
of Film Technique (1935; reprint, University of California, 1962).
8. Andrew Sarris, Confessions of a Cultist: On the Cinema, 1955/1969 (New York:
Simon 8c Schuster, 1971), 65.
9. Sarris, American Cinema, 190.
10. I discuss these traditions in On the History of Film Style (Cambridge, Mass.: Harvard
University Press, 1997), chs. 2-3.
11. Andrew Sarris, The Primal Screen: Essays on Film and Related Subjects (New York:
Simon & Schuster, 1973), 25-26.
12. Ibid., 27.
13. Sarris, American Cinema, 50.
14. Ibid., 100.
15. Ibid., 78.
16. Ibid., 106.
17. Despite its title, Andrew Sarris' "You Ain't Heard Nothin' Yet": The American Talking
Film: History and Memory, 1927-1949 (New York: Oxford University Press, 1998)
is not a systematic survey history but rather a collection of critical essays on films,
genres, and stars. Still, it bears out Sarris' commitment to explaining the phenom-
enon of 1930s and 1940s Hollywood as an open-textured plurality of forces. Here is
just one example:
Fortunately, the studios were never a monolithic bloc of crass commercialism,
but rather many raging torrents of conflicting tastes and aspirations. The official
religion of the studios upheld collective craftsmanship over personal artistry,
as did many of the practical requirements of the motion picture medium. At
some point, however, craft lifted itself up into art, and an individual personal-
ity here and there found its own inimitable voice in the group-sing of studio
film-making, (p. 18)
18. Sarris, American Cinema, 27.
19. Sarris, Primal Screen, 12..
20. Andrew Sarris, ed., Interviews With Film Directors (Indianapolis, Ind.: Bobbs,
Merrill, 1967), vi.
21. Sarris, American Cinema, 58.
22. Ibid., 54.
23. Ibid., 46.
24. Ibid., 47. I'd nominate this single passage as the one that most vividly sets Sarris
apart from his contemporaries. Kael, Simon, and Kauffmann, whatever their virtues,
seemed to have no capacity to notice things like this. Granted, after generations of
Film 101 courses and volumes of shot-by-shot exegesis, this observation may seem
tame, but I still can feel the electric jolt of Sarris' passage, describing, in plain and
moving words, the sheer artistic tact of this sequence.
Notes 461
25. Sarris, Primal Screen, 38-53.
26. Ibid., 51.
27. Sarris, American Cinema, 196.
28. Sarris, Primal Screen, 79.
29. Alexandre Astruc, "L'Evolution du cinema americain," in his Du Stylo a la camera...
et de la camera au stylo: Ecrits (1942-1984) (Paris: L'Archipel, 1992), 291.
30. See, for example, Noel Burch, Life to Those Shadows, trans, and ed. Ben Brewster
(Berkeley: University of California Press, 1990); Barry Salt, Film Style and Tech-
nology: History and Analysis, 2nd ed. (London: Starword, 1992); David Bordwell,
Janet Staiger, and Kristin Thompson, The Classical Hollywood Cinema: Film Style
and Mode of Production to 1960 (New York: Columbia University Press, 1985);
and David Bordwell, Ozu and the Poetics of Cinema (Princeton, N.J.: Princeton
University Press, 1988).
31. Sarris, Primal Screen, 50.
32. Ibid., 50.
33. Ibid., 50.
34. For a more detailed defense of the director as "coordinator" of disparate contri-
butions, see V. F. Perkins, Film as Film (Harmondsworth, UK: Penguin, 1973),
158-86.
35. Sarris, Primal Screen, 167-68; and Sarris, American Cinema, 118-19.
36. Sarris, American Cinema, 110.
37. I understood the phrase this way myself in my Making Meaning, 49. Now I'm not
so sure.
38. Sarris, Primal Screen, 51.
39. Ibid., 51.
40. Ibid., 63.
41. "The 20 Greatest Movie Directors of All Time," supplement to Empire, July 2005.
42. Sarris, Primal Screen, 130.
Chapter 9
1. For a discussion of these traditions, see my On the History of Film Style (Cambridge,
Mass.: Harvard University Press, 1997), chs. 2-3.
2. Examples are Ben Brewster, "Deep Staging in French Films 1900-1914," in Early
Cinema: Space Frame Narrative, ed. Thomas Elsaesser (London: British Film Insti-
tute, 1990), 45-55; Yuri Tsivian, "Portraits, Mirrors, Death: On Some Decadent
Cliches in Early Russian Films," Iris, nos. 14-15 (Autumn 1992): 70-78; Barry Salt,
Film Style and Technology: History and Analysis, 2nd ed. (London: Starword, 1992);
Tom Gunning, "Notes and Queries About the Year 1913 and Film Style: National
Styles and Deep Staging," in "L'Annee 1913 en France," 1895 hors serie (1993):
195-204; Yuri Tsivian, "Cutting and Framing in Bauer's and Kuleshov's Films,"
Kintop, no. 1 (1993): 103-13; Richard Abel, The Cine Goes to Town: French Cinema
1896-1914 (Berkeley: University of California Press, 1994); Kristin Thompson, "The
International Exploration of Cinematic Expressivity," in Film and the First World
War, ed. Karel Dibbets and Bert Hogenkamp (Amsterdam: Amsterdam University
Press, 1995), 73-74; Kristin Thompson, "Stylistic Expressivity in Die Landstrasse,"
in A Second Life: German Cinema's First Decades, ed. Thomas Elsaesser (Amsterdam:
Amsterdam University Press, 1996), 256-63; Yuri Tsivian, "Two 'Stylists' of the
 Notes
462
Teens: Franz Hofer and Yevgenii Bauer," in Elsaesser, A Second Life, 264-76; and
Ben Brewster and Lea Jacobs, Theatre to Cinema: Stage Pictorialism and the Early
Feature Film (Oxford: Oxford University Press, 1997).
3. I discuss these strategies in ch. 5 of my On the History of Film Style and ch. 2 of my
Figures Traced in Light: On Cinematic Staging (Berkeley: University of California
Press, 2005).
4. Barry Salt suggests that the slow cutting rates and more distant framings of
German films from the period 1912 to 1917 made them less "exciting, gripping, and
entertaining" than American imports. See Salt, "Early German Film: The Stylistics
in Comparative Context," in Elsaesser, A Second Life, 236.
5. On Feuillade's cutting, see Richard Abel, "Before Fantdmas: Louis Feuillade and
the Development of Early French Cinema," Post Script 7, no. 1 (1987): 4-26; and my
Figures Traced in Light, 43-82.
6. Jan-Christopher Horak, "Robert Reinert: Film as Metaphor," Griffithiana, nos. 60-61
(October 1997): 181-89.
7. See the brisk dismissal in Siegfried Kracauer, From Caligari to Hitler: A Psychological
History of the German Film (Princeton, N.J.: Princeton University Press, 1947), 45.
8. According to Horak, Sterbende Volker is preserved in negative by Gosfilmofond;
judging by Reinert s surviving films, it definitely deserves restoration. Four reels
of Die vier letzten Sekunden des Quidam Uhl are also preserved at the Munich
Film Museum.
9. Salt claims that most German filmmakers adhered to the four-meter line ("Early
German Film," 231). Kristin Thompson discusses German approaches to editing and
performance in chs. 4 and 5 of her Herr Lubitsch Goes to Hollywood: German and
American Film After World War I (Amsterdam: University of Amsterdam Press, 2005).
10. Horak, "Robert Reinert," 183.
11. Hans Schmidt, Kino-Taschenbuch fur Amateure und Fachleute (Berlin: Union
Deutsche Verlagsgesellschaft, 1921), 32.
Chapter 10
This essay is especially indebted to Schawn Belston, director of film preservation
at the Twentieth Century Fox Film Archive, for access to films and information
about CinemaScope. Thanks as well to Roy Wagner, American Society of Cinema-
tographers, and Tak Miyagishima of Panavision. I'm also grateful to John Belton
and Sheldon Hall for comments and corrections.
1. Figures from "All-Time Adjusted Domestic B.O. Champs," in The Variety Insider
(New York: Penguin, 1999), 66.
2. Spyros P. Skouras, "CinemaScope: Industry Boon," in 1953 Yearbook of Motion
Pictures (New York: Film Daily, 1953), 157, 159.
3. Quoted in Skouras, "CinemaScope," 157.
4. John Belton, "CinemaScope and the Widescreen Revolution," Cinegrafie, no. 16
(2003), 244-53.
5. Jacques Rivette, "The Age of metteurs en scene" (1954), in Cahiers du cinema: The
1950s: Neo-Realism, Hollywood, New Wave, ed. Jim Hillier (Cambridge, Mass.:
Harvard University Press, 1985), 278.
Notes
463
6. For discussions of the Cahiers directors' enthusiasm for CinemaScope, see Marc
Cerisuelo, "Au premier rang: La Nouvelle Vague des Cahiers du Cinema et le Scope
comme avant-garde," in Le CinemaScope entre art et industrie / CinemaScope
Between Art and Industry, ed. Jean-Jacques Meusy (Paris: Association fran^aise de
recherche sur l'histoire du cinema, 2003), 263-68; and Michel Marie, "Le Cinema-
Scope et la Nouvelle Vague, un parcours esthetique entre quelques film fondateurs,'
in ibid., 269-76.
7. There is a vast technical library on widescreen cinema. The indispensable refer-
ence book is Robert E. Carr and R. M. Hayes, Wide Screen Movies: A History and
Filmography of Wide Gauge Filmmaking (Jefferson, N.C.: McFarland, 1988). This
volume has been revised by Daniel J. Sherlock; see the Film-Tech website, http://
www.film-tech.com/tips/wsmc.html. See also Martin Hart's informative website,
"The WideScreen Museum," http://www.widescreenmuseum.com. The most analy-
tical and incisive history of U.S. processes is John Belton's Widescreen Cinema
(Cambridge, Mass.: Harvard University Press, 1992), to which my thumbnail history
here is indebted at many points. Information on the relations between CinemaScope
and Technicolor is available in Richard W. Haines, Technicolor Movies: The History
ofDye-Transfer Printing (Jeffer son, N.C.: McFarland, 1992), 74-81. A compact survey
highlighting French contributions is Valerie Peseux, La Projection grand spectacle:
Du Cinerama a I'Omnimax (Paris: Dujarric, 2004). See also The Velvet Light Trap,
no. 21 (Summer 1985), an issue devoted wholly to widescreen cinema.
On CinemaScope specifically, two Ph.D. dissertations are of great value: James
Spellerberg, "Technology and the Film Industry: The Adoption of CinemaScope"
(University of Iowa, 1980); and Richard Emil Hincha, "Twentieth Century-Fox's
CinemaScope: An Industrial Organization Analysis of its Development, Market-
ing, and Adoption" (University of Wisconsin-Madison, 1989). See also Stephen
Huntley, "Sponable's CinemaScope: An Intimate Chronology of the Invention of
the CinemaScope Optical System," Film History 5, no. 3 (1993): 298-320. Two infor-
mative anthologies are Helga Belach and Wolfgang Jacobsen, eds., CinemaScope:
Zur Geschichte der Breitwandfilme (Berlin: Spiess, 1993); and the Meusy collection,
Le CinemaScope entre art et industrie, cited above. CinemaScope's role in Fox's
business operations is discussed in Aubrey Solomon, Twentieth Century Fox: A
Corporate and Financial History (Metuchen, N.J.: Scarecrow, 1988), 70-145.
8. "All of the studios are convinced that the old 3/4 picture is gone and that the wider
aspect ratio is here to stay," wrote Merle Chamberlin in "Past, Present, and Future (?),"
International Projectionist 29, no. 9 (September 1954): 23.
9. They played relatively few venues in true Cinerama, however, and were wide-released
in 70mm and CinemaScope versions.
10. For information on Soviet systems, see Michael Z. Wysotsky, Wide-Screen Cinema
and Stereophonic Sound, trans. A. E. C. York, ed. Raymond Spottiswoode (New
York: Hastings House, 1971), 152-243.
11. For more on these processes, see Carr and Hays, Wide Screen Cinema, 67-89; and
Rick Mitchell, "The Widescreen Revolution: Expanding Horizons—the Spherical
Campaign," from The Operating Cameraman, (Summer 1994), http://www.soc.org/
opcam/04_s94/mg04_widescreen.html.
12. See E. I. Sponable, H. E. Bragg, and L. D. Grignon, "Design Considerations
of CinemaScope Film," Journal of the Society of Motion Picture and Television
Engineers 63, no. 1 (July 1954): 1-4.
464 Notes
13. Confused about all the names of these widescreen systems? The promoters used
a pretty limited vocabulary. Out of a few prefixes—cine(ma), vista, techni(o),
and pana—and a few suffixes—scope, vision, and rama—ingenious companies
dreamed up a baker's dozen of brand names: CinemaScope, Cinevision, Cinerama,
Cinepanoramic, Vistascope, VistaVision, Vistarama, Techniscope, Technovision,
Technirama, PanaScope, Panavision, and Panoram. ( Panorama, unfortunately,
could not be trademarked.) To any of these, Super, Ultra, and 70 may be added
as desired.
14. Anonymous, "CinemaScope—What It Is; How It Works, American Cinematographer
34, no. 3 (March 1953): 113.
15. Attendance figures are taken from "Average Weekly Motion Picture Attendance
in U.S., 1922-1965," in Christopher H. Sterling and Timothy R. Haight, The Mass
Media: Aspen Institute Guide to Communication Industry Trends (New York: Praeger,
1978), 352; and box office information comes from Joel Finler, The Hollywood Story
(London: Octopus, 1988), 288.
16. "New Dimensions Perk Up Hollywood," Business Week, March 14, 1953,122.
17. See Douglas Gomery, Shared Pleasures: A History ofMovie Presentation in the United
States (Madison: University of Wisconsin Press, 1992), 230-46; Kira Kistopanidou,
"The Widescreen Revolution and 20th Century-Fox's Eidophor in the 1950s," Film
History 15, no. 1 (2003): 32-56; and John Belton, "Fox and 50mm Film" (paper given
at Widescreen conference, Bradford, UK, July 12,2003).
18. Bob Mintz, "The Big Changeover," American Cinematographer 34, no. 10 (October
1953): 481, 497-98.
19. Tino Balio, "Retrenchment, Reappraisal, and Reorganization, 1948-," in The
American Film Industry, 2nd ed., ed. Tino Balio (Madison: University of Wisconsin
Press, 1985), 431.
20. For a useful chronology, see Herbert Bragg, "The Development of CinemaScope"
(1953), Film History, no. 2 (1988), 339-71, helpfully annotated by John Belton.
21. The ratios of these 1953 releases were 1.66:1 for Paramount, RKO, and Republic;
1.75:1 for MGM and Disney; and 1.85:1 for Allied Artists, Columbia, and Universal.
Warner Bros., toying with its own anamorphic system (WarnerScope), came out
for 1.75:1 for flat films, whereas Fox released its flat films in 1.66:1. Most exhibitors
seemed to favor 1.66:1, perhaps because it was the least variation from the orthodox
ratio. Later (though it's not clear exactly when), the new Academy ratio was stan-
dardized at 1.85:1. See Rick Mitchell, "The Origins of Ratios," American Cinematog-
rapher 75, no. 5 (May 1994): 9-10.
22. Quoted in Aubrey Solomon, Twentieth Century Fox, 87.
23. Rudy Behlmer, ed., Memo From Darryl F. Zanuck: The Golden Years at Twentieth
Century Fox (New York: Grove, 1995), 230.
24. Solomon, Twentieth Century Fox, 88-89,113; and Hincha, "Twentieth Century Fox s
CinemaScope," 221.
25. Spellerberg, "Technology and the Film Industry," 187. Another Skouras brother,
George, was president of the United Artists Theatre Circuit. It should be noted that
the cour t-decreed divestiture of the Majors' theatre holdings proceeded very slowly,
with some studios taking several years to spin off their circuits. See Michael Conant,
Antitrust in the Motion Picture Industry (Berkeley: University of California Press,
1960), 107-11.
26. On exhibitor resistance to Scope and especially its stereophonic sound package,
see Hincha, "Twentieth Century Fox's CinemaScope," 207-21.
Notes 465
27. Quoted in Mel Gussow, Don't Say Yes Until I Finish Talking: A Biography of Darryl
F. Zanuck (Garden City, N.Y.: Doubleday, 1971), 176.
28. See David W. Samuelson, "The Golden Years," American Cinematographer 84, no. 9
(September 2003): 71.
29. See Walter Wallin, "Patent Application for Anamorphosing System," Patent no.
2,890,622, filed August 11, 1954, granted June 16, 1959, cols. 2-4. I thank Tak
Miyagishima of Panavision, Inc. for providing me access to this material.
30. Hincha, "Twentieth Century Fox's CinemaScope," 263; Samuelson, "The Golden
Years," 76; and interview with Roy Wagner, American Society of Cinematographers,
Los Angeles, California, May 15, 2005.
31. Huntley, "Sponable's CinemaScope," 318-19. See also Rick Mitchell, "Widescreen
Revolution: Fourth in a Series," The Operating Cameraman 7, no. 1 (July-December
1997): 53-63.
32. Bob Allen, "Wide Screen Production Picking Up," International Photographer 33,
no. 11 (November 1961): 215. For helpful accounts of Panavision's development,
see Hincha, "Twentieth Century Fox's CinemaScope," 255-85; and Ellen Shatkin,
"Solid Gold: Panavision Celebrates 50 Years of Excellence," American Cinematogra-
pher 75, no. 11 (November 2004): 12-18. Adriaan Bijl has made his informative 1991
master's thesis, "The Importance of Panavision," available at http://www.in70mm.
com/newsletter/2002/67/panavision/index.htm.
33. On Fox's history at this period, see Solomon, Twentieth Century Fox, 91-135.
34. See figures assembled in David Pratt, "Widescreen Box Office Performance to 1959,"
The Velvet Light Trap, no. 21 (Summer 1985): 65-66.
35. Although Gigi's credits carry the CinemaScope trademark, it's likely that like most
late 1950s MGM releases, it used Panavision optics.
36. Quoted in Solomon, Twentieth Century Fox, 86.
37. Quoted in Ronald L. Davis, Just Making Movies: Company Directors on the Studio
System (Jackson: University of Mississippi Press, 2005), 18.
38. Compare the optimistic discussion of depth of field following the October 1953
presentation of the system for professional film engineers, recorded as a follow-up
discussion to James R. Benford, "The CinemaScope Optical System," Journal of the
Society of Motion Picture and Television Engineers 62, no. 1 (January 1954): 70.
39. These distortions were partly caused by the anamorphic squeeze being applied
differentially across the visual field. See Samuelson, "The Golden Years," 71.
40. Interview with Roy Wagner, American Society of Cinematographers, Los Angeles,
California, May 15, 2005.
41. Joseph Ruttenberg, "Souvenirs d'un directeur de la photographie", Positif, no. 142
(September 1972): 78.
42. Interview with Roy Wagner, American Society of Cinematographers, Los Angeles,
California, May 14, 2005.
43. See Huntley, "Sponable's CinemaScope," 312-16; and the "CinemaScope" entry in "The
Widescreen Museum," http://www.widescreenmuseum.com/widescreen/wingcs5.htm.
44. Gayne Rescher, "Wide Angle Problems in Wide Screen Photography," American
Cinematographer 37, no. 5 (May 1956): 300-1, 322-23.
45. Quoted in Peter Bogdanovich, Who the Devil Made It (New York: Knopf, 1997), 356.
Fritz Lang gives a similar opinion on p. 224, noting that of great paintings, only
DaVinci's Last Supper is in CinemaScope proportions.
 Notes
466
46. See Richard Overstreet, "Interview With George Cukor," Film Culture, no. 34 (Fall
1964): 7; Vincente Minnelli with Hector Arce, I Remember It Well (New York:
Doubleday, 1974), 280.
