Figures Traced in Light: On Cinematic Staging 9780520938007, 9780520241978

Citation preview

FIGURES TRACED IN LIGHT

figures traced

DAVID BORDWELL

in light On Cinematic Staging

UNIVERSITY OF CALIFORNIA PRESS

/

Berkeley Los Angeles London

University of California Press Berkeley and Los Angeles, California University of California Press, Ltd. London, England © 2005 by the Regents of the University of California Chapter 2 is adapted from David Bordwell, “La Nouvelle Mission de Feuillade; or, What Was Mise-en-Scène?” from Velvet Light Trap, no. 37 (1996): 10–29. © University of Texas Press. All rights reserved. Library of Congress Cataloging-in-Publication Data Bordwell, David. Figures traced in light : on cinematic staging / David Bordwell. p. cm. Includes bibliographical references and index. ISBN 978-0-520-24197-8 (pbk. : alk. paper) 1. Motion pictures—Production and direction. 2. Auteur theory (Motion pictures). I. Title. PN1995.9.P7B635 2005 791.4302'33—dc22

2004001354

Manufactured in the United States of America 14 10

13 9

12 11 10 09 8 7 6 5 4

08 3 2

The paper used in this publication meets the minimum requirements of ANSI/NISO Z39.48–1992 (R 1997) (Permanence of Paper).

For Sally

The publisher gratefully acknowledges the generous contribution to this book provided by the General Endowment Fund of the University of California Press Associates.

We have long come to realize that art is not produced in an empty space, that no artist is independent of predecessors and models, that he no less than the scientist and philosopher is part of a specific tradition and works in a structured area of problems. The degree of mastery within this framework, and at least in certain periods, the freedom to modify these stringencies are presumably part of the complex scale by which achievement is being measured.

ERNST KRIS

I create neither for the audience nor for myself. I create for solutions. GYÖRGY LIGETI

Contents

Plates ( follow page 180) •

Acknowledgments •

xi

1

Staging and Style

2

Feuillade, or Storytelling

3

Mizoguchi, or Modulation

4

Angelopoulos, or Melancholy

5

Hou, or Constraints

6

Staging and Stylistics Notes

•

271

Index

•

301

1

•

•

43 •

186 •

238

83 •

140

Acknowledgments

What might seem to be a fairly straightforward project—a plain thesis, four directors, a few dashes of theory—has turned out to be complicated. Most important, it involved getting access to 35 mm prints of movies from filmmakers who aren’t as widely known as they ought to be. So I must thank, as usual, the Royal Film Archive of Belgium, its hardworking staª, and in particular its director, Gabrielle Claes. Without Gabrielle’s generous cooperation, this book could never have been written. The project also benefited from screenings of Feuillade restorations and silent Japanese films at Italy’s annual silent-film event Le Giornate del Cinema Muto (formerly held in Pordenone, now in Sacile); so thanks to the Pordenone posse for their ongoing and exuberant event. Back in the mid-1970s, my work on Mizoguchi was helped enormously by Ken Wlaschin and the staª of the National Film Archive of the British Film Institute and by Charles Silver of the Museum of Modern Art, New York; in the 1980s, by Paul Spehr, Pat Loughney, and Katherine Loughney of the Library of Congress, Washington, DC; by Kato Michiro, who kindly arranged for me to visit the 1997 Kyoto Film Festival; by Matsuda Yutaka of Matsuda Film Productions; and by Ohba Masatoshi, Okajima Hisaichi, and Saiki Tomonori of the National Film Center in Tokyo. My research on Hou Hsiao-hsien was aided in the earliest days by Wendy Lidell’s US distribution of two Hou masterworks. In 1999 I was lucky enough to receive an invitation to the Taipei Film Festival; for this I am indebted to my hosts, Robert Chen and Gene Fon-Liao of the National University of the Arts. At the festival Hou Hsiao-hsien and Yee Chin-hyen obligingly answered my questions about Taiwanese filmmaking practice. Many students there also made my visit unforgettable. In the United States Rodney Hill and

