Bordwell. David. Poetics of cinema [PDF] | Online Book Share (2024)

510 Pages • 240,519 Words • PDF • 36.5 MB

+ Cinema+ Poetics

Uploaded at 2021-09-21 18:11

Report DMCA

PREVIEW PDF

POETICS of CINEMA

Davi d B o r d w e l l

New York London

Routledge Taylor & Francis Group 270 Madison Avenue New York, NY 10016

Routledge Taylor & Francis Group 2 Park Square Milton Park, Abingdon Oxon 0X14 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 informa­ tion storage or retrieval system, without written permission from the publishers. Trademark Notice: Product or corporate names may be trademarks or registered trademarks, and are used only for identification and explanation without intent to infringe. Library of Congress Cataloging-in-Publication Data 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 Yisit the Taylor & Francis Web site at littp://www.taylorandfrancis.com ind the Routledge Web site at hittp://www.routledge.com

2007004036

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

Contents

Acknowledgments

xi

Introduction

I

1

Questions of Theory 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

2. Convention, Construction, and Cinematic Vision Shot/Reverse Shot: A Convention? 57 Primary Theory and a Continuum of Conventions A Package Deal 66 Contingent Universals and Us 73 Afterword 75 V ll

43

57 61

Contents

viii

II

Studies in Narrative 3. Three Dimensions of Film Narrative Some First Moves 88 Protagonists and Their Problems 90 Narration 93 Plot Structure 102 The Narrative World 110 Afterword: Narrators, Implied Authors, and Other Superfluities

85

121

4. Cognition and Comprehension

Viewing and Forgetting in Mildred Pierce Narrative Norms 137 Two Methods of Murder 138 The Partial Replay 143 Secrets and Ties, and Narration

135

149

5. T he A rt Cinema as a Mnde of Film Practice Realism, Authorship, Ambiguity The Art Cinema in History 157 Afterword 158

151

152

6. Film Futures

171

Rules of the Game 174 Some Sources 185

7. M utual Friends and Chronologies of Chance Protagonists and Projects 192 This Particular Web 194 Major Players 198 Familiars, Strangers, and Random Walks Only Connect, or at Feast Collide 204 Narration Makes Networks 207 Compare and Contrast 211 Criss-Crossers Cross Over 214 Four Small Worlds 218 Nashville (1975) 221 Magnolia (1999) 227 Favoris de la lune (1984) 233 Les Passagers (1999) 237

189

200

Contents Sustaining the Network 242 Network Narrative: A Working Filmography

III

245

Studies in Style 253

8. Cinecerity Style and Film History 255 Style and the Critic 257 Envoi 261

9. Taking Things to Extremes H allucinations Courtesy of Robert Reinert Revelations at the Cinema Verdi Close-Up Depth 269 Problems and Solutions 276

263

266

10. CinemaScope The ISAodern JVLiracle You See Without (jlasses

281

The Big Picture 284 Hollywood Cadillac 287 A Lack of Scope 290 Taming a New Technology 301 Some Virtues of Clotheslines 307 Reinventing the Tableau 312 The End of Screen Ratios? 320

11. Who Blinked First?

327

The Tightrope 327 The Strength of the Stare 331 Streamlined Behavior 334

12. V isual Style in Japanese Cinema, 1925-1945 A Classical Cinema? 338 Voices in the Dark 345 Exercising the Eye 347 Constructing and Reconstructing Japaneseness 352 Stylistic Trends 356 Blending and Refining 363 The Pacific War: Toning Down Technique 369 Afterword 372

337

Contents

X

13. A Cinema of Flourishes Decorative Style in 1920s arid 193 Os Japan ese Film Some Functions of Style 377 Games With Vision 380 A Tradition of Ornamentation

388

14. Aesthetics in Action Kung-Fu, Cjunplay, and Cinematic Expression Hollywood Action: The 1980s and After The Mechanics of Movement 399 Toward an Ecstatic Cinema 406

395 396

15. Richness Through Imperfection King D u and the Cjlimpse Inherited Norms 414 A Problem and Some Solutions Cuts and Slashes 423 Glimpses of Marvels 429

375

413

416

N otes

431

Index

481

Acknowledgments

Some of the essays in this book originally appeared in the journals and anthologies listed below. The author is grateful for permission to reprint them in revised form. “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 Noël 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 CinémaScope 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, and Cinematic Expressivity.” In Law Kar, ed. Fifty Years of Electric Shadows. Hong Kong: Urban Council and Hong Kong International Film Festival, 1997, pp. 81-89. “Richness through Imperfection: King Hu and the Glimpse.” In Law Kar, ed. Transcending the Times: King Hu and Eileen Chan. Hong Kong: Urban Council and Hong Kong International Film Festival, 1998, pp. 19-24.

Introduction

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

Poetics of Cinema

2

political processes of the “trans” in transplanting, transmitting, transculturating, 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.1If 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 selfcongratulating 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 technophilic 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 intercultural 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 filmmaking 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 to fit the 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 editor.. . . 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.7That 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. Who 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

than Lacan could imagine (and he had a capacious imagination). Film studies, even with its professed “historical turn,” continues to emphasize “methods” over questions, catchwords over concepts, and doctrines over free-ranging inquiry.9 Too many film scholars promote a limited conception of interdisciplinarity, borrowing ideas only from trends that fall within the Continental hermeneutic tradition (theories of literary interpretation, Lacanian psychoanalysis, postmodern anthropology, and the like). By asking questions from a broader purview, we open ourselves to ideas from comparative narratology, cognitive psychology, Darwinian theoretical programs, network theory, and other progressive trends in the human sciences. We needn’t look to them for ultimate answers—we don’t need another dogmatism—but we should canvass widely in seeking out help in answering the research questions we pose. Most of the essays collected here have appeared in print before. All have been revised, some a lot. Now I appreciate why so many authors prefer reprinting old articles to recasting them. It’s harder to patch up an old piece than to weave a wholly new one. Although I haven’t tried to summarize developments in any exhaustive way, a few essays in this book include codas that develop some themes in the light of recent research. New to this volume are “Poetics of Cinema,” “Three Dimensions of Film Narrative,” “Mutual Friends and Chronologies of Chance,” and “CinemaScope: The Modern Miracle You See Without Glasses,” although the first and last have their roots in older pieces.10 Further reprints of essays will, I hope, be appearing on my Web site, http://www.davidbordwell.net, along with new material from time to time. 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, Noël Carroll, Kelley Conway, Beth Corbett, Jim Cortada, Don Crafton, Jim Danky, John Davis, Susan Davis, Maxine Fleckner-Ducey, Tom Flynn, Kevin French,

Introduction

7

Doug Gomery, Erik Gunneson, Meg Hamel, Debbie Hansen, Dave Hellenbrand, Linda Henzl, Boyd Hillestad, the late Chris Hoover, Lea Jacobs, the late Nietzchka Keene, Vance Kepley, Jared Lewis. J. J. Murphy, the late Ordean Ness, Beth Onosko, the late Tim Onosko, Sandy Rizzo, Matt Rockwell, Mary Rossa, Paddy Rourke, Lrank Scheide, Ben Singer, and Andrew Yonda. While conducting my education in public, Tve learned a lot from people in adjacent wings of our department: Julie D’Acci, Michael Curtin, Michele Hilmes, Shanti Kumar, Joe Cappella, Joanne Cantor, Mary Anne Litzpatrick, Zhangdong Pan, the late Ed Black, Tim Haight, Jim Dillard, Blake Armstrong, Ray McKerrow, Lloyd Bitzer, Michael Leff, Steve Lucas, the late Michael McGee, and Sue Zaeske. Chuck Wolfe, Tom Gunning, Richard Maltby, Pete Parshall, and many other visitors to Vilas Hall over the years have taught me more than they probably realize. So, too, have the friends I’ve made in graduate school and at archives, universities, film festivals, conferences, and the normal networking of academic life. And so as well have the thousands of students (no kidding) I’ve had the pleasure of teaching here at Wisconsin. A teacher is said to be the only person who talks in somebody else’s sleep; happily my students have been awake, demanding, and good-humored. My 30-some dissertators, from Brian Rose (1975) to Jonathan Lrome (2006), have been particularly patient with me. Linally, I must single out Bill Germano, a friend of long standing who made this anthology possible, and Kristin Thompson, whose loyalty and love have sustained me for even more years than this collection spans.

I Questions of Theory

9

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 René 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

12

Poetics of Cinema

Hollywood?—a set of up-to-date approaches constituted intellectual bona tides. 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 Gérard 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 maybe 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

André 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 Lumières and Georges Méliès. 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 intersubjectively 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 noncinematic, 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 overall systemic norms.13 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 ephemerality 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 cliché, 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 roughedged police detective, and Waldo Lydecker, the effete bon vivant who has been the patron of the murdered Taura. 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 cliché. The text elaborates this donnée 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. Noël 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 Mécanique, 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 prograpis 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 Mukafovsky, 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 André Bazin, the Soviet filmmakers of the 1920s, and Noël 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 concretesay, 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 discontinued. 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 laymans 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 François 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 policy. . . 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 greater generality.34 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 Détective (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 practicebased 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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

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.37We 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 perspectivai space

26

Figure 1.1 In Victor Sjostrom’s Ingeborg Holm (1913), the clerk in the family’s store slips perfume to a pretty customer as Inge­ borg comes in from the back room. The cam­ era position lets us see that the cash register blocks the clerk’s theft from Ingeborg’s view, an action that wouldn’t be evident from every seat in a stage theater; indeed, from some sightlines, the register would conceal Inge­ borg’s arrival from the spectator.

P o e tic s o f C i n e m a

Figure 1.2 Ingeborg Holm-. Later, a similar camera angle shows Ingeborg’s crisis. The bill collector presses her at the cash register, while the guilty clerk pops into a patch of space behind them. Again, this shot wouldn’t work on the stage because the clerk’s expression wouldn’t be visible from most seats. Sjostrom realized that cinematic space, unlike theatri­ cal space, is resolutely monocular; only the 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 Ffollywood 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

P o e tic s o f C i n e m a

27

Figure 1.3 Jeanne Dielman, 23 Quai du 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

28

P o e tic s o f C i n e m a

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 of this sort. In the Hollywood studio system of the 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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

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 of the 1950s, and sooner or later you’ll invoke the reign of blandness that was the

P o e tic s o f C i n e m a

31

Eisenhower administration. But why do we assume that Americas 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

P o e tic s o f C i n e m a

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 Is damaging effects on European film industries, surely served as preconditions, but those didn’t directly cause the aesthetic changes. More proximate causes included the renewal of national film industries, the saliency of new templates of cinematic storytelling (such as the feature-length fictional narrative), 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 remained constant in its essentials for about a century, in the face of profound social, political, and economic changes—largely, we suppose, because it continues to fulfill functions that filmmakers deem worthwhile. This isn’t to say that society has no impact on films. Of course it does. But that impact isn’t single or simple.53 I’m proposing that causal explanation in poetics can best proceed in steady steps, moving from the artwork to the proximate conditions of production (agents, institutions, and communal norms and practices). These in turn may be influenced by both immediate social causes and longer-term preconditions; we have to look and see exactly how. That Japanese films of the Pacific War period were shaped by political demands is undeniable, but the works of Ozu, Mizoguchi, and their peers reworked the assigned materials in distinctive ways. Likewise, the vague demands of socialist realism in the USSR were fulfilled in ways that both reject and rework Soviet montage norms of the 1920s. In any instance, a social command will be mediated by the film industry, existing traditions, and the varied ingenuity of filmmakers. Similarly, long-standing social attitudes, such as racism or homophobia, supply stereotypes, but those can be transformed by the process of production and the dynamics of the particular film. Poetics is in a good position to show how that works. It reminds us that themes will be recast, form and style will transform social givens, and filmmakers will still choose among importantly different ways to tell the story. Ideology doesn’t switch on the camera.

From Shriek to Shot I’ve suggested that all the constructional factors are connected, that inquiring into particulars and patterns can lead us to principles, purposes, and practices. To illus­ trate this, let me provide a tentative example. Before the arrival of videotape, Alfred Hitchcock’s Rope (1948) was a mythical beast, unavailable in 16mm and seldom shown on television. On a trip to a film archive in the mid-1970s, I watched it, eager to enjoy what had become celebrated as one of Hitchcock’s greatest experiments. The film tells

P o e tic s o f C i n e m a

Figure 1.4 Rope (1948): The first “invisible” cut from Brandon’s back. . .

33

Figure 1.5 . . . disguises a change in the camera reels, which will be spliced together for projection on a bigger reel.

how two gay college men, Bfandon and Phillip, garrote their friend David and hide his body in a decorative chest. They then invite his friends and relations to a dinner party at which the buffet is arrayed, ghoulishly enough, on the chest. The killers’ former prep school teacher Rupert eventually realizes their crime and reveals it to the police. What made everyone keen to see Rope was its reputation as a technical tour de force. Before production Hitchcock had announced that each shot would last nearly a full camera reel. A camera reel normally held 1,000 feet of film, but the Technicolor camera’s maximum capacity was 952 feet, and some of that had to be wasted getting the camera up to speed, so the maximum length of a take for Rope would be around 10 minutes. The publicity for the film had made it seem that every shot was about the same length. “Each take averaged 925 feet,” proclaimed an article in American Cinematographer, which also mentioned that the scenes were rehearsed in sections of about 9 minutes each.54 For decades thereafter, the 80-minute film was described as consisting of eight 10-minute takes. Things are more complicated than that, because the film contains 11 shots, and their lengths vary considerably. Just as important are the ways in which the shots are connected. Rope begins with a comparatively brief high-angle shot establishing the street outside, seen under the credits, before the camera pans to a window and we hear a cry. The next shot is a very long take, but the shots that follow aren’t all the same length, and most run significantly shorter than 10 minutes. More striking is the 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 matches: when a character looks offscreen, we get a cut to another character. Is there a pattern to these particulars? After the external establishing shot, there’s a cut inside to the parlor, where Brandon and Phillip are strangling David. This shot plays out for several minutes while the two young men jam their victim into the chest, close it, and start to talk. At the shot’s end, we get a “hidden cut” as Brandon’s back blocks the frame and a new shot starts (Figures 1.4-1.5). But at the end of that shot, number 3, there’s a conventional eyeline-match cut. Kenneth looks off left—cut—a

P o e tic s o f C i n e m a

34

Figure 1.6 K e n. ..

Rope: The first visible cut, from

Figure 1.7

. . . to the two killers leaving.

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 91 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

P o e tic s o f C i n e m a

Figure 1.8 Rope: Phillip’s wounded reply that he wasn’t afraid to strangle a chicken yields a cut to . . .

35

Figure 1.9 . . . Rupert watching thoughtfully. It’s the beginning of his suspicions, 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

Figure 1.10 Rope: Phillips hand clutching the revolver . . .

P o e tic s o f C i n e m a

Figure 1.11 . . . marks another high point, 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 myself.”59 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

P o e tic s o f C i n e m a

37

Figure 1.12 Rope: Rupert lifts the lid and discovers Davids 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 view.. . . 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 découpage.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

P o e tic s o f C i n e m a

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 planséquence 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 pseudosubjective 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

P o e tic s o f C i n e m a

Figure 1.13 In Notorious (1946), Hitchcock works a variant on the aggressive foregrounds of Wyler and Welles.

39

Figure 1.14 Striking frontality and compact arrangement in Lifeboat (1944).

enough continuous action.”66 Hitchcock likewise defended Ropes 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 François 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, Rope's 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 outré technique. “The

40

P o e tic s o f C i n e m a

Figure 1.15 Rope: In the absence of overthe-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 1.17 Rope: At the climax, Rupert unfolds the rope while commenting that he won’t ride with his pupils.

Figure 1.18 Rope: The searching camera scans the room as we hear Rupert speak of suspense . . .

audience must never be conscious of it.. . . The 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 cameras 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

P o e tic s o f C i n e m a

41

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).

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 shriek! . . . 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

P o e tic s o f C i n e m a

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 M for 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 découpage 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

P o e tic s o f C i n e m a

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.