47. See Leo McCarey's remarks, quoted in Bogdanovich, Who the Devil Made It, 432;
Sidney Lumet's remarks, quoted on 804; and Fritz Lang's, quoted on 224.
48. Henry Koster, "Directing in CinemaScope," in New Screen Techniques, ed. Martin
Quigley Jr. (New York: Quigley, 1953), 171.
49. Ibid., 172.
50. Ibid., 173.
51. JeanNegulesco,"NewMedium—New Methods," in Quigley, New Screen Techniques,
175.
52. Leon Shamroy, "Filming 'The Robe,'" in Quigley, New Screen Techniques, 178.
53. Ibid., 180.
54. This line of argument is usually associated with Andre Bazin and the critics of
Cahiers du cinema. Despite the mediocrity of early Scope releases, Francois Truffaut
believed that the new process was as important as the coming of sound in shifting
cinema toward realism: "Every film is more or less the story of a man walking, and
a man walks more beautifully in CinemaScope"; Francois Truffaut, "Le Cinema-
scope," Cahiers, no. 25 (July 1953): 23. Bazin was more cautious, but he did indi-
cate that widescreen film would hasten the decline of cinematic "expressionism"
and move closer to cinema's real vocation, that of revealing reality; Andre Bazin,
"Le Cinemascope sauvera-t-il le cinema?" Esprit 12, nos. 10-11 (October-November
1953): 683. Bazin also argued that the wide format was in accord with a general
trend in world cinema to minimize editing: "[CinemaScope] has come to once and
for all destroy montage as the major element of cinematic discourse"; Andre Bazin,
"Fin du montage," Cahiers, no. 31 (January 1954): 43. For translations of the Bazin
pieces, see Andre Bazin, "Three Essays on Widescreen Film," The Velvet Light Trap,
no. 21 (Summer 1985): 8-17.
In my On the History of Film Style (Cambridge, Mass.: Harvard University
Press, 1997), I proposed that Bazin's 1940s pronouncements on deep-focus tech-
nique had been primed by studio promotional campaigns (225). Publicity for
Citizen Kane and other films called attention to the sharp-focus depth, the long
takes, and the absence of cutting. The same thing seems to have happened with
the arrival of CinemaScope, with the Cahiers circle taking their cue from Fox
publicity. The study of film poetics can alert us to the ways in which craft innova-
tions consciously pursued by filmmakers were announced to prepare audiences to
notice them. These announcements can in turn become the basis of more abstract
or theoretical discussions by critics.
The French were not the only critics to echo and recast Fox's official line on Scope.
Gilbert Seldes made several nuanced comments on the subject in his The Public Arts
(New York: Simon & Schuster, 1956), 46-60. The most thoroughgoing defense of the
idea that Scope favored an aesthetic based on intrashot effects rather than cutting is
Charles Barr's landmark essay, "CinemaScope: Before and After," Film Quarterly 16,
no. 4 (Summer 1963): 4-24. Elsewhere, I argue that the critics around Cahiers and
the British journal Movie (e.g., Barr and V. F. Perkins) revised Bazin's realist aesthetic
into something more self-consciously formal. See David Bordwell, "Widescreen
Aesthetics and Mise en Scene Criticism," Velvet Light Trap, no. 21 (Summer 1985):
118-25, available at http://www.davidbordwell.net/articles.
Notes 467
Interestingly, the Bazin-Cahiers-Movie line of argument had already cropped
up during Hollywood's earlier cycle of widescreen films. King Vidor commented
that thanks to the sharpness of 70mm, his Fox Grandeur production of Billy the Kid
(1930) could clearly show in a single shot a hold-up in the foreground and a rescue
party far in the distance. "The oncoming party does not know what is happening
in front, but the audience observes every moment of both with more suspense than
could be possible in any system of'cut-backs' [crosscutting]"; quoted in Harry Alan
Potamkin, "Reelife: Wide Screen Projection" (1930), in The Compound Cinema:
The Film Writings of Harry Alan Potamkin, ed. Lewis Jacobs (New York: Columbia
University Teachers College Press, 1977), 111. Unhappily, the version of Billy the Kid
I've seen doesn't contain the scene Vidor describes.
55. Charles G. Clarke, "CinemaScope Photographic Techniques," American Cinema-
tographer 36, no. 6 (June 1955): 336-37, 362-64; reprinted with minor alterations as
CinemaScope Photographic Techniques (Los Angeles: Twentieth Century Fox Studios,
1955). I'm grateful to Roy Wagner for sharing his copy of this pamphlet with me.
56. Clarke, "CinemaScope Photographic Techniques," 362.
57. Ibid., 363.
58. Ibid., 362.
59. Belton, Widescreen Cinema, 185-96.
60. See "CinemaScope," International Sound Technician 1, no. 2 (April 1953): 2-4, 22;
and Lorin D. Grignon, "Sound for CinemaScope," in Quigley, New Screen Tech-
niques, 159-70.
61. For example, Angels Over Broadway (1940), The Fatal Hour (1940), The Little Foxes
(1941), The Grand Central Murder (1942), A Guest in the House (1944), The Late
George Apley (1947), and Manhandled (1949).
62. See Kristin Thompson, Herr Lubitsch Goes to Hollywood: German and American
Film After World War I (Amsterdam: Amsterdam University Press, 2005), 71-89.
63. On recessive staging, see my On the History ofFilm Style (Cambridge, Mass.: Harvard
University Press, 1997), 168-69.
64. See David Bordwell, Janet Staiger, and Kristin Thompson, The Classical Hollywood
Cinema: Film Style and Mode of Production to 1960 (New York: Columbia University
Press, 1985), 346-52; and ibid., 221-37.
65. Quoted in Edward L. de Laurot and Jonas Mekas, "An Interview With Boris
Kaufman," Film Culture 1, no. 4 (Summer 1955): 5.
66. Quoted in Overstreet, "Interview With George Cukor," 7.
67. Quoted in Ronald Haver, A Star Is Born: The Making of the 1954 Movie and Its 1983
Restoration (New York: Knopf, 1988), 133.
68. Quoted in Todd McCarthy, Howard Hawks: The Grey Fox of Hollywood (New York:
Grove, 1997), 532.
69. Fred Zinneman, A Life in the Movies: An Autobiography (New York: Scribners,
1992), 153.
70. On the patterning of the screen surface in Scope, see Marshall Deutelbaum, "Basic
Principles of Anamorphic Composition," Film History 15, no. 1 (2003): 72-80.
71. Rescher, "Wide-Angle Problems," 301.
72. Data that follow, as well as examples used throughout this essay, come from view-
ing 185 English-language films released between 1953 and 1960. This set constitutes
about 43% of the maximum of 431 A releases of the period carrying the Cinema-
Scope trademark. I say a "maximum" because several films from 1956 to 1960,
chiefly those released by MGM, bear the trademark but were shot with Panavision
 Notes
468
lenses. Where I could confidently determine that such was the case, I didn't consider
the title a CinemaScope film and so did not count it here. It's likely that many more
films I haven't yet seen will turn out to be "pseudo-Scope," so the percentage of true
Scope films sampled here will be correspondingly higher.
The list of Scope releases I took as a point of departure was that published in Carr
and Hayes, Widescreen Movies, 91-103, but this list includes several Panavision titles.
A slightly fuller list, and the source of my total, was culled from the American Film
Institute online catalogue, available by subscription or from a subscribing library at
http://afi.chadwyk.com. To my knowledge, a complete list of films in true Cinema-
Scope (using Bausch & Lomb lenses or their clones) has not yet been compiled.
I aimed to see a minimum of one third of the Scope (i.e., non-Panavision)
releases from each year considered. My sample runs as follows: 1953, 100% of 5
Scope releases; 1954, 77.8% of 36 releases; 1955, 53% of 76 releases; 1956, 40.6% of
64 releases; 1957, 33.8% of 80 releases; 1958, 34.3% of 70 releases; 1959, 35.7% of
56 releases; and 1960, 34% of 44 releases. RegalScope films are scarce in archives
and on video, so my survey includes only three films released under that rubric.
73. Examples are There's No Business Like Show Business (1954), at 23.2 seconds; Night
People (1954), 27.3 seconds; and Pete Kelly's Blues (1955), 26.1 seconds.
74. On this point, see also Barry Salt, Film Style and Technology: History and Analysis,
2nd ed. (London: Starword, 1992), 246.
75. For example, The Badlanders (1958), 5.9 seconds; The Fly (1958), 6.7 seconds; Journey
to the Center of the Earth (1959), 6.3 seconds; and North to Alaska (1959), 5.9 seconds.
76. Oklahoma! (1955), the earliest Todd-AO title, has an ASL of just over 17 seconds,
but South Pacific of 3 years later comes in at 8.8 seconds. The ASLs of Panavision
films from 1957 range from 14 seconds (Jailhouse Rock) down to 7.3 seconds (House
Chapter 11
of Numbers), a range that holds good in later years (with, predictably, the exception
of Otto Preminger). VistaVision films exhibit a comparable flexibility. Technirama's
first U.S. release, Night Passage (1957), averages 7.8 seconds per shot; and the first live-
action Super Technirama 70 production, King Vidor's Solomon and Sheba (1959), is
cut even faster, at an ASL of 6.9 seconds.
77. On the move toward faster ASLs, see my The Way Hollywood Tells It: Story and Style
in Modern Movies (Berkeley: University of California Press, 2006), 121-24,141-47.
78. See for example Peter von Bagh, "Breadth of View," Cinegrafie, no. 16 (2003): 219,
where he speaks of "the basic flat, non-existent style so typical of the first period
of CinemaScope."
79. Hincha, "Twentieth Century Fox's CinemaScope," 266.
80. Quoted in Andy Marx, "Techie Achieves Honored by Academy," Variety, January
10-16,1994, 22.
81. Quoted in Behlmer, Memo from Darryl F. Zanuck, 235.
82. Ibid., 236.
83. Jacques Rivette, "The Age of metteurs en scene," Cahiers du cinema, no. 31 (January
1954); reprinted in Jim Hillier, ed., Cahiers du Cinema: The 1950s: Neo-Realism,
Hollywood, New Wave (Cambridge, Mass.: Harvard University Press, 1985), 279.
84. Minnelli with Arce, I Remember It Well, 280.
85. Rivette, "The Age of metteurs en scene" 278.
86. It seems likely that some such focus point was what Clarke had in mind, because he
remarks that setting the diaphragm at f/4.5 allows "quite extreme ranges of focus"
("CinemaScope Photographic Techniques," 363).
Notes 469
87. Richard Fleischer, 20,000 Leagues Under the Sea, DVD commentary, Disney DVD
no. 27327, 1:15-1:34, 79:09-79:40.
88. Interestingly, Jack Webb initially resisted CinemaScope, refusing Warner's offer to
make Dragnet (1954) in the format. "Normal size," he is said to have replied, "is plenty
big enough for men, and besides how do you fill up the sides of that long ribbon of
film?" See Michael J. Hayde, My Name's Friday: The Unauthorized but True Story of
Dragnet and the Films of Jack Webb (Nashville, Tenn.: Cumberland, 2001), 82.
89. Tonino Valerii, "Italy on Widescreen," Cinegrafie, no. 16 (2003): 299.
90. Dwight Macdonald, "The Preminger Problem," in his On Movies (New York: Da
Capo, 1981), 154. In fairness, Huston isn't seen in the sort of "extreme close-up" that
disturbs Macdonald, and the shot of Schneider—"a billboard view . . . of the area
between Miss Schneider's chin and eyebrows"—is the tightest framing in the film
and comes at the story's climax.
91. Bordwell, The Way Hollywood Tells It, 117-89.
92. Quoted in Roger Ebert and Gene Siskel, The Future of the Movies: Interviews With
Martin Scorsese, Steven Spielberg, and George Lucas (Kansas City, Mo.: Andrews
and McNeel, 1991), 73.
93. Paula Parisi, "Larger Than Life, Widescreen Rules Film," The Hollywood Reporter,
April 18, 1995, 1, 8.
94. I'm referring here to virtually all the Movie Brats, including Francis Ford Coppola
and Martin Scorsese—but not Steven Spielberg, one of the few of his cohort to
practice more or less classical mise-en-scene.
1. Paul Ekman and Maureen O'Sullivan, "Facial Expression: Methods, Means, and
Moues," in Fundamentals of Nonverbal Behavior, ed. Robert S. Felderman and
Bernard Rime (Cambridge: Cambridge University Press, 1991), 185.
2. For a useful review of these factors, see Alain Brossard, La psychologie du regard:
De la perception visuelle aux regards (Lausanne, Switzerland: Delachaux et Niestle,
1992), 185-91.
3. Primatologists debate whether chimpanzees and our other near relatives have
deictic gaze, which seems to entail a folk theory of mind. Some researchers think it
likely, whereas others disagree. See B. Hare, J. Call, B. Agnetta, and M. Tomasello,
"Chimpanzees Know What Conspecifics Do and Do Not See," Animal Behavior 59
(2000): 771-86, for one position; and, for a contending one, see Michael Povinelli,
"Theory of Mind: Evolutionary History of a Cognitive Specialization," Trends in
Neuroscience 18, no. 9 (1995): 418-24. A useful introduction to the debate is Daniel
J. Povinelli, Timothy J. Eddy, R. Peter Hobson, and Michael Tomasello, What Young
Chimpanzees Know About Seeing (Oxford: Blackwell, 2000).
4. The importance of the mother-child mutual gaze is pointed out in Jeffrey Cohn and
Edward Tronick, "Specificity of Responses to Mother's Affective Behavior," Journal
of the American Academy of Child Adolescent Psychiatry 28, no. 2 (1989): 242-48.
More generally, see the wide-ranging survey in Simon Baron-Cohen, Mindblind-
ness: An Essay on Autism and Theory of Mind (Cambridge, Mass.: MIT Press, 1995),
especially ch. 7.
 Notes
470
5. Noel Carroll, "Toward a Theory of Point-of-View Editing: Communication, Emotion,
and the Movies," in his Theorizing the Moving Image (Cambridge: Cambridge
University Press, 1996), 127-29.
6. Michael Argyle, Bodily Communication, 2nd ed. (London: Methuen, 1975), 159. For
a more complete treatment of the subject, see Michael Argyle and Mark Cook, Gaze
and Mutual Gaze (Cambridge: Cambridge University Press, 1976), chs. 3-5.
7. Janet Beavin Bavelas, Linda Coates, and Trudy Johnson, "Listener Responses as
a Collaborative Process: The Role of Gaze," Journal of Communication 52, no. 3
(September 2002): 569.
8. A functional magnetic resonance imaging (fMRI) study revealed evidence that
direct gaze elicits activity in the amygdala, a brain center bearing on emotion, either
positive or negative. See Elizabeth A. Hoffman and James V. Haxby, "Distinct Rep-
resentations of Eye Gaze and Identity in the Distributed Human Neural System for
Face Perception," Nature Neuroscience 3 (January 1, 2000): 83.1 thank Brian Boyd
for supplying this reference.
9. An exception is the scene of two characters conversing while engaged in a com-
mon task, such as working on a machine or riding in a vehicle. Even in car dialogue
scenes, though, I suspect that the person who isn't driving will look at the driver far
more often than we'd expect in real life. Drivers in movies seem to look away from
the road recklessly often as well.
10. Since this essay was first published, the point is further supported by Daniel Nettle's
observation that conversation is central to social life and that drama captures our
interest by offering an "intensified version of the concerns of normal conversation,"
a "supernormal" display that elicits a correspondingly stronger reaction in us. See
Daniel Nettle, "What Happens in Hamlet?" in The Literary Animal: Evolution and
the Nature of Narrative, ed. Jonathan Gottschall and David Sloan Wilson (Evanston,
IL: Northwestern University Press, 2005), 66.
11. Tom Kingdon, Total Directing: Integrating Camera and Performance in Film and
Television (Los Angeles: Silman-James, 2004), 315.
12. Kingdon, Total Directing, 316.
13. Ed S. Tan, Emotion and the Structure of Narrative Film: Film as an Emotion Machine
(Mahwah, N.J.: Erlbaum, 1996), 85-119.
14. I cannot find any passage in V. I. Pudovkin's Film Acting (Eng. orig., 1937; reprint,
New York: Grove, 1960) that alludes to blinking or not blinking.
15. Michael Freedland, Michael Caine (London: Orion, 1999), 37-38. Perhaps Martin
Amis learned of Caine's regimen, for in his novel Yellow Dog (New York: Vintage,
2003), we read, "On screen actors blink only when they mean to; and when Xan
decided he wanted to be an actor he had spent a lot of time practicing not blinking.
'Stop staring!' his mother used to say. 'I'm not staring. I'm practicing not blinking!"'
(240-41).
16. Noel Carroll suggests to me that this may be the source of the sense of strength
conveyed by dark glasses; they seem to present an unbroken stare.
17. Anthony Hopkins, interview with Barbara Walters, Arts and Entertainment Mundo,
broadcast March 8, 2001, in Buenos Aires, Argentina.
18. Quoted in Celebrity News, Internet Movie Database, November 30,2001, http://www.
imdb.com/news/wenn/2001-11-30. Thanks to Jonathan Frome for the reference.
19. Joseph Anderson, personal communication, July 22, 2002.
20. Quoted in Benjamin Bogery, "Digital Cinema, by George," American Cinematogra-
pher 82, no. 9 (September 2001): 71.
Notes 471
21. Daniel McNeil, The Face: A Natural History (Boston: Little, Brown, 1998), 29. See
also Argyle and Cook, Gaze and Mutual Gaze, 26.
22. Patricia Delaney and Sean Smith, "Tecce Analysis Catches Media's Eye," Boston College
Chronicle, October 31,1996, http://www.bc.edu/bc_org/rvp/pubaf/chronicle/v5/031/
tecce.html. See also J. A. Tecce, "Body Language in 2004 Presidential Debates,"
http://www.bc.edu/schools/cas/meta-elements/html/teece_analysis__2004.html.
23. Yasuko Omori and Yo Miyata, "Eyeblinks in the Formation of Impressions," Percep-
tual and Motor Skills 83, no. 2 (October 1996): 591-98.
24. Hideoki Tada, "Eyeblink Rates as a Function of the Interest Value of Video Stimuli,"
Tohoku Psychologica Folia 45 (1986): 107-13; Kenroku Tsuda and Naoto Suzuki,
"Effects of Subjective Interest on Eyeblink Rates and Occurrences of Body Move-
ments," Japanese Journal of Physiological Psychology and Psychophysiology 8, no. 1
(June 1990): 31-37; and Niina Kobayashi et al., "The Effect of Appreciating Music
Videos on Spontaneous Eyeblink," Japanese Journal for Physiological Psychology and
Psychophysiology 17, no. 3 (1999): 183-91.
25. Quoted in Hugh Kenner, Chuck Jones: A Flurry of Drawings (Berkeley: University of
California Press, 1994), 53.
26. Quoted in Delaney and Smith, "Tecce Keeps His Eye on Presidential Debates,"
http://www.bc.edu/bc_org/rvp/pubaf/chronicle/v9/019/tecce.html.
27. "Acting by Numbers," Entertainment Weekly, September 8, 2002, 89.
28. Ed McBain, Calypso (1979; reprint, New York: Avon, 1988), 73.
29. W. S. Condon and W. D. Ogston, "A Segmentation of Behavior," Journal ofPsychiatric
Research 5 (1967): 229.
30. A few such culture-specific rules are listed in Argyle and Cook, Gaze and Mutual
Gaze, 27-32.
31. See my Ozu and the Poetics of Cinema (Princeton, N.J.: Princeton University Press,
1988), 173.
32. On today's use of singles, see my The Way Hollywood Tells It: Story and Style in
Modern Movies (Berkeley: University of California Press, 2006), 129-34.
33. See Matt Ridley, The Red Queen: Sex and the Evolution of Human Nature (New York:
Penguin, 1993), 33-34.