ACKNOWLEDGMENTS

xi

Suzanne Fedak arranged access to Hou’s films circulated by Wellspring Media. More recently, the Taipei Film Archive and the Information Division of the Taipei Economic and Cultural O‹ce in Chicago went to extraordinary lengths to help me. I am very grateful to Ben Shiao and Carol Li of the Information Division for their friendly assistance. Mr. Shiao and Ms. Li also made it possible for our Center for Film and Theater Research to acquire over 120 Taiwanese films, a magnificent gift that will enable researchers to study this important cinematic tradition. I must also thank those who have shared insights, helped me get materials, and provided critical feedback: Barbara Anderson, Joseph Anderson, Günter A. Buchwald, Michael Campi, Jerry Carlson, Peggy Chiao, Julie D’Acci, Fujiwara Toshi, Torben Grodal, Tom Gunning, Hasumi Shiguehiko, Sam Ho, Andy Horton, Nicole Huang, Charlie Keil, Kitatani Kenji, Brent Kliewer, András Bálint Kovács, Li Cheuk-to, Makino Mamoru, Bill Paul, Tony Rayns, Donald Richie, Shu Kei, Jeª Smith, Peter Tsi, Malcolm Turvey, Edward Yang, and Yeh Yueh-yu. Here in Madison many members of the Film Studies unit—J. J. Murphy, Lea Jacobs, Ben Brewster, Kelley Conway, Tino Balio, Erik Gunneson, Maxine Ducey, Joe Beres, Andrew Yonda, and Paddy Rourke—aided me greatly. My colleagues Vance Kepley and Ben Singer, and my longtime friend Peter Rist, oªered detailed critiques of the manuscript, and the book is better for their eªorts. Jake Black lent his digital expertise to sprucing up my stills. I’m particularly pleased to acknowledge four students. I learned a great deal from directing the dissertation work of Don Kirihara on 1930s Mizoguchi, Darrell Davis on interwar Japanese cinema, and Jim Udden on Hou Hsiao-hsien. I also benefited from supervising the master’s degree of Fujiki Hideaki, who discussed Mizoguchi at length with me. All four also located and translated useful sources of information. I will be happy if my chapters on Mizoguchi and Hou reflect some of the excitement I felt in working with these gifted young scholars. Jonathan Frome assisted me substantially by building and maintaining my Web site, www.davidbordwell.net. On this site the reader will find supplementary essays exploring various issues raised in this book. This book would not have been as wide-ranging as it is without the financial support of a Hilldale Professorship in the Humanities at the University of Wisconsin–Madison. I thank the university for this generous honor. At the University of California Press Eric Smoodin, Sheila Levine, and Mary Francis were kind enough to support the project. A version of chapter 2 appeared in The Velvet Light Trap, no. 37 (spring 1996), and chapter 4 is based on an essay for The Last Modernist: The Films of Theo Angelopoulos, ed. Andrew Horton (New York: Praeger, 1997). I am grateful for permission to republish this material. Kristin Thompson read every chapter painstakingly, and she encouraged me to keep going when other demands seemed overwhelming. She, too, is a figure of light for me. Sally Banes did not comment directly on the work oªered here, but she, like her husband, Noël Carroll, has been a dear friend and a model of intellectual passion for me and many others. The book is dedicated to her.

xii

ACKNOWLEDGMENTS

1 STAGING AND STYLE

You are a film director. Today the script requires four of your characters to have a conversation around a dinner table. How might you stage and shoot it? You could place the camera a fair distance back and record the scene in one continuous shot. This can be e‹cient if your actors know their lines well; no need to change camera positions and relight several shots. But people tend to arrange themselves on all sides of a table, so unless you set up a Last Supper composition, at least one person will be facing away from us. What if that person has some important lines? Even if that person merely reacts to others, don’t we want to see his or her face occasionally? Finally, what if during postproduction you decide that the pace is too sluggish? If you want to eliminate some dead spots, you’ll have to cut, but if you film one long take you have no shots to insert. So you consider presenting your dinner-table scene in several shots. You might give each character a separate shot—what American filmmakers call a “single”—whenever that character speaks or reacts. This was a common solution in the 1920s, and it worked well for silent film, since a single could specify the source of a dialogue intertitle, and facial expression added important information about how we are to understand the line (see Figs. 1.1–1.3). In Japan Ozu Yasujiro happily refined this approach well into the sound era (Figs. 1.4–1.5). Still, cutting together many singles may create a choppy feel, and it doesn’t cover every eventuality. When two characters seated side by side start to chat, you may want a shot that includes both. Then singles can be reserved for underscoring a key line or facial expression. Sometimes, too, you will want to remind the audience of where

STAGING AND STYLE

1

1.1 A prototypical solution to the dinner-table problem in the silent era: a long shot shows the entire family.

1.2 Then “singles” take over, first isolating the daughter . . .

1.3 . . . then her father (Are Parents People? 1925).