What 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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

46

A Model of Viewer Activities Concept-Driven processes |r

^

< |r

^ |r

Appropriation

v .. v ;i >r ; >r Comprehension

Perception

V

1 11

A' A * A A

t t t t Data-Driven processes Phenomenal film

/W W A A A A A .

Figure 1.20 A schematic model of the spectator’s activities. Continuous arrows indicate the primary direction of processing, with dotted arrows indicating a degree of feedback among 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 Touise don’t die won’t make it so. Perception evolved in large part to give us reliable information about the threedimensional 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.

P o e tic s o f C i n e m a

47

The perception of film as a representation of the world emerges very early in human development. Many of the experiments testing babies’ reactions to facial expressions use televised images of the mother, which indicates that children spontaneously identify an 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 to be shared with other primates. Experimenters routinely use videos to test chimps, monkeys, and their cousins, and the results indicate that these creatures identify their 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 of other monkeys shrinking from a snake, they begin to show fear themselves.87 This isn’t to say that the processes of filmic perception are innate, as if a newborn 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 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 momentby-moment effects of composition, lighting, cutting, and the like. From the Russian montagists through Rudolf Arnheim, André Bazin, and Noël 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

P o e tic s o f C i n e m a

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 largescale form of the film is designed to create a flow of cues that ask viewers to apply schemas for typical situations and human 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 topdown. 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 Eagle’s 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

P o e tic s o f C i n e m a

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

P o e tic s o f C i n e m a

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 tragicomedy, are to be found throughout the world’s literatures.91 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

P o e tic s o f C i n e m a

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. Tess 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

52

P o e tic s o f C i n e m a

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 non-Asians do.96 More obvious are the emotions that fund comprehension. As we come to under­ stand a narrative, we begin to run scenarios that require “emotional intelligence”— good guesses about how characters will react to the story situations. At the same time, we gauge a character’s personality or current attitude on the basis of their emotional responses. Our inferential elaboration of the cues were given is guided by the emo­ tions that characters register. At the same time, the emotions we feel shape our sense of the film’s macro-action. If we feel that a character has been wronged, we may mimic, in weakened form, her anger and self-righteousness. Screenwriters provide strong prompts for sympathy, such as making sure that the protagonist is treated unfairly, and many screenplay manuals argue that the skilful filmmaker evokes both hope (for the character’s success) and fear (of the character’s failure).97 Again, there may be considerable cross-cultural regularities in these emotions, most of which depend upon recurrent social situations that people in most cultures encounter—sympathy for children, anger at being wronged, and a sense of fairness or justice. In comprehension, emotion and thought mesh. Greg Smith argues that narra­ tive films tend to sustain moods and then punctuate them with bursts of emotion proper.98 These in turn can focus our attention on story developments. In Rope, our knowing where the corpse is hidden generates suspense, a mood that in turn makes us hyperattentive to every movement toward the chest. When Rupert lifts the lid, the mood prepares us to concentrate on his face, which betrays his shock and distress. Emotion also affects memory; in real life, a traumatic event becomes sharply etched into our minds. Films exploit this tendency by making the most vividly emotional scenes crucial for the plot—a death, a separation, a reunion. Ben Singer has proposed a catalogue of prototype scenarios, drawn from intense emotional experiences in 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 India (1957), when the mother, in the aftermath of a flood, digs desperately through 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 salient certain emotion-based prototypes of happiness.100 There’s a great deal to be studied about how emotion works within our cinematic experience, including the bonding effects of watching a film with others. In some of the essays that follow, I propose that narrative films often model social intelligence,

P o e tic s o f C i n e m a

Figure 1.22 The first shot of Ohayo (Ozu Yasujiro, 1959) shows a neighborhood and its clotheslines squatting under electrical towers.

53

Figure 1.23 Ohayo: The last shot presents the same configuration, but from the oppo­ site angle. Ozu’s typical reshuffling of shot elements is here applied to landscapes, empha­ sizing the washed underpants of a boy who can’t control his bowels

and that modeling in turn rests upon sensitivity to emotional signals. Here all I want to indicate is that my last p, processing, can support systematic studies of the range of emotional effects that a film can have. More broadly, the cognitive perspective I’ve sketched here has informed several, quite different research programs, and many prac­ titioners wouldn’t endorse all the claims I’ve made.101 My efforts are simply to show how a historical poetics of cinema of the sort exemplified in the essays that follow can usefully adopt some cognitive principles as a way to chart a film’s manifold effects. Sometimes, however, we can point to patterns and principles and purposes with­ out being able to specify effects very well. This is largely because we don’t have a tight theory allowing us to trace the consequences of artistic choices. Occasionally we may find constructional principles that don’t correlate with any plausible effect. The begin­ ning shot and final shot of Ozu’s Ohayo (1959) are reverse-angle views of one another (Figures 1.22-1.23). Given Ozu’s well-known interest in rejuggling objects and view­ points and his use of 180-degree cutting patterns, it’s likely that these two shots were planned to have the sort of regularity we find here.102 But when I point this out, some­ one will say quite reasonably that no one could perceive this pattern while viewing. Only by geeky concentration on Ohayo’s overall architecture can we detect it. My response is that a few filmmakers build their films as objects as much as experiences, as patterned constructs that may or may not fit snugly in a viewing unfolding across time. Like a poet who plants hidden acrostics or numerical codes, Ozu expects some 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. More generally, I’d suggest that we adopt the circuitparticulars-patterns-purposesprinciples-practices-effects as a default to guide our inquiry. Most of the time it will serve us well. In a few cases, we may have to register the possibility that a film’s orga­ nization can outrun its effects on any viewer. Yet by pointing out this state of affairs, the poetician may help the artist realize his or her design. Now that I’ve highlighted

54

P o e tic s o f C i n e m a

the reverse-angle shots separated by an entire film, viewers are free to be alert for them when they see Ohayo, or other Ozu films. At this point, a connoisseur’s appre­ ciation for tiny felicities becomes the most relevant effect, and of course this too can be fruitfully investigated as part of an art tradition to which Ozu belongs. Let me sum up. I’ve argued that broadly speaking, the central question of film poetics, posed as a methodological point of departure, can be understood in this way: 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, largescale 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 sixstep 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,

P o e tic s o f C i n e m a

55

empirical research, and theoretical explicitness, has worn fairly well. Despite errors of fact, thought, and judgment—not least my own—the conversation has advanced. By concentrating on particular questions and then comparing our reasoning and research with that of others asking congruent questions, we have begun to produce reliable knowledge about film.103 A historical poetics of cinema isn’t the only vehicle for this enterprise, but the essays that follow try to show that it can be a sturdy one.

fl

2

.

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

P o e tic s o f C i n e m a

58

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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

Figure 2.3 Class Relations (Jean-Marie Straub and Danièle Huillet, 1983).

Figure 2.4

59

Class Relations.

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

P o e tic s o f C i n e m a

purely artistic considerations, we can ask if they are not simply conventions. Any artistic device as widely used as shot/reverse shot, if not significantly motivated by perceptual equivalences, is likely to be seen as a stylistic convention. This presumption, I think, dominates film studies today. From this perspective, shot/reverse-shot cutting is an arbitrary device, having no privileged affinities with natural perception. But what is a convention, on this view? Minimally, I suppose, most contemporary scholars would say that shot/reverse shot is a convention because it is a piece of artifice, and because it must be learned. Most theorists are content to leave the matter there, but neither point really blocks the naturalist position. The naturalist position does not have to claim that shot/reverseshot editing is not artificial in some sense. After all, it is an invention; it was not present at the birth of cinema, and people decided to use it. Nor does the naturalist position have to deny the role of learning. Once we have learned to perceive the world, the naturalist might argue, we can learn to grasp artistic devices that provide equiva­ lents to the world. Accordingly, our ability to grasp those devices ought to ride upon the appropriate sorts of perceptual skills. To this, a contemporary theorist might reply that the typical convention is arbi­ trary. Here, arbitrariness must mean something like this: In principle, an indefinitely large number of other representations would serve as well; the one chosen is simply assigned that task by the rules of the langue or code in force. A dog might as easily have been called a chien or a hund. There are some problems with extending to nonlinguistic phenomena a conception of arbitrariness derived from the lexical items in a language. Consider the turn signals on a car. To an observer on Sirius, the fact that I flash the right signal when I intend to turn right might appear arbitrary. Other options are logically possible: People could signal a right turn by activating the left signal. But in fact such mechanisms are designed to fit our propensities to signal rightward movement by something that stands in a rightward relation to our body. Our nonverbal symbol systems, like our technical gadgets, are engineered to our fixed dispositions, including innate ones, and the choice among all possible options is not indifferent.7 A similar case can be made about the shot/reverse-shot technique. If the director seeks to represent two people looking at each other, it is less arbitrary to show them looking at each other than to show them, say, looking away from each other, or at the moon. A visual “code” that showed figures looking at each other in order to signify that they are not looking at each other would be bizarre in the extreme. We would, I think, be inclined to call that alternative code “arbitrary,” but not the normal case, which reflects naturalistic assumptions about the image’s representation of the state of affairs. For creatures like us, the two options are not equiprobable. Nevertheless, the naturalist’s position on shot/reverse shot remains problematic because of the undeniably “unrealistic” qualities present in orthodox uses of the device. And something theoretically stronger is probably required to allay the conventional­ ist’s worries. At this point, I want to suggest a middle way between the two positions, one that captures the intuition that such visual devices are constructed and signifi­ cantly artificial while also preserving the idea that they are not utterly arbitrary.

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

61

Primary Theory and a Continuum of Conventions In contrasting the two views of shot/reverse shot, I followed precedent in distin­ guishing something called nature from something called convention. The first step in forging a more comprehensive theory, I believe, is to discard these notions and offer some more flexible concepts in their place. The term nature comes to us fraught with connotations. To most film theorists, it suggests either biologically innate capacities or universal laws operating in the physical world generally. It also suggests the realm of necessity, that which cannot be changed by human will or skill. Such conceptions of the “natural” have been frequently attacked by structuralist and poststructuralist theorists, who insist that all signification is constructed, conventional, and culture bound. Still, only dogmatists would deny that representation, especially visual representation, relies at least partly on the perceiver’s psychophysical capacities. It seems very unlikely that our ability to perceive humans and objects in images owes nothing to our biologi­ cal heritage. Our understanding of images could hardly be unconnected to our capaci­ ties to move through a three-dimensional environment and to recognize conspecifics. The individual’s development of language, according to the most powerful theories now available, is as much a biological capacity as the inclination to grow arms rather than wings.8 Certain relevant abilities may not even be species specific: Pigeons and monkeys respond to photographs as if recognizing the sorts of things represented.9 Nonetheless, I propose that we can make some progress if we bypass the natureculture couplet for the moment and concentrate upon some “contingent universais” of human life. They are contingent because they did not, for any metaphysical reasons, have to be the way they are; and they are universal insofar as we can find them to be widely present in human societies. They consist of practices and propensities that arise in and through human activities. The core assumption here is that given certain uniformities in the environment across cultures, humans have in their social activities faced comparable tasks in surviving and creating their ways of life. Neither wholly “natural” nor wholly “cultural,” these sorts of contingent universais are good candi­ dates for being at least partly responsible for the “naturalness” of artistic conventions. Paradigm cases of contingent universais would seem to be practical skills such as the ability to use language for communication, to divide labor tasks, to distinguish between living and nonliving things, and so on. I have stated these rather generally; it is an empirical question as to whether there are not much more specific contingent universais, such as recognizing focal colors or taking turns during conversation.10 I have stressed contingent universais as involving behavior, but it seems likely that they constitute a conceptual frame of reference as well. The anthropologist Robin Horton calls such a framework “primary theory” and characterizes it as follows: Primary theory gives the world a foreground filled with middle-sized (say between a hundred times as large and a hundred times as small as human beings), enduring, solid objects. These objects are interrelated, indeed, interdefined, in terms of a ‘push-pull’ conception of causality, in which spatial and temporary contiguity are seen as crucial to the transmission of change. They

62

P o e tic s o f C i n e m a

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 universal, 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

63

appropriate to certain ends than others. If I am a film director and I want spectators to study an actors 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 universais, 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

P o e tic s o f C i n e m a

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 universais, instances of Horton’s “primary theory.” As in our case of face-to-face interaction in the shot/reverse shot, these contingent universais 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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.19Although 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

P o e tic s o f C i n e m a

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 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 naturalist-conventionalist couplet. In order to achieve certain effects, artists may tap biological propensities and contingent universais; in order to achieve other effects, artists may invoke more localized and recondite skills. Does all this ascribe too little a role to culture? I think not. For one thing, as Gombrich has often pointed out, it is culture that generates the tasks and interests that shape the ways in which visual effects are manipulated. If conventions are relations of means and ends, the social purposes of a representation necessarily govern how the activity is conducted, how the first region of effects is formed into more complex ones. Moreover, the centrality of artistic schemas—those inherited patterns and formulas through which the artist achieves effects in the medium—assures that culture plays a central role. “Only where there is a way is there also a will,” Gombrich notes. “The individual can enrich the ways and means that his culture offers him; he can hardly wish for something that he has never known is possible.”21

A. Package Deal One advantage of tracing out this gamut of factors is that it allows us to see that the most intuitively obvious phenomena of visual representation seldom fall neatly on one point of the continuum. It is unlikely that a single image, or a design system like linear perspective, or even a technical device like shot/reverse-shot editing will constitute a pure instance of any sort of effect along this spectrum.22 As critics and historians we inherit a language of practical craft, and this set of terms has other ends in view than the distinctions that produce the continuum I’ve sketched. We ought to expect that a typical representational device will present a bundle of effects of different sorts. Any given technique may call upon sensory triggers (such as color contrast to indicate contours), contingent universais such as identification of the human figure, easily learned effects such as outline drawing, and more complex skills, such as that of identifying allegorical figures. Because most of the technical devices we encounter package many sorts of appeals 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 specific and more difficult to pick up. That’s part of what we mean by understanding something “because of its context.” In the representational package we’re offered, the

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

Figure 2.5

Yaaba (Idrissa Ouedraogo, 1989).

Figure 2.6

67

Yaaba.

more contingently universal cues lead us to make sense of more esoteric cues in partic­ ular ways. This would obviously facilitate learning: Not only do we need little exposure to certain effects, but also, in each image, the universal factors reinforce our hypoth­ eses about the proper reaction we should have to the more culturally specific ones. I suggest that shot/reverse shot is best considered along these lines—as a compos­ ite phenomenon, drawing on features from various regions of the continuum. I can’t itemize all the relevant cues here, but let me make a start. In its prototypical form, shot/reverse shot is predicated on a two-person, face-toface encounter. This phenomenon would seem to be a good candidate for a contingent universal of social intercourse—something that would be intelligible across cultures and periods. This consideration is so rudimentary that neither the naturalist nor the conventionalist position on shot/reverse shot deems it necessary to weigh it, but in my 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 between the young protagonist and his elder, and they do so through a prototypical shot/reverse-shot construction. The pattern may capture other contingent universais at work as well. Conversa­ tional turn taking, with its interchanging role of speaker and listener, might furnish an approximate structure for the alternation of images we get in shot/reverse shot. Indeed, it would seem likely that historically this alternating editing grew out of an effort to capture the turn-taking phenomenon in cinematic form. Another important cue, at least in the prototypical instance, is the glance of the persons represented on the screen. Noël Carroll has suggested that the informational saliency of eye move­ ments in primates gives filmmakers a powerful opportunity to engage audiences cross-culturally.23 We don’t lack testimony from filmmakers that eyes matter. Here is J. J. Abrams, director of Mission: Impossible III (2006): “No matter how many trucks and trailers are at base camp, it’s ultimately about those few actors—those eyes, what’s being conveyed emotionally.”24 In the terms I have proposed, the direction of the glance would function as a sensory trigger, informing us of the object of the person’s attention. It stands as another cross-cultural regularity of human activity that can elicit effects in beholders.