34. Torben Grodal, Moving Pictures: A New Theory of Film Genres, Feelings, and
Cognition (Oxford: Oxford University Press, 1997), 21.
35. Michael Newman has argued persuasively that social science research into facial
expression can help us understand film narrative. See chs. 3 and 4 of his "Character-
ization in American Independent Cinema" (Ph.D. diss., University of Wisconsin-
Madison, 2005).
Chapter 12
1. The original version of this essay drew upon a sample of 163 films, nearly all viewed
on editing tables for close analysis. I am very grateful to the following archives and
their staffs for cooperation in this project: the Library of Congress, Washington,
D.C.; the National Film Archive, London; the Cinematheque Royale de Belgique,
Brussels; the Museum of Modern Art Film Library, New York; and Matsuda Eigasha,
Tokyo. I also thank John Dower, John Gillett, Peter High, Don Kirihara, and espe-
cially Komatsu Hiroshi for helping me obtain out-of-the-way material. Since the
essay's publication, I've viewed another 41 titles, on video and in projection, and
472 Notes
in revising the piece I've inserted a few new mentions. Thanks again to Komatsu
Hiroshi, as well as to Gunter A. Buchwald, to Okajima Hisashi and the staff of the
Japan Film Center, and to Kato Michiro, who invited me to the Kyoto Film Festival
of 1997. A special thanks to the organizers of the Giornate del Cinema Muto festival
in Pordenone and Sacile, Italy, who arranged festival screenings of restored classics
in 2001 and 2005.
2. Okajima Hisashi reports that the Japan Film Center holds only 500 feature titles
produced before World War II. He estimates the survival rate for Japanese films of
the prewar period to be about 4%. See Okajima Hisashi, "Japan's Case: Hopeful or
Hopeless?" Bulletin FIAF, no. 45 (1992): 2. The causes of the low survival rate include
the transience of many companies, the bigger studios' neglect of their holdings, the
fact that few prints were made of any title, and the ravages of war (e.g., the allied
firebombing of Tokyo) and natural disasters (e.g., the destruction of many Tokyo
distribution houses in the 1923 earthquake).
3. Figures 12.1 and 12.2 may not be from the same film version of the play. What has
survived seems to be a compilation of two or more films from between 1913 and
1917, all directed by Makino Shozo. Even with their uncertain dating, however, the
shots indicate some compositional variety across the general period I'm discussing.
4. For a detailed discussion of this period, see Hiroshi Komatsu, "Some Character-
istics of Japanese Cinema Before World War II," in Reframing Japanese Cinema:
Authorship, Genre, History, ed. Arthur Noletti Jr. and David Desser (Bloomington:
Indiana University Press, 1992), 229-58.
5. See Hiroshi Komatsu, "The Fundamental Change: Japanese Cinema Before and
After the Earthquake of 1923," Griffxthiana, nos. 38-39 (October 1990): 191-92.
Joanne Bernardi discusses the modernizing tendencies of this period in detail in her
Writing in Light: The Silent Scenario and the Japanese Pure Film Movement (Detroit,
Mich.: Wayne State University Press, 2001).
6. Cited in Peter B. High, "Japanese Film and the Great Kanto Earthquake of
1923," International Studies (Research Institute for International Studies, Chubu
University) 1, no. 3 (1985): 72.
7. See High, ibid., 83-84.
8. Noel Burch argues, "The Western codes had impinged upon Japanese perception,
but Japan was on the whole not interested in them as a system; they were merely used
on occasion to produce special dramatic effects" (To the Distant Observer: Form and
Meaning in the Japanese Cinema [Berkeley: University of California Press, 1979],
82). By contrast, standard Hollywood continuity, though at times a bit awkward in
handling, pervades and undergirds the fiction films from the 1920s I have been able
to see.
9. Burch, To the Distant Observer, 148-60.
10. See E. H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Repre-
sentation (Princeton, N.J.: Princeton University Press, 1961), 146-77.
11. Ozu's films won such esteem for the Shochiku company that he influenced the studio
style. It's probable that the Ozu-like cuts we occasionally find in other directors'
home dramas are one-off imitations. A New Family (1939; Figures 12.9-12.10),
directed by Ozu's pupil Shibuya Minoru, would seem to be an example.
12. Burch, To the Distant Observer, 79.
13. Eileen Bowser, The Transformation of Cinema 1907-1915, vol. 2 of History of the
American Cinema (New York: Scribners, 1990), 18-19; and Rick Altman, Silent Film
Sound (New York: Columbia University Press, 2004), 134-44.
Notes 473
14. See the essays collected in "Le bonimenteur de vues animees / The Moving
Picture Lecturer," Iris, no. 22 (Autumn 1996); and see Tom Gunning, "The Scene
of Speaking: Two Decades of Discovering the Film Lecturer," Iris, no. 27 (Spring
1999): 67-79.
15. Burch is here influenced by Roland Barthes' account of the bunraku puppets, which
stresses the physical separation of speaker and performer and finds in it a Brechtian
alienation effect. See Roland Barthes, Empire of Signs, trans. Richard Howard
(New York: Hill and Wang, 1982), 48-55.
16. For cogent criticisms of this illusionist argument, see Noel Carroll, Mystifying
Movies: Fads and Fallacies in Contemporary Film Theory (New York: Columbia
University Press, 1988), chs. 1-3; and Noel Carroll, The Philosophy of Horror; or,
Paradoxes of the Heart (New York: Routledge, 1990), 60-87.
17. Much of what follows relies upon Joseph L. Anderson's exhaustive and exuberant
study "Spoken Silents in the Japanese Cinema; or, Talking to Pictures: Essaying the
Katsuben, Contextualizing the Texts," in Noletti and Desser, Reframing Japanese
Cinema, 259-311. See also Hiroshi Komatsu and Charles Musser, "Benshi Search,"
Wide Angle 9, no. 2 (1987): 73-90; and Donald Kirihara, "A Reconsideration of the
Institution of the Benshi," Film Reader, no. 6 (1985): 41-53.
18. There is even some evidence that while early jidai-geki were shot, a prompter read
out the lines so that actors could time their movements and speech to the script
that would eventually be followed in the theatre during projection. See Anderson,
"Spoken Silents," 269.
19. Quoted in ibid., 292.
20. Quoted in Komatsu and Musser, "Benshi Search," 73. The text, from the Japanese
magazine Kinema Record (August 10, 1915), is in English in the original.
21. Cited in Anderson, "Spoken Silents," 288.
22. I discuss these options further in my Ozu and the Poetics of Cinema (Princeton, N.J.:
Princeton University Press, 1988), 66-68.
23. Quoted in "Industry: Beauty From Osaka," Newsweek, October 11, 1955, 116.
24. A lively semifictional account from the period is Yasunari Kawabata's novel
The Scarlet Gang ofAsakusa (Berkeley: University of California Press, 2005).
25. See Donald Roden, "Taisho Culture and the Problem of Gender Ambivalence," in
Culture and Identity: Japanese Intellectuals During the Interwar Years, ed. J. Thomas
Rimer (Princeton, N.J.: Princeton University Press, 1990), 58.
26. For a discussion of the "return" debates, see Tetsuo Najita and H. D. Harootunian,
"Japanese Revolt Against the West: Political and Cultural Criticism in the Twentieth
Century," in The Cambridge History of Japan, vol. 6, ed. Peter Duus (Cambridge:
Cambridge University Press, 1988), 711-74.
27. See Peter N. Dale, The Myth of Japanese Uniqueness (New York: St. Martin's, 1986),
56-76.
28. Ian Buruma, Behind the Mask: On Sexual Demons, Sacred Mothers, Transvestites,
Gangsters, Drifters and Other Japanese Cultural Heroes (New York: Pantheon, 1984),
13-14.
29. See Carol Gluck, Japan's Modern Myths: Ideology in the Late Meiji Period (Princeton,
N.J.: Princeton University Press, 1985).
30. This process has been most fully explored by Miriam Silverberg, particularly in the
following works: "The Modern Girl as Militant," in Recreating Japanese Women,
ed. Gail Lee Bernstein (Berkeley: University of California Press, 1991), 239-66;
"Constructing the Japanese Ethnography of Modernity," Journal of Asian Studies
474 Notes
51, no. 1 (February 1992): 30-54; "Remembering Pearl Harbor, Forgetting Charlie
Chaplin, and the Case of the Disappearing Western Woman: A Picture Story,"
Postions: East Asia Cultures Critique 1, no. 1 (Spring 1993): 24-76; and "Constructing
a New Cultural History of Prewar Japan," in Japan in the World, ed. Masao Miyoshi
and H. D. Harootunian (Durham, N.C.: Duke University Press, 1993), 115-43. See
also the essays collected by Carol Gluck and Stephen R. Graubard, eds., in Showa:
The Japan ofHirohito (New York: Norton, 1992).
31. Kokutai no hongi: Cardinal Principles of the National Entity of Japan, trans. John
Owen Gauntlett, ed. Robert King Hall (Cambridge, Mass.: Harvard University
Press, 1949), 52.
32. "Appendix A: Uie Way of Subjects," in Otto D. Tolischus, Tokyo Record (New York:
Reynal and Hitchcock, 1943), 417.
33. Admittedly, Komatsu Hiroshi describes an early film of a Kabuki episode filmed
from a high angle. Yet apparently this was done to show the audience that the
space was that of a stage, not to recall the high-angle perspective of graphic art. See
Komatsu, "Some Characteristics," 235.
34. See my Ozu and the Poetics of Cinema, 143-59. See also Keiko I. McDonald's
overview, "Popular Film," in Handbook of Japanese Popular Culture, ed. Richard
Gid Powers and Hidetoshi Kato (Westport, Conn.: Greenwood, 1989), 97-126. This
volume as a whole offers rich information about the peculiar mix of Western and
Japanese elements in the century's urban culture.
35. Julia Hutt, Understanding Far Eastern Art: A Complete Guide to the Arts of China,
Japan, and Korea (New York: Dutton, 1987), 35.
36. An interesting collection of such efforts may be found in Linda C. Ehrlich and David
Desser, eds., Cinematic Landscapes: Observations on the Visual Arts and Cinema of
China and Japan (Austin: University of Texas Press, 1994), 149-321.
37. See my Ozu and the Poetics of Cinema, 49.
38. This film has an original French title because it was financed and written by
Belgians, presumably to proselytize for Christianity. Nonetheless, it was shot at
Nikkatsu by two Japanese, T. Ikeda and D. M. Hirayama, and its cast list is entirely
Japanese. Because of these circumstances, and because of the film's strong similarity
to the visual style of Japanese films of the period, I consider it a Japanese film.
39. Yoda Yoshikata, "Souvenirs (2)," Cahiers du cinema, no. 169 (August 1965): 34.
40. Kido Shiro, Nihon eiga den: Eiga sesakuska no kiroku [A Japanese Film Biography:
Chronicle of a Film Producer] (Tokyo: Bungei-Shunjusha, 1956), 40.
41. See Darrell Davis, Screening Japaneseness: Monumental Style, National Identity,
Japanese Film (New York: Columbia University Press, 1996).
42. In my sample, 67% of the 1.925—1933 films had average shot lengths of fewer than
6 seconds. Because Japan converted to talking pictures quite gradually, we should
expect a rapid cutting rate for inserted titles. Hollywood cinema likewise extended
Chapter 13
shot length with the coming of sound, but the very long take did not become as
normalized there as it apparently did in Japanese films.
43. As David Desser puts it, "To be a judo master is to be free of ego, vanity, and the need
for glory. The true way of judo is non-violent" (The Samurai Films ofAkira Kurosawa
[Ann Arbor, Mich.: UMI Research Press, 1983], 63).
44. See my Ozu and the Poetics of Cinema, 73-142.
45. Discussion of Shimizu's work in English is almost nonexistent. Burch considers
Star Athlete in his To the Distant Observer, 247-56. In Japanese, an annotated
filmography appears in Film Center, no. 28 (August 28, 1974):
Notes
475
46. I discuss aspects of this scene in the following works: "Mizoguchi and the Evolu-
tion of Film Language," in Cinema and Language, ed. Stephen Heath and Patricia
Mellencamp (Los Angeles: American Film Institute, 1983), 107-17; On the History of
Film Style (Cambridge, Mass.: Harvard University Press, 1998), 215-16; and Figures
Traced in Light: On Cinematic Staging (Berkeley: University of California Press,
2005), 114-15. See also Don Kirihara, Patterns of Time: Mizoguchi and the 1930s
(Madison: University of Wisconsin Press, 1992), 109-11.
47. Otsuka Kyoichi, cited in Peter B. High, "Japanese Film Theory and the National
Policy Film Debate: 1937-1941," International Studies (Research Institute for Inter-
national Studies, Chubu University) 2 (1986): 142.
48. For discussions of the cultural impulses behind this genre, see High, "Japanese Film
Theory," 142; and Davis, Screening Japaneseness.
49. A detailed discussion of these genres and the context of their production may be
found in Shimizu Akira, "War and Cinema in Japan," in Japan/America Film Wars:
WWII Propaganda and Its Cultural Contexts, ed. Abe Mark Nornes and Fukushima
Yukio (New York: Harwood, 1994), 7-58.
50. For an excellent commentary on Five Scouts, see Peter B. High, "A Propaganda of
'Real Human Emotions': The 'Humanist' War Films of 1938-1940," International
Studies (Research Institute for International Studies, Chubu University) 4 (1987):
82-92.
51. For a discussion, see High, "A Propaganda of'Real Human Emotions,"' 82-83.
52. See John Dower, War Without Mercy: Race and Power in the Pacific War (New York:
Pantheon, 1986), 203-32.
53. "'As long as I don't fight, I'll make it home': Suzuki Murio," in Haruko Taya Cook and
Theodore F. Cook, Japan at War: An Oral History (New York: New Press, 1992), 132.
54. John Dower, "Japanese Cinema Goes to War," in his Japan in War and Peace: Selected
Essays (New York: New Press, 1993), 36-40.
55. Cited in High, "A Propaganda of'Real Human Emotions,"' 78.
56. See Donald Richie, A Hundred Years of Japanese Film, rev. ed. (Tokyo: Kodansha,
, 2005), 54-56.
57. Hiroshi Komatsu, "The Foundation of Modernism: Japanese Cinema in the Year
1927," Film History 17, nos. 2-3 (2005): 375.
58. Jeffery A. Dym, Benshi, Japanese Silent Film Narrators, and Their Forgotten Narra-
tive Art of Setsumei: A History of Japanese Silent Film Narration (Lewiston, Wales:
Mellen, 2003).
59. Tokugawa Musei, quoted in Dym, Benshi, 151.
1. I discuss how academic film interpretation treats film style in my Making Meaning:
Inference and Rhetoric in the Interpretation of Cinema (Cambridge, Mass.: Harvard
University Press, 1989), chs. 7 and 8.
2. E. H. Gombrich, The Sense of Order (Ithaca, N.Y.: Cornell University Press, 1979),
134-52.
3. Heinrich Schenker systematized this concept most fully in his ideas of Stufen
(underlying harmonic regions) and Ursatz (fundamental tonal structure). An infor-
mal discussion may be found in Joseph Kernan, Contemplating Music: Challenges
to Musicology (Cambridge, Mass.: Harvard University Press, 1985), 79-90; a more
476 Notes
technical survey is in Nicholas Cook, A Guide to Musical Analysis (New York:
Braziller, 1987), 27-66. A cognitive reworking of Schenkerism is offered by Fred
Lerdahl and Ray Jackendoff in A Generative Theory of Tonal Music (Cambridge,
Mass.: MIT Press, 1983), 105-123.
4. I don't want to play down the abstraction that is probably inherent in early stages ofvisual
processing. See David Marr, Vision (San Francisco: Freeman, 1982), 91-98, 296-328.
It's just that judgments of decorative patterning are relatively high-level and top-down,
applicable only after what Marr calls the "3-d sketch" has been constructed.
5. David Bordwell, Narration in the Fiction Film (Madison: University of Wisconsin
Press, 1985), 30-33.
6. This might be one source of the form-content distinction that is so hard to shake.
One can, it seems, always imagine the same "plain version"—the same action or
state of affairs—but embellished in different, more or less straightforward ways.
7. Viktor Shklovsky, "Poetry and Prose in Cinematography," Russian Poetics in Transla-
tion, no. 9 (1982): 88.
8. Gombrich, The Sense of Order, 239.
9. Ibid., 240.
10. Bordwell, Narration, 280-81.
11. Ibid., 49-62.
12. Ibid., 281.
13. See David Bordwell, Ozu and the Poetics of Cinema (Princeton, N.J.: Princeton
University Press, 1988), 73-142.
14. Konishi Jin'ichi, "The Art of Renga," Journal of Japanese Studies 2, no. 1 (Autumn
1975): 47.
15. I develop this case in my Ozu and the Poetics of Cinema, chs. 7 and 8; and in David
Bordwell, Figures Traced in Light (Berkeley: University of California Press, 2005),
96-98.
16. See, for a brief but penetrating discussion, Donald Keene, Japanese Literature:
An Introduction for Western Readers (New York: Grove, 1955), 10-13.
17. A brief but suggestive discussion is in Earl Jackson Jr., "Elaboration of the Moment:
The Lyric Tradition in Modern Japanese Narrative," in Literary History, Narrative,
and Culture: Selected Conference Papers, ed. Wimal Dissanayake and Steven
Bradbury (Honolulu: University of Hawaii and the East-West Center, 1989), 5-13.
18. See Joseph Anderson and Donald Richie, The Japanese Film: Art and Industry
(Princeton, N.J.: Princeton University Press, 1982), 48-62; David Bordwell, "Our
Dream-Cinema: Western Historiography and the Japanese Film," Film Reader, no.
4 (1979): 46-48; and Bordwell, Ozu and the Poetics of Cinema, 19-21,151-59.
19. See Masumura Yasuzo, Profilo storico del cinema giapponese, trans. Guido Cincotti
(Rome: Bianco e nero, 1955), 19-20; and Anderson and Richie, Japanese Film,
323-24.
20. For this information, I'm indebted to Eric Crosby's fine research essay, "A
Comparative Look at Japanese and Hollywood Widescreen Films," unpublished
paper (University of Wisconsin-Madison, December 20, 2004).
21. The Kowa company, which had begun producing optics only in 1946, supplied these
anamorphic elements. "Put behind an off-the-shelf 10:1 zoom lens, they didn't look
too bad—it was probably the failings of the zoom lens that hid the shortcomings of
the rear anamorphoser" (David Samuelson, "Golden Years," American Cinematog-
rapher 84, no. 9 [September 2003]: 76). Samuelson also remarks that many anamor-
Notes 477
phic processes in Europe used Kowa lenses. Kowa also manufactured zoom lenses
for movie cameras and projectors in several gauges. Eventually the firm came to
specialize in lenses for binoculars and telescopes.
Kowa lenses were used for Hong Kong productions as well. Even after Salon
Films acquired the regional Panavision franchise in 1973, many productions con-
tinued to employ the Kowa system. Thanks to Charles Wang Cheung Tze, managing
director of Salon Films, for this information in a November 2003 interview.
Chapter 14
1. Stefan Hammond and Mike Wilkins, Sex and Zen and a Bullet in the Head (New
York: Fireside, 1996), 11.
2. Since this essay was published in 1997, several sources have appeared that help
clarify the trends that I trace. Foremost are the books published by the Hong Kong
Film Archive and the Hong Kong International Film Festival: The Cathay Story
(Hong Kong: Hong Kong Film Archive, 2002) and The Shaw Screen: A Preliminary
Study (Hong Kong: Hong Kong Film Archive, 2003) are indispensable for research
into the 1960s and 1970s action cinema. In addition, the archive has published
Chang Cheh: A Memoir (Hong Kong: Hong Kong Film Archive, 2004), a collec-
tion of the director's writings; and Sam Ho and Ho Wai-leng, eds., The Swordsman
and His Jiang Hu: Tsui Hark and Hong Kong Film (Hong Kong: Hong Kong Film
Archive, 2002). The festival, often a copublisher on these ventures, has also produced
Li Cheuk-to, ed., A Tribute to Action Choreographers (Hong Kong: Hong Kong Film
Archive, 2006), with many valuable interviews and research articles.