1.4 Late Autumn (1960): conversation around the dinner table . . .

characters are sitting by angling the camera to show a bit of one, say a shoulder or an elbow, in the foreground while the shot favors the character opposite. So you’ve decided to break your dinner-table scene into a variety of shot scales, some showing several characters, others picking out one or two. In eªect, you’ve come up with the solution developed in the “classical” age of sound filmmaking in most countries around the world (Figs. 1.6–1.8), and it remains one salient option.1 Whether you shoot only singles or a mixture of singles and more distant framings, you confront some problems. For the sake of speed and economy it’s best to film shots taken from a single camera setup all at once, regardless of where they’ll eventually cut into the scene. If we are going to use several two-shots of Anna and Bart in the course of the scene, we should shoot all of those in a batch. Then we shift the camera angle to film all our shots of Anna alone before shifting again to shoot all our singles of Bart. In sum, we will shoot each actor’s lines, gesture, and glances “out of continuity.”

2

STAGING AND STYLE

1.5 . . . handled in singles in the US silent manner.

1.6 Meet Me in St. Louis (1944): the irate father, flanked by his shamed son and daughter . . .

1.7 . . . confronts the womenfolk who have deceived him. Minnelli covers the scene by splitting the table in half.

1.8 A closer view emphasizes the father’s protest: “Just when was I voted out of this family?”

Filming around a table in this fashion consumes a lot of production time. Kevin Bacon, actor turned director, complains: “It’s a real pain in the ass to shoot dining-room scenes because you’ve got to go all the way around the room with coverage.”2 This solution is also hard on the actors, who must break their performances into pieces taken out of chronological order. Moreover, as the scene progresses, people are eating, drinking, putting down napkins, and indulging in other bits of business. If an actor waves her fork in a two-shot, she ought to wave it the same way in the over-the-shoulder shot, which might be filmed days later. If an actor drinks half a glass of wine in one shot, the glass had better be half-empty in the next shot (even if that shot’s setup was taken days before). “If you film such scenes without preparation,” notes Kitano Takeshi, “the moment you change camera position, the hands, for instance, will be positioned diªerently, and you risk not being able to match the diªerent takes.”3 The script supervisor (formerly known as the script girl) is charged with keeping track of shot-to-shot continuity, but there are

STAGING AND STYLE

3

1.9 Reservoir Dogs (1992): as the camera arcs around the table, it picks up one robber explicating a Madonna song. Shoulders gliding by in the foreground help mask cuts.

1.10 Another robber, still seen in a lateral tracking shot, reacts skeptically.

so many details involved that most dinner-table scenes will contain mismatches.4 “If you’ve got a lot of people around a table,” remarks Anna Campion, “it can honestly drive you mental.”5 The problems of matching eyelines, props, and movements are so great that contemporary filmmakers often take a diªerent tack: filming the meal with a camera circling the table to pass from character to character at the proper moment. You might use rather lengthy arcing movements, as Woody Allen does in Hannah and Her Sisters (1985), or you might try for the sort of briefer moving shots that Quentin Tarantino uses in the opening of Reservoir Dogs (1994, Figs. 1.9–1.10). One contemporary directing manual recommends the circling camera as the best way to dodge the numerous matching problems raised by “the dinner-table conundrum.”6 Yet, like all solutions, this one closes oª other options. If you handle the conversation in lengthy spiraling shots, the pace of the scene can’t later be tightened or slowed by cutting. If you cut up the roaming shots, then the old di‹culties of continuity reappear. Reservoir Dogs avoids some of these problems by tightly framing the gang members’ faces so that we can’t see the food on their plates, and their hands become visible only sporadically. Tarantino has explained that he used the moving camera so that “you don’t quite get a bearing on anybody, who’s talking, who are these people, you’re slightly confused.”7 In making these choices, however, he has relinquished actors’ byplay with food and drink, an important component of Ozu’s and Minnelli’s scenes. Every choice eliminates certain possibilities. The dinner-table conundrum can stand as an emblem for many types of scenes. You face the same decisions if you’re shooting a poker game, or a gang planning a heist, or o‹cers in GHQ poring over a map of enemy territory, or mourners gathered around a grave. Whenever three or more characters surround a common site, the problems of staging and shooting and editing will be similar. To solve these problems, filmmakers have developed fairly standardized craft practices—stylistic routines that I’ll be calling schemas.8 Classical continuity filming provides two principal schemas: the silent-era approach emphasizing singles, and the more mixed sound-cinema strategy. The circular tracking shots now common for such scenes provide a third schema, also grounded in assumptions about spatial continuity.