68

P o e tic s o f C i n e m a

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 human 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 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 filmmaking 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 from her. We shouldn’t think of this camera position as providing 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). As for the instantaneous change of view that is said to create the “ubiquitous” or “ideal” observer, this would seem to be a special case of the immediate leap in time or space caused by any cut, of any sort. And once spectators, presumably from a very young age, have acquired the skill of taking a cut to signal such a shift in orienta­ tion, the other cues present in shot/reverse shot may suffice to motivate the distinct changes of angle.26 There are doubtless other cues that are ingredient to the shot/reverse-shot device, such as the more localized norm that the figures will be observed from the same side 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 human 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

Figure 2.7

Suvorov (Vsevelod Pudovkin, 1941).

Figure 2.8

Figure 2.9

The Land (Youssef Chahine, 1969).

Figure 2.10

69

Suvorov.

The Land.

be impossible to grasp the purpose of the camera positions and editing. In a meta­ phorical sense, the prototype of shot/reverse shot is constructed out of such contin­ gent universais: It is a refined elaboration of them, a piece of artifice serving cultural and aesthetic purposes. The multiplicity of those purposes can best be grasped, I think, if we turn to a final issue. Gombrich has argued that the history of style in the visual arts is usefully understood as a process of schema and revision. The artist takes an available pattern and recasts it in the light of the capacities of the medium, the purposes she or he has 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).

P o e tic s o f C i n e m a

70

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). John 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, René 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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

Figure 2.20

The Cloak.

Figure 2.23 1931).

Figure 2.24

À Nous la liberté.

À Nous la liberté (René Clair,

71

P o e tic s o f C i n e m a

72

Figure 2.25

Nosferatu (F. W. Murnau, 1922).

Figure 2.26

Nosferatu.

Figure 2.27 Muriel (Alain Resnais, 1963): The medium shot of Hélène s client is 17 frames long (0.7 seconds).

Figure 2.28 Muriel: The next shot, show­ ing the client’s coat button, is 12 frames long (0.5 seconds).

Figure 2.29 Muriel: The next shot, of her hat, lasts only 11 frames (0.45 seconds).

Figure 2.30 Muriel: The “reverse shot” of 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

73

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 universais as its bridgehead. At the same time, it presupposes some norms, however easily learned they may be.

Contingent Universais and Us 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 IVe 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 human 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 threedimensional 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 universais. 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

P o e tic s o f C i n e m a

74

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 nurture.. . . 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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 human 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 universais, 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 universais 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 2izek 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 Touis Feuillade, is not talking about crosscultural 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

P o e tic s o f C i n e m a

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 “human 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.37 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 crosscultural 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 universais.” 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 universais 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 universais (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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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 universais 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,

78

P o e tic s o f C i n e m a

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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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 Fictions 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 universais 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 universais of human makeup and experience, but most scholars in the humanities tend to doubt the existence (or the

80

P o e tic s o f C i n e m a

importance) of empirical universais. 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, rationalempirical 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 “naïve 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

C o n v e n tio n , C o n s tr u c tio n , a n d C in e m a tic V is io n

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. }. 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

P o e tic s o f C i n e m a

to the ecological view, cinematic representation relies on a great many nonconventional 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 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 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.

II Studies in Narrative

83

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

Figure 3.1 A Trobriand Island string figure: The headhunters face one another.

P o e tic s o f C i n e m a

Figure 3.2 When one fighter wins, he departs 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).1In Australian Aboriginal sand paintings, what might seem to outsiders to be abstract squiggles and whorls represent mythical events or incidents from daily life.2 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.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 him 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 him 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 identihed 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 him, 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 hgure 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

P o e tic s o f C i n e m a

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 its 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?” Narratologists 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 actioncentered 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.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

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.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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.5No 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 presented from the point

92

P o e tic s o f C i n e m a

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? Tots 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?

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 Ive 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 openended. 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

P o e tic s o f C i n e m a

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 presumably contribute a lot to your inquiry. Indeed, this is just what’s happening. After I floated this comprehension-centered account of narrative in the mid-1980s, several scholars who wanted to pursue questions about emotional response built upon narrational 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? Moreover, because comprehension involves such features as tracking psychological states, causality, time shifts, and the like, the onus is on the critic to show that women 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 settlers. Even this apparently apocryphal anecdote, however, doesn’t damage my case. I assume that the audience understood the story—that the settlers were crossing Indian land, that the Indians wanted to wipe out the settlers—and that the viewers took sides in a way not anticipated by the film’s makers. To say that there’s conver­ gence in understanding is not to say that all spectators act upon their understanding in the same ways. By focusing on comprehension from a mentalistic perspective, I hope to adhere to other conditions I set out earlier. In accord with my layout of spectatorial activi­ ties, I assume that there’s a fair amount of convergence in viewers’ understanding of 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 that viewers bring to films. But divergences in comprehension aren’t anything like as wide and varied as we’d find in interpretations, for reasons I’ve already suggested. Again, in accord with the sort of poetics I’m proposing, this study of narrative treats films holistically. My conceptions of narration, plot structure, and the story world try to take into account the overall form of a film. The assumption here is that regularities we find across the whole artifact allow us to make inferences about the purposes of its makers and the activities coaxed from its viewers. Take the openings we find in ordinary movies. Very often we get an expository title giving time and place, along with one or more long shots of an area of action. Cuts or camera move­ ments may carry us into a scene, with characters moving toward us, or tracking shots

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

Figure 3.3 The opening of The Silence of the Lambs (1991): Agent Starling comes out of the story world to meet us.

95

Figure 3.4 The Silence of the Lambs: She turns and runs into the forest; the camera follows and carries us into the story action.

Figure 3.5 At the end of The Silence of the Lambs, Flannibal Lecter turns from us and follows his prey into the story world, but we stay behind. This is a conventional mark of closure.

that follow a character from behind in exploratory fashion (Figures 3.3-3.4). On the soundtrack, music sets a mood, and dialogue rises to audibility. Clearly all these tac­ tics are blended to engage the spectators interest, parceling out information needed to understand the action. The cut-ins or forward camera movements also suggest that we are being drawn gradually into the story world. Strikingly, the ending of an ordinary movie often reverses these devices. The camera pulls back, characters turn away and we don’t follow them, doors and gates may shut, the music rises again, and titles may appear (Figure 3.5). The opening literally opens up the movie and lets us in; the closing shuts it down and expels us. The best explanation for these regularities onscreen is that they’re manifesting principles that filmmakers share, perhaps tacitly, and they function to shape our experience of the story. The symmetries between openings and closings suggest that narration is a system that’s put into motion across the whole film. All the factors we normally associate with narration—play with the order of events, shifts in point of view, and voice-over commentary—shape our overall experience. They’re not just one-off tactics; they play roles in larger patterns running across the entire movie. So once we’ve identi­ fied a passage of omniscient narration or optical point of view, we should go on to look at how that functions in the broader organization of the narration. Why shift to optical point of view here? How does it shape the experiential logic of the overall film? Narration is more than an armory of devices; it becomes our access, moment by moment, to the unfolding story. A narrative is like a building, which we can’t grasp

P o e tic s o f C i n e m a

96

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.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 boyIwoman!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

P o e tic s o f C i n e m a

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 offabula, 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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 offabula events, providing a flashback or flashforward. It can manipulate/a&M/a 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-ofview 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.”8No 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.

P o e tic s o f C i n e m a

100

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 Gérard 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. Tikewise, 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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 héros, 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

P o e tic s o f C i n e m a

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 thefabula 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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 syuzhefs pattern of actions according to some macrostructural 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.17Alternatively, 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

P o e tic s o f C i n e m a

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 parts.20 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.21 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

T H r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

the goals. The Climax is often followed by an Epilogue, which asserts that a stable situation has been achieved. Thompsons 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,

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

and falls among aimless neighborhood kids. This portion shows the father, Furious Styles, trying to keep Tré 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, Tré 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. Tré is working as a salesman, Ricky is a football star, and Doughboy flits in and out of prison. New goals are established. Tré, Ricky, and Très girlfriend Brandi hope to go to college. Ricky is recruited by the University of Southern California, Tré tries to convince Brandi to have sex, and Tré 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 Tré, 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. Très 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, Très 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 Tré 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 Tré has a change of heart and leaves his friends, who go on to wipe out the gang members. Next morning, Doughboy cracks his tough façade to confess to Tré 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 Tré 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,

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 threeact 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-Tara,

110

P o e tic s o f C i n e m a

François Truffaut, and Andre Téchiné 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 threeact / 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.

The Narrative AX/orld 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 tion—restricted point of view, flashback construction, and the like.31 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 isn’t 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-

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 him 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 “naïve realism” or “illusionism,” but it isn’t. Sophisticated artifice is responsible for these effects of easy inference. It’s just that

112

P o e tic s o f C i n e m a

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 excited imagination of the hearer to cooperate with them, creating for itself.37

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

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 hero’s 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 middleaged), 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.41 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 plotdriven movies that have to announce the heroes and villains swiftly? Not necessarily.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

115

Michael Newman has shown that even the Uncalled 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

P o e tic s o f C i n e m a

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 Danièle 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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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.51This 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

P o e tic s o f C i n e m a

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 something.. . . Hollywood is sustained on the illusion that human 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 Brunetière. “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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 Ive 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

P o e tic s o f C i n e m a

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 lepers5 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.

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

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. André Gaudreault and François 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 been wounded, if they wanted to shoot him.”63 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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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

P o e tic s o f C i n e m a

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 some principles 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,

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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 well 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 recountedflashback 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

P o e tic s o f C i n e m a

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 didn’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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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—

P o e tic s o f C i n e m a

128

Narrative text Real author

►(Narrator)^, (Narratee), ■h Implied author 1

Implied ' reader

Real “T reader ï

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 agent.. . . 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)

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

129

and image-based ones (mise-en-scène, 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 communiqués. 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 plainspoken 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

P o e tic s o f C i n e m a

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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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?79My response in the 1985 book remains: These terms are shorthand metaphors and constitute merely a façon de parler. 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

P o e tic s o f C i n e m a

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

T h r e e D im e n s io n s o f F ilm N a r r a tiv e

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.

í

h.

Cognition and Comprehension V ie w in g a n d F o rg e ttin g in

M ild r e d P ie rc e

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 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

P o e tic s o f C i n e m a

psychoanalysis have returned to the “classic” structuralist semiotics of the 1960s and early 1970s, or even to traditional film aesthetics.3 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 nowconscious 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 genrebased or stylistic conventions are learned and applied through processes exercised

C o g n it io n a n d C o m p r e h e n s io n

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 filmmaking 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

P o e tic s o f C i n e m a

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 the story action. (In some films, such as Fritz Lang’s You Only Live Once [1937] and Otto Preminger’s Fallen Angel [1945], such ellipses are later revisited and reveal that the narration skipped over important information.) Finally, I suggest that all these factors vary historically and culturally. We ought to expect that different filmmak­ ing traditions, in various times and places, will develop particular norms, schemas, and cues. Correspondingly, the inferences and hypotheses available to spectators will vary as well. My outline is very skeletal, so I’ll try to put some flesh on the bones by considering a concrete case. My specimen is Mildred Pierce, an instantiation of that vast body of norms known as the classical Hollywood cinema.8 I’ll be concentrating on its system of narration, which involves not only its construction of a plot and a diegetic world, but also its use of film technique. First, I’ll try to show that the film utilizes norms of narration so as to encourage not one but two avenues of inference and hypothesis testing; both of these would seem to have been available to contemporary audiences. Second, I want to show that the film assumes that in the viewer’s effort after gist, she or he will ignore or forget certain stylistic norms. That is, Hollywood norms posit a hierarchy of impor­ tance, with narrative gist at the top and local stylistic manipulations subordinated to that. In Mildred Pierce, this hierarchy allows the filmmakers to conceal crucial narrational deceptions.

Two Methods of Murder Because Mildred Pierce opens with a murder, it’s profitable to start our inquiry with a norm-based question. What kinds of options were open to filmmakers in the 1940s who wished to launch their plot with such a scene? In the early 1940s, the options were essentially two. One is exemplified by the second scene of The Maltese Falcon (1941). Here, the murder of Sam Spade’s partner, Miles Archer, is rendered in a way that conceals the killer’s identity (Figure 4.1). We see the victim from over the killer’s shoulder, but a reverse-shot view of the murderer

C o g n itio n a n d C o m p r e h e n s io n

Figure 4.1 The Maltese Falcon (1941): the unseen killer.

Figure 4.2 victim.

139

The Letter (1940): the unseen

isn’t supplied. The film thus poses the question of who killed Archer, and this creates one strand in the overall mystery plot. A second normative option is exemplified at the very start of The Letter (1940). Here the shooting of a colonialist is plainly committed by the Bette Davis character (Figure 4.2). The question posed is now that of why she killed him. What, if any, circumstances justify the crime? The first two scenes of Mildred Pierce, however, offer a more complex case. In a lonely beach house at night, with a car idling outside, a man is shot by an unseen assailant. As he dies, he murmurs, “Mildred.” We glimpse a woman driving off in the car. In the next scene, our protagonist, Mildred Pierce, is seen wandering along a deserted pier. We couldn’t ask for a better example of a film that lures us down inferential pathways. In a remarkably brief time—the murder scene lasts only 40 seconds—the spectator has accomplished a great deal. She has perceptually constructed a diegetic world—a beach house at night, peopled by two characters. Further, she infers that a homicide has taken place; that gist is central to understanding this narrative. Only a little less probable is the inference that the killer has fled by car. And the viewer may also have inferred that the murderer is the woman named in the film’s title. Yet such inferences are not one-time-only products. They form the basis of hypoth­ eses, which lead in turn to further inference making. As Meir Sternberg points out, 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 that there are prior reasons for the murder of Monte and that the film will reveal them in its progressive unfolding. As a mystery film, Mildred Pierce will, so to speak, create suspense hypotheses about how curiosity hypotheses will be confirmed. We can specify two primary inference chains that this opening prompts. One is that Mildred is the killer. Most critics have assumed that the average spectator comes to this conclusion, and they characteristically take the opening as carefully directing us to form this inference. First, like the Maltese Falcon sequence, the scene does not show who fires at Monte; this poses the question of the murderer’s identity. Moreover,

140

P o e tic s o f C i n e m a

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 just killed the man we duly learn is her second husband.11 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

C o g n it io n a n d C o m p r e h e n s io n

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. Turing 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.14Roughly, then, a character consists of some person-like features plus the social roles that she or he fills. This

142

P o e tic s o f C i n e m a

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 selfsacrificing 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.

C o g n it io n a n d C o m p r e h e n s io n

143

Confirming that Mildred committed the murder would clinch the trusting viewers 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.

The 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

P o e tic s o f C i n e m a

144 TABLE 4.1

Mildred Pierce: The Opening Scene and Its Replay

Opening Shots (A)

Replay Shots (B)

1. 5 seconds: (extreme long shot) A beach house at night; a car is visible alongside (Figure 4.3). Dissolve to:

1. 12 seconds: (medium shot) Mildred goes to the car and tries to start the engine (Figure 4.12).

2. 4 seconds: (long shot) House and car. Two pistol shots are heard (Figure 4.4).

2. 4 seconds: (medium close-up): Mildred slumped over the steering wheel. Two shots are heard (Figure 4.13).

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).

3. 5 seconds: (medium shot) Veda fires four times (Figure 4.14).

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).

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) Empty parlor, with Monte’s corpse in the firelight. Doorway is empty (Figure 4.10).

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. (long shot) Car outside pulls off (Figure 4.11).

6-10. In the parlor, Veda tells Mildred lies about shooting Monte. Near the 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 option for the ordinary film viewer (at least, until the arrival of home video, and even 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 highly redundant. But in learning this, filmmakers have also learned how to prompt m«remembering. Given our effort after gist and our inability to turn back to check a point (especially one made 90 minutes earlier), the film can introduce both redundant cues and highly nonredundant, even contradictory, ones. Table 4.1 aligns the two sequences, the opening murder, (labeled A) and the climactic replay of the shooting (B).

C o g n it io n a n d C o m p r e h e n s io n

145

•h #

Figure 4.3 Mildred Pierce (1945): the opening scene, shot 1.

Figure 4.4 shot 2.

Figure 4.5 shot 3.

Figure 4.6 Mildred Pierce: the opening scene, shot 3 (continued).

Mildred Pierce: the opening scene,

Figure 4.7 Mildred Pierce: the opening scene, shot 3 (concluded).

Figure 4.8 shot 4.