Outside Hong Kong, the interest has proven intense. See Miles Wood, Cine East:
Hong Kong Cinema Through the Looking Glass (Guildford, UK: FAB, 1998); Charles
Tesson, Claudine Paquot, and Roger Garcia, eds., LAsie a Hollywood (Locarno:
Festival International du Film de Locarno, 2001); Daniel O'Brien, Spooky Encoun-
ters: A Gwailo's Guide to Hong Kong Horror (Manchester, UK: Headpress, 2003);
and Leon Hunt, KungFu Cult Masters: From Bruce Lee to Crouching Tiger (London:
Wallflower, 2003). See also the essays in Heroic Grace: The Chinese Martial Arts
Film, ed. David Chute (Los Angeles: UCLA Film and Television Archive, 2003).
Useful sources on individual directors are Lisa Morton, The Cinema of Tsui Hark
(Jefferson, N.C.: McFarland, 2001); Tim Youngs, ed., Black Roses and Sentimental
Swordsmen: The Cinema ofChor Yuen (Udine, Italy: Far East Film Festival, 2004);
and Robert K. Elder, ed., John Woo Interviews (Jackson: University of Mississippi
Press, 2005). Stephen Teo surveys the martial arts film tradition in "Fight and
Flight: The Wuxia Film in Chinese Cinema" (Ph.D. thesis, Royal Melbourne Insti-
tute of Technology, 2002).
John Charles' The Hong Kong Filmography 1977-1997 (Jefferson, N.C.: McFarland,
2000) is an indispensable reference work. Another in-depth resource is Alberto
Pezzotta, Tutto il cinema di Hong Kong: Stili, caratteri, autori (Milan, Italy: Baldi and
Castoldi, 1999).
A fine scholarly study of the wuxia novel is provided in John Christopher Hamm,
Paper Swordsmen: Jin Yong and the Modern Chinese Martial Arts Novel (Honolulu:
University of Hawai'i Press, 2005). See also Margaret Baptist Wan, "Green Peony
as New Popular Fiction: The Birth of the Martial Romance in Early Nineteenth-
Century China" (Ph.D. diss., Harvard University, 2000).
478 Notes
I develop this essay's argument in greater detail in my Planet Hong Kong: Popular
Cinema and the Art of Entertainment (Cambridge, Mass.: Harvard University Press,
2000), ch. 8.
3. Indeed, Pudovkin's example is an automobile accident, broken into several shots to
give the impression of the event. See V. I. Pudovkin, Film Technique and Film Acting,
trans. Ivor Montagu (New York: Grove, 1960), 95-100.
4. Quoted in Eric Rudolph, uThe Rock Offers No Escape," American Cinematographer
77, no. 6 (June 1996): 65.
5. Andre Bazin, "The Virtues and Limitations of Montage," in his What Is Cinema? ed.
and trans. Hugh Gray (Berkeley: University of California Press, 1967), 50.
6. For excellent discussions of the martial arts film as a genre, see' A Study of the Hong
Kong Martial Arts Film, program book for the Fourth International Hong Kong Film
Festival, April (Hong Kong: Urban Council, 1980), particularly Sek Kei's overview,
"The Development of'Martial Arts' in Hong Kong Cinema," 27-38.
7. Sergei Eisenstein and Sergei Tretyakov, "Expressive Movement," in Meyerhold,
Eisenstein and Biomechanics: Actor Training in Revolutionary Russia, ed. Alma Law
and Mel Gordon (Jefferson, N.C.: McFarland, 1996), 189.
8. Sergei Eisenstein, "Beyond the Shot," in Writings, 1922-34, vol. 1 of Selected Works,
ed. and trans. Richard Taylor (London: British Film Institute, 1988), 148. See also
Sergei Eisenstein, "An Unexpected Juncture," in Writings, 1922-34, 115-22; and
Sergei Eisenstein, "To the Magician of the Pear Orchard," in Writings, 1934-47,
vol. 3 of Selected Works, ed. Richard Taylor, trans. William Powell (London: British
Film Institute Publishing, 1996), 56-67.
9. Author's interview with Yuen Woo Ping, Hong Kong, November 23,1996.
10. Eisenstein and Tretyakov, "Expressive Movement," 187.
11. I discuss this aspect of Eisenstein's theory, as well as his conception of how
mise-en-scene and other techniques can create the effect of ecstasy, in The Cinema
of Eisenstein (New York: Routledge, 2005), chs. 3-5.
Notes
479
9. He commented that whereas Western art often strongly separated the imitation of
nature from the expression of other worlds, Chinese art lived between these poles,
showing imaginary things in realistic ways (Tesson, "Calligraphic," 21).
10. See the previous essay, "Aesthetics in Action: Kung-Fu, Gunplay, and Cinematic
Expression."
11. Cited in Teo, "The Dao of King Hu," 34. Ann Hui recalled that early 1970s experi-
mental director Tang Shuxuan told her that with his eight-frame shots, "King Hu
invented editing for Chinese movies" (Ann Hui, interview, Madison, Wisconsin,
February 12,1997).
12. Perhaps this is why Hu called the forest battle "a stylistic breakthrough in cinema"
(quoted in Teo, "The Dao of King Hu," 34).
13. Quoted in Peter Bogdanovich, Who the Devil Made It (New York: Knopf, 1997), 250.
14. Commentators have long noted this tendency toward abstraction in his work.
The Valiant Ones, Tony Rayns observes, constitutes "a suite of highly formalized
action scenes which move progressively toward abstraction" (Rayns, "Director: King
Hu," 105). See also Tony Rayns, "King Hu: Shall We Dance?" in A Study of the Hong
Kong Martial Arts Film, ed. Lau Shing-hon (Hong Kong: Hong Kong International
Film Festival / Urban Council, 1980), 103-6; and Olivier Assayas, "King Hu: Geant
Exile," Cahiers du cinema, nos. 360-61 (September 1984): 17-19.
15. Quoted in Tesson, "Calligraphic et simulacres," 22.

Chapter 15
1. Several of these topics are pursued in the volume in which this essay originally
appeared: Law Kar, ed., Transcending the Times: King Hu and Eileen Chang (Hong
Kong: Hong Kong International Film Festival / Urban Council, 1998).
2. Quoted in Stephen Teo, "The Dao of King Hu," in A Study of Hong Kong Cinema in
the Seventies, ed. Li Cheuk-to (Hong Kong: Hong Kong International Film Festival /
Urban Council, 1984), 34. This essay is available in revised form in Stephen Teo,
Hong Kong Cinema: The Extra Dimensions (London: British Film Institute, 1997),
87-96.
3. Michel Ciment, "Entretien avec King Hu," Positif, no. 169 (May 1975): 34-35.
4. Quoted in Charles Tesson, "Calligraphic et simulacres," Cahiers du cinema,
nos. 360-61 (September 1984): 24.
5. Tony Rayns, "Director: King Hu," Sight and Sound 45, no. 1 (Winter 1975-1976): 10.
6. See his remarks in Tesson, "Calligraphic et simulacres," 21.
7. For a broader survey of stylistic norms in Hong Kong cinema, see chs. 5 - 8 of my
Planet Hong Kong: Popular Cinema and the Art of Entertainment (Cambridge, Mass.:
Harvard University Press, 2000).
8. See Ciment, "Entretien," 34.
Index

+1-1 (Plus One Minus One, 1998), 186 Adler, Buddy, 289, 305
13 rue de Percebe, 195 Adventures of a 10-MarkNote, The (Die Abenteur
180-degree rule, 53, 340, 341, 366 eines Zehnmarkscheines, 1926), 202
2 Days in the Valley (1996), 228 Advise and Consent (1962), 192
2 ou 3 choses queje sais d'elle (1967), 158 Akerman, Chantal, 116
2:37 (2006), 243 Albert, Barbara, 217
20 30 40 (2003), 211,212,230 Alexander the Great (1956), 291
20,000 Leagues under the Sea (1955), 288, 315 Alfie (1966), 153
200 Cigarettes (1999), 204, 208 Alice in Wonderland, 220
2001 (1968), 157, 259 Aliens (1986), 399
2046 (2004), 161 Alila (2003), 203
21 Grams (2003), 216, 217, 219, 458n36 All for Love (Nae Saengae Gajang Areumdaun
3 Women (1977), 157 Iljuil/My Lovely Week, 2005), 243
3-D, 286, 287, 290, 293 All Quiet on the Western Front (1930), 370
400 coups, Les (1959), 156, 157 All the King's Men (Tianxia diyi, 1983), 413
70mm film, 285, 289, 290, 463n9, 467n54 Allen, Gene, 300
71 Fragments of a Chronology of Chance (1994), Allen, Woody, 262
197, 221 Almodovar, Pedro, 159
8 1/2 (Otto e mezzo/Eight and a Half, 1963), 151, Alphaville (1965), 217
153, 154, 156, 162 Altman, Robert, 157,191, 197,198, 202, 212,
221-228, 234, 243
Amateur Club (Amachua kurabu, 1920), 390
A ambiguity, 50,101, 152,156-158,169, 219, 241, 261
American Cinematographer, The (magazine), 33,293
A bout de souffle (1960), 27,152, 154, 217 American Gun (2005), 191
A nous la liberte (1931), 71 Amoresperros (2000), 206-209, 457
Abe Clan, The (Abe ichizoku, 1938), 356, 362 Amour, L' (Aibu, 1933), 382
Abe Yukata, 390 analytical editing, 14, 48, 95, 264-6, 273, 278, 364
Abrams, J.J., 67 anamorphosis, 282, 283, 285-292, 304, 312,
Academy ratio, 285, 310, 323, 464 321-325, 391, 392, 422, 464, 465, 476
Acmeism, 65 Anastasia (1956), 321
act structure, 29, 89, 102-110, 219, 437n58, 446n30, And God Created Woman (Et Dieu ... crea lafemme,
458n36 1956), 153
Actor's Revenge, An (Yukinojo henge, 1935), 377, Anderson, Barbara Fisher, 78
384,385 Anderson, Joseph, 78, 346, 473
Adam's Rib (1949), 301 Anderson, Paul Thomas, 163, 191, 228, 229, 231,
Adaptation (2002), 63 232,243
Adieu, plancher des vaches! (Farewell, Home Sweet Anger (Nu), segment of film Hsi nou ai lueh (1970),
Home!, 1999), 234 419, 424

481
482 Index Index 483

animated film, 186, 408
Annee derniere a Marienbad, L' (1961), 158
Antonioni, Michelangelo, 25, 155-157, 161, 452
Any Way the Wind Blows (2003), 219, 220
Aoyama Shinji, 161
Apart From You (Kimi to wakarete, 1933), 380,
382-384
Apartment, The (1960), 120
aperture framing, 308, 316, 348, 355,
apparent motion, 63
Apple, The {Sib, 1998), 162
appointments, 199, 219, 225, 240
April (Aprili, 1961), 234
Apu trilogy (1955-9), 153,155
arbitrariness in representation, 57, 60, 62-64, 73,
80, 136, 381, 441n7, 442nl4
Argent, L' (1983), 203
Argyle, Michael, 328
Aristotle, 12,13, 89, 98, 101-103, 105, 445
Armor of God II (Feiying Jihua, 1991), 408
Arnheim, Rudolf, 47, 255
Around the World in Eighty Days (1956), 31, 289,
293
art cinema, 27, 66, 81, 101, 117, 151-70, 178-80,
210, 218, 219, 221, 238, 241, 242, 244
Ashes and Diamonds (Popiol i diament, 1958), 151
Ashes of Time (Dung che sai duk, 1994), 104,161
aspect ratio, 282, 284-286, 324, 463, 464n21
Assunta Spina (1915), 295
Astaire, Fred, 292
Astruc, Alexandre, 259, 260
Au hasard Balthazar (1966), 202
Auberge espagnole, L' (2002), 192
Auden, W.H., 13-15, 17
Autant-Lara, Claude, 109
authorship, 152-160,166, 169, 241, 254-262, 282
Autumn Afternoon, An (Samma no aji, 1963), 157
avant-garde, 14,19, 66, 157, 337, 338, 390
average shot lengths (ASL), 37, 38, 42, 294, 301,
304, 307, 315, 359, 361, 363, 397, 399,
401, 424, 437n61,468n76, 474n42
Avventura, L' (1960), 25, 152-154,156
Awful Truth, The (1937), 257
Ayckbourn, Alan, 185, 455nl5
B Bonjour Tristesse (1958), 311, 312
Back to the Future (1985), 186, 454nl0
Back to the Future II (1989), 186, 455nl0
backstory, 105,169, 208, 226
Bad Day at Black Rock (1955), 313, 314, 317, 318
Badlanders, The (1958), 310, 468
Badlands (1973), 129
Bahrani, Ramin, 163
Bakhtin, Mikhail, 4, 87
Ball, Lucille, 141
Ballet Mecanique (1924), 19
Balzac, Honore de, 194, 213, 220, 244
Barr, Charles, 466
Barthes, Roland, 141, 143, 473nl5
Battle Cry (1955), 288, 303
Battle of Russia, The (1943), 87
Battleship Potemkin (Bronenosets "Potemkin," 1925),
104, 357
Bauer, Yevgeni, 265, 278
Baum, Vicki, 194, 195
Bausch & Lomb, 287-289, 291, 292, 305, 468n72
Bay, Michael, 398, 399, 401
Bazin, Andre, 14-16,19, 41, 42, 47, 154, 256, 268,
263, 282, 401, 466n54
Beautiful People (1999), 193, 196, 213
Becker, Josh, 42
Beijing Opera, 401, 406, 409, 413
Belmondo, Jean-Paul, 254
Belton, John, 282, 294, 438
Belvaux, Lucas, 244
Beneath the 12-Mile Reef(1953), 287, 288, 304, 317
Ben-Hur (1959), 289
Benjamin, Walter, 75, 76
Benning, James, 45
benshi, 339, 340, 345-347, 353, 373
Bergman, Ingmar, 152, 155, 161, 254, 452
Bertolucci, Bernardo, 254
Better Tomorrow 2 (Yinxiong bense er, 1987), 410
Betty Fisher et autres histoires (2001), 215-217
Between War and Peace (Senso to heiwa, 1947), 391
Bicycle Thieves (Ladri di biciclette, 1948), 152
Big Chill, The (1983), 203
Big Fisherman, The (1959), 285
Big Heat, The (Cheng shi tejing, 1988), 408
Big Parade, The (Dayue bing, 1986), 161
Billy the Kid (1930), 467n54
biology, 4, 45, 61, 62, 66, 73-76, 78-81, 448n52
Birdcage, The (1996), 118
Birth of a Nation, The (1915), 283
Blackboards (Takhte siah, 2000), 162
Blind Chance (Przypadek, 1987), 101,172-176,
178-185
blinking, 24, 79, 327, 331-335,470nl5
Blow-Up (1966), 153, 155
Blue Gardenia, The (1953), 98
Bodyguard from Beijing (ZhongNan Hai bao biao,
1984), 404, 405,411
Boe, Christoffer, 163
Bogart, Humphrey, 113, 254
Bonnie and Clyde (1967), 396
Borges, Jorge Luis, 171,172, 174,175, 187
Botelho, Joao, 191
Boyz N the Hood (1991), 107,108
Brakhage, Stan, 19, 66
Breathless (1960). See A bout de souffle
Brecht, Bertolt, 158, 473
Bresson, Robert, 163,202,203,262,279,334,448n55
Bridge of San Luis Rey, The, 194
Bridge on the River Kwai, The (1957), 281, 289,
303, 316
Brigadoon (1954), 291, 303, 304
Brigands, Chapitre VII (1996), 234
Brighter Summer Day, A (Gulingjie shaonian sha
ren shijian, 1991), 161
Brokeback Mountain (2005), 50
Bronson, Charles, 334
Brooks, Richard, 256
Bruner, Jerome, 136
Brunetiere, Ferdinand, 117, 118
Bug (2002), 193,213,227
Building Stories, 196
Bullitt (1968), 396
bunraku puppet theatre, 339, 473
Bunuel, Luis, 155, 452
Burch, Noel, 19, 20, 47, 341, 345-347, 472n8,
473nl5, 474n45
Burden of Life (Jinsei no onimotsu, 1935), 384
Buruma, Ian, 354, 373
Bus Stop (1956), 301, 302, 305
Busey, Gary, 397, 398
Butch Cassidy and the Sundance Kid (1969), 157,
192,396
Butor, Michel, 195
Butterfly Effect, The (2004), 457n23
Bwana Devil (1952), 286
c 419,420, 426,429, 446n38, 458n38
Cabinet of Dr. Caligari, The (Das Cabinet des
Dr. Caligari, 1920), 269
Cache (2005), 169
Cafe Lumiere (Kohijiko, 2003), 161
Cahiers du Cinema, 155, 258, 283, 312, 466n54
Caine Mutiny, The (1954), 289
Caine, Michael, 330-334, 470
Calvino, Italo, 189, 193
Camera, 65, 286, 289
camera movement, 19, 20, 38-41, 65, 68, 94, 95,
148, 155, 163,165, 233, 255-258, 294,
296, 324, 348, 357, 363, 364, 368, 370,
379, 384, 397, 398, 428, 438n73
Cameron, James, 262
Cannibals, The (Os Canibais, 1988), 164
Capra, Frank, 87, 256
Caprices of Marianne, The, 195
Carabiniers, Les (1963), 217
Carax, Leos, 164
Cardiff, Jack, 438n74
Cardinal, The (1963), 322
Carello, Claudia, 81
Carmen Jones (1954), 304, 452
Carousel (1956), 286
Carroll, Noel, 18, 67, 150, 328, 433nl3, 442nl4,
470nl6
Carrosse d'or, Le (1953), 283
Cat on a Hot Tin Roof (1958), 229
Cathay (film production company), 430
Caught (1949), 297
cause-effect structure in narrative, 15, 20, 27,
152-8, 166, 176, 179, 180, 187, 191, 193,
195, 199, 204, 211, 212, 216, 217, 219,
225, 241, 388, 452n6
Cellular (2004), 113, 114
censorship, 338, 400
Centipede Horror (Wu gongzhou, 1984), 395
Ceremony, The (Gishiki, 1971), 158
Chahine, Youssef, 69
chambara, (swordplay film), 359, 360, 371, 372, 377,
380, 385
Chan, Jackie, 43, 47, 404, 407, 408, 414, 430
Chang Cheh. See Zhang Che
Chang, Sylvia, 190,195, 211
Changing Lanes (2002), 214
chaos theory, 198, 455
Chaplin, Charles, 260, 262, 338, 359
characters and characterization, 52, 87-92, 94,
95, 97-101, 103, 105-7, 109,112-21,
124-7, 131, 133, 137-9,141-3, 150,
152-69,172,173, 176-8, 180, 182, 186,
187, 189-97,199-220, 222-37, 239-43,
255, 268, 269, 271-3, 276-8, 280, 283,
293-8, 300, 301, 303, 307, 308, 310,
313-15, 322, 327, 329-33, 335, 341, 342,
346-9, 359, 363, 366, 368, 373, 377,
380-2, 384, 388, 392, 398, 407, 413, 415,
Chasse auxpapillons, La (ChasingButterflies, 1992),
234
Chatman, Seymour, 102, 112, 127-133
Chen Kaige, 161
Chiao, Roy, 427
Child and the Soldier, The (Koudak va sarbaz, 2000),
162
Children in the Wind (Kaze no naka no
kodomotachi, 1937), 348, 349
China Gate (1957), 322
China Night (Shina noyoru, 1940), 370