4

STAGING AND STYLE

Are these the only schemas available? Perhaps we passed over one option too quickly. Suppose we shot our dinner-table scene with a stationary camera, set somewhat back, in a single take, as in the South Korean film The Virgin Stripped Bare by Her Bachelors (a.k.a. Oh, Su-jeong! 2000). Our three main characters are at a drinking party, with everyone gathered around one low table. The entrepreneur Jae-hoon and the television director Yeong-su are both carrying on aªairs with Yeong-su’s assistant, Su-jeong. Late in the evening we get a shot showing all three ranged along one side of the table, with another woman in the left foreground and a young man turned from us in the right foreground (Fig. 1.11). Director Hong Sang-soo presents the entire scene in an unmoving take lasting ninetysix seconds. He thereby forfeits the advantages of editing (tightening up the pace in postproduction, cutting to closer views to underscore a line or reaction) and camera movement (highlighting one reaction after another, “giving energy” to the scene). And although the composition turns two characters away from us, it favors three other figures rather than only one, so Hong will have to rely almost entirely on ensemble performance. At the start of the scene the somewhat fussy Jae-hoon, who has invested in Yeong-su’s film, asks how preproduction is going, and the drunken Yeong-su replies sarcastically, “Oh, now you’re interested in me?” (Fig. 1.12). Since the first part of the scene is about the two men’s disintegrating friendship, the composition gives them the privileged position, with Jae-hoon in the center of the frame and Yeong-su right beside him, facing us. Su-jeong becomes a secondary presence, her eyes lowered and her face shielded by strands of hair, but one can sense some embarrassment in her posture. Jae-hoon lowers his eyes before Yeong-su’s onslaught, leaving Yeong-su the most visible face at the table. He demands that Su-jeong serve him, as if flaunting his power over her, and Jae-hoon quietly rises and leaves the table (Fig. 1.13). This is an attention-getting movement, centered in the frame and underscored by Yeong-su’s momentary lowering of his eyes. Su-jeong watches Jaehoon go out of frame, and the woman in the left foreground also turns her head, in the process making Su-jeong’s response more visible (Fig. 1.14). Su-jeong slides into the center slot (Fig. 1.15). This marks a second phase of the scene, in which she’ll display distress at Yeong-su’s aggressive drunkenness. He urges her to drink, and as if echoing his demand, the woman on the left picks up a cognac bottle and inspects it before pouring herself another (Fig. 1.16). Su-jeong protests softly that she doesn’t want any more. As she does, the young man in the right foreground puts a restraining hand on the cognac bottle (Fig. 1.17), but by then the other woman has emptied it. Finally, Yeong-su accepts Su-jeong’s unwillingness to drink and says, “Okay, my turn,” before gulping down another glassful (Fig. 1.18). The scene ends with Su-jeong pouring him a fresh drink (Fig. 1.19). Hong’s manner of handling requires some fairly careful choreography, most apparent when Su-jeong’s face becomes visible just as Jae-hoon leaves (Fig. 1.14) and when the youth in the right foreground shifts rightward slightly to let us see Yeong-su drain his glass greedily (Fig. 1.18). The trouble that Hong and his players have taken yields some