Mildred Pierce: the opening scene,

Mildred Pierce: the opening scene,

Many actions are reiterated in the second version, and the redundancies suggest that we are seeing a straight replay. In the opening scene, over the second long shot of the beach house (A2, Figure 4.4), we hear two gunshots. The cut inside to Monte facing the killer (A3, Figure 4.5) comes right on the third pistol shot. In the flashback version, the cut from Mildred in the car (B2, Figure 4.13) comes at exactly the same point. The next shot (B3, Figure 4.14) replaces the image of Monte with that of Veda firing the revolver. No time can be said to be omitted here. More subtly, the screen time

P o e tic s o f C i n e m a

146

Figure 4.9 Mildred Pierce: the opening scene, shot 4 (concluded).

Figure 4.11 shot 6.

Mildred Pierce: the opening scene,

Figure 4.13 Mildred Pierce: the final flashback, shot 2.

Figure 4.10 Mildred Pierce: the opening scene, shot 5.

Figure 4.12 Mildred Pierce: the final flashback, shot 1.

Figure 4.14 Mildred Pierce: the final flash­ back, shot 3.

that elapses between the third gunshot and Monte’s dying word is virtually identical in both versions (9 seconds and 10 seconds respectively). Finally, the slamming door we hear in the opening scene (A4, over Figure 4.9) is revealed to be not the killer leav­ ing, as we initially inferred, but rather the sound of Mildred entering to find Veda in the living room (B5, Figure 4.18). These are what the mystery novelist might consider fair misdirections of the spectator’s attention. They suggest that the second version is identical with the first, except that the former fills in certain details of the latter.

C o g n itio n a n d C o m p r e h e n s io n

Figure 4.15 Mildred Pierce: the final flashback, shot 4.

147

Figure 4.16 Mildred Pierce: the final flash­ back, shot 4 (continued).

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 view of the first scene, the apparent emptiness of the car suggests that the important action occurs inside the house. If anyone should recall that scene 100 minutes later, the 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 and speaking her name. 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). There is no close-up, and no pause either. He says nothing when he is in the position he assumed in the earlier scene (just as earlier, he said nothing when he rolled over). The second version produces a different effect. By speaking when he is not looking toward his killer, he no longer seems to be naming the culprit but rather recalling Mildred. In this flashback, she no longer seems guilty. The narration gets two distinct cues out of the two versions, and it is able to do so because it counts on our remembering only

148

Figure 4.17 Mildred Pierce: the final flashback, shot 4 (concluded).

P o e tic s o f C i n e m a

Figure 4.18 Mildred Pierce: the final flash­ back, shot 5.

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

C o g n itio n a n d C o m p r e h e n s io n

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 scenes 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

P o e tic s o f C i n e m a

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 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 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 Noël Carroll, Murray Smith, Ed Tan, and others suggest that a cognitive perspective can enrich our understanding 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. In Narration in the Fiction Film, I try to show that the activity of inferential elabo­ ration is prompted by melodramas (In 7his Our Life, Say It With Songs), Westerns (Wild and Woolly), comedies (His Girl Friday), and straight dramas (Heaven Knows, Mr. Allison). Every narrative of any complexity withholds some story information from both viewers and characters. This creates gaps in our knowledge, disparities among various characters’ states of knowledge, and mismatches between a character’s knowledge and the viewer’s knowledge, all the while generating Meir Sternberg’s response trio of curiosity, suspense, and surprise. Every film’s narration depends upon regulating the flow of information, and we don’t have perfect information until the end (if then). In this respect, every narrative harbors secrets. There is much more to understand about how viewers understand films. The line of inquiry sketched here puts a priority on studying particular films in the light of how narrational and stylistic processes are designed to elicit certain spectatorial effects. In this research program, Mildred Pierce exemplifies key features of the classical Holly wood film. There are, of course, other traditions that call on differ­ ent sorts of narrational cues, schemas, and norms.24 Comparative inquiry into these traditions can contribute to that research program I’ve called a film poetics. By avoid­ ing misplaced conceptions of codes or slippery analogies between film and language, the cogn itive perspective offers a robust account of the viewer’s activity, one that can guide a historical poetics of cinema.

5

.

The Art Cinema as a Mode of Film Practice

La Strada (1954), 8 1/2 (1963), Wild Strawberries (1957), The Seventh Seal (1957), Persona (1966), Ashes and Diamonds (1958), Jules et Jim (1962), Knife in the Water (1962), Vivre sa vie (1962), Muriel (1963): Whatever else one can say about these films, cultural fiat gives them a role altogether different from Rio Bravo (1959) on the one hand and Mothlight (1963) on the other. They are “art films,” and, ignoring the tang of snobbishness about the phrase, we can say that these and many other films constitute a distinct branch of the cinematic institution. My purpose in this essay is to argue that we can usefully consider the “art cinema” as a distinct mode of film practice, possess­ ing a definite historical existence, a set of formal conventions, and implicit viewing procedures. Given the compass of this paper, I can only suggest some lines of work, but I hope to show that constructing the category of the art cinema is both feasible and illuminating. It may seem perverse to propose that films produced in such variable cultural contexts might share fundamentally similar features. Yet I think there are good reasons for .believing this, reasons which come from the films’ place in history. In the long run, the art cinema descends from the early film d art and such silent national cinema schools as German Expressionism and Neue Sachlichkeit and French Impressionism.1 (A thorough account of its sources would have to include literary modernism, from Proust and James to Faulkner and Camus.) More specifically, the art cinema as a distinct mode appears after World War II when the dominance of the 151

152

P o e tic s o f C i n e m a

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, Authorship, Ambiguity 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 causeeffect 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 L’Avventura (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

T R e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

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 Mépris, 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 (LAvventura; Blow-Up, 1966; High and Low, 1963), even the making of a film (8 1/2; Le Mépris; 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

P o e tic s o f C i n e m a

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), L’Avventura, Deserto rosso (1964), and Une femme mariée (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 8 1/2 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 À 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. André 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.

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

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 cinéma 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 artcinema 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.7In 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 Mépris—are the least capable of realistic justification. Contrariwise, to push the

156

P o e tic s o f C i n e m a

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.

T A e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

157

The Art Cinema 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

P o e tic s o f C i n e m a

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 L’Année dernière â Marienbad (1961), Resnais dissolved causality altogether and used the very conventions of art cinema to shatter the premise of character subjectivity. In Nicht versöhnt (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 que je sais d ’elle [1967] can be seen as a critique of Deserto rosso, or even of Une femme mariée.) 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

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

159

prefer the deaner 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 cant 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 nations 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 François 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, Tars 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 make 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

P o e tic s o f C i n e m a

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 Fédération Internationale de la Presse Cinématographique (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

T h e A r t C i n e m a as a M o d e o f F ilm P r a c tic e

161

on Bergman and Antonioni. Brazil’s Cinema Nôvo 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 Tam’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 cinéastes, 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 cinéphile, 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 Café Lumière, 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

162

P o e tic s o f C i n e m a

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 girls 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 2/2 (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

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

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 halfhour 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 cinéma 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 Danièle Huillet and Jean-Marie Straub, such as Vie 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

164

Figure 5.2 Manuel de Oliveira uses the planimetric 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.

P o e tic s o f C i n e m a

Figure 5.3 The flattened image suggests a family portrait gone wrong in Terence Davis’ Distant Voices, Still Lives (1988).

Figure 5.5 In Maborosi (1995), Hirokazu Kore-eda mutes the drama of a drunken hus­ band returning home. The distant, perpen­ dicular framing accentuates the slight gesture 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

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

165

Figure 5.7 A Time to Live and a Time to Die (1985): With the family gathered, the griev­ ing mother buries her face in the dead father’s blankets.

Figure 5.8 A Time to Live and a Time to Die: She is drawn up and back by her relations, her movement and the panning camera revealing one of the sons against the back wall.

Figure 5.9 A Time to Live and a Time to Die: The mother is pulled back into the adjacent room, and while she’s visible in the distance, her sons kneel in the foreground. Hou Hsiaohsien’s long lens creates layers of space that are dynamically blocked and revealed by small movements of the characters.

Figure 5.10 In Peacock (GuChangwei, 2004, China), the telephoto full shot recalls not only the planimetric image but also the staggered staging arrangements of Fiou Hsiao-hsien.

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,

166

Figure 5.11 Although it contains nearly 1,000 shots, Simon Staho’s Day and Night (2004, Denmark) employs only two camera positions, one showing the car’s driver, and the other showing the passenger seat. Any action taking place outside the car is viewed from those fixed angles.

P o e tic s o f C i n e m a

Figure 5.12 Vagabond (aka Sans toit ni loi; 1985): The most peculiar of the initial “reverse shots” of the police investigation.

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 This game of form isn’t present in every film that belongs to the mode, but it does reappear.28 A striking instance is Agnès 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 her.. . . 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

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

167

Figure 5.13 Vagabond: The garage owner begins his tale.

Figure 5.14 Vagabond: The garage owner ends his tale, annoyed.

Figure 5.15 housework.

Figure 5.16 Vagabond: Yolande recalling seeing Mona and David.

Vagabond: Yolande halts her

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 reverseangle 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

168

P o e tic s o f C i n e m a

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 of her life, Mona’s gaze brushes across the camera axis.

Figure 5.20 Vagabond: Assoun after kissing 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.

T h e A r t C i n e m a as a M o d e o f F i l m P r a c t i c e

Figure 5.21 Caché (Michael Haneke, 2005): The central family’s apartment house, seen on video. But who is watching the tape?

169

Figure 5.22 Caché: A nearly subliminal shot follows. Where does this image belong in the story chronology?

At the dose, the tale of Mona’s wanderings retains its episodic quality, and her motives remain enigmatic. We construct this art cinema fabula 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, Varda 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 planimetrically 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 Caché (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.

fi.

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 fork.. . . Tn 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 Ts’ui 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

172

P o e tic s o f C i n e m a

moral or ideological position seem inevitable. Conventional techniques like fore­ shadowing and “backshadowing,” he suggests, build a grim determinism into a plots 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 gam 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.

F ilm F u tu re s

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,

174

P o e tic s o f C i n e m a

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’t fit that 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 réintroduction of viewer-friendly devices in the context of what might seem to be ontologically or epistemically radical possibilities.7

Rules of tKe Game Here are seven conventions of forking-path tales. 1. Forking paths are linear. In principle, as Borges’ Tsui 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

F ilm F u tu re s

175

where our fate lies.”8 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 girl. . . 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 roJ?s 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

P o e tic s o f C i n e m a

Figure 6.1 Run Lola Run (1998): Lola brushes against a protesting woman as she races along.

in Hong Kong and avoiding Bos 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 wouldn’t 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 that. . 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

F ilm F u tu re s

Figure 6.2 Wong’s wristwatch: the opening shot of Too Many Ways to Be No. 1 (1997).

177

Figure 6.3 Too Many Ways to Be No. T. 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 forkingpath 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

P o e tic s o f C i n e m a

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, Helens 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 Blind 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 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 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 not been shown in any of the replays of his race through the station—and in the third story they end up marrying. Finally, the three stories in Blind Chance are linked by certain pervasive social conditions. In each future, Witek is involved directly or indi­ rectly with the unofficial student movement and their underground publications. In his Communist career, he turns a blind eye to the movement; in his Catholic career, he is an activist within it, helping print the leaflets; and during his medical career, he must replace his mentor, who is fired because his son is involved with the movement. Recurring characters and background conditions render widely divergent futures more cognitively coherent. 4. Forking-path tales are unified by traditional cohesion devices. By cohesion devices, I mean formal tactics that link passages at the local level— from scene to scene or from one group of scenes to another. The classical narrative cinema of Hollywood and the narrational strategies characteristic of art cinema have developed many such tactics to aid the viewer’s comprehension. We find them in the forking-path tales as well, usually serving to tighten up linear cause and effect. Two primary cohesion devices of mainstream cinema are appointments and deadlines, and our forking-path movies provide these aplenty. Run Lola Run is built around a looming deadline: If Lola doesn’t meet Manni by noon, he’ll try to rob a supermarket to get 100,000 marks. Sliding Doors is structured around a cascade of appointments: in one line of action, the appointments necessary to find Helen a new job; and in the other, the dates she makes with James, the man who attracts her after she leaves Jerry. In Too Many Ways to Be No. 1, each of Wong’s alternative futures

F ilm F u tu re s

Figure 6.5 Blind Chance (1981): Witekscreams in the first shot.

179

Figure 6.6 Blind Chance: The second shot: Casualties in the hospital.

Figure 6.7 Blind Chance: The last shot, linking to the first (Figure 6.5).

hinges on appointments (with the Mainland and Taiwanese gangs) and deadlines (chiefly, in the second plotline, the one pushing Matt to kill rival triad bosses). As we might expect, our “art movie” Blind Chance is somewhat looser at this level, relying more on the sheer successiveness of events and leaving appointments and deadlines offscreen. In the second story, for instance, Witek’s childhood friend Daniel appears at a meeting of the underground students’ organization, along with his sister Vera. Witek’s subsequent romance with Vera is shown in brief scenes of them meeting on the street, or spending time together in his apartment. These scenes, like Vera’s departures by train, aren’t set up by explicit appointments, though such arrangements must have been made. Indeed, when the couple split up, it’s because of not making an appointment (Witek is told she’s gone to Lodz, but actually she’s waited outside his apartment for hours before finally leaving). As I suggested in the previous essay, this loosening of causal and temporal bonds is characteristic of much ambitious filmmaking in Europe after World War II. Yet in Blind Chance, cohesion operates from another angle. The film opens with an enigmatic prologue showing Witek sitting in a train or airplane seat, facing us and starting to scream (Figure 6.5). The credits unroll over his howl. After the credits, we see an enigmatic image of a hospital emergency room, with a woman’s leg in the foreground and a bloody corpse hauled away in the background (Figure 6.6). Only at the end of the film will these images make sense: In the final story, Witek is aboard a plane to Paris and it explodes in midair; this is the last image of the film, over which the final credits appear (Figure 6.7). Now we can place the opening shot of his shriek—presumably, his last moments—and we can understand that it is apparently

180

P o e tic s o f C i n e m a

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 were 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, Tola 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.

F ilm F u tu re s

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 café 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 café 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

P o e tic s o f C i n e m a

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.

F ilm F u tu re s

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 plots 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

P o e tic s o f C i n e m a

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 of JerryHelen-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 consummatory 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.

F ilm F u tu re s

185

Some 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 hanahi, 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 off them. The plot may initially set out two parallel futures but then concentrate on just one, bring­ ing the other one onstage at intervals (Me MyselfI [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. Tuxuriant 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 Schrodinger’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

P o e tic s o f C i n e m a

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 story “Quenby and Ola, Swede and Carl.”16 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 café 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 timetravel 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

F ilm F u tu re s

187

idea was implicit in Schrödingers 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.201 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 Mu t ua 1Friends 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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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, João Botelho, Alejandro González Inárritu, 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! (Elungary), 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

P o e tic s o f C i n e m a

Protagonists and 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.”4Aren’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’auberge 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,

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

193

Tip IIIIb IA; “A. i

i

m

IT

*

ill

‘ " “ -A-.* »JsP

w

i»awr

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

P o e tic s o f C i n e m a

This 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 convergingfates 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 Cela’s The Hive (La Colmena, 1953) probably constitutes the record, at least in sweep, with its 300 characters passing through a neighborhood café. (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 Were 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-

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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

P o e tic s o f C i n e m a

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.14Social 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 naive 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

M u t u a l F r ie n d s a n d C H r o n o lo g ie s o f C h a n c e

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 tirne. 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 story lines that cross borders. The new

198

P o e tic s o f C i n e m a

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

Major 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.

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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 multipleprotagonist 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

200

P o e tic s o f C i n e m a

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, and Random AX^alks 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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

201

cubide 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 Ywva, 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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

203

entire structure but only a part of it, as counterfeit banknotes do in early sections of Bressons L’Argent (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, cafés, 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

P o e tic s o f C i n e m a

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.