Chinatown (1974), 329, 330, 333, 334
Ching Siu-tung, 430
Chion, Michel, 110
Chomsky, Noam, 5, 93
Choose Your Own Adventure, 187
Chor Yuen, 402, 411
Chow Yun-Fat, 399, 401, 404, 407, 409
Chretien, Henri, 287, 288, 291, 292
"Christmas Carol, A," 172, 174, 182, 186
Chromophobia (2005), 191
Chronicle of Anna Magdalena Bach (Chronik der
Anna Magdalena Bach, 1968), 163, 164
Chungking Express (Chung Him sam lam, 1994),
103, 197
Chushingura (compilation, with footage from
1913-1917), 339
Cincinnati Kid, The (1965), 253
Cinema City (production company), 104
cinema du look, 163
Cinema Novo, 161

484
cinema pur, 451
CinemaScope, 3, 25, 29, 281-325, 392
average shot lengths (ASL) in, 294, 301, 304,
307, 468
black-and-white, 289, 305, 306, 321
close-ups in, 288, 290-295, 301, 305, 306, 309,
311,321-324, 469
clothesline staging in, 300, 301, 303, 307, 308,
313-315, 320, 321,323
as cropping of traditional frame, 311, 312, 323
depth in, 291, 293, 294, 303, 305, 310-317, 320,
322
distortion in, 288, 291-294, 301-303, 306, 312,
465
edge framing in, 292, 302, 309, 321, 402, 415,
420, 421,423,427
focus in, 288, 290-293, 298, 300, 305, 306, 308,
312, 313, 315, 317, 319, 468n86
movement of the camera, 154, 292-295, 301,
304, 307, 323, 324, 401, 428
"mumps" caused by, 288, 291, 301, 306, 308
partitioning of the screen in, 310, 311, 316, 317,
319, 321
staging principles in, 294, 295, 300, 301, 307,
308, 311-313, 320, 321, 323-325
recessive staging in, 303, 306, 323
stereophonic sound, 285, 287, 290, 292, 307,
464n26
CinemaScope, 55, 286
cinematography, 19,129,152, 292, 307, 399,401,
402,404. See also camera movement;
close-up; deep-focus; depth in image;
depth of field; fast-motion; focus; lenses;
lighting; long take; over-the-shoulder
shot; pan shot; plan americain;
planimetric image; reverse shot;
three-quarter view of action; zoom lens
Cinerama, 282, 284, 286, 287, 293, 294, 463n7
Citizen Kane (1941), 14,15, 27, 38, 166, 255, 257,
294, 297, 298,466n54
Citizens Band (1977), 228
City of Hope (1991), 228
City of Sadness (Beiqing chengshi, 1989), 161, 201
Clair, Rene, 70, 71,117
Claire Dolan (1998), 117,120, 448
Clarke, Charles G., 293, 294, 300, 301, 307, 313,
468n86
Class Relations (Klassenverhaltnisse, 1984), 59
classical Hollywood cinema, 14, 24, 75, 76, 101,
109, 138, 144, 149,150,152, 153,
155-158,163, 178,180, 218, 259, 342
classical style in cinema, 14, 76, 78, 107,117,
153-156, 218, 219, 225, 230, 233, 242,
283, 284, 306, 312, 321, 325, 338, 342,
357, 359, 364, 365, 372, 375, 359, 364,
365, 372, 375, 376, 380, 389, 393, 452n6
Index
climax, 25, 26, 36, 37, 79, 92, 94, 102, 103,
105-108, 117, 126,140, 142-144, 184,
202, 206-208, 210, 215, 217, 224-226,
230, 233, 261, 265, 269, 273, 276, 278,
321, 347, 357, 359, 364, 368, 384, 385,
397, 399, 404, 406, 408, 414, 419, 420,
425, 427, 430, 469n90
Cloak, The (Shinel, 1926), 70, 71
Clodhopper, The (1917), 265
Close Encounters of the Third Kind (1977), 290
Close Up (Nema-ye Nazdik, 1990), 162
close-up, 25, 36-40, 63, 147, 177, 221, 234, 259,
269, 276-8, 280, 294, 295, 298, 323,
324, 334, 335, 342, 357, 359-61, 384,
392, 398, 401, 404, 407, 416, 424
Close-Up Long-shot (1996), 162
clothesline staging, 296, 299, 301, 307, 308, 321, 323
Clowes, Daniel, 196
Clowns, The (I Clowns, 1971), 153
Cocteau, Jean, 18, 41
Code inconnu (2000), 213, 221
codes, 22, 53, 60, 63-65, 93,136, 137, 150, 155
Coen, Ethan and Joel, 262
cognitivism, 4-6, 44, 45, 51, 53, 54, 75, 76, 81,
101,137,149, 150, 174,178, 378, 438,
440nl01, 476n4
cohesion, 178-180, 187, 225
Come Drink With Me (Da zui xia, 1966), 413, 415,
424-428
comedy, 24, 31, 97,107, 150, 157,159, 184,192,
202, 210, 217, 219, 233, 234, 243, 244,
287, 325, 338, 339, 343, 350, 359, 376,
379, 395
comic strips, 58, 65, 87, 195, 353
communication model of narration, 61, 99, 124,
125, 127-131, 133,149, 328, 448n65,
449n77
Compulsion (1959), 302
Confidence (2002), 323
constructive editing, 397, 401, 415-417, 419, 423,
425, 426, 429, 430
constructivism, social, 57, 73, 74, 80, 335, 379
Contempt (1963). See Mepris, Le
continuity editing, 27, 28, 32, 72, 76, 77, 79, 137,
152,157, 283, 294, 296, 306, 312, 323,
339-343, 351, 356, 363-365, 368, 372,
388, 390, 408, 425, 429,472n8
conventionalism, 15, 60, 61, 63, 65-68, 74, 79, 81,
82,335, 441n7
Index
conventions, artistic, 3, 15, 17, 22, 24, 25, 27-30,
48, 57, 59-63, 65, 66, 74, 77, 79, 81, 82,
86, 99,113, 127,136,137,140,142,151,
152, 155, 157,158, 161, 163, 173, 174,
180,184-187, 191, 195, 202, 203, 206,
209-212, 214, 215, 217-219, 224, 225,
227-229, 231, 233, 241, 243-245, 273,
276, 333-335, 341, 344, 355, 366, 370,
372, 375, 377, 379, 380, 382, 384, 388,
390, 392, 396, 409, 415, 425, 442nl4,
443n33, 448n65
Conversation, The (1974), 157
Conversations on a Sunday Afternoon (2005), 243
Cook, Pam, 149
Coover, Robert, 186
Coppola, Francis Ford, 157, 290, 469
craft skills, 1, 3, 12, 15, 22, 29, 42, 54, 60-66, 68,
76-79, 86-88, 91, 109, 112, 123, 137,
160, 180,196,197, 244, 259-261, 265,
284, 320, 324, 333, 347, 351, 353, 385,
389, 396, 411, 436n47, 460nl7, 466n54
Crash (2004), 29, 195, 201, 205, 206, 213, 221,243,
245
Crash Test Dummies (2005), 191, 213
Crawford, Joan, 141
Crews, Frederick, 2
crosscutting, 65, 97, 99, 152, 206-208, 211, 225,
229, 233, 236, 339, 357, 467n54
Crossfire (1947), 215
Cross-Harbour Tunnel (Guo hai sui dao, 1999), 211
Crossroads (2006), 243
Crow of the Moonlight Night (Tsukiyo karasu,
1939), 345, 360, 361
Cubism, 65
Cuckoo (Hototogisu, 1922), 382
cues, 137, 138,140,144,147, 149,150, 168, 229, 278,
328, 335, 342, 347-349, 380, 409, 410
Cukor, George, 38, 262, 283, 292, 300, 301, 304,
308, 309, 321
cultural studies, 4,11, 20, 23, 30, 262, 433, 450
culture, 3,16-18, 21, 23, 29-31, 44, 45, 47, 49, 51,
52, 54, 57, 61-69, 73-82, 85-87, 94, 109,
116, 117, 138, 142,159, 191,197,200,
244, 262, 334, 335, 338, 351-356, 372,
373, 388, 390, 401, 443n33, 446n33,
447n47
popular culture, 185, 187, 195, 198, 255, 262,
338, 352, 354, 355, 372, 390, 410
curiosity, 100, 101,125,139,140, 150, 206-208, 210
Curtis, Richard, 219
D
Dance of the Drunken Mantis (Nan bei zui quan,
1979), 404
Dangan Runner (Dangan ranna, 1996), 392
Dangerous Corner, 185
Daniel, Frank, 105
485
Dardenne, Jean-Pierre and Luc, 162
Darwinism, 6
Daves, Delmer, 283, 290, 292, 302, 305
Davies, Terence, 164
Davis, Bette, 139, 141
Dawkins, Richard, 5
Day and Night (Dag och natt, 2004), 166
Day for Night (La Nuit americaine, 1973), 153
Days of Being Wild (A Fei zheng chuan, 1991), 161,
220,458n44
D-Day the Sixth of June (1956), 302
De Sica, Vittorio, 152
deadlines, 25, 26, 28,178,179, 199, 219, 225
Death by Hanging (Koshikei, 1968), 158
Decalogue, The (Dekalog, 1989), 203, 457
decorative technique, 311, 347, 356, 359, 362, 368,
372, 375, 377-385, 387-392, 476n4
deep focus, 27, 154, 263, 274, 298, 300, 301, 305,
312, 392,466n54
Demetrius and the Gladiators (1954), 301, 313, 316
DeMille, Cecil B„ 195, 281
Denis, Claire, 162
denotation, 347, 377-380, 388, 392
denouement, 37, 102, 104, 105, 343
depth in image, 14, 32, 263-266, 269, 271, 273,
276-280,282, 283, 295-299, 305,
308, 310, 312, 320,382, 392, 414, 430,
466n54
depth of field, 14, 273, 91, 296, 299, 303, 305, 306,
312-314,321, 465n38
Deserto rosso, II (1964), 154,156,158
Desperation, 185
Detective (1985), 22, 218
Detective Story (1951), 42
Deutsche Bioscop, 266
deviation from norms, 27, 28,155-157, 341, 342
Dial Mfor Murder (1954), 42, 438
Dickens, Charles, 194, 213, 244
Die Hard (1988), 48, 90, 118, 399
Die Hard 2 (1990), 399
diegesis, 125, 138, 139,143, 226, 230, 345-347
dissolve, 25, 65, 79, 149,157, 380, 382
dissolve-in-place, 380
Distant Voices, Still Lives (1988), 164
Dmytryk, Edward, 305
Do the Right Thing (1989), 197
documentary, 110, 121, 154,162,166, 223, 224,
226,234
Dog Days (Hundstage, 2001), 220
Dogme, 95, 162, 163, 453n21
Dolce vita, La (1960), 153
Doll's House, A, 119
Dolls (2002), 161, 307
Donen, Stanley, 314
Donner, Richard, 397-399, 401, 407
Donnie Darko (2001), 187
Doody, Margaret Anne, 195
Dower, John, 371, 373, 374
486 Index Index 487

Downfall ofOsen (Orizuru Osen/Osen of the Paper
Cranes, 1935), 366
Dr. Kinuyo (Joi Kinuyo sensei, 1937), 381
Dragnet (1954), 469
Dragon Gate Inn (Long men ke zhen/Dragon Inn,
1966), 415, 416, 419-421, 424, 426-428
Dreyer, Carl Theodor, 158, 254, 260
Drive-In (1976), 228
Drury, Allen, 185
Dulac, Germaine, 70
Dunbar, Robin, 196
Dunne, John Gregory, 262
Dym, Jeffrey A., 373, 374
E expressionism, 268, 353
Early Summer (Bakushu, 1951), 117, 120
East of Eden (1955), 305
Eastmancolor, 313
Eastwood, Clint, 262
Eclisse, L' (1962), 153
editing 3, 15, 19, 25, 26, 28, 29, 37-39, 41, 48, 58,
67, 69, 129,136, 169, 223, 257, 259, 283,
292, 293, 295, 304, 307, 329, 339-341,
364, 370, 374, 375, 377, 382, 397, 400,
408,410, 418,421, 423-427, 429,462n4,
466n54, 479nll. See also 180-degree
rule; analytical editing; average
shot length; constructive editing;
continuity editing; crosscutting; jump
cuts; Kuleshov effect; master-shot
editing; matching shots in editing;
montage; overlapping cuts; piecemeal
decoupage; point-of-view editing;
reverse shot; shot/reverse shot editing
Egoyan Atom, 197
Egri, Lajos, 118,119
Eikhenbaum, Boris, 19, 22, 23
Eisenstein, Sergei Mikhailovich, 14, 18, 20, 28,
104,158, 256, 337, 396, 403, 406,407,
410,411,425, 427, 478nll
Eitzen, Dirk, 81
Ekman, Paul, 327
Electra (Szerelmem, Elektra, 1974), 35
Eliot, George, 195, 244
Elvira Madigan (1967), 153
e-makimono picture scrolls, 351, 355
emotions, 19, 30, 45, 47, 51-53, 67, 78, 79, 88, 93,
94,100,101, 112-114,116,118, 119, 123,
124, 126, 150, 184, 237, 242, 259, 276,
294, 308, 330, 331, 333-335, 338, 347,
354, 359, 364, 393, 399, 400, 406,407,
409-411,447n47,470n8
Empire (1964), 42
empiricism, 3, 15
Empson, William, 197
Enemy Below, The (1957), 311, 321
Enfants terribles, Les (1950), 42
Eno, Brian, 174, 175
Enoken (Enomoto Ken'ichi), 338
Erdos, Paul, 198
ero-guru-nansensu, 353
Escort Over Tiger Hills (Hu shan lang, 1969), 430
Eternal Heart (Fue no shiratama, 1929), 359
ethology, 63
Eureka (Yureka, 2000), 161
Europa '51 (1952), 154
evolutionary theory, 4, 43,45-47, 51, 52, 62, 63,
75, 76, 78, 79, 81,115, 196, 328, 334,
378, 432n6, 442nll, 443n33, 447n47
Exotica (1994), 197, 208
exploitation, 401
exposition, 99,102,117,120,138,199-201, 208-210,
218,231,235,237,238
Expressionism, German, 14,151, 268
expressive amplification, 347, 406-408
eye behavior in films, 24, 36, 81, 115, 239, 241, 327,
329-333, 335
Eyes of Youth (1919), 186
Eyes of Youth (1920), 455nl7
F 381
fabula and syuzhet, 98-103,110,120,165,169, 206,
207, 388, 389
facial expressions, 18,19, 25, 63, 73, 73, 76, 81, 115,
116,147, 230, 276, 295, 328, 333, 335,
447n47, 471n35
Fahrenheit 9/11 (2004), 19
Fairbanks, Douglas, 410
fairy tales, 87, 243
Faithful Servant Naosuke (Chuboku Naosuke, 1939),
383
Fallen Angel (1945), 38, 98,138, 257, 304
Fallen Angels (Duo luo tian shi, 1995), 200
Fallen Blossoms (Hana chirinu, 1938), 347, 348,
382,383
Falls, The (1980), 87
Family Man, The (2000), 185
Fan, The (1949), 304
Fanny and Alexander (Fanny och Alexander,
1982), 254
fantasy, 112, 163, 186, 419
Fassbinder, Rainer Werner, 159
fast-motion, 376, 377, 404, 409
Fate of Lee Khan, The (Ying chun ge zhi Fengbo,
1973), 413-415, 420-422, 424, 425, 427,
428,430
Favoris de la lune, Les (1984), 221, 233-235, 237
Feher, Gyorgy, 162
Fellini Roma (Roma, 1972), 153
Fellini, Federico, 152,154-156
Femme est unefemme, Une (1961), 217
Femme mariee, Une (1964), 154, 158
Festival (2005), 191, 243
festivals, 155, 160-1, 165, 191, 219, 243-5, 266, 395, Formalism, Russian, 12, 19, 54
453nl6 Forrest Gump (1994), 30, 436n51
Feuillade, Louis, 75, 221, 265, 266, 278, 462 Forster, E.M., 103
fiction, 18, 65, 87, 99, 102-4,112, 117, 122, 128, Fort, Charles, 228, 231, 232, 321
140,141,171, 194, 195, 220, 254, 327, Forty Guns (1957), 316, 321, 322
333, 352, 353, 355, 370, 437 Four Moods (Hsi nou ai lueh, 1970), 424
detective, 13-15, 96, 99, 118, 155 Four Shades of Brown (Fyra nyanser av brunt, 2003),
forking paths in, 172, 174, 183, 455 243
mystery, 89, 101, 140, 144, 146 Fox Grandeur, 467
Field, Syd, 105, 110 Frank, Ann, 133
Fifth Generation in mainland Chinese film, 161 Free Radicals (Bose Zellen, 2003), 198,203, 217, 220
film criticism and analysis, 4, 11,12, 13,16, 17, 18, freeze frame, 155-157, 176
20, 23, 24, 40, 41, 44, 49-51, 54, 57, 66, French Connection, The (1971), 396
92, 129, 132, 139,140,149,150, 154, French impressionism, 151, 357
160, 182,193, 221, 253-257, 259-261, French poetic realism, 337
273, 282, 292, 322, 337, 377, 379, 389-9, Freudianism, 136
413, 466n54 Freytag, Gustav, 103,104, 112
film noir, 15, 31, 99, 163, 215, 217, 228, 305 From Here to Eternity (1953), 289
film schools, 159,160, 396 Frozen Land (Paha maa, 2005), 191, 210, 213
film studies, 3-6,11, 12,16, 51, 54, 60, 79-81, 105, Fugitive, The (1993), 399
136, 263, 396, 449n2 Full Contact (Xia dao Gao Fei, 1993), 359
Fireworks, Should We See It from the Side or the Fuller, Samuel, 262, 283, 316, 321, 322
Bottom? (Uchiage hanabi, shita kara
muruka?yoko kara miruka?, 1993), 185
First Steps Ashore (Joriku dai-ippo, 1932), 349, 351,
G
First Sword (Diyijian, 1967), 425 Gad, Urban, 103, 264
Fist of Fury (Jing wu men, 1972), 395 Gaghan, Steven, 219
Five Dedicated to Ozu (2003), 162 Garcia, Rodrigo, 191, 200
Five Easy Pieces (1970), 157, 262 "Garden of Forking Paths," 171
Five Fingers of Death (Tian xia diyi quan/King Garden State (2004), 219
Boxer, 1972), 395 Garland, Judy, 49, 331, 335
Five Scouts (Gonin no sekkohei, 1938), 370, 475n50 Garon, Sheldon, 373
Five Senses, The (1999), 243 Garrel, Philippe, 162
Flaming Star (1960), 307 Gaudreault, Andre, 122
flashback, 98-100, 102,103,110, 125-7, 142, 143, Gay, Cesc, 185, 202
145,147, 154, 158, 166, 167, 191, 199, gendai-geki (contemporary-life drama), 339, 340
206, 207, 209, 210, 215, 219, 223, 238, Genette, Gerard, 13, 100
239, 269, 448n65, 451nl8, 457n31 genre, 43, 77,136, 142, 152, 154, 157,159, 161,185,
flashforward, 99,155 202, 215, 216, 219, 224, 227, 233, 321,
Flaubert, Gustave, 90 337, 338, 355, 363, 370, 377, 379, 380,
Flavors (2003), 210, 211 388, 395, 396, 408, 434nl5, 460nl7,
Fleischer, Richard, 283, 315, 321 478n6. See also comedy; fantasy;
Flourishes, stylistic, 375, 385, 387-389 horror; martial arts; musical; policier;
Flowers of Shanghai (Hai shanghua, 1998), 161 science fiction; swordplay films;
Flunky, Work Hard!. See Little Man, Do Your Best thriller; war films; Westerns
Fly, The (1958), 468 Genroku Chushingura (1941), 362, 363
focus 29, 274, 275, 278, 298, 347, 348, 377, 380, Gentlemen Marry Brunettes (1955), 292
382, 384, 387, 388, 399, 404, 466n4 Gertrud (1964), 158
folk psychology, 88,112,113,115,126,173-175, 335 Giant (1956), 31, 289
FongSai Yuk (1993), 407 Gibson, James J., 78, 81, 82
For a Few Dollars More (1965), 323 Gigerenzer, Gerd, 81
For Ever Mozart (1996), 217 Gigi (1958), 289, 465
Ford, John, 157,257-259,262, 281,299, 342,460n24 Girl Can't Help It, The (1956), 292
Foreign Correspondent (1940), 37, 437n61 Giulietta degli spiriti (1965), 156
foreshadowing, 172, 225, 232, 241 Glass Key, The, novel, 96
Forest, The (Rengeteg, 2003), 203 Glory Guys, The (1965), 254
forking paths, 3, 173-178, 180-187 Gluck, Carol, 373
formalism, 1, 19, 23 Go (1999), 215
488 Index Index 489

Godard, Jean-Luc, 22, 27, 28,158, 163, 217, 218, Hart, William S., 410 Hu, King (Hu Jinquan), 43, 233, 413-417, 419-430, Iwai Shunji, 185