STAGING AND STYLE

5

1.11 The Virgin Stripped Bare by Her Bachelors (2000).

1.12 The Virgin Stripped Bare by Her Bachelors.

1.13 The Virgin Stripped Bare by Her Bachelors.

1.14 The Virgin Stripped Bare by Her Bachelors.

distinct advantages. By placing the camera somewhat far back, the framing builds the texture of the scene out of both faces and gestures—the repeated acts of pouring, of checking the cognac bottle, of trying to restrain the second woman. The episode benefits from the simultaneous presence of action and reaction: while Yeong-su lashes out, we can see both Jae-hoon’s shame and Su-jeong’s discomfiture, reactions that would have been rendered as successive events if they had been presented in single shots. Cutting in to singles would also not have permitted Hong to include the sort of small bodily echoes that researchers tell us are a basic part of human interaction. Conversing people execute a dance of mirrored postures and gestures.9 Here the echoes underscore the moment-by-moment turns of action. Yeong-su lowers his head as Jae-hoon leaves; the woman in the left foreground turns her head in a mimicry of Su-jeong’s head movement (Fig. 1.14); the same woman slides aside just a little bit at the same moment Sujeong slips into Jae-hoon’s place (Fig. 1.15). Similarly, while Yeong-su urges a reluctant Su-jeong to drink, the young man in the foreground tries to keep the other woman from drinking more (Figs. 1.16–1.17). The nearest couple’s byplay remains subsidiary because of the way the frame emphasizes the three protagonists. At first, by pivoting two characters’ backs toward us, Hong denies the couple the emphasis that cutting or tracking shots might give them; but he then makes a virtue of exactly their secondary status. If Hong had cut around the table to show a two-shot of the woman inspecting the bottle and of the boy trying to keep her from drinking, he would have underscored actions that here form a subdued counterpoint to the thrust of the scene. The long take from a middle-

6

STAGING AND STYLE

1.15 The Virgin Stripped Bare by Her Bachelors.

1.16 The Virgin Stripped Bare by Her Bachelors.

1.17 The Virgin Stripped Bare by Her Bachelors.

1.18 The Virgin Stripped Bare by Her Bachelors.

1.19 The Virgin Stripped Bare by Her Bachelors.

distance position has produced what Charles Barr calls “gradation of emphasis,” a scale of dramatic overtones resonating with the shot’s main action.10 Like all stylistic choices, Hong’s approach closes oª other possibilities, but the options he embraces belong to an important tradition in cinematic staging, one whose implications go far beyond dinner-table situations. That tradition, running back to the earliest days of cinema, has become more and more neglected. Few filmmakers today subscribe to it, and few film scholars probe its possibilities. It is what this book is about.

Over most of cinema’s history, staging was crucial to moviemaking. From the early 1900s to the 1970s, directors working in film industries on every continent were expected to

STAGING AND STYLE

7

turn the script into scenes, and that task involved plotting, moment by moment, the dramatic interactions of characters in space. In studio systems, staging was the foundation of the director’s craft. There were specialists in lighting and costume, set design and cutting, but the director was responsible for shaping the performances across the arc of the scene. There were no film schools to teach this skill. Directors picked it up through observation, informal advice, and trial and error. Some mastered only the basics, whereas others refined their craft to startling levels of subtlety. They seldom talked about it to outsiders; one finds almost no written record of the schemas guiding staging. We can reconstruct them only through watching the movies. An unassuming art, it passed unnoticed by audiences. What do we remember from a movie? Some settings, the actors’ faces, flashes of pathos and humor, a striking line or gesture. Yet those vivid instants burst out of a quiet, insistent pressure that prepares us for them, like the small movements and prolonged stillnesses that set into relief Jae-hoon’s abrupt departure from the table (Figs. 1.12–1.14). It is indeed hard to watch for such things. We don’t know how the action will unfold, so we cannot usually spot the minor-key preparatory work. For reasons I’ll sketch later, we rightly concentrate on the faces, the words, and the gestures, testing each one for its relevance to the ongoing story action. Yet the faces (and the bodies), the words (and the reactions to the words), and the gestures (and the interplay of gestures) are all working together. At every instant, in most storytelling cinema, cinematic staging delivers the dramatic field to our attention, sculpting it for informative, expressive, and sometimes simply pictorial eªect. We don’t notice it, but it aªects us. Nor have critics helped us appreciate this skill, and there are historical reasons why. As film journalism developed in the 1910s and 1920s, many of the best writers sought to prove that cinema was a distinct art, owing nothing of consequence to theater. Film staging seemed “theatrical,” whereas the emerging technique of editing seemed “purely cinematic.” Thanks to editing, the filmmaker could flash us from one locale to another, skip over stretches of time, draw metaphorical comparisons, and generate a rhythm purely out of shots of changing lengths. Small wonder that the most thoughtful critics, as well as the most imaginative theorists, usually preached the virtues of cutting and ignored staging. It took another generation of critics, those who grew up with the sound cinema, to try to right the balance. But their revolution was an unfinished one. To this day critics praise a virtuoso piece of editing while remaining oblivious to far more subtle and powerful passages of staging. Some film theorists call cutting “invisible,” but for both ordinary moviegoers and film experts, cinematic staging remains a truly imperceptible art. Which means that we are very far from understanding the robust artistic resources of staging-driven cinema. In Hong’s drinking scene the filmmaker uses small glances, gestures, and changes of position to give the sustained shot a dramatic arc. There is suspense as we wait for a character’s reaction, surprise when a new piece of information comes to light in the frame. There is also the gradation of emphasis, the possibility that secondary aspects of the image can reinforce or play oª the major action. A shot can also