Only Connect, or at Least Collide A very small cause that escapes our notice determines a considerable effect that we cannot fail to see, and then we say that the effect is due to chance. Henri Poincaré29 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 in shops, on the street, in the Métro, in cafés, and the like. Nearly as often, though, 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 Métro platform, then glimpsed by yet another character from a passing train. A cookie discarded by one character is nibbled by a pigeon; the pigeon flies away, and its dropping spatters a snapshot taken by another character. The photographer takes the snapshot into a shop, where the clerk, wiping the photo clean, recognizes her old boyfriend. Wind, water, and other elemental forces can be important vehicles for chance in network narratives, and in Happenstance they bring together two people who do not speak a word to one another in the film. A man gives a pebble to his mistress. She drops it out the window. It hits a taxi windshield, and when the cab lurches, the passen­ ger Irene whacks her nose. Later the young man, Younès, given a yellow slicker by a well-meaning passerby, is mistaken for the pickpocket who discarded the slicker and winds up with a dislocated nose too. In the first scene, Irene and Younès had sat opposite one another in a Métro car without speaking a word (Figures 7.2—7.3). Now in the final scene, they sit back to back outside the hospital, each unaware of the other (Figure 7.4). The immigrant, having reached his family, has brought some sand from the desert, but a stray breeze blows the sand out the window and onto the two young people below. Turning to wipe their eyes, they recognize each other, each with identi­ cal nose splints (Figure 7.5). On this wryly symmetrical image of merged destinies, the film ends. The most common chance-based convergence, as conventional as a Main Street shootout in a Western, is the traffic accident. It seems that a network movie can’t do without a traffic jam, smashup, fender-bender, felled pedestrian, or brake-squealing

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

Figure 7.2 Happenstance (2000): When Irène gives her birthdate, Younès in the seat oppo­ site glances abruptly at her.

Figure 7.4 Happenstance: Irène and Younès, having left the hospital separately, sit outside.

205

Figure 7.3 Happenstance: She returns his look. They won’t meet again until the conclu­ sion.

Figure 7.5 Happenstance: The final shot’s symmetry announces the artifice of uniting the broken-nosed couple.

near-miss. I had thought that Crash (2004), by lifting this conceit to virtual self-parody, would purge it from our screens, but now I believe that it will never die. Not because car crashes figure forth that lust for speed and perceptual overload characteristic of modernity, or because they embody that morbid fascination with spectacle character­ istic of The Postmodern Moment—in other words, not because of some all-purpose explanatory Zeitgeist. The car crash conceit will survive because as a device it snugly suits degrees-of-separation storytelling. For one thing, traffic accidents are plausible within a story world. We know that they happen all too often. Moreover, they’re the most obvious chance encounter that can have grave consequences. Bump me with your shoulder, and we’ll probably move on and forget about it. Dent my car with yours, and we have to halt to sort things out. Smash into my car, and our lives can change forever. Claude Telouche’s II y a des jours . . . et des lunes (1990) juggles several plotlines, some situated in a small town, others in an airport, and others on the road. Throughout the bulk of the film, one character in particular doesn’t directly link to the rest. Gérard is a short-tempered truck driver who quits his job and steals a sports car. He picks up a runaway bride, and-the two make their way toward the French Riviera. A 30-car pileup in a tunnel forces them to detour through the village, where they run into a traffic snarl created by an ambulance bearing victims of another car crash. Gérard pulls an evasive maneuver that makes his car block that of a doctor, who’s driving his mistress from the airport. Each man refuses to give way. In a rage, the doctor stabs Gérard. As he

206

P o e tic s o f C i n e m a

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 Amoves 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, Amoves 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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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 clichés to the bursting point. Traffic accidents are just the most extreme instance of dramatic but contingent nodes, the unlikely points of connection within the films 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 cinemas narrational resources. As ever, our curiosity, suspense, and surprise are created by manipulating what we know and when we know it.

Narration Makes Networks 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 The 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

P o e tic s o f C i n e m a

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 Terrorizes 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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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,

210

P o e tic s o f C in e m a

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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

211

character relationships that most films would provide at the start. Flavors uses all the self-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 a n d 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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

213

films 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

P o e tic s o f C i n e m a

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 émigré 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 Gérard, 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. Gérard 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 Over 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.

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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 une femme, 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

218

P o e tic s o f C i n e m a

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 Détective (1985), set in and around a luxury hotel, employs the confinedarena 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 actionbased 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 AX/orlds 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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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 lowbudget 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 IVe 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 threeact (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

P o e tic s o f C i n e m a

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 Barmans film is typical of European and Asian network movies. In Ulrich Seidls 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 Frédérique 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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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).

IS ia s h v ille (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

P o e tic s o f C i n e m a

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 clichés 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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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 Walkers 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

P o e tic s o f C i n e m a

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 these people’s everyday dealings, but it also sketches in a social truth we might have 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 Fraser. Tom taunts Kelly for being a soldier, and Kenny carries around a mysterious violin case. A series of scenes introduces an ominous tone. Opal chatters to Triplette about political assassination, claiming that a gun-carrying culture encourages inno­ cent but confused people to “pull the trigger.” On that line, we cut to Kenny talking to his mother on the phone. We’ve already seen Kenny pass the Walker campaign headquarters; it’s a throwaway, but it’s distinctive for being one of the few occasions 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 behavior. Later at the Opry Bell, we zoom in on Kenny raptly staring at Barbara Jean, but the camera quickly pans to Opal, asking Private Kelly what he did in Vietnam. When the crowd starts to boo Barbara Jean after her breakdown onstage, Private Kelly stands up in defiance of them. So the narration equivocates. Kenny is an obvious suspect, so perhaps the wor­ shipful Private Kelly will turn out to be a crazed Vietnam vet? In fact, the narration has tricked us. It’s the target, not the shooter, who will come as a surprise. Altman sustained this narrational ploy in his interviews. “We also condone political assas­ sinations,” he remarked. “Of course we don’t shoot our pop stars or sports heroes, because they don’t arouse divisive feelings.”51 In his film, we do shoot our pop stars, for reasons he leaves to the viewer to decide. “Nashville’s a musical, really,” Altman notes.52 The film contains over an hour of performances, and this isn’t counting all the background tunes we hear in dubs or the bursts of nondiegetic music now and then. Most of the big numbers were shot as actual shows with roving cameras and the 16-track recording system used for rock concerts, a major departure at the period.53 As often happens in the genre, the songs comment on the action, as when after Mary learns that Tom is bedding other women, she must sing, “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 Café before its all-out, hypnotically repetitive treatment in the Parthenon epilogue makes it virtually a new national anthem. For all the traces of documentary capture, Nashville also trades on the traditional musical’s self-conscious stylization. This is most evident in the brilliantly odd open­ ing credits, a parody of late-night TV pitches for compilation LP albums (Figure 7.11). This rapid-fire pseudo-commercial introduces the music industry, the premise of the ensemble cast, the faces of the principal players, and lines from the songs we’ll hear, the

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

227

Figure 7.11 Nashville’s credits, selling the movie as an album on late-night TV.

whole contrivance framed by a whirl of fake albums cut by the characters. The jokes pile up. The lead-in has indicated that Nashville was financed by ABC Entertainment, a television subsidiary, and the ad concludes with the announcer explaining that the film will be brought to us “without commercial interruption.” The ad quickly shrinks to a pinpoint, as if a television were switched off. Even the scratched-up Paramount logo at the very start recalls a weather-beaten 16mm print running in the small hours on a local TV station. The hyperbolic, direct-address prologues of Bug and Magnolia owe a lot to Altman’s recognition that a network narrative, although apparently close to life in some respects, is no less an exercise in artifice than any other movie. Altman claims he had long been tempted by the converging-fates format. In the 1960s, he imagined a film featuring four strangers whose lives intersect only in a yoga class.54 He’d also thought about shooting two films based on Michener’s Tales of the South Pacific, one centered on the nominal hero, and the second film putting the subsidiary plotlines to the fore and making the first film’s hero a minor character.55 Along these lines, he considered releasing two features, Nashville Red and Nashville Blue. Each would cover the same story but would favor different characters. Venture­ some ideas like these merely confirm the evidence on the screen: Altman brought a new vitality to cinematic storytelling. One could argue that Nashville is more com­ plicated than complex, that scrambling together lots of one-off encounters is easier than sculpting everything into a coherent design. Yet the film’s openness to accident and digression, its urge to create a fresco of American life, and its counterweighting of sharp social realism by stylization made it hugely influential. Just as influential was the realization that familiar conventions of genre and style could help audiences pick their way through a teeming profusion of incident. Nashville pointed the way toward a renewal of narrative form in American cinema.

Is A a g n o lia (1999) About six years before the disappearance of Ambrose Small, Ambrose Bierce had disappeared. Newspapers all over the world had made much of the mystery of Ambrose Bierce. But what could the disappearance of one Ambrose, in Texas, have to do with the disappearance of another Ambrose, in Canada? Was some­ body collecting Ambroses?. . .

228

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

Figure 7.12 Magnolia (1999): A suicide attempt and the portentous coiled hose.

231

Figure 7.13 Magnolia: As Jim the cop drives 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

P o e tic s o f C i n e m a

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 miseen-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 happened’ . . . 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 knows.. . . 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.

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

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 Magnolias narrational 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.

F a v o r i s d e la lu n e (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

Figure 7.14 Play Time (1967): In early por­ tions of the film, Jacques Tati stages his gags at an immense distance. Arriving in the airport, Barbara glances back to see a woman stroking her bag and to hear a dog softly yelping. Still farther back, the silhouettes of airline workers are revealed as cardboard cutouts.

P o e tic s o f C i n e m a

Figure 7.15 Play Time-. The dense restaurant scene packs several gags into a shot. Here the owner drinks what he thinks is a headache remedy in the middle ground while confu­ sion reigns in the distance.

alert, Where s-Waldo frame of mind (Figure 7.15). Were 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 and separate and never see one another again.’”58 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 Lesfavoris de la lune (Favorites of the Moon; 1984), La chasse aux papillions (Chasing Butterflies; 1992), Brigands, Chapitre VII (1996), and Adieu, plancher des vachesl (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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

Figure 7.16 Favoris de la lune (1984): Less extreme than Tati, Otar Iosseliani still lets his gags play out at a distance.

235

Figure 7.17 In Favoris, the characters are identified as full-length figures, character­ 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

236

P o e tic s o f C i n e m a

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 typed bits of behavior—a man picking up a woman in his car, and a man presenting a woman with a rose before taking her to a hotel. Romantic partners are shuffled, but 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 Iosseliani, who reflected more theoretically on his chosen form than Tati did, seems to have understood that criss-crossing can create some fruitful uncertain­ ties. It can hint at a relationship visually without pinning it down. (Is the hooker a lover of one of the burglars, or just their confederate?) The throwaway dialogue, which we overhear rather than attend to, doesn’t specify exactly how characters are linked. There’s a more insidious effect as well. After two characters intersect, the next moment can attach us to either one, and then, when the one to whom we’re attached brushes past someone else, that’s another opportunity to deflect the narrative line. String together several brief encounters like this, and we can be quite unsure exactly whom the narration might follow next. Were thrown back on simply registering each moment in itself, however opaque it might seem. This effect is magnified if no char­ 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 us back, through character overlap or through crosscutting, to familiar terrain. The effect is to treat the social world as a series of spreading and overlapping trajectories that we study from a distance. One short scene soon after the prologue furnishes a good example. The bomb maker has helped his blonde girlfriend get ready for work, and she goes out, passing the prostitutes clustered at her doorway. She crosses the street to a waiting car, where a man (whom we’ll later learn is the police inspector) greets her with a kiss. The car drives past the camera, which follows it to pick up an old man starting to cross the street. He greets the prostitutes, and one goes with him to the building opposite. In the next shot, as the old man and the hooker arrive at her landing, the bomb maker passes them. He’s the one we follow next as he descends to the street, where 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 burglary team come out of their next-door apartment and give her keys, adding, “Tonight.” (The stingy Iosseliani won’t supply a medium shot of this gesture, so the 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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

Figure 7.18 The tourist Barbara receives a sprig of plastic flowers as a parting gift from M. Flulot (Play Time).

237

Figure 7.19 Play Time: Glancing out as her bus moves off, she sees them as echoing the shape of streetlights.

bedroom or the thieves? Instead, Iosseliani cuts to the machine shop where the bomb maker starts his day, as the arms dealer’s wife comes to give him his assignment. This rhythm of expanding adjacency functions as cryptic exposition. By register­ ing the isolated gestures and tiny intersections, we gradually—very gradually—come to grasp a set of circumstances and a narrative action. The dominant impression is the fluid, ephemeral hookups among individuals. “What’s important to me is this possibility of treating the characters or the actions as [musical] themes, which permits everything in the film to flow, all the streams to join and separate.”62 The pattern that finally results is as wayward as the protagonists. Iosseliani fills his networks with clumsy characters and awkward accidents. The paths intertwine, he seems to be 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. Tati, who was fascinated by clichés, concluded Play Time by investing kitsch with genuine emotion, matching a sprig of plastic flowers to curved street lights as the young woman leaves Paris (Figures 7.18-7.19). Iosseliani likewise makes common­ places the pretext for ingenious formal play. In Favoris de la Lune, we wander in an impersonal city (has Paris ever looked more drab?), and we register the familiar tactic of cross-class comparisons: Both the bourgeoisie and the underclass live by theft. The painting and crockery pass from rich to poor and turn from objets d’art into bric-à-brac. This motif, Iosseliani tells us, implies that the search for possessions is a vain thing. The cyclical exchanges and traffic among the characters invoke the theme of the eternal return, “the old, banal notion, nothing new under the sun.”63 As usual, however, ingenious cinematic form can turn even overfamiliar materials into invigorating exercises for the eye, ear, and mind.

Les P a s s a g e r s (1999) The prow of our tram moves through the streets and into a tunnel, and most of the credits emerge from the darkness. Once we pass back into daylight, minor incidents develop. One rider has no ticket because the vending machines aren’t working. A man

238

P o e tic s o f C i n e m a

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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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 were 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

240

Figure 7.21 Les Passagers: The ranting busi­ nessman, declaring that there are three sexes. At first it seems to be another direct-address monologue, like that of the Narratrice, but it’s interrupted b y . . .

P o e tic s o f C i n e m a

Figure 7.22 . . . a shot of the youth, appar­ ently watching him, perhaps the target of the rant.

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

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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, unapologetically, 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 redressed 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 filmmaker-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

Figure 7.25 Les Passagers: The Narratrice as nurse, talking to another nurse offscreen, in frontal shot/reverse shot like that of Figure 7.7. The more traditional device is introduced after the unconventional usage, to revise the film’s narrational norm.

P o e tic s o f C i n e m a

Figure 7.26 The final scene of Les Passagers: The Narratrice bids goodnight to those sleep­ ing in their graves.

a woman refuses to give money to a beggar on the tram, she looks at us. During Isabellas 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 were 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)

M u t u a l F r ie n d s a n d C h r o n o l o g ie s o f C h a n c e

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 citywide 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.67A 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 crisscrossers were threatening to become passé, 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 filmmaking 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

P o e tic s o f C i n e m a

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.”74If 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 networkbased 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 crisscrosser 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

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

245

unheard of even a decade ago. The fact that Nine Lives (2005) consisted of nine long takes was spotlighted in the films 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 risk— always remembering that balancing it with familiarity on other fronts will help the viewer grasp the innovations more confidently.