254, 262, 281, 283, 311, 312, 319 Hartley, Hal, 163, 279, 453n22 479n9, 479nll Iwami Jutaro (1937), 380, 384
Godfather, The (1972), 91, 93, 109, 157,445n5 Hathaway, Henry, 304, 321 Huckleberry Finn, 121, 132
Godfather: Part II, The (1974), 157, 285 Hawks, Howard, 155, 258, 262, 281, 292, 302, 427, Hui, Ann, 161, 479
Goldfinger (1964), 396 452n6, 460n4 Huillet, Daniele, 59, 116, 158, 163, 164, 279
Goldfish Memory (2003), 195, 202 Hearty Response (Yigaiyun tian, 1986), 399-401, humanities, 1, 3, 11, 12, 21, 22, 57, 73, 79, 80, 261,
J
Goldsman, Akiva, 109 403, 407 432, 440 Jackie Brown (1996), 215
Gombrich, E.H., 19,48, 63, 64, 66, 69, 342, 378, 385 Heat (1995), 192, 399 Hung, Sammo, 404, 407, 430 Jackson, Peter, 262
Gonzalez Inarittu, Alejandro, 191,206,207,216,243 Heat of the Earth (Ginetsu, 1938), 349, 351 Hunters, The (1957), 283 Jackson, Samuel L., 331, 334
Good Men, Good Women (Haonan Haonu, 1995), Heaven (2002), 150, 159 Huston, John, 323, 469n90 Jacob, Francis, 21
161 Heaven's Doors (2006), 243 Hyman, John, 442n22 Jade Bow, The (Yun hai yu gong yuan, 1966), 419,
Gosford Park (2001), 203 Hegelianism, 3, 14 Hyman, Stanley Edgar, 11 425
Gosho, Heinosuke, 343, 359, 363, 374, 382, 384 Heights (2005), 191, 201, 209 Jailhouse Rock (1957), 306, 468n76
Gossette (1923), 70, 72 Heiress, The (1949), 42 Jakobson, Roman, 19
Gould, Stephen Jay, 21 Heiroku's Dream Story (Heiroku yume-monogatari, I James, Henry, 91, 101, 103, 151
Grand Canyon (1991), 213 1943), 338 Jancso, Miklos, 35, 163
Grand Hotel (1932), 194, 195, 197, 203, 219 Helen of Troy (1956), 303, 304 I Live in Fear (Ikimono no kiroku, 1955), 154 Japaneseness, 352, 353, 355, 356, 369, 370, 372
Great Locomotive Chase, The (1956), 316 Hell and High Water (1954), 321 Ichikawa Kon, 337 Jaws (1975), 290, 396
Greenaway, Peter, 87 Help! (1965), 254 iconography, 17, 65, 268 Je vous salue, Marie (Hail Mary, 1985), 218
Greimas, Algirdas Julien, 141 Henry V (1944), 41 ideology, 16, 18, 23, 32, 44, 54, 135, 158, 172, 354, Jeanne Dielman, 23 Quai du Commerce,
Grice, Paul, 447, 448 371 1080-Bruxelles (1976), 26, 27,116
Henry, O., 22, 23, 172,181
Griffith, D.W., 14, 25, 122, 256, 260, 262, 263 Heroine White Rose (Nu xia bai meigui, 192,9), 416 Iemitsu and Hikozaemon {Iemitsu to Hikozaemon, Jerry Maguire (1996), 99, 120, 125
Grodal, Torben, 51, 81, 335 1941), 356, 362 Jewison, Norman, 253
Hidden Fortress, The (Kakushi-toride no
Groom Talks in His Sleep, The (Hanamuko no I f . . . (1968), 160 Jezebel (1938), 119, 120, 296
san-akunin, 1958), 392
negoto, 1935), 343, 345 Ikeda Yoshinobu, 382 JFK (1991), 262
High and Low (Tengoku tojigoku, 1963), 153
Groundhog Day (1993), 187,455 Ikeda, T„ 474n38 Jia Jiang-ke, 161
High and Mighty, The (1954), 302
Group, The (1966), 253 Ikiru (1952), 154 jidai-geki (historical film), 355, 357, 359, 362, 364,
High, Peter B., 374
Gu Changwei, 165 IIy a des jours ... et des lunes (1990), 205, 206 370, 414, 473nl8
Hill Street Blues, 195
Guare, John, 198 Imax, 285 Johnson, Samuel, 13
Hino, Ashihei, 370
Guerre estfinie, La (1966), 165 In Cold Blood (1967), 106 jokes, 51, 85-90, 110, 112,113
Hirayama, D.M., 474
Guiguet, Jean-Claude, 221, 238-241 In the Bedroom (2001), 219 Joki (2001), 203
Hiroshige (Utagawa Hiroshige), 356, 390
Gun Crazy (Deadly Is the Female, 1949), 298 In the City (2003), 202, 203, 208 Jones, Chuck, 332
Hiroshima mon amour {1959), 154
Gunfighter, The (1950), 122 In This Our Life (1942), 150 Jost, Francois, 122
His Girl Friday (1940), 150, 192, 456
Gunning, Tom, 122 Inagaki Hiroshi, 346 Jubal (1956), 305
Hitchcock, Alfred, 32-35, 37-42,104, 157, 158,
Guns ofNavarone, The (1961), 333 Indiana Jones and the Temple of Doom (1984), 406 Jules et Jim (1962), 121, 151, 153
258, 262, 262, 281, 294, 300,424,
Guys and Dolls (1956), 31, 307 Indiana Jones film series, 399 July, Miranda, 208, 216, 217
437n57,438n73 Ingeborg Holm (1913), 26, 32 jump cut, 154,157, 364, 371, 377, 428
Hofer, Franz, 265, 271,278 Inn of the Sixth Happiness, The (1958), 321
Hogan, Patrick, 50, 52 inpei (veiling), 355
II Home from the Hill (1960), 229 Intelligent Design, 432
Homunculus (1916), 266 Interlude (1957), 305
K
Hagen, Margaret, 65
Haggis, Paul, 195, 201 Hong Sang-soo, 161 Internet, 29, 198, 244, 454n7 Kabuki theatre, 338, 346, 352, 354, 355,406,474n33
haiku, 371, 390 Honky Tonk Freeway (1981), 228 interpretation, 6, 12,14, 16-18, 20, 24, 49, 50, 76-8, Kael, Pauline, 253, 254, 460
"Haircut," 131,132 Hope, Bob, 338 80, 81, 94,150, 156, 255, 377,433,475nl Kaeriyama Norimasa, 352
Hamilton, Patrick, 39,437n58 Hopkins, Antony, 331, 334 intertitles, 94, 95, 121, 127, 176, 197, 206, 208-211, Kaigari Ippei, Part One (1930), 340
Han, Yingjie, 419 Horak, Jan-Christopher, 266, 273, 462n8 217, 221, 229, 236, 240, 269, 346, 347, Karmitz, Marin, 159
Hand That Rocks the Cradle, The (1992), 116 Horror 51,137, 325 359, 370, 380, 384 Kato Tai, 392, 416
Handle with Care (1977). See Citizens Band Horse (Uma, 1941), 370 dialogue, 339, 347, 359, 377, 380, 384 Kaufman, Boris, 300, 325
Haneke, Michael, 159,169,197, 213, 221 Horton, Robin, 61, 62, 64, 111, 116 expository, 94, 103, 339, 380 Kawate Jiro, 350
Hannah and Her Sisters (1986), 201, 209 Hotel Room (1998), 185,203 Intolerance (1916), 180 Kazan, Elia, 283, 305, 311
Hansen, Miriam Bratu, 75-8 Hou Hsiao-hsien, 25,161, 164-166, 201, 453 Iosseliani, Otar, 191, 197, 221, 234-237 Keating, Patrick, 439n97
Happenstance (Le Battement d'ailes dupapillon, House of Bamboo (1955), 321 Island in the Sun (1957), 314 Keaton, Buster, 260-2
2000), 198, 204, 205, 207, 210, 213 How Green Was My Valley (1941), 257 Isle, The (Seom, 2000), 161 Kerrigan, Lodge, 117
Happiness (1998), 200, 201, 228 How the West Was Won (1962), 284 isocephaly, horizon-line, 271 Key, The (Kelid, 1987), 162
Happy Endings (2005), 29,191, 206, 209, 210, 219, How to Marry a Millionaire (1953), 212, 281, 287, Istanbul Tales (2005), 191, 243 Kiarostami, Abbas, 162, 165
221,243 290,301,304, 321 It Happened One Night (1934), 15, 192 Kido Shiro, 359
Hard Luck Hero (2005), 206, 207, 209 Howe, James Wong, 311 It's a Wonderful Life (1946), 186 Kieslowski, Krzystof, 162, 172, 174, 197, 203
Hard-Boiled (Lashou shentan/Lat sau san taam, Howitt, Peter, 172 Ito Daisuke, 357, 359 Killer, The (Da sha shou/Lightning Swords of
1992), 409 Hrdy, Sarah Blaffy, 196 Ivory, James, 90 Death, 1972), 402
490
Killer, The (Diexue Shuangxiong, 1989), 71, 404,
405, 407
Killers, The (1946), 298
Kim Ki-duk, 161
King and I, The (1956), 31, 321, 325
King Boxer. See Five Fingers of Death
King Kong (1933), 338
King of the Children (Haizi Wang, 1987), 161
King of the Khyber Rifles (1953), 287, 304
King Richard and the Crusaders (1954), 317, 318
King, Stephen, 185
Kinoshita Keisuke, 391
Kinugasa Teinosuke, 384, 385
Kiss Them for Me (1957), 314, 315
Kitano (Beat) Takeshi, 161, 163, 411
Klahr, Lewis, 48, 49
Klapisch, Cedric, 192
Kleptomaniac, The (1905), 212
Klercker, Georg af, 265, 271
Klute (1971), 157
Km. 0 (2001), 194
Knife in the Water (Noz w wodzie, 1962), 151, 156
Knight, Arthur, 153, 452
Knights of the Round Table (1953), 287, 304
Kochiyama Soshun (1936), 345
kokusaku ("national policy"), 370
Komatsu, Hiroshi, 372, 373, 474
Komatsu Riyuzo, Part II (1932), 386, 387
Konishi Jin'ichi, 389
Kore-eda Hirokazu, 161, 164
Koster, Henry, 291, 292, 302, 307, 313
Kowa Company, The, 476,477
kowairo (vocal imitation), 346, 347, 373
Kozintsev, Grigori, 70, 71
Kozloff, Sarah, 130
Kramer, Stanley, 262
Krauss, Werner, 267
Kubrick, Stanley, 259, 262
Kuhn, Thomas, 432
Kuleshov effect, 401
kung-fu films, 43, 396, 401, 407,409, 414,429
Kurosawa Akira, 152, 159, 262, 337, 363-365, 372,
390,392,416
Kushner, Barak, 374
L
L.A. Confidential (1997), 329, 330, 332-334
Lacan, Jacques, 6
Lacanianism, 11, 136, 432
Lady and the Tramp (1955), 281, 304
Lady in the Lake, The (1947), 38, 294
Lady Vanishes, The (1938), 39
Laffay, Albert, 122
Lam, Ringo, 359, 406
Lancelot du Lac (1974), 157
Land of Cherry Blossoms (Sakura no kuni, 1941),
359, 360
Index Index 491
Land of the Pharaohs (1955), 281, 311 Locket, The (1946), 103 Marrying Kind, The (1952), 300
Land, The (al-Ard, 1969), 69 Logan, Joshua, 283, 301, 311 martial arts, 104, 359, 364,385,401,407-409,478n6
Lang, Fritz, 58, 98,103,138, 157, 262, 268, 281, Loggerheads (2005), 191 Marxism, 11, 12
283,465n45 Lone Wolf and Cub (Kozure Okami, 1972), 411, 416 master-shot editing, 48, 265
langue, 60 Long Gray Line, The (1955), 281 matching shots in editing, 340-342, 357, 389
lantern slides, 58 Long Hot Summer, The (1958), 229 eyeline matching, 435, 443n26
Lardner, Ring, 131, 132 long take, 14, 25, 33-42, 154, 163-165, 234, 245, graphic matching, 28, 70, 342, 349, 366
LaShelle, Joseph, 38 257, 294, 295, 304, 308, 315, 321, 322, matching action, 425, 427
Last Action Flero (1993), 290 339, 360, 363, 368, 371, 373, 397, 424, matching direction, 428
Last Movie, The (1971), 153 430, 437n63, 438n74, 466n54, 474n42 Mauvais Sang (1986), 164
Late George Apley, The (1947), 38, 467n61 Look Both Ways (2005), 191, 216, 217 McCarey, Leo, 256, 262
Late Spring (Banshun, 1949), 94 Lord of the Rings, The (2001-3), 50 Me and You and Everyone We Know (2005), 191,
Lau Kar-leung, 404, 430 Love Actually (2003), 202, 203, 219, 245, 458n41 216, 217, 219
Laura (1944), 17,18, 38, 257, 304 Love and Sacrifice (Ninin Shizuka/The Two Me Myself I (1999), 185
Laurents, Arthur, 39 Shizukas, 1922), 340 meaning, 12, 14, 16, 22, 47,48, 63, 77, 80, 124, 131,
Law Man (Norman Law), 399-401 Love for a Mother (Mabuta no haha, 1962), 416 133, 136,149, 158, 260, 261, 328, 333,
Le Fanu, Mark, 374 Love Is a Many-Splendored Thing (1955), 301,308,309 377-380, 388, 406
Lean, David, 262 Love Massacre (Aisha, 1981), 161 Meiji period, 352-355, 389, 390
Leave Her to Heaven (1945), 125, 126, 299 Love Me Tender (1956), 288, 306 Melies, Georges, 14
Lee Chang-dong, 161 Love ofSunya, The (1927), 455nl7 Melinda and Melinda (2004), 186
Lee, Ang, 50 Low, Bobbi, 115 melodrama genre, 44, 52, 150,159, 184, 192, 202,
Lee, Conan, 408 Lubitsch, Ernst, 195, 257, 262, 296, 322, 359, 390 216, 219, 228, 229, 232, 241, 243, 244,
Lee, Danny, 404 Lucas, George, 290, 324, 332 266, 268
Lee, Spike, 71, 228 Lumiere, Auguste and Louis, 14 Melville, Jean-Pierre, 42, 262
Legend of the Mountain (Shan-chung ch'uan-ch'i, Lust for Life (1956), 304 Memento (2000), 110,163, 215, 219, 244, 446n30
1979), 413 Mepris, Le (Contempt, 1963), 153, 155, 281
Legendary Weapons of China (Shi ba ban wuyi, Merchant, Ismail, 90
1982), 404, 408 M Messaris, Paul, 47, 443n26
Lelouche, Claude, 191, 203, 205, 206,457 Metro-Goldwyn Mayer Pictures, 31, 194, 286-289,
Leninism, 18 M. Hulot's Holiday (Les Vacances de M. Hulot, 464n21, 465n35,467n72
lenses, 26, 29, 58,152, 163, 164, 263, 269, 274-278, 1953), 203, 233 Metropolis (1927), 58, 268
285, 287-293, 296, 298, 305, 306, M/Other (1999), 161 Metsys, Quentin, 65
308, 313, 324, 380, 392, 397, 400, 401, Maborosi (Maboroshi no hikari, 1995), 161, 164 metteur en scene, 254
468n86, 476n21 Macbeth (1948), 38, 437 Metz, Christian, 135, 136
wide-angle, 278, 285, 305, 306, 380, 392, 399 Macdonald, Dwight, 322, 469 Meyer, Leonard, 19
Leone, Sergio, 322, 406 Mackendrick, Alexander, 192 Middlemarch, novel by George Eliot, 194
Lerski, Helmar, 273-275 Mad Mad Ghost (Huang jin dao shi, 1992), 395 Mildred Pierce (1945), 135,138,139,143-50, 229,
Lethal Weapon (1987), 397-401, 403, 406, 407 Madam and Wife {Madamu to nyobo, 1931), 353 451nl8
Letter, The (1940), 139, 140,143 Madden, John, 29 Milestone, Lewis, 260
Leung Chiu-wai, Tony, 409 Made in USA (1966), 217, 311, 312 Milgram, Stanley, 198
Li, Jet, 404, 405,407,411 Magnificent Ambersons, The (1942), 38, 294 Millar, Gavin, 59
Hang hsiang ("bright appearance"), 406 Magnificent Obsession (1954), 120 Miller, Claude, 215
Lieutenant-Colonel Tachibana (Gunshin Magnolia (1999), 191,197,219,221,227-234,241,244 Mills, C. Wright, 21
Tachibana chusa, 1926), 357, 371 Makhmalbaf, Mohsen, 162, 185 mimesis, 446
Lifeboat (1944), 37, 39, 300,437 Makino Masahiro, 357, 359, 362 Mind the Gap (2004), 207, 209
lighting, 24, 26, 47, 64, 123,152, 155, 260, 335, Makino Shozo, 339, 472 Minghella, Anthony, 262
349, 368, 382, 400, 438 Malcolm in the Middle, 185 Minnelli, Vincente, 38, 155, 262, 283, 292, 303,
Lily of the Valley (Tsuruganeso, 1935), 350, 384 Maltese Falcon, The (1941), 138, 139, 299 304,311,321
linguistics, 5, 60, 62, 93, 135 Man in the Gray Flannel Suit, The (1956), 309 Miramax, 159
Linklater, Richard, 455 Man Push Cart (2005), 163 Mirror, The (Ayneh, 1997), 162
Lippert, Robert, 289 Man Who Never Was, The (1956), 291, 309, 310 Mirror, The (Zerkalo, 1975), 162
Little City (1997), 195 Mandler, Jean, 141 Mischief Night (2006), 242
Little Foxes, The (1941), 14, 297, 324, 350, 467n61 Mankiewicz, Joseph, 38, 304 mise en cadre, 410
Little Man, Do Your Best (Koshigen gambare/Flunky, Mann, Aimee, 230, 232, 233, 241 mise en scene, 15, 129, 152,232, 255, 256, 278, 281,
Work Hard, 1931), 381, 385 Mann, Anthony, 342 312, 325, 400, 401, 410, 469n94, 478nll
Little Women (1933), 296 Manual of Love, The (Manuale d'amore, 2005), 191 Mission: Impossible III (2006), 67
Littlewood, Ian, 373 Manuli, Guido, 186 Mister Roberts (1955), 281, 288
Livingston, Paisley, 449 Mark ofZorro (1920), 295, 296 Misume Kenji, 416
Lloyd, Harold, 262, 322, 338, 353 Marriage Circle, The (1924), 195, 257, 359 Mitchell, David, 194
492
Mito Komon Part II (Ishoku mito komon oshuhen,
1932), 384-386
Mizoguchi Kenji, 25, 32,110, 254, 262, 337, 359,
362, 363, 366, 368, 369, 372, 374, 391
mobo ("modern boys"), 353
Moby Dick (1956), 31
modernism, 65, 72, 75, 76, 151,157, 158, 161, 220,
353
modernity, 3, 76, 81, 205, 353, 373, 436
moga ("modern girls"), 353
Moment of Innocence (Nun va Goldoon, 1996), 162
montage, 14, 158, 163, 176, 211, 242, 256, 263, 379,
380,385,410, 466n54
Monte Carlo Story, The (Monte Carlo, 1957), 286
Moonfleet (1955), 281
Morning for the Osone Family (Osone-ke no ashita,
1946), 391
Morning Mist in Edo (Edo no asagiri, 1935), 360, 362
Morson, Gary Saul, 171-174, 183
Most Beautiful, The (Ichiban utsukushiku, 1944),
364,390, 391
Motel Cactus (Motel Seoninjang, 1997), 203
Mother (Mat, 1926), 357
Mother India (1957), 52
Mother Is Strong, A (Haha wa tsuyoi, 1939), 347, 348
Mothlight (1963), 151
Movie (magazine), 452, 466n54
Mr. Destiny (1990), 186
Mr. Smith Goes to Washington (1939), 257
Mr. Thank You (Arigato-san, 1936), 366, 368
Mud and Soldiers (Tsuchi to heitai, 1939), 370
multiple-camera shooting, 283, 284
Mulvey, Laura, 81
Murata Minoru, 339
Murder, Czech Style (Vrazdapo cesku, 1967), 186
Muriel ou le temps d'un retour (1963), 72, 151
Murnau, F.W., 72, 254, 256, 257, 262, 283
Musashi Miyamoto IV (Miyamoto Musashi:
Ichijoji no ketto, 1964), 414
musical genre, 15, 31, 217, 224-226, 241, 379
Mutiny on the Bounty (1962), 286
My Darling Clementine (1946), 299
Mysterious Edogawa Ranzan, The (Kaiki Edogawa
Ranzen, 1935), 342, 343
N On the Waterfront (1954), 287
Nagata Masaichi, 352
Naniwa Elegy (Naniwa ereji, 1936), 368
Napoleon Dynamite (2004), 163
narration, 13, 27, 90-102,107,108, 110-112, 116,
117, 120,122, 123, 125-33, 138,140,
143, 144,147, 149, 150, 152,155-158,
165-69, 172, 176, 178, 181,190-2, 196,
199-201, 206-11, 215, 216, 218-21, 225,
226, 228, 231-3, 236, 238, 239, 241, 242,
244, 333, 347, 372, 380, 387-9, 449n77.