8

STAGING AND STYLE

develop its own distinct expressive dimensions (lassitude, melancholy, or delicacy) and its own pacing (fleet or lugubrious). More strikingly, the most adept filmmakers in this tradition exploit all these possibilities while showing that one need not forfeit some of the resources of cutting. The cutting/staging dichotomy is too coarse to be of great critical use. Admittedly, directors working in the tradition with which I’m concerned have found ingenious equivalents for close-ups and cut-ins. But editing can cooperate, in subtle ways, with the sustained shot and the dense image. Indeed, a sustained shot gives the cut a particular saliency, as the change of shot can be reserved for maximum eªect. Very often, patterns of staging are bound up with patterns of editing. This book’s central question, then, is, What aesthetic possibilities emerge when cinematic staging is employed in such ways? This is not exactly to ask about acting or ensemble performance, although both impinge on the phenomenon I’m considering.11 So, too, does camerawork, particularly the act of framing. Perhaps Eisenstein captures the idea best in speaking of mise-en-shot, the presentation of an action through the film image. Mise-enshot is at once theatrical, involving staging, and pictorial, since the screen, like a painting, presents the viewer with a framed vertical plane. By analyzing mise-en-shot I hope to disclose a treasury of quite fine-grained options, some so delicate that they normally escape our notice. The impulse for this inquiry stems from my earlier book On the History of Film Style. That was an exploration of the historiography of style, and in the last chapter I sketched a history of staging in depth that sought to avoid some of the drawbacks of earlier accounts. That chapter tried to show that we can usefully consider film techniques as solutions to concrete problems of representation and that filmmakers inherit many solutions from the trial-and-error eªorts of their predecessors. In thinking further about the ways one could move actors around the frame, I was drawn to examine four filmmakers who worked in the staging-centered tradition. So I decided to examine in detail the staging strategies of Louis Feuillade, a master of silent-cinema storytelling; Kenji Mizoguchi, melodramatist of Japan; Theo Angelopoulos, the Greek filmmaker who synthesizes several features of European modernism; and Hou Hsiao-hsien, who explored staging strategies within a Taiwanese context. Far from merely illustrating my earlier account, however, the study of these directors reveals that the tradition is even more supple and subtle than I had realized. Figures Traced in Light, then, not only oªers concrete illustrations of ideas proposed in On the History of Film Style; it also refines and extends them. Certainly we could learn from scrutinizing many other filmmakers, from the 1910s masters through Keaton, Dreyer, and Eisenstein and on to Chantal Akerman, Otar Iosseliani, and Béla Tarr. I select my quartet because they both exemplify some typical norms and display some unusual exploitations of them. By focusing on a few outstanding exponents of the style we won’t exhaust it, but we can probe certain possibilities in depth. I confess, as well, that I was drawn to less-examined figures. There’s still a lot to say about Renoir, for example, but his staging has been studied a good deal more than that of Mizo-