Network Narrative: A kX/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

P o e tic s o f C i n e m a

American Gun (Aric Avelino, 2005) Among Adults (Stephane Brise, 2007) Amoresperros (Alejandro Gonzalez Inarritu, 2000) Amourfou, V (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)

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

Dead Girl, The (Karen Moncrieff, 2006) Decalogue, The (Krzysztof Kieslowski, 1989) Distant Tights (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 d e ..., 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 (Doug Liman, 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 d ’ailes du papillon; 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)

247

248

P o e tic s o f C i n e m a

11y a des jours . . . et des lunes (Claude Lelouche, 1990) Ingorgo—Una storia impossibile, V (Luigi Comencini, 1979) In the City (Cesc Gray, 2003) InterMission (John Crowley, 2004) Istanbul Tales (Umit Unal, Kudret Sabanci, Selim Demirdelen, Yucel Yolcu, and Omur Atay, 2005) Joki (The River; Jarmo Lampela, 2001) Km. 0 (Yolanda Garcia Serrano and Juan Luis Iborra, 2000) Lawless Heart (Neil Hunter and Tom Hunsinger, 2001) Leave on Word of Honor (Karl Ritter, 1938) Life as It Comes (Stefano Incerti, 2003) Loggerheads (Tim Kirkman, 2005) Look Both Ways (Sarah Watt, 2005) Love Actually (Richard Curtis, 2003) Love and Happiness (Jordan Allan, 1995) Magnolia (Paul Thomas Anderson, 1999) Manual of Love, The (Giovanni Veronesi, 2005) Me and You and Everyone We Know (Miranda July, 2005) Mind the Gap (Eric Schaeffer, 2004) Monsoon Wedding (Mira Nair, 2001) Motel Cactus (Park Ki-yong, 1997) Nashville (Robert Altman, 1975) New Age of Living Together, The (aka In Between; Samson Chiu, Yeung Fan, and Sylvia Chang, 1994) Nicotina (Hugo Rodriguez, 2003) Night of the Sunflowers (Jorge Sanchez-Cabezudo, 2006) Night Shapes (Andreas Dresen, 1999) Nine Lives (Rodrigo Garcia, 2005) Nines, The (John August, 2007) Noel (Chazz Palminteri, 2004) Nosotras (Women; Judith Colell, 2000) No Sweat (Eoin Moore, 2006) Oktoberfest (Johannes Brunner, 2006) Once Upon a Time in Triad Society 2 (Cha Chuen-yee, 1996) Opening (Rob Nilsson, 2006) Out 1: Noli me tangere (Jacques Rivette, 1970) Parisiens, Les (Claude Lelouche, 2004) Paris nous appartient (Jacques Rivette, 1960) Park (Pat Holden, 2006) Passagers, Les (Jean-Claude Guiguet, 1999) Pastorale (Otar Iosseliani, 1976) Play Time (Jacques Tati, 1967) Prairie Home Companion, A (Robert Altman, 2006) Premiere juillet, le film (Moving; Philippe Gagnon, 2004)

M u t u a l F r ie n d s a n d C h r o n o lo g ie s o f C h a n c e

Private Fears in Public Places (Alain Resnais, 2006) Pulp Fiction (Quentin Tarantino, 1994) Rave Fever (Mak Siu-fai, 1999) Ready to Wear (aka Prêt-à-porter; Robert Altman, 1994) Rush (Lee Sang-in, 1999) Sad Movie (Kwon Jong-kwan, 2005) Safety of Objects, The (Rose Troche, 2001) Sauve qui peut (la vie) (Jean-Luc Godard, 1980) Scenes of a Sexual Nature (Ed Blum, 2006) See You in Space! (Jozsef Pacskovszky, 2005) Selon Charlie (Nicole Garcia, 2006) Sexual Life (Ken Kwapis, 2004) Ship of Fools (Stanley Kramer, 1965) Situation, The (Philip Haas, 2007) Standing Still (Matthew Cole Weiss, 2005) Strawberry Shortcakes (Yazaki Hitoshi, 2007) Street of Shame (Mizoguchi Kenji, 1956) Sud Express (Chema de la Pena; Gabriel Velazquez, 2005) Sunshine State (John Sayles, 2002) Tales of Manhattan (Julien Duvivier, 1942) Tapas (Jose Corbacho and Juan Cruz, 2005) Terrorizers, The (Edward Yang, 1986) They All Laughed (Peter Bogdanovich, 1981) Things to Do Before You’re 30 (Simon Shore, 2004) Things You Can Tell Just by Looking at Her (Rodrigo Garcia, 2000) Three Palm Trees (João Botelho, 1994) Three Seasons (Tony Bui, 1999) Tic Tac (Daniel Alfredson, 1997) Top Banana Club (Anthony Wong, 1996) Tout ça . . . pour ça! (All That.. .for This?!; Claude Lelouche, 1993) Toute une vie (And Now My Love; Claude Lelouche, 1974) Trafico (João Botelho, 1998) Trilogy (Lucas Belvaux, 2002) Two Days in the Valley (John Herzfeld, 1996) Uchoten Hoteru, The (aka Suite Dreams; Mitani Koki, 2006) Une pour toutes (Claude Lelouche, 1999) Uns et les autres, Les (Claude Lelouche, 1981) Up and Down (Jan Hrebejk, 2004) Utopia—Nobody’s Perfect in the Perfect Country (Morten Tyldrum et al., 2002) Voisins voisines (Malik Chabane, 2005) Way Off Broadway (Daniel Kay, 2001) Weapons (Adam Bhala Lough, 2006) Wedding, A (Robert Altman, 1978) Weirdsville (Allan Moyle, 2007)

249

250

P o e tic s o f C i n e m a

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 (André Téchiné, 2007) Wonderland (Michael Winterbottom, 1999) Yacoubian Building, The (Marwan flamed, 2005) Year Zero (Joseph Pitchade, 2005) Yellow Rolls Royce, The (Anthony Asquith, 1964) Yuva (Mani Ratnam, 2004)

Studies in Style

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

P o e tic s o f C i n e m a

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 mediums 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, R 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

C in e c e r ity

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 “auteurstructuralists” 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.51 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-scène, 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 décor, 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

P o e tic s o f C i n e m a

a tower after Arkadin’s death supersedes the script and quite naturally escapes the attention of our literary film critics.”8And 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 André 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-scène. 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

C im e c e r ity

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 masterworks—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

Figure 8.1 The Searchers (1956). Drinking his coffee, the Ward Bond character glances toward the wife of the household.

P o e tic s o f C i n e m a

Figure 8.2 The Searchers: His point of view reveals her gently stroking the uniform of her 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 liitchcock 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.22Ford, 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

C in e c e r ity

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 director’s 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 américain [the knees-up framing] was the most efficient shot, permitting the greatest economy of editing.

260

P o e tic s o f C i n e m a

This technique may have lacked ambition, but it was faultless and sure. Today it would still be interesting to analyze its smallest details.29 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

C im e c e r ity

261

assign to a film or oeuvre.37 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 règle du jeu (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 cinéphiles 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

P o e tic s o f C i n e m a

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 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 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 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 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 hardly a living presence for them. We shouldn’t be surprised. Today’s tastes are the legacy of the 1970s. Directors attuned to psychological and pictorial subtlety couldn’t make much headway then, when Hollywood, groping for a new business model, lurched between megapictures and heavy-handed “personal” films. Neither The Poseidon Adventure nor Five Easy Pieces has room for moments like that of The Searchers. The midrange program pic­ ture was the arena of so many of Hollywood’s greatest creators, and its elimination set the menu for most popular film consumption today, the choice between the relentless steamrolling of the blockbusters and the preening quirkiness of independent cinema. For those of us who learned from Sarris, then, there remains much to do. In partic­ ular, although “style” as sensory bombast has become part of a movie’s packaging and marketing (X-Men and JFK order up Technique in 40-gallon drums), viewers remain almost completely unaware of style in any rigorous sense, let alone of its nuances. “The auteurists,” Sarris remarked over 25 years ago, “are still fighting an uphill battle to make movie audiences conscious of style.”42 La lutte continue, as the French used to say.

9. Taking Things to Extremes H a llu c in a t io n s C o u r t e s y o f R o b e r t R e in e r t

If anyone doubts that progress in film studies is possible, we need only point to our increasing understanding of the extraordinary tradition of staging in European films of the 1910s. The first historians of silent film, smitten by montage, criticized the staging tradition as too theatrical and set it against the more progressive and properly “cinematic” cutting of D.W. Griffith. Bazin in turn speculated that the films of this period harbored a “primitive” deep focus that was eventually to find fulfillment in the work of Jean Renoir, Orson Welles, and William Wyler.1Across the 1990s, a host of studies showed that this tradition constituted a robust aesthetic in its own right.2 Certainly the 1910s filmmakers’ art remained “theatrical” in many respects, but it was also “cinematic” by virtue of its reliance on the perspectivai qualities of the camera. Thanks to projective optics, the playing space of any shot consists of a pyramid with the apex resting at the lens. The filmmaker can guide our attention within the depths of that fanned-out space by centering figures or props, moving figures to and from the camera, turning their faces toward and away from us, creating discrepancies of size or texture or illumination, and choreographing performers so that they block and reveal crucial actions at the proper moment.3

263

264

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.

P o e tic s o f C i n e m a

Figure 9.2 The Traitress: Von Mallwitz lifts his head to clear a space for the Marquis’ servant, who enters and announces Yvonne.

Figure 9.3 The Traitress: As the servant steps aside, the Lieutenant rises and goes to frame right, clearing still more space for Yvonne’s entrance and establishing a remote­ ness between them that will have dramatic consequences.

Take for example the 1912 German production The Traitress (Die Verräterin), 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

T a k in g T h in g s to E x tre m e s

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.

265

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 9.6 The Traitress: After a final glance at von Mallwitz, she leaves, the servant bowing as she passes.

these simple principles, Louis Feuillade, Evgenii Bauer, Franz Hofer, Georg af Klercker, Victor Sjostrõm, 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

P o e tic s o f C i n e m a

266

Figure 9.7 In a later scene of The Traitress, Yvonne and her father are dining, and she starts to rise from the table.

Figure 9.8 The Traitress: Her movement is carried over into a slightly more distant 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 tine 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

T a k in g T h in g s to E x tre m e s

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.

P o e tic s o f C i n e m a

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), clichés 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,

T a k in g T h in g s to E x tre m e s

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 limits 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. Fie 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),

P o e tic s o f C i n e m a

270

Figure 9.12

Opium.

Figure 9.13

Opium.

F ig u r e 9 .1 6

O p iu m .

F i g u r e 9 .1 7

O p iu m .

T a k in g T h in g s to E x tr e m e s

Figure 9.20

271

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, and the palanquin emerges in the distance (Figure 9.21). He ducks out of the 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 scene in Figures 9.10-9.16. He lines up the heads in considerable depth and clusters

P o e tic s o f C i n e m a

272

Figure 9.23

Opium.

Figure 9.25

Opium.

Figure 9.24

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 protagonists too. When the old Dr. Armstrong accuses Gesellius of killing his son, we get an intense 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

T a k in g T h in g s to E x tre m e s

Figure 9.28

Opium.

273

Figure 9.29

Opium.

guiding the viewers 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

P o e tic s o f C i n e m a

274

I iyure 9.34

Opium.

Figure 9.35

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 f/9 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

T a k in g T h in g s to E x tre m e s

Figure 9.40

275

Nerven.

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 frontmost 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 RolofFs 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.

P o e tic s o f C i n e m a

276

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

T a k in g T h in g s to E x tre m e s

Figure 9.47

Nerven.

277

Figure 9.48

Nerven.

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

P o e tic s o f C i n e m a

278

Figure 9.49

Nerven.

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-américain 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 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-enscène. 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 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 up with an image consisting of what are in effect close-ups . .. without editing. An important consequence of this strategy is to remove some of those cues for spatial stability that had been prized by the 1910s directors. Feet and legs are cut off, even in establishing shots. Similarly, Feuillade, Bauer, Hofer, and most other directors of the time use rear doors or entryways to let us know how characters enter or leave the scene. With such close foregrounds as Reinert’s, however, characters may leave the frame but not necessarily the locale, and this creates an uncertainty we seldom find in the work of his peers. Where does Marja go when she darts out of view in Figure 9.44—to another part of the room, or to another room? Where is the cottage

T a k in g T h in g s to E x tre m e s

Figure 9.54

279

Nerven.

door through which Johannes must have come in Figure 9.47? Reinert’s urge to inten­ sify the diagonal thrusts of the 1910s approach led him to an in-between aesthetic position: His typical depth shot is neither a tableau composition that surveys the entire locale nor a part of an American-style editing pattern that builds a consistent overall space out of several complementary camera positions. A Reinert “full shot” may present a relatively small patch of the scene’s space, and we may never be properly introduced to the overall arena of action. As European directors were adopting Hollywood découpage, Reinert comes up with a fragmentary scenography that looks ahead to the strategically incomplete establishing shots employed by Robert Bresson, Jean-Marie Straub and Danièle Huillet, and Hal Hartley.

P o e tic s o f C i n e m a

280

Figure 9.55

Nerven.

Figure 9.56

Nerven.

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 (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’être. 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.

CinemaS cope T h e lSA odern B A iracle T o n S e e V T ith ou t G la s s e s

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-Tuc 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

P o e tic s o f C i n e m a

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 cinéphiles 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.”3But 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 presentation, its original ability to excite spectators.”4 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 André Bazin had taught the younger generation the virtues of pictorial depth in the work of Jean Renoir, Orson Welles, and

C in e m a S c o p e

Figure 10.1 In situations when some actors were sitting and some were standing, CinemaScope close views posed problems, and some directors opted to chop out a body, leaving a head at the lower frame line (The Hunters, 1957).

283

Figure 10.2 Godard seems to take this schema as the visual premise for this shot from Pierrot lefou (1965), which features one head at far left and a clone (actually a sculpted head) at far right.

William Wyler, the Young Turks of Cahiers du cinéma 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, François 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

P o e tic s o f C i n e m a

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.

C in e m a S c o p e

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. Tess 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,”

286

P o e tic s o f C i n e m a

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 (SleepingBeauty, 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.

C in e m a S c o p e

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 Chrétiens anamorphic lens system. The first tests of the lens convinced Darryl F. Zanuck, head of Fox production, to adopt the system immediately. Chrétiens 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 Chrétien 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

288

P o e tic s o f C i n e m a

Figure 10.3 Richard Egan suffers CinemaScope mumps in Love Me Tender (1956).

$25 million worldwide, making it the top grosser of the year and one of the highestearning 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 highend industry standard.26 “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 Chrétien 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 Chrétiens 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

C in e m a S c o p e

289

its method of correcting lens astigmatism.29 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

290

P o e tic s o f C i n e m a

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 Tucas 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

C in e m a S c o p e

Figure 10.5 Steps that should be parallel are bulging in this shot from The Young Lions (1958).

291

Figure 10.6 CinemaScope made classic columns look bloated, although some set 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 Chrétien 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.38Directors 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 closeups, 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

292

P o e tic s o f C i n e m a

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.46 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 Roster’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 closeup, 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

C in e m a S c o p e

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 closeups, 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 closeup, 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

P o e tic s o f C i n e m a

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 Scopes 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 Ambersons (1942) called attention to their long takes, as did The Lady in the Lake (1947) with its 24 shots, and ITitchcock’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

C in e m a S c o p e

Figure 10.9 Although lining up characters in a row perpendicular to the camera was the 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.

295

Figure 10.10 A standard medium two shot from The Mark ofZorro (1920).

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-américain, 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

P o e tic s o f C i n e m a

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 10.13 Several characters grouped so that their bodies, facing us, fill out the frame on a single plane (Little Women, 1933).

Figure 10.14 Stacking rows: A crowded scene in The Thin Man (1934) still obeys the clothesline principle, in layers.

in not one but two rows, one behind the other but still more or less 90 degrees to the lens (Figure 10.14). These “clothesline” arrangements, spreading several players across 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.

C in e m a S c o p e

Figure 10.15 In the 1940s, recessional stag­ ing became more common, setting characters into considerable depth (Caught, 1949).

297

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 10.17 Depth-oriented directors could take,advantage of the nearly square propor­ tions of the Academy ratio to build up vertical compositions (Ihe Little Foxes, 1941).

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

298

Figure 10.18 A shot from Gun Crazy (1949) exemplifies the extremes to which the deepspace staging of the 1940s could go.