See also voiceover
Index
narrative, 3, 4, 15,18-20, 22, 32, 47-50, 52, 62,
69, 85-96, 98, 100-103, 110, 112-118,
120-125,127-133, 138-142,149, 150,
158, 172, 173, 180,181, 187, 192, 195,
196, 199, 203, 207, 233, 295, 345-7, 349,
357, 364, 368, 370, 371, 375, 377-80,
388-90, 448n65, 451n22. See also
act structure; backstory; cause-
effect structure; climax; deadlines;
denouement; exposition; flashback;
flashforward; foreshadowing; forking
paths; narration; narratology;
narrator; network narrative;
order of story events; parallel worlds;
parallels, narrative; parametric
narration; plot; protagonists; reels as
structural divisions
narratology, 6, 85-88, 90, 98, 100,110,127, 141
action-centered theories, 88, 89
agent-centered, theories, 88
narrator, 87, 88, 99,121-33,155,156,158,169,232,
238
Naruse Mikio, 262, 359, 363, 380-382, 384, 387
Nashville (1975), 191, 219, 221-228, 243, 245
Naturama, 285
Neame, Ronald, 309, 310
Negulesco, Jean, 292, 307
Nelson, Joyce, 140, 149
neoclassicism, 13, 89
neoformalism, 23
Neorealism, 14, 42,152, 153, 160, 161, 338
Nerven (Nerves, 1919), 266-269, 272, 275-280
Nettle, Daniel, 470
Network (1976), 228
network narrative, 3, 29, 103,189-250
art cinema in, 219-21, 235-7, 238-42
causality in, 190,193, 230
circulating objects in, 202-3, 235, 237
classical cinema in, 218-219, 221, 230
characterization in, 197, 200-2
and genre, 201-2, 221, 224-7, 228-9
historical sources of, 194-8
mixed cases of, 214-8
narration in, 199-200, 206, 207-11, 225-226,
231-3, 233-5, 237-42
parallelism in, 212
plot structure in, 199, 202-7, 212, 229-30, 243
protagonists in, 192-3, 198-9, 200-4
role of chance in, 204-7, 230-3
story world of, 200-4, 222-4, 228-9, 235-6,
238-40
themes in, 211-14, 231-2, 241
traffic accidents in, 204-7
New Age of Living Together, The (San tunggui shut
doi, 1994), 215
New Family, A (Atarashiki kazoku, 1939), 341, 359,
472
New One-Armed Swordsman, The (Shin du bei dao,
1971), 419
Index 493
New Taiwanese Cinema, 161 orthochromatic film stock, 284
NewWave, French, 152-5, 157, 158 Osen of the Paper Cranes. See Downfall ofOsen, The
New Wave, Japanese, 372 Oshima Nagisa, 158
Newman, Michael, 115, 471n35 Ouedraogo, Idrissa, 67
Ng See-yuen, 414 Our Mutual Friend, 194
Nicholson, Jack, 333, 334 Our Town (1940), 238
Nicht versdhnt (1965), 158 Out 1: Noli Me Tangere (1971), 220, 221
Nick of Time (1995), 24 overlapping editing, 377, 408, 427
Nielsen, Asta, 264 over-the-shoulder shot, 39, 48, 58, 59, 68, 293, 295,
Night Shapes (Nachtgestalten, 1999), 212 297, 307, 321, 323, 340
Nightly Dreams (Yogoto noyume, 1933), 380, 384 oyama (female impersonator), 340
Nights ofCabiria (LeNotte di Cabiria, 1957), 153 Oyuki the Virgin (Maria no Oyuki, 1935), 366
nihon kaiki ("return to Japan"), 353 Ozu Yasujiro, 17, 25, 28, 32, 53, 54, 117, 158, 161,
Nikkatsu, 339, 361, 366, 474n38 260, 262, 334, 337, 342, 343, 356, 359,
Nilsson, Leopoldo Torre, 160 365,371,372, 389, 391,472nll
Nine Lives (2005), 189, 191, 200, 245
Niven, David, 333
No Sweat (2006), 243
P
noh theatre, 339, 352, 389
Nolletti, Jr., Arthur, 374 Painted Skin (Hua pi zhi: Yinyangfa wang, 1993),
Nomad (Lie huo qingchun, 1982), 161 424
Nomura Hiromasa, 359 painting, 42, 58, 65, 86,112, 268, 312, 351, 352,355,
norms, filmic, 1,3, 15, 16, 20, 22, 25-29, 32, 35, 36, 373, 378, 389,465
38, 42, 43, 47, 48, 54, 62- 65, 68, 73, 74, abstract expressionist, 66
89, 90, 104, 116,125, 131-133, 137-139, pan shot, 36, 59, 68, 235, 236, 294, 384, 427
149, 150, 155-157, 167, 169, 174, 180, Panavision, 282, 286, 288, 289, 306, 322, 324, 325,
187, 191, 200, 217, 219, 220, 259, 269, 464, 465,467, 468, 477n21
280, 306, 322, 323,329, 330, 333-335, panchromatic film stock, 284
338, 339, 342, 343, 347, 355, 365, 366, Panofsky, Erwin, 19
372, 375, 378-380, 384, 388-390, 401, Paradine Case, The (1947), 37-39, 41,437
406,414-416, 430,474, 478n42 Parallax View, The (1975), 225
Nosferatu, eine Symphonie des Grauens (1922), parallel worlds, 171, 173, 175,182, 184-187, 215,
70, 72 454n3
Nostalghia (1983), 162 parallels, narrative, 152, 180,182, 184-187, 191,
Notorious (1946), 37, 39, 437n61 211-213, 216-218, 225, 242, 458n33
parametric narration, 98
Parenthood (1989), 201
o Parents terribles, Les (1948), 41
Paris nous appartient (1960), 214, 220
October (Oktyabr, 1928), 157, 427 Park Chan-wook, 161
Odyssey, The, 207 Pasolini, Pier Paolo, 152, 154, 254
Ohayo (1959), 53, 54 Passagers, Les (1999), 101, 210, 221, 237-42
Okajima Hisashi, 472 Passenger, The (Professione: reporter, 1975), 153
Oklahoma! (1955), 285, 468n76 Passing Fancy (Dekigokoro, 1933), 367
Oktoberfest (2005), 243 Passion (Szenvedely, 1998), 162
Oliveira, Manoel de, 164 Passion de Jeanne dArc, La (1928), 157
Olivier, Laurence, 41 Pastorale (Pastorali, 1976), 234
Pat and Mike (1952), 308
Once Upon a Time in Triad Society 2 (Hui ba! pause/burst/pause construction of action scenes,
Ja "fit"yan bingtuen, 1996), 200 403,404,406,408,409,411
One Day in Europe (2005), 243 Peacock (Kong que, 2004), 165
One Hour with You (1932), 257 Peckinpah, Sam, 425, 427
One-Armed Swordsman (Dubei dao, 1967), 416, Peking Opera. See Beijing Opera
419, 428 Peppermint Candy (Bakha satang, 2000), 161
Ophuls, Max, 253, 261, 379 perceptual effects of cinema, 48, 50, 52, 63, 66-68,
Opium (1919), 266-274 72, 79, 90, 91, 93, 101,116, 118,123,
Opposite of Sex, The (1998), 219 124, 126, 127, 130-132
order of story events, 100, 110, 183, 194, 197, 198, Perec, Georges, 195
209,215,237,244, 378, 388 Perkins, V. F„ 436n55, 438n73, 452n6,466n54
494 Index Index 495

Persona (1966), 151, 153, 155, 156, 254
perspective, linear, 65, 66, 442n22
Pete Kelly's Blues (1955), 319, 468n73
Peyton Place (1957), 289
Picnic (1956), 31, 311
pictorialism (Japanese), 360-363, 366, 372
piecemeal decoupage, 359, 363, 365, 366, 371, 372
Pierrot lefou (1965), 153, 283
Pincus, Leslie, 373
plan americain, 259, 278, 295, 323
planimetric image, 163, 169, 453n25
Plantinga, Carl, 51, 116
Play Time (1967), 157, 200, 203, 218, 233, 234, 237,
315
plot, 13,14, 17, 18, 27, 29, 49, 52, 87, 89-91, 94, 98,
101-8,110, 114, 118, 120,122, 125, 135,
138, 139,141, 142,154, 155, 172-176,
178, 179,181, 183-187, 190, 191, 194,
195, 199, 200, 202-20, 222, 224-9, 233,
235, 238, 255, 257, 258, 261, 267-69,
346, 355, 364, 390, 392, 401, 413, 445n7,
4, 458n36
poetics of cinema, 1, 4, 12-25, 29, 32, 38, 42, 44,
48, 51, 53-5, 88, 90, 92-4, 100-102,
105, 109,120, 133, 150, 163, 169,191,
217, 325, 338, 388, 434, 440nl03
Poetics, by Aristotle, 89, 445
point of view, narrative, 35, 95, 99, 111, 121, 130,
172,174,209, 215, 259, 292, 389,413,417
point-of-view editing, 88, 99, 100, 102, 110, 158,
183, 328, 340, 359, 366, 424
Polanski, Roman, 159
Police {Keisatsukan, 1933), 375-377, 379, 382, 384,
387, 388
Police Story (Jingcha gushi, 1985), 408
policier genre, 359, 396, 404
Pom Pom and Hot Hot (Shen Qiang Shou yu Ga Li Ji,
1992), 410
Pons, Ventura, 185
Pool, Ithia de Sola, 198
Poppy, The (Gubijinso, 1935), 366
Poseidon Adventure, The (1972), 192, 262
postmodernism, 81, 94, 262
Posto, II (1961), 154
poststructuralism, 4, 11, 57, 61, 66, 449
Pot-Bouille, 194, 195
Power of Kangwon Province, The (Kangwon-do ui
him, 1998), 161
Prairie Home Companion, A (2006), 243
Prehistoric Women (1967), 282
Premier juillet, lefilm (2004), 208
Preminger, Otto, 38, 98, 138,155, 257, 262, 304,
320-22, 342, 452n6, 468n76
Pride and Prejudice, 121,122, 129
Priestley, J.B., 185
primacy effect, 100, 115, 182, 183
Primer (2004), 163
Prince Valiant (1954), 304
prints, Japanese woodblock, 351, 352, 355, 356
Project A (A Jihua, 1983), 408
Propp, Vladimir, 141
protagonists, 16, 52, 90-3, 96, 99, 101, 104-6, 108,
113, 117, 124, 131, 139, 190-3,199, 202,
203, 208, 209, 212, 214, 216, 217, 220,
222, 224, 229, 233, 237, 153, 157, 158,
163, 165, 173, 178, 182-4, 186, 187, 272,
375, 384,414,428, 456n5
Prunella (1918), 70
Psycho (1960), 92, 93, 157
psychoanalysis, 6, 11, 12, 54, 80, 135, 136
psychology, cognitive, 4-6, 44, 45, 51, 54, 75, 78,
81, 82,143, 379, 438n81, 443n33
Pudovkin, Vsevolod, 18, 58, 59, 69, 331, 357, 397,
401, 425, 470nl4, 478n3
Pulp Fiction (1994), 79, 197, 209, 228, 244, 245
Punch-Drunk Love (2002), 163
Puppetmaster, The (Hsimeng Jensheng, 1993), 25,
161, 166
Purple Heart, The (1944), 299
Q rising action, 102, 223, 229
Quiet Man, The (1952), 113
R Robe, The (1953), 281, 287-292, 302, 304, 308, 313,
Ragtime (1981), 228
Rainer, Yvonne, 66
Raining in the Mountain (Kongshan lingyu, 1979),
415, 419, 420, 424, 425, 427, 428, 430
Raintree County (1957), 286
Rancho Notorious (1952), 283
Rashomon (1950), 160
rationality, 3-5, 14-16, 22, 28, 29, 54, 80, 81, 173,
354, 355,454n5
Ratnam, Mani, 201
Rave Fever (Chow moot kwong hyn, 1999), 215
Ray, Nicholas, 283, 310, 321
Ray, Satyajit, 153, 155,159
Rayns, Tony, 479
realism, 14,15, 17,18, 80, 81, 111, 152-157, 161-163,
194, 227, 256, 282, 292, 352, 371, 379,
398,406,466n54
rear projection, 297
Rebecca (1940), 40, 42, 131, 437n61
Rebel Without a Cause (1955), 310, 321
Reconstruction (2003), 163
Record of My Love (Waga ai no ki, 1939), 351, 368,
370
Red Peony Gangster: Flower Cards Match (Hibotan
bakuto: hanfuda shobu, 1969), 392
Red River (1948), 29
Red Violin, The (Le Violon rouge, 1998), 202
redundancy, 68, 79, 143-145, 147,187, 342, 455
reels as structural divisions, 33-35, 38, 39, 42,
103-105,269, 436n56,437n57
reflectionism, 30, 31, 262, 373 Sanshiro Sugata (Sugata Sanshiro, 1943), 364, 365,
Regal Pictures, 289 390
Regeneration (1915), 120 Sanshiro Sugata II (Zoku Sugata Sanshiro, 1945),
Regie dujeu, La (1939), 261 364
Reinert, Robert, 27, 28, 263-80 Sansho the Bailiff (Sansho daiyu, 1954), 110
Reisz, Karel, 59 Sarris, Andrew, 253-262
renga (linked verse), 389, 390 Satanstango (Satantango, 1994), 162
Renoir, Jean, 28, 70,158,253,254, 261-263, 282, 338 Saussure, Ferdinand de, 5, 62,441n7
rensageki ("chain drama"), 338 Sauve quipeut (la vie) (1979), 217
Republic Pictures, 285, 464 Say It With Songs (1929), 150
Resnais, Alain, 72, 152, 158, 163, 165,185, 254, Sayles, John, 191, 212, 228
452n6 Scarlet Bat, The (Beni komori, 1931), 382, 384
Respect for the Emperor (Sonnojoi, 1928), 377 schemas, cognitive, 48, 93, 94, 97, 99, 101, 112-114,
reverse shot, 39, 42, 53, 54, 59, 138, 167, 239, 265, 123, 126,127, 137, 138,140-3, 149, 150,
295, 342, 399, 401 187, 195, 236, 349, 379, 446n38, 450n7
rhetoric, 2, 12, 18,19, 87, 119 schemas, stylistic, 48, 66, 69, 70, 72, 73, 163, 243,
Richardson, Tony, 254 283-284, 295-8, 300, 301, 303, 307,
Richie, Donald, 372 312, 321, 342, 351, 365, 368, 372, 392,
Ride Lonesome (1959), 316 410,428
Riegl, Alois, 19 Schenker, Heinrich, 475, 476
RifFaterre, Michael, 18 Schlegel, A.W., 89
Rio Bravo (1959), 151 Schneider, Romy, 322, 469n90
School Daze (1988), 71
Schrodinger, Erwin, 187
Rist, Peter, 374
River of No Return (1954), 320 science, 3-5, 15, 16, 21, 22, 80, 86, 174, 198, 231,
432n8, 443n33, 471n35
Rivette, Jacques, 197, 214, 220, 283, 309, 312
science fiction, 113, 160, 186, 187, 396
"Roads of Destiny," 172, 174, 181
Scorsese, Martin, 262, 469
Scott, Ridley, 262
321, 391
Screen magazine, 155
Rock, The (1996), 398, 399
Searchers, The (1956), 31, 258, 259, 262
Rogers, Sheena, 81
Second Circle, The (Krug vtoroy, 1990), 162
Rohmer, Eric, 155
Secret Rivals (Nan quan bei tui, 1976), 414
Ronde, La (1950), 121, 127, 238
Secret, The (Fengjie, 1979), 161
Roos, Don, 210
See You in Space (Egveled!, 2005), 191
Rope (1948), 32-42, 52, 104, 294
seishinshugi ("spirit-ism"), 370
Rosen, Charles, 19, 26
Seldes, Gilbert, 466n54
Rosen, Morris, 38
Selon Charlie (2006), 243
Rossellini, Roberto, 158
Selznick, David 0., 37-39, 41, 437n65
Run Lola Run (Lola rennt, 1998), 172-174, 176-178, semiotics, 22, 80, 86, 135, 136, 441n7
182,183, 185, 187 Sennett, Mack, 410
Running Time (1997), 42 Serpent, The (Orochi, 1925), 339
Ruskin, John, 257 Serpent's Egg, The (1977), 152
Russian Ark (Russkiy kovcheg, 2002), 42 Seven Brides for Seven Brothers (1954), 321
Ruttenberg, Joseph, 291 Seven Year Itch, The (1955), 288
Ryan, Marie-Laure, 112, 113, 116, 130 Seventh Seal, The (Det Sjunde inseglet, 1957), 151,
153
sex, lies, and videotape (1989), 197, 219
s Shame (Skammen, 1968), 154
Shamroy, Leon, 291, 292
SlmOne (Simone, 2002), 332 Shaw Brothers, 419
Saboteur (1942), 37, 437 Shcheglov, Yuri, 18
Sabu. See Tanaka Hiroyuki Sherfig, Lone, 159
Saburi Shin, 381 Shibukawa Bangoro (1922), 339
Sacred Knives of Vengeance. See Killer, The (1972) Shibuya Minoru, 472
Sad Movie (2005), 243 Shimazaki Toson, 390
Saegusa Genjiro, 357 Shimazu Yasujiro, 349, 359, 381
Salt, Barry, 58, 462n4, 462n9 Shimizu Hiroshi, 348, 359, 366, 371
Sanjuro (Tsubaki Sanjuro, 1962), 392, 416 shimpa (new drama), 352, 355
496
shi-shosetsu ("I-novel"), 390
Shklovsky, Viktor, 19, 20, 383
Shochiku, 352, 359, 363, 366, 472
shomin-geki (middle-dass films), 359
Shopgirl (2005), 99
Short Cuts (1993), 191, 197, 212, 221, 230, 244, 245
shot/reverse shot editing, 14, 28, 48, 57-61, 64,
66-70, 72, 73, 76, 163, 221, 297, 307,
335, 340, 342, 357, 368, 381, 425
Siegfried (Die Nibelungen: Siegfried, 1924), 103
Silence of the Lambs, The (1991), 95, 124
Silence, The (Tystnaden, 1963), 153
silent cinema, 14, 151, 157, 186, 235, 236, 255, 256,
258, 263, 266, 282-284, 294, 295, 297,
304, 321, 322, 346, 347, 351, 425
Silk Stockings (1957), 292, 307
Simon, John, 460
Simpsons, The, 196
Singer, Ben, 52
Singer, Peter, 52, 75
Sirk, Douglas, 157, 283, 305
Sirocco (Sirokko, 1969), 35
Sisters ofGion (Gion no shimai, 1936), 368
Sitney, P. Adams, 19
Sjostrom, Victor, 26, 265
Skouras, Charles, 288
Skouras, Spyros, 282, 287-289
Slacker (1991), 197, 215, 455nl8
Sleep (1963), 42
Sleeping Beauty (1959), 286
Sleepless in Seattle (1993), 48
Sliding Doors (1998), 172,174-176,178,180,182,184
slow-motion, 136,137, 364,402,407-409, 417,427,
428
Smiley, Jane, 102
Smith, Greg, 51, 52
Smith, Murray, 51, 101, 150, 447
Smoking/No Smoking (1993), 185, 187
Snake in the Eagle's Shadow (Shexxing Daoushou,
1978), 43, 47, 48, 50, 404
soap opera, 195,196, 244
Socialist Realism, 32
Soderbergh, Steven, 163
Soigne ta droite (1987), 218
Sokurov, Aleksandr, 42,162
Solaris (2002), 163
Solaris (Solyaris, 1972), 163
Solondz, Todd, 200, 228
Sons of the Good Earth (Da di er nu, 1965), 424
Souls on the Road (Rojo no reikon, 1921), 339
sound cinema, 14, 19,104,129, 152,177, 221, 232,
256-8, 282-285, 294, 295, 304, 353,
363, 397, 409, 466n54, 474n42
South Pacific (1958), 468
Soviet film, 14, 18, 401, 416
montage style in, 14, 18, 27, 32, 254, 255, 337,
357,4125,451
spatial representation, 45, 64, 65, 111, 112,154,
156,235, 278,283, 341,342
Index
spectator, 20, 26, 43, 44, 46-51, 54, 58, 59, 63, 68,
93-95,101,124,130,133,135-42,146,
147,149,150,152,187, 200, 235, 282,
283, 293, 320, 345-7, 349, 350, 373,
379, 380, 382, 384, 388, 389, 410, 411.