STAGING AND STYLE

9

guchi, who is no less great a director. Of course, these four directors are not exactly unknowns. Feuillade is by common consent one of the most important directors of the 1910s, and critical opinion of Mizoguchi could hardly be higher. Angelopoulos and Hou have both backers and detractors, but if critical acclaim counts for anything, they are among the half-dozen most-praised directors working today. The fact that these directors’ works are for the most part unavailable in adequate video copies says less about their stature than about the tastes of the market. All that granted, though, this is not a set of auteur essays; there are many aspects of each director’s technique and worldview that I set aside. Indeed, by looking at how these filmmakers exploit and explore traditions of staging, we open our inquiry to matters not usually broached in auteur studies. Chapter 2’s examination of Feuillade also shows that many silent-era filmmakers revealed dazzling resources in the fixed shot. Considering Mizoguchi (chapter 3) not only demands that we consider the Japanese filmmaking context but also raises issues of how certain staging strategies became integrated with sound filming and camera movement. In chapter 4 I treat Angelopoulos as a figure who synthesizes trends of the 1950s and 1960s, bending certain options to the purposes of austere modernism. A study of Hou Hsiao-hsien (chapter 5) allows us to examine how kindred stylistic options became a defining feature of Asian filmmaking. Apart from tracing an important and somewhat neglected line of film artistry, I have other goals in view. Most abstractly, I hope to show the viability of a poetics of cinema, the study of how films work (narratively, stylistically) to shape the audience’s experience. Cinematic staging oªers a rich repertory of techniques that answer nicely to analysis from the standpoint of poetics. I don’t intend to make this examination of “cinepoetics” a dry treatise, however; I hope that this book will further film appreciation for the interested moviegoer. Most scholars concerned with particular films have concentrated on interpreting them at a fairly high level of generality. The scrutiny of style that is commonplace in art history or musicology has still not become well established in film studies. In particular, film scholars have not helped viewers appreciate what Svetlana Alpers and Michael Baxandall have called “pictorial intelligence.”12 The demand is all the more pressing in that filmmakers expend enormous energies on making things look a certain way and not others. A poetics of cinema can make us more sensitive to the films we see, encouraging us to recognize the skill, and sometimes the brilliance, of the people who make them. The study of cinematic staging is an ideal way to achieve just this sensitivity. Once we start to look closely at even a simple scene like Hong’s drinking bout, our awareness of filmmaking craft and the capacities of the medium deepens. Most optimistically, I hope that filmmakers—particularly young ones—will here learn about traditions flagrantly ignored in most movies on view today.

Several thinkers have laid a foundation for understanding the traditions with which I am concerned. The central figure is the great French critic and theorist André Bazin, who

10

STAGING AND STYLE

formulated a conception of cinema based on the photographic properties of the medium, its ability to record events in continuous space and time. In tracing out a history of film style Bazin distinguished between “directors who put their faith in the image” and “directors who put their faith in reality.” Image-based directors built their style around painterly manipulations of the image (such as in German Expressionism) or juxtapositions of images (such as the intellectual montage of Soviet filmmakers). By contrast, directors who put their faith in reality made cinematic art out of certain phenomenal features of the world, such as temporal continuity and spatial adjacency. Bazin believed that F. W. Murnau, Jean Renoir, Orson Welles, William Wyler, and the Italian neorealists built very diªerent styles out of cinema’s power to capture the concrete relations of people and objects knit into the seamless fabric of reality.13 Bazin’s writings of the 1940s and 1950s furnish us a generous array of analytical concepts, most notably bearing on certain techniques that earlier aestheticians had neglected: the long take, camera movement, and composition in depth (profondeur de champ). The reality-based directors tended to rely on these techniques, rather than editing and closeups, to capture the spatiotemporal continuity of the world in which we live. Very shortly, though, Bazin found his preferred techniques championed by a slightly younger generation of critics in love with Hollywood. To Eric Rohmer, Jacques Rivette, François Truªaut, and other writers at Cahiers du cinéma, a filmmaker’s respect for reality was less important than an ability to create expressive artifice. Bazin’s key filmmakers, such as Murnau and Renoir, became stepfathers to Nicholas Ray, George Cukor, Otto Preminger, and above all Alfred Hitchcock and Howard Hawks. These Hollywood directors also knew how to use long takes, camera movement, and composition in depth but not principally as a way of capturing phenomenal reality; now the techniques were vehicles for each director’s unique conception of the world. These were auteurs, presenting their personal visions on film, and their chief means of doing so was something called mise-en-scène. Few terms in film aesthetics are as polyvalent as this one. In French it chiefly means what we in English call “direction,” and its sources lie in the theater. Mettre en scène is to “arrange the action onstage,” and it implies controlling performance, lighting, setting, costume, and the like. Since Bazin, some French critics have treated mise-en-scène as simply the entire process of directing a film, including everything from staging to editing and adding music. But the Cahiers critics tended to narrow the term: following Bazin, they counterposed mise-en-scène to editing (conceived as either aggressive montage or more standardized continuity cutting). A mise-en-scène director tended to avoid cutting, creating significance and emotion chiefly by means of what happened within each shot. The young critics pushed beyond Bazin, treating mise-en-scène as both process and product. Mise-en-scène comprised all the factors that the director could control during shooting—the performances, the blocking, the lighting, the placement of the camera. Thus Hollywood directors, who may not have worked on the film’s script and might have no say in editing, could still decisively shape the film at the mise-en-scène phase. The term also referred to the result onscreen: the way the performers fitted into the frame,