P o e tic s o f C i n e m a

Figure 10.19 The “baroque” side of reces­ sional staging {The Killers, 1946). Such com­ positions can be found throughout world cinema in earlier decades, but they became far more common during the 1940s.

compensating maneuver is to turn the foreground actor toward us, motivating this as the characters 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 camera than in the earlier era. In many shots the foreground figure is presented in looming close-up, and the background figures can be either distant or fairly close, packing the frame (Figure 10.18). In the recessive strategy, the amount of playing 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 times find in 1940s cinema (Figure 10.19). As directors began experimenting with recessive staging in the 1930s and early 1940s, it became clear that planes so far from one another could not be kept easily in focus, especially if the foreground needed to be quite close. Most directors learned to live with this, either letting one plane drift a little out of focus or keeping the foreground fairly far from the camera. The “deep-focus” style heralded by Citizen Kane (1941) provided technical solutions: lots of light, faster film, coated lenses, exact diaphragm stops, and special effects trickery. Now one could have very deep shots and full focal range. Welles’ flamboyant staging schemas would be toned down and normalized throughout the 1940s and 1950s.64 Once filmmakers started exploring diagonal staging schemes, they seemed to have realized that standard establishing shots became less obligatory. A scene could start right on a deep composition, then reveal the set as necessary. Although depth staging is

C in e m a S c o p e

299

Figure 10.20 Small and rather cheap Warner Bros, sets are well concealed by recessive layouts of figures in The Maltese Falcon (1941).

Figure 10.21 A tight composition from The Purple Heart (1944). The precise alignment of props and players, particularly the foreground judge’s head, is typical of post-Kane American cinema.

Figure 10.22 John Ford favored bold depth compositions from the very beginning of his career. Here Wyatt Earp in the foreground is about to slide his pistol down the bar to his brother under the nose of Doc Holliday (My Darling Clementine, 1946).

Figure 10.23 Films shot on color stock, which needed more light and didn’t allow for great depth of field, continued to favor the clothes­ line layout (Leave Her to Heaven, 1945).

usually associated with big sets, having the camera hug the axis of action also permitted sets to be more compact (Figure 10.20). In general, 1940s compositions became tighter, as the urge to fill the frame created layouts that click neatly into place (Figures 10.21-10.22). Because the 1.33 frame was firmly established as the standard, directors could count on their clenched compositions being retained in most movie houses. For all its popularity, recessive staging remained a secondary option for 1940s filmmakers. Most films continued to use variants of clothesline arrangements (especially in color, which did not permit great depth of field; Figure 10.23). Many directors adopted a moderate approach to depth, avoiding the most outré composi­ tions and blending depth staging with more lateral layouts within a single shot. A film shot predominantly in the clothesline manner might include a few deeper com­ positions. Even directors w7ho began in the 1920s and 1930s became accustomed to diagonal staging options, if only as an occasional resource (Figure 10.24).

300

Figure 10.24 Alfred Hitchcock experimented with recessional compositions after coming to America; this striking shot is from Lifeboat (1944).

P o e tic s o f C i n e m a

Figure 10.25 The classic two shot remains a basic resource into the 1950s, as shown here in George Cukor’s The Marrying Kind (1952).

Such was the menu of staging schemes in force when Scope appeared in 1953, and the 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: “Pie 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

C in e m a S c o p e

301

Figure 10.26 In Adam’s Rib (1949), Cukor also accedes to the new impulse: A wife trail­ ing her husband is caught in a sharply angled depth shot on location.

Figure 10.27 the “intimate” two shot in Demetrius and the Gladiators (1954). Compare Figures 10.10,10.12, and 10.25.

Figure 10.28 A first meeting, rendered in another two shot (Love Is a Many-Splendored Thing, 1955) that leaves large areas to be filled with props and décor.

Figure 10.29 Clothesline staging with a vengeance in the soda-fountain epilogue to How to Marry a Millionaire (1953).

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

302

P o e tic s o f C i n e m a

Figure 10.30 An unusually big close-up for early Scope (Bus Stop, 1956).

Figure 10.31 Bus Stop: A more typical ensemble shot.

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 of the thieves stands poised at the right side of the 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 10.34 As the officer in the center real­ izes that his rival has been promoted, were expected to notice the new star on the mans shoulder, at extreme left (D-Day the Sixth of June, 1956).

Figure 10.35 Aboard the airplane in The High and the Mighty (1954), the characters are framed by the seat on the left and the stewardess’s body on the right.

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

C in e m a S c o p e

303

Figure 10.36 The lines of rowing sailors create vectors that culminate in Paris and Aeneas, lounging in the foreground (Helen of Troy, 1956).

Figure 10.37 The Scope frame broken into two symmetrical chunks by windows fram­ ing the characters in The Bridge on the River Kwai (1957).

Figure 10.38 Recessive anamorphic com­ positions could be built around tables and other set elements (Battle Cry, 1955).

Figure 10.39 In Brigadoon (1954), the shots spread out the mythical Scottish town in a fresco of dance and color.

Figure 10.40 Brigadoon: Once the adman protagonist is back in New York, however, the compositions are cramped and closed-off, suggesting stifling metropolitan life.

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 Scopes 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).

304

P o e tic s o f C i n e m a

Cutting rates also indicate how differently filmmakers responded to Scopes pro­ claimed limitations.72 Early on, several directors accepted the challenge of the anamorphic 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 this is a higher proportion than we find in flat films of those years. Nine more films have ASLs running over 20 seconds, a proportion not seen since the early silent years.73 Evidently, directors who had made the long take integral to their style found no rea­ son to change. Otto Preminger had already become Hollywood’s principal long-take 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. Significantly, in their Scope long takes, these directors often make use of extensive camera movements, regardless of Fox’s warnings to the contrary. The lateral tracking 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 6.9 seconds for Knights of the Round Table. For 1954-1955, between 30 and 40% of the ASLs I examined run between 8 and 12 seconds. Most surprisingly, nearly a fifth of the films sampled for these years have ASLs shorter than 8 seconds. If quick cutting on the big screen made viewers uncomfortable, nobody told Henry Hathaway (Prince Valiant, 1954,6.6 seconds ASL), Robert Wise (Helen of Troy, 1956, 5.4 seconds), or the animators of Lady and the Tramp (1955, 4.5 seconds).74 The dynamic of recovery continued during Scope’s life span, as directors absorbed the format into more normal cutting rhythms. Really long takes become increasingly rare. For the 1956-1960 period, the center of gravity in my sample shifts, and two thirds of the films’ ASLs fall between 7 and 13 seconds. Fewer films average longer takes than 13 seconds, but more ASLs come in at less than 7 seconds.75 It seems that when Scope was introduced, long takes offered a line of least resistance, particularly 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, incidentally, were available in most other widescreen systems.76 As the years passed, filmmakers working in both flat and anamorphic formats tended to accelerate their editing pace. In the 1960s, double-digit shot lengths began to become almost extinct in all Hollywood movies.77 As in the early sound era, artisans struggled to normalize the new technology, to throw off its constraints and restore earlier options. For many directors, this entailed recovering the look they had come to prize in the 1940s: tightly composed images, taken from high or low angles and yielding striking differences of scale

C in e m a S c o p e

305

Figure 10.41 The brilliant southwest sun­ light of Bus Stop permits striking depth compositions, such as this one showing the cowpoke protagonist tying on the scarf of the woman he loves, while she watches, in frame center, from the stands.

Figure 10.42 Elia Kazan cants the Scope frame throughout East of Eden (1955). Here it creates a dynamic depth composition, with Cal on the porch while his father and Abra wait at the party inside.

Figure 10.43 A low-key, low-angle framing recalling film noir for Jubal (1956).

Figure 10.44 The Tarnished Angels (1958): A wide-angle composition reminiscent of 1940s style made possible by black-and-white Scope.

between planes. These were easiest to accomplish outdoors, where brilliant sunlight permitted even Bausch & Lomb lenses to achieve robust depth of field. Many of the most remarkable shots in Scope can be found in Westerns, in adventure yarns, in ancient world sagas shot in Italy or Spain, and in contemporary dramas set in the blasting daylight of the Southwest (Figure 10.41). Even under studio illumination, though, some depth was achievable. In the canted framings in East of Eden (1955), shot both on location and in the studio, Elia Kazan seemed to be trying to become the Orson Welles of Scope (Figure 10.42). More discreetly, Delmer Daves relied on the deep-focus look in both exteriors and interiors for his Western Jubal (1956), built out of deep shots reminiscent of film noir (Figure 10.43). Two other factors helped directors recover the 1940s depth aesthetic. One was Buddy Adler’s decision to permit Scope films to be shot in black and white. Because black-and-white film required much less light than color, cinematographers could stop down the lens diaphragm and get sharper images with better depth of field. With color, Scope filmmakers shot most close-ups with long, shallow-focus lenses, but black and white allowed freer use of wide-angle lenses. The crisp, wide-angle imagery of Douglas Sirk’s The Tarnished Angels (1958; Figure 10.44) would have been virtu­ ally impossible in color (as his Interlude, 1957, shows). The bivouac scenes of Edward Dmytryk’s The Young Lions (1958) present compositions as tensely jammed as any­ thing from the 1940s (Figure 10.45). Black-and-white Scope has a special following among cinephiles, perhaps because images like these announce the triumph of aggressive style over the academic blandness promoted by Fox’s spokesmen.78

P o e tic s o f C i n e m a

306

Figure 10.45 A composition even more tightly articulated than Figure 10.21, in blackand-white Scope (7he Young Lions).

Figure 10.46 Me Tender).

Elvis gets Scope mumps (Love

Figure 10.47 Elvis is cured, thanks to Panavision optics (Jailhouse Rock, 1957).

Figure 10.48 Panavision lenses permit a close and full-frame composition in Under­ 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.”80A bonus, however, was the ability of Panavision lenses to handle recessive staging without loss of focus. Early (and uncredited) Panavision films, mostly blackand-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

C in e m a S c o p e

Figure 10.50 Guys and Dolls (1955): Rows of figures fill out the frame and highlight the cen­ tral information as Nathan Detroit challenges Sky Masterson to date Salvation Army Sarah.

307

Figure 10.51 The “Glorious Technicolor” sequence of Silk Stockings (1957) mocks the impossible ratio.

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

P o e tic s o f C i n e m a

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 10.54 Pat and Mike: As the reenactment starts, the changed positions reveal a waiter in the back tier . . .

Figure 10.55 . . . and he slips into a gap between characters to put in his version of 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-

C in e m a S c o p e

309

Figure 10.56 Love Is a Many-Splendored Thing: A clothesline array leaves a vacancy on the righ t. ..

Figure 10.57

. . . to be filled.

Figure 10.58 The Man in the Gray Flannel Suit (1956): Scope breadth to dramatize a couple’s fraying marriage.

Figure 10.59 A Star Is Born (1954): Television versus cinema, balanced at opposite ends 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: 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 r 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

P o e tic s o f C i n e m a

Figure 10.60 A Star Is Born: Later, when Vicki Lester receives her award, her image on the television monitor swells. Why?

Figure 10.61 A Star Is Born: The TV camera coasts in at frame left, recalling the earlier scene’s interplay of film and television.

Figure 10.62 A father decides whether to give his son’s remains to the war effort (The Man Who Never Was).

Figure 10.63 Rebel Without a Cause (1955) puts Judy in the center, framing Jim outside 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).

C in e m a S c o p e

Figure 10.65 Abstract masses in Land of the Pharaohs (1955), with the toiling slaves as verticals balanced by striated horizontals.

311

Figure 10.66 Bracing for a hit, the German submarine crew is fanned out like a poker hand (The Enemy Below, 1957).

Figure 10.67 The Scope frame as an abstract slice of space (Picnic, 1955).

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 Fogan 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 j ust 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

P o e tic s o f C i n e m a

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 10.70 Bonjour Tristesse (1958): Cécile and her new boyfriend are talking about Anne, who has started a romance with Cécile’s father.

Figure 10.71 Bonjour Justesse: Carried away by the idea of Anne’s happiness, Cécile flings herself into a kiss with Philippe.

Figure 10.72 Bonjour Tristesse: The beach umbrella he’s holding drops forward, momentarily filling the frame with brilliant blue before revealing the couple behind.

Figure 10.73 Bonjour Tristesse: They tumble to the floor, with the umbrella now masking their faces and throwing all the emphasis on 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

C in e m a S c o p e

313

Figure 10.74 Midrange depth in Demetrius anã 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 10.76 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 frame center as other characters rearrange themselves.

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

P o e tic s o f C i n e m a

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 quarrel, characters advance, cross the paths of others, and turn from the camera . ..

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 10.82 Bad Day at Black Rock: Soon Hector comes into frame center to bully the old man 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.

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 Themfor Me (1957). In the party scenes, Stanley Donen uses several strategies to shift attention from one line of action to another. Fie 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

C in e m a S c o p e

315

Figure 10.84 Somewhat like Jacques Tati’s restaurant scene in Play Time (1967; Fig­ ure 7.15), the party in Kiss Themfor Me (1957) creates layers of distinct actions.

Figure 10.85 Kiss Them for Me: When new­ comers arrive, the clothesline layers become slightly more recessive.

Figure 10.86 Kiss Them for Me: And as the group breaks into conversing pairs, their mild recession is broken by a woman towing a man through along a slightly opposing diagonal.

Figure 10.87 At the end of one shot of 20,000 Leagues Under the Sea (1954), all look off, awaiting the entrance of Captain Nemo.

Figure 10.88 20,000 Leagues Under the Sea: 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.

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

P o e tic s o f C i n e m a

316

Figure 10.89 Distant depth reminiscent of the 1910s in Demetrius and the Gladiators. Compare Figure 10.9.

Figure 10.90 A distant silhouette framed for saliency (Ride Lonesome, 1959).

Figure 10.91 The 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 Forty Guns (1957).

Figure 10.93 The Bridge on the River Kwai: Saito’s order is passed down the chain of command, farther and farther into the dis­ tant buildings.

,

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 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 Fullers 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).

C in e m a S c o p e

317

Figure 10.94 Beneath the 12-Mile Reef (1953): As the policemen thread their way through the bar, a young Greek boldly dances with his rival’s girlfriend.

Figure 10.95 Beneath the 12-Mile Reef: The dancing couple drifts further back, and the jilted fisherman steps to frame center, turned from us.

Figure 10.96 Beneath the 12-Mile Reef: Our view of the couple is blocked first by a cop, then by the rival’s brawny back, but then we see them moving toward the faraway door­ way as he reaches for his knife.

Figure 10.97 Beneath the 12-Mile Reef: The couple rush outside in the distance, but when the fisherman rushes to pursue them down the left aisle, the boy’s father halts him in the 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. To take advantage of it, a young Greek dances out the door with another man’s girl­ friend, and the drama develops in striking depth (Figures 10.94-10.97). Our vision 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 the viewer helps energize the experience of long-shot views. King Richard and the 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 café 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

P o e tic s o f C i n e m a

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 10.99 King Richard and the Crusad­ ers: A cut to a new angle shows Sir Giles, whom we know is behind the assassination attempt.

Figure 10.100 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 10.101 . .. before stopping in a gap to confront Sir Kenneth’s suspicions.

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 10.103 Bad Day at Black Rock: Coley douses Macreedy’s food with ketchup, with Reno barely visible over his shoulder.

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.

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.

G in e m a S c o p e

319

Figure 10.106 A tight Scope two shot, but in the center there is some out-of-focus move­ ment in the doorway behind (Pete Kelly’s Blues, 1955).

Figure 10.107 Pete Kelly’s Blues: Pete and Ivy turn, creating a very unusual shot for Holly­ wood anamorphic, one more reminiscent of the opening of Godard’s Vivre sa vie (1962).

Figure 10.108 Pete Kelly’s Blues: Pete advances, with the composition still not revealing why.

Figure 10.109 Pete Kelly’s Blues: The camera creeps forward, and Ivy moves slightly aside 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 the opposite point in space—letting Coley be flung out from the doorway into the audiences face—but Sturges’ arrangement activates our awareness more keenly, forcing us to attend to a small parcel of the screen surface. A more overt instance of “spatial suspense” is exploited by Jack Webb in Pete Kelly’s Blues (1955). Pete and his girlfriend Ivy are on the club’s balcony, where they discuss marriage in a prolonged, profiled two shot. They kiss, but then they separate abruptly (Figure 10.106). Why? There is a dim, out-of-focus figure shifting behind them. They turn from us in a startlingly Godardian planar shot (Figure 10.107) as Pete says, “What’d you get, a bleacher seat?” Pete walks straight to the rear, and the camera moves forward, throwing Ivy gradually out of focus (Figure 10.108). She steps aside to reveal the terminally disheveled cop George, who has been pressuring Pete to give details on the shooting of his young drummer (Figure 10.109). As Pete draws up to George, the cop tells Pete that another of his friends may have been killed by the mob (Figure 10.110). The forward tracking shot and Ivy’s sidestep flout the Fox aesthetic, but Webb treats the entire scene as an exercise in overt, purely visual teasing.88

320

P o e tic s o f C i n e m a

Figure 10.111 River of No Return (1954): Matt, the gun, and its holster.