See also curiosity; emotions; perceptual
effects of cinema; suspense; surprise
Speed (1994), 399
Spellbound (1945), 42, 437n61
Spider-Man (2002), 30
Spielberg, Steven, 262, 290, 323, 410, 469n94
split screen, 97, 99,100, 185
Spottiswoode, Raymond, 255
Springtime in a Small Town (Xiao cheng zhi chun,
2002), 161
Stagecoach (1939), 94, 257
staging, 3,14,19, 25-7, 38, 39, 42, 3, 163, 164,
180, 256, 263, 265, 269, 277, 283, 295,
296, 298-300, 307, 321, 324, 325, 334,
338-40, 342, 347, 349, 350, 353, 355,
357, 364-366, 371, 375, 379, 382, 389,
399, 402, 415, 453n22
recessive, 164, 297-300, 323, 467n63. See also
tableau
Staho, Simon, 166
Staiger, Janet, 49
Standing Still (2005), 191
Standish, Isolde, 373
Star Athlete (Hanagata senshu, 1937), 366, 368, 474
Star Is Born, A (1954), 49, 300, 304, 309, 310, 321
Star Trek television series, 50
Star Wars (1977), 290
stars, 31, 39, 47, 92, 93, 113,154, 159,160,197, 217,
288, 292, 294, 306, 322, 331, 338, 355,
396, 401, 411, 460nl7
Stella Dallas (1937), 229
Sterbende Volker (1921-2), 462
stereotypes, 17, 32,110,173,196, 236
Sternberg, Josef von, 262, 382
Sternberg, Meir, 19, 100, 101,139,150,181,199,
440nl01, 448n62
Stone, Oliver, 262
Stormy Days of the Sanada Clan (Sanada
fuunroku, 1963), 416
Story of Tank-Commander Nishizumi, The
(Nishizumi senshacho-den, 1940), 370
storyboards, 406
Stout, Rex, 96
Strada, La (1954), 151,153
Straits of Love and Hate, The (Aien kyo, 1937), 369
Stranger Knocks, A (Enfremmed banker pa, 1959),
153
Strangers When We Meet (1960), 306
Straub, Jean-Marie, 59, 116, 158, 163,164, 279
Street Without End (Kagirinaki hodo, 1934), 359,
371, 361,362,384,
Strike (Stachka, 1925), 104
Stroszek (1977), 152
structuralism, 57, 61, 98, 255, 449
Index
studio system, 14, 22, 28, 41, 105, 126, 137, 152,
219, 254, 255, 262, 282, 283, 286-288,
290, 337, 353, 355, 373, 460
Sturges, John, 283, 313, 319
Sturges, Preston, 261
style in cinema, 1, 3,14, 15, 18, 19, 23, 24, 27, 32,
37, 39, 47, 48, 50, 54, 65, 69, 75-78,
97-99, 129, 137, 149,152, 157, 159, 164,
227, 234, 255-262, 265, 276, 282, 284,
295, 304, 305, 322-325, 334, 335, 337,
338, 342, 347, 351, 353, 356, 357, 359,
363, 365, 366, 368, 369, 371-4, 375-82,
387-390, 396, 398, 399, 401, 413, 426,
427, 429, 430, 472n8, 474n38, 475nl.
See also cinematography; editing;
spatial representation; staging
stylistics of film, 14, 15, 17,19, 25, 27, 28, 36, 48,
54, 58-60, 65, 77, 78, 97-9, 110,129,
136, 138,144, 149,150,152, 154, 155,
163, 219, 221, 233, 254, 256-8, 260, 261,
273, 283, 284, 303, 327, 355, 356, 359,
363, 365, 366, 368, 371, 372, 376-81,
384, 385, 387-9, 396, 399, 406, 408, 411,
413-416, 428,430, 452n6
Sud Express (2005), 191
Sunshine State (2002), 203, 212, 213
Super Panavision, 285
Super Technirama, 70, 286, 468n76
Super-35mm, 324
superimposition, 280
surprise, 51, 90, 99-101, 150, 200, 201, 207, 226,
271, 332
Surrealism, 24, 353, 435, 451
suspense, 13, 39-41,44, 52, 89, 100, 101, 125, 139,
150, 206-8, 210, 216, 236, 273, 298, 317,
319, 320, 370, 467n54
Suspicion (1941), 437
Suvorov (1941), 69
Suwa Nobuhiro, 161, 165
Suzuki Daisetsu, 353, 354
Suzuki Seijun, 161, 392
swordplay films, 337, 339, 342, 344, 357, 363, 377,
385, 401,406,419
Swordsman II (Xiao aojianghu zhi Dong Fang Bu
Bai, 1992), 419, 428
Sympathy for Mr. Vengeance (Boksuneun naui
geot, 2002), 161
Syriana (2005), 219, 242
syuzhet. Seefabula and syuzhet
T
tableau, 32, 58, 78, 279, 295, 298, 312
Tabu (1931), 283,
taishu bunka (middle-class culture), 352
Take Off Your Clothes and Live (1964), 282
Taki no shiraito (1933), 366
Tale of Guidance (Shido monogatari, 1941), 370
497
Tales of Manhattan (1942), 202
Tales of the South Pacific, 227
Tam, Patrick Kar-ming, 161
Tan, Ed, 51, 81, 116, 150, 331, 447n47
Tanaka Hiroyuki (Sabu), 191, 206, 392
Tanaka Kinuyo, 381
Tang Shuxuan, 479
Tanizaki Junichiro, 353, 354, 390
Tapas (2005), 191, 203
Tarantino, Quentin, 197, 198, 209, 262
Tarkovsky, Andrei, 162, 163
Tarnished Angels, The (1958), 305
Tarr, Bela, 162
Tasaka Tomotaka, 370
Tashlin, Frank, 292
Tati, Jacques, 28, 163, 197, 200, 203, 233-237, 262,
315
Tecce, Joseph, 332
Techine, Andre, 110
Technicolor, 33, 286, 292, 307
Technirama, 286, 289, 464n3, 468n76
Techniscope, 286, 322, 323, 464nl3
technology of cinema, 28, 29, 34,42, 58, 282, 284,
290, 293, 301, 304, 320, 324
television, 32, 58,152, 195, 219, 221, 227, 228, 244,
284, 287, 309, 332, 456n9
Ten (2002), 162, 165
Ten Commandments, The (1956), 31, 289
Ten North Frederick (1958), 99,126
Ten Skies (2004), 45
Terrorizers, The (Kongbu fenzi, 1986), 161, 207,
244, 208
Tewksbury, Joan, 222
text, 12, 15, 49, 127, 128, 130,154-157
That Day, on the Beach (Haitan deyitian, 1983), 161
The White Balloon (Badkonake sefid, 1995), 162
thematics, 13, 17, 18, 28, 31, 54, 108, 152,153,156,
157, 159, 169,173, 200, 213, 218, 225,
238, 255, 303, 377-81, 389,413
theorizing, 1-4, 12, 19, 20, 24, 58, 60, 61, 81, 89,
93, 123,124, 129, 135, 136, 255, 256,
292, 377, 388, 441n7
Theory, Grand, 1-5, 21, 22, 73
theory, primary, 61, 62, 64, 111, 116
There Was a Father (Chichi ariki, 1942), 372
Thief (1981), 396
Thin Man, The (1934), 296, 323
Things You Can Tell Just by Looking at Her (2000),
202,208,209
Thirteen Conversations about One Thing (2001), 213
This Is Cinerama (1952), 287
Thompson, Kristin, 28,105,106,108,109,159,434,
462n9
Three Ages, The (1923), 180
Three Beauties (San renka, 1935), 384
Three Coins in the Fountain (1954), 215, 287
Three Colors trilogy (Trois couleurs/Trzy kolory,
1993-4), 162
Three Kings (1999), 325
498
Three Men from the North (Kita no sannin, 1945),
340
Three Times (Zui hao de shi guang, 2005), 161
three-quarter view of action, 59, 68, 79
thriller, 215-217, 219, 244, 298, 321, 379, 396, 399
Through a Glass Darkly (Sasom i en spegel, 1961),
153
THX1138 (1971), 324
Tian Zhuangzhuang, 161
Tiger on the Beat II (Laohu Chugeng, 1998), 408
Time of Love (Nobat e Asheghi, 1990), 185
Time to Live and a Time to Die, A (Tong nien wang
shi, 1985), 165
To Die (or Not) (Morir (o no), 2000), 185
To Kill a Mockingbird (1962), 120
Todd-AO, 285, 286, 289, 293,468n76
Todorov, Tzvetan, 12
Toho,353, 359
TohoScope, 391
Tokyo Chorus (Tokyo no gasho/Tokyo no korasu,
1931), 343, 344, 366
Tolstoy, Lev, 172
Toni (1935), 70
Too Many Ways to Be No. 1 (Yi ge zi tou de dan
sheng, 1997), 172,174,175, 177,178,
180-183, 186
Top Banana Club (Jin zhuang xiang jiao ju le bu,
1996), 202
Toto le heros (1991), 101, 454
Touch of Zen, A (Hsia nu, 1969), 413, 415, 417-419,
421,424-426, 428-430
Tourneur, Maurice, 70
Toutva bien (1972), 158
Toute une vie (1974), 203
Traffic (2000), 219
Traitress, The (Die Verraterin, 1912), 264-266,273,
278
Trauberg, Leonid, 70, 71
Trial ofVivienne Ware, The (1932), 376
Trier, Lars von, 159, 163, 453
triggers, sensory, 45, 51, 63, 64, 66, 67, 74, 328
Trilogy (Apres la vie, 2002), 244
Truffaut, Francois, 39,110,154, 155,159, 283, 466
Tsui Hark, 409, 430
Tu Quangqi, 425
Turn Back the Clock (1933), 186
Twain, Mark, 22,121, 132
Twentieth Century-Fox, 281,282,285-289,292-294,
300, 302, 304, 305, 307, 319, 321
Twenty Bucks (1993), 202, 213
Twenty-Six Japanese Martyrs, The (Les vingt-six
martyrs japonais, 1930), 357, 358, 371,
380, 381, 474n38
Two Minutes to Zero Trilogy, The (2004), 48, 49
Tykwer, Tom, 159, 172,176,177,183
Tyler, Parker, 140, 149
Tynianov, Yuri, 19
Index Index 499
u Wang Yu (Jimmy Wang Yu), 416
Wang, Joey, 399-401, 403, 407
Wong, Lawrence, 211
Woo, John, 70, 71, 395, 404-407, 409, 410,427
Uchida Tomu, 375, 376, 379, 414 War and Peace (1956), 31 Wood, Robin, 155
Ultra Panavision, 70, 286 War and Peace, novel by Leo Tolstoy, 194 World, The (Shijie, 2004), 161
Under Capricorn (1949), 41, 437, 438n74 War at Sea from Hawaii to Malaya, wuxia pian, 401, 416, 419
Underwater Warrior (1958), 306 The Hawai mare oki kaisen, 1942), 370 Wyler, William, 14,42, 262, 263, 283
Underworld Beauty (Ankokugai no bijo, 1958), 392 war films, 217, 321, 370, 371, 396 Wyndham, John, 187
Unfaithfully Yours (1948), 186 Ware, Chris, 185, 196 Wynne-Jones, Diana, 194
United Artists, 288, 289, 464 Warhol, Andy, 42, 116, 162
Universal Pictures, 289, 464n21 Warner Bros., 285, 288, 289, 299, 375, 408,
universals, 13, 61-64, 66-69, 73-77, 79-81, 86, 464n21, 469n88
127, 256,443, 447, 448 WarnerScope, 285, 464
X
contingent, 61-64, 66-69, 73-77, 81, 86, 244, Wayne, John, 258, 335 X-Men (2000), 262
443n33, 448n52 Webb, Jack, 283, 319, 321, 469n88 Xu Feng, 415, 420
Untouchables, The (1987), 92 Wedding, A (1978), 202
Urlaub aufEhrenwort (Leave on Word of Honor, Weir, Peter, 262
1938), 456 Welcome to the Dollhouse (1996), 115
Usual Suspects, The (1995), 101, 219, 244 Wellek, Rene, 11
Y
Welles, Orson, 14, 15, 27, 38, 257, 262, 263, 282, Yaaba (1989), 67
298, 305, 379, 437n63 Yamanaka Sadao, 343, 345, 363
Westerns, 94, 150, 204, 207, 243, 305, 321, 406
V Yang, Edward, 161, 207, 244
What a Wonderful Place (Eiza Makom Nifla, 2005), Year Zero (Shnat Ejfes, 2005), 191
Vagabond (Sans loi ni toit, 1985), 117, 118,166-169, 191 Yellow Earth (Huang tu di, 1984), 161
242 What's Wrong With This Picture? (Tid til Yellow Rolls-Royce, The (1969), 202
Valiant Ones, The (Chung lieh tu, 1975), 415, 419, Forandring, 2004), 190-193, 199, 212 Yoda Yoshikata, 359, 361
422-429, 479nl4 Wheel of Life, The (Da lunhui, 1984), 424, 425 Yojimbo (1961), 416
Vanoye, Francis, 109, 448 Whispering Pages (Tikhiye stranitsy, 1993), 162
Yoshi Castle (Fuum Yoshi, 1928), 340, 341
Varda, Agnes, 117, 166, 169, 454 Who's Camus Anyway? (Kamyu nante shiranai,
You Only Live Once (1937), 98, 138
Variety (trade newspaper), 191, 243, 261 2005), 191
You've Got Mail (1998), 106,107,120
Veidt, Conrad, 267 Why Change Your Wife? (1920), 195
Young and Innocent (1937), 37
verisimilitude, 152, 154-156, 233 Wiedemann, Vinca, 159
Young Lions, The (1958), 291, 305, 306
Vernet, Marc, 114 Wife, Be Like a Rose! (Tsumayo bara no yo ni, 1935),
Young Master (Shi di chu ma, 1980), 414
Victory Song of the Orient (Toyo no Gaika, 1942), 384
370 Yuen Biao, 407
Wilbur Wants to Kill Himself (2002), 159
video, 32, 42, 47,144, 160, 162,187, 244, 261, 289, Yuen Kuei, 430
Wild and Woolly (1917), 150, 265
332, 395, 396 Yuen Woo-ping, 43, 404,410, 430, 478n9
Wild Bunch, The (1969), 396, 425
video games, 187 Yuva (2004), 201, 212, 457
Wild River (1960), 311, 312
Vidor, King, 260,467n54, 468n76 Wild Strawberries (Smultronstallet, 1957), 151, 153
Vier letzten Sekunden des Quidam Uhl, Die Wilder, Billy, 262, 304
(1924), 266, 462n8 Wilder, Thornton, 194 z
Violent Men, The (1955), 324 Willeford, Charles, 194
Zanuck, Darryl F„ 122, 287, 289, 290, 302, 307
Violent Saturday (1955), 302, 321 Wilson, Robert Anton, 185
Zatoichi film series, 411,414
VIPs, The (1963), 194 Winchester '73 (1950), 202
Zavattini, Cesare, 42
Vistarama, 285, 464nl3 Wind and the Lion, The (1975), 396
Zeitgeist, 205
VistaVision, 286, 288, 289, 464nl3, 468n76 Winter Camellias (Kantsubaki, 1921), 339, 340
Zellweger, Rene, 334
Vivre sa vie (1962), 151, 153, 319 Wise, Robert, 304
Zen Buddhism, 353, 413
voiceover, 40, 95, 99, 125, 152, 169, 216, 219, 223, Wolfflin, Heinrich, 19, 297
Zhang Che (aka Chang Cheh), 416, 419, 424
230-232,238,426 Wollen, Peter, 460n4
Voisin Voisines (2005), 191 Zhang Xinyan, 425
Woman of Paris, A (1923), 359
Woman Who Touched the Leg, The (Ashi ni Zillmann, Dolf, 81
sawatta onna, 1926), 390 Zinnemann, Fred, 303
w Wonderful World of the Brothers Grimm, The (1962),
284
Zizek, Slavoj, 75
Zola, Emile, 194, 213
Wakamatsu Kojiro, 161 Wong Kar-wai, 104, 161,197, 200, 220,409,453nl9 zoom lens, 110, 221, 226, 235, 380, 385, 392, 398,
Wald, Jerry, 282 Wong, Anthony Sau-chang, 202 402, 408, 409, 476n21
Walsh, Raoul, 410
Walt Disney World, Florida, 284
Wandering Gambler, The (Horo zanmai, 1928), 339
Wang Xinglei, 430