STAGING AND STYLE

11

the way the action exfoliated over time. Alexandre Astruc, a contemporary of Bazin’s who was an important influence on the younger critics, noted: “We have come to realize that the meaning which the silent cinema tried to give birth to through symbolic association exists within the image itself, in the development of the narrative, in every gesture of the characters, in every line of dialogue, in those camera movements which relate objects to objects and characters to objects.”14 For François Truªaut, mise-en-scène comprised “the camera position, the angle selected, the shot’s length, an actor’s gesture,” in sum, “at once the story that is being told and the manner of telling it.”15 (See box.) The Cahiers critics celebrated Rossellini, Visconti, Cocteau, Bresson, Tati, and other European filmmakers, but their most controversial strictures on mise-en-scène were provoked by American studio films. Some critics argued that Hollywood resembled a Renaissance capital, where experienced craftsmen pooled their talents to create brilliant art.16 At the same time, the technological innovations of color and widescreen seemed to favor spreading out bodies and objects within an all-encompassing space. “Yes,” remarked Rivette, “ours is the generation of CinemaScope, the generation of metteurs en scène finally worthy of the name: moving the creatures of our mind around the unlimited stage of the universe.”17 Above all, an auteur’s work yielded to the sort of thematic interpretation that had proven valid for literature.18 Mise-en-scène became a vehicle of abstract meaning; if Sartre had shown that every style conceals a metaphysic, then it was possible to see Hawks and Hitchcock as conveying profound thoughts about human conduct. A rival school, centered around the magazine Présence du cinéma (1959–1967), objected to Cahiers’ canon but celebrated the powers of mise-en-scène in an even more delirious fashion. Michel Mourlet was the head of the MacMahonists, so called because of the theater the group frequented. Mourlet’s 1959 manifesto, “Sur un art ignoré,” reiterated Bazin’s conception of cinema as an art based in recording and pushed to a limit certain Cahiers writers’ belief that the cinema was destined to be a “classical” art. For Mourlet flagrant editing, bizarre camera angles, and strained lighting eªects ran against film’s inherent bias toward transparent rendering. Cahiers favorites like Hitchcock and Welles, even Hawks, relied too much on tricks of camerawork and cutting. “The arrangement of actors and objects, their displacements within the frame—this must express everything.”19 Fritz Lang, Preminger, and very few others had mastered that sober lucidity that could hypnotize the spectator by gestures, looks, and “the tiniest movements of faces and bodies.”20 The broader Cahiers conception passed into US film culture during the 1960s, when Andrew Sarris argued for the worth of mise-en-scène, a technique that placed “the emphasis on the spectacle itself rather than on the dialectical relationship of separate shots and images.”21 In Britain the journal Sequence (1946–1952) had disseminated some ideas resembling Bazin’s, but it was the critics around Movie magazine (founded 1962) who pioneered the detailed analysis and interpretation