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 10.113 River of No Return: Only when Matt reaches for it out of habit. . .

Figure 10.114 . . . does Preminger reveal 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.

The 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.

C in e m a S c o p e

Figure 10.115 California sunshine allows Nicholas Ray to produce one of the most exciting compositions in early CinemaScope, as the gang members switchblade rises omi­ nously up from the lower frame edge (Rebel Without a Cause).

321

Figure 10.116 House of Bamboo (1955): The climax on the slowly spinning rooftop globe.

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

Figure 10.117 A Leone-like extreme closeup in Forty Guns.

P o e tic s o f C i n e m a

Figure 10.118 Griff steps out into the back 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). Fuller’s 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 anamorphic 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]

C in e m a S c o p e

Figure 10.120 Lee Van Cleef, the man with Techniscope eyes, in For a Few Dollars More (1965). (But compare Figure 10.30, which anticipates this framing.)

323

Figure 10.121 A similarly cropped shot from Confidence (2002).

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-américain 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

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).

P o e tic s o f C i n e m a

Figure 10.125 The long lens already abstracts space, but in THX 1138 (1971), George Lucas designs his anamorphic shots to flatten and compartmentalize his futuris­ tic locale further.

Anamorphic filming never went away, and it enjoyed a resurgence in the 1990s that continues unabated.93 Today any movie can be comfortably shot or released in 2.40:1. Indeed, any image can be repackaged in any ratio. Filmmakers working with Scope could be fairly confident that their images would be shown more or less as they wished. Many of our examples from the 1930s to the 1960s are so precisely composed that careless projection would ruin them. But today’s directors must frame loosely, know­ ing that many shapes and sizes will be carved out of their images (megaplex projec­ tion at anything from 2.4 to 1.85, full-frame TV at 4:3, widescreen TV at 16:9, and the peephole displays on handheld devices like cell phones). The success of Super-35mm, which slices a variety of ratios out of a single square picture, is an acknowledgment that at some basic level, compositional precision is just less important. The acreage afforded by Scope challenged directors to fill it up, and some found thrilling ways to shift bodies around the screen space. By the end of the 1960s, however, most directors had no interest in articulating a scene through staging. Cutting and camera movement were enough, aided of course by close-ups of gripping performances. There emerged a generation of talented directors who loved movies, who could spin engaging yarns and elicit memorable performances, and who had 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.

C in e m a S c o p e

Figure 10.126 Filling the anamorphic frame in Three Kings (1999).

325

Figure 10.127 Getting to know her stu­ dents, the hired tutor Anna spreads her vast dress across the CinemaScope screen (The King and 1,1956).

As with every innovation, the fact that anamorphic imagery is no longer a problem has its benefits as well as its costs. Panavision opened new possibilities for all-over composition in the anamorphic ratio, that maximal use of the screen format that Boris Kaufman valued. Many directors have taken advantage of it, assuming that their densely composed images would be displayed in full (Figure 10.126). And in fairness, I have to say that most directors today face problems no less pressing than the Scope format. How can one develop computer-driven spectacle? Or make full use of digital sound? Or endow children’s fantasy, high school comedy, teenage horror, and comic book superheroics with freshness, beauty, and intelligence? Most of these weren’t on the agenda for filmmakers of the 1950s. For the student of film poetics, though, the Scope era can be seen as giving a cluster of classical staging options one final run-through. A system deplored for its tech­ 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 (like me) find Scope movies so intriguing. Their images invite us into realms where people have bodies and move in real time, and the shape and size of the screen encour­ aged sheer graphic gamesmanship as well. Who among today’s filmmakers would risk the nuttiness of spreading Deborah Kerr’s crinoline across 60 feet (Figure 10.127)? Screen proportions may persist, but styles can change.

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.

The 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

P o e tic s o f C i n e m a

jaw, perhaps aided by a flushed complexion or a loud tone of voice. A Fantômas-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, Noël 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 B l i n k e d F ir s t?

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

P o e tic s o f C i n e m a

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 ing intently at one another.10 “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

" W h o B l i n k e d F ir s t?

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. Our 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 Tans 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.

The 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 1 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

P o e tic s o f C i n e m a

he was directing a film, his editor felt he had to make a certain cut in order to elimi­ nate an actor’s blink.19 George Lucas, pointing out the postproduction advantages of digital video, employs a telling example: “If someone blinks right where I’m making 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 fairly often and fulfills purposes beyond the enhancement of Michael Caine’s career. 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 ingly, in conversation, playing the role of speaker tends to raise the blink rate, whereas 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 her gaze wanders much less. And Caine is right to worry about looking flustered. We blink faster when we are excited or in other states of arousal, such as feeling anxious, addressing a large crowd, or telling lies. One psychological researcher, Joseph Tecce, has specialized in studying U.S. presidential candidates’ eyeblinks during televised debates, on the presumption that stress and anxiety increase the blink rate. In 1996, Robert Dole set a recent record by blinking an average of 147 times per minute.22 Do observers pick up on such inadvertent signals? Another study found that people rated a frequently blinking person to be more nervous and less intelligent than one who blinks rarely.23 When do we not blink? It seems that absorption in a visual task creates longer intervals between blinks. Several Japanese researchers have found that blink rates slow down when people are engrossed in television watching.24 In movie dialogue scenes, the absence of blinking is a very direct way to convey each partner’s attentiveness and mutual interest. If my L.A. Confidential scene were an everyday conversation, we should find around 160 blinks in total (20 blinks a minute x 2 minutes x 4 men). Yet I can find only 34 blinks shown onscreen. Each man seldom blinks when he is speak­ ing, especially when he is the subject of a “single,” or a shot framing only him. Actors whom I’ve asked seldom report that they decide to avoid blinking, but they do invoke 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 blinks may become a by-product. Yet blinking isn’t outlawed altogether. A few blinks make our characters human. Even the “limited animation” of Japanese anime needs to present eye movement and blinks. Chuck Jones points out that Bugs Bunny and Daffy Duck “blink when some­ one’s talki ng. That’s big stuff. To establish that the character is alive.”25 Consequently, the eyes that never blink may be thought inhuman. Tecce commented that during the 1992 presidential debates, Ross Perot came across as having a “reptilian stare.”26 In shooting SlmOne (aka Simone, 2002), the director told the actress playing a computer-generated performer not to blink; when she inevitably did, the blink was digitally removed.27 An occasional blink humanizes, but the trick is to make it significant. In everyday life, a blink is often read as a signal of surprise, concern, or bafflement. This tendency

W h o B l i n k e d F ir s t?

333

seems to have become a convention of fictional narrative generally. No novel reports a characters every blink, but when the narration mentions one, its important. “I don’t understand,” Meyer said, puzzled. “Did she look thirty, or did she . . . ? ” “Well, how would I know how she looked, man?” Meyer blinked. “What do you mean?” he said. “She was wearing a mask,” Bones said. “A mask?” Meyer said, and blinked again. “At a wedding?” “Oh,” Bones said. “Yeah.” He blinked, too. “Maybe I got something mixed up, huh?” he said.28 Our L.A. Confidential policemen blink very little. They blink most while shifting the angle of their gaze, and occasionally they blink to register a reaction to what is said; Exley, for example, keeps his face severely composed, but he blinks in response to chastisement from his captain. My Chinatown example, where only Gittes’ and the purported Mrs. Mulwray’s faces are discernible, consumes 118 seconds, about the same length as the L.A. Confidential scene. Here I count 53 blinks, making the average (about 13 blinks per person per minute) a little closer to the real-life norm. Again, though, the blinks tend to be dramatically meaningful. Mrs. Mulwray blinks when she talks of her husband’s infidelity; as with Exley, her facial expression appears unconcerned, but her blinking shows her to be agitated. Similarly, Gittes’ efforts to avoid taking the case are registered by several blinks that convey not only hesitation and avoidance but also an elaborate effort to be polite. At one point, as Gittes strains to charm Mrs. Mulwray, Jack Nicholson makes Gittes positively flutter his eyelashes. , Thus the demands of film acting build upon normal patterns of blinking but functionalize them: Actors strive to make this natural, necessary act a tool of their craft. One study has indicated that in ordinary life blinks occur with greater fre­ quency at the start of an utterance or word,29but actors tend to blink when they want 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 context. Just as a change of eye direction will not be read onscreen as the gaze drift characteristic of normal conversation, a series of blinks is likely to be taken not as natural lubrication of the eye but rather as betraying a particular emotional state—all those variants of Caine’s “weakness” we can call apprehensiveness, anxiety, remorse, fatigue, or sadness (blinking back tears). In The Guns ofNavarone, when the David Niven character confronts a member of the demolition team with damning evidence that she is a spy, he holds the screen with unblinking force. The other team members watch her warily, and their vigilance is expressed through a pronounced absence of blinks. By contrast, the suspected spy blinks to indicate that her façade is cracking. Yet in the next scene, when Niven is waiting for the moment to launch the raid on the guns, he blinks mightily. He’s afraid.

334

P o e tic s o f C i n e m a

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., Renée 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 B l i n k e d F ir s t?

335

characters who won’t look others in the eye, thereby reinforcing and sometimes finetuning 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

338

P o e tic s o f C i n e m a

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 wellrecognized 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

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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

Figure 12.1 Chushingura (1913-1917): The classic play staged planimetrically, with pockets of space on the top and sides.

339

Figure 12.2 Chushingura: Other shots are rendered along diagonals reminiscent of 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 longer than it did in the West. In Japan, this meant freeing the film from the need for 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 {Horn zanmai; 1928), adheres

P o e tic s o f C i n e m a

340

Figure 12.3 In an early scene of Love and Sacrifice (1922), the hero Teruo meets Namiji. The establishing shot is followed b y . ..

Figure 12.4 . . . a cut-in to Namiji (played by an oyama, a male actor impersonating a woman). ..

Figure 12.5 Teruo.

Figure 12.6 Love and Sacrifice: A later shot of Namiji in the conversation is even closer than Figure 12.4, emphasizing her reaction. Elsewhere in the film, dialogue scenes are pre­ sented in lengthy over-the-shoulder shots.

. . . and then a reverse shot of

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 découpage (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, découpage 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 Flollywood 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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

341

Figure 12.7 Yoshi Castle (1928): A swordfight, filmed in axial shot/reverse shot.

Figure 12.8

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.

Yoshi Castle.

or frame entrances and exits are inconsistent (Figures 12.9-12.10). What should we make of these apparently deliberate violations of conti nuity? 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. Noël 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

P o e tic s o f C i n e m a

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 to baseline norms of classical staging and continuity cutting. This conclusion obliges us to question any claims that these films are sui generis, radically other to the Western mainstream. Yet employment of norms doesn’t entail lockstep conformity. 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 struggle to escape from a gambler. But the director works a startling variation on 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 protocols of continuity, show how novel effects can be built out of received devices. What E. H. Gombrich calls stylistic schemas, the traditional patterns that form the 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 découpage out of the premises of Hollywood continuity. Ozu retains all the devices of the American approach: establishing shot, analytical editing, reverse angles, eyeline matches, matches on action, and enlargement of the scene from distant views to close-ups. He also grants the need for spatial redundancy, so character placement and movement are signaled through real-world cues like eyelines and our knowl­ edge of how people sit and stand in certain situations. Ozu absorbs these devices into a unique set of cinematic coordinates. By setting his camera consistently lower than what it films, he brings the bottom half of the shot into greater prominence and creates a striking graphic similarity across images. He intensifies this similarity by compositional devices, such as arranging figures so that they match in position and contour. Above all, he conceives of dramatic space as a circle rather than as a line,

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

343

Figure 12.11 Instead of cutting from one swordsman to a reverse angle of his ninja antagonist, The Mysterious Edogawa Ranzan (1937) cuts . . .

Figure 12.12 . . . to a gambler attacking a woman, seen from an angle that answers to the one of Figure 12.11. This is followed by...

Figure 12.13 ... a reverse angle of the woman, and only then...

Figure 12.14 . . . doe» ihe director supply a shot of the ninja that replies to the first image (Figure 12.11).

as in Hollywood’s conception of the axis of action. Ozu can then place his camera at various spots around the circle, but not just any old spots; he typically shifts the angle of view in multiples of 45 degrees.11 An early example of his ambitions, from Tokyo Chorus (1931), presents the rudiments of a system that he would tweak and refine across 30 years (Figures 12.15-12.18). The result is a unique creation, a world that invites the viewer to explore cinematic space as never before. Yet it remains completely legible for viewers accustomed to classic continuity norms. It’s worth mentioning another recasting of a standard device. Even after the triumph of continuity in the American manner, Japanese directors cultivated unusually distant and dense establishing shots. In Figure 12.19, Sadao Yamanaka arrays a remarkable number of faces and bodies across and into the space, each element precisely posi­ tioned. Most directors don’t play out entire scenes in such framings, preferring to cut into and around them as per analytical editing strategies; but such packed shots are more salient than their parallel framings in Western cinema. The denouement of The Groom Talks in His Sleep (1935), a charming marital comedy by Gosho Heinosuke,

344

P o e tic s o f C i n e m a

Figure 12.15 lok\o ( horu> (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 12.17 Tokyo Chorus: A closer view of the husband. . .

Figure 12.18 . . . is followed by a compara­ bly framed view of the wife, with her head in almost the same spot of the frame as his. In later films, Ozu’s graphic matching of bodies and locales would become eerily exact.

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). Flousehold 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.

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

Figure 12.19 InYamanakaSadao’sKochiyama Soshun (1936), as in Ozu’s shots, a low height accentuates foreground elements like the jar on the right and brings out layers of depth.

345

Figure 12.20 A dense establishing shot in The Groom Talks in His Sleep (1934).

Figure 12.21 A slightly high angle show­ ing samurai dotting the frame is a common image schema of the period (Crow of the Moonlight Night, 1939).

Voices in the Dark 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 Noël 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

P o e tic s o f C i n e m a

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.14The 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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —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-titlespeaker 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 s 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

P o e tic s o f C i n e m a

Figure 12.22 Fallen Blossoms (1938): The heads of the geisha in the foreground are turned from us or blocked, so as to concentrate attention on the centered, frontal women in the background as they talk.

Figure 12.23 A Mother Is Strong (1939): A wide-angle lens shows Ryuku entering the house,

Figure 12.24 A Mother Is Strong: The cam­ era pans left with her as she looks in and finds her mother ill. The doctor looks up.

Figure 12.25 A Mother Is Strong: The cam­ era pans still further left as she pauses in the 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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

349

Figure 12.26 Heat of the Earth (1938): Decorative use of aperture framing.

Figure 12.27 As the father is taken away through the gate in the far distance, the boys ask their mother what is happening (Children in the Wind, 1937).

Figure 12.28 First Steps Ashore (1932): Characters slotted into a decorative screen’s portholes.

Figure 12.29 First Steps Ashore: Later, the same divider yields a new array of figures.

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.

P o e tic s o f C i n e m a

350

Figure 12.30 Partial disclosure and decen­ 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 12.31

Lily of the Valley (1935).

Figure 12.32

Figure 12.33

Lily of the Valley.

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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —1945

Figure 12.34

Record of My Love (1939).

Figure 12.35

Figure 12.36

Record of My Love.

Figure 12.37

351

Record of My Love.

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

P o e tic s o f C i n e m a

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 bunko). 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 (bunko) 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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —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

P o e tic s o f C i n e m a

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 Japan’s 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 twentiethcentury Japanese culture. We should also consider that when an artist draws on preMeiji 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

V i s u a l S t y l e i n J a p a n e s e C i n e m a , 1 9 2 5 —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

Figure 12.38 Massive and geometrical crowd scenes characterize The Abe Clan (1938).

P o e tic s o f C i n e m a

Figure 12.39 Iemitsu and Hikozaemon (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 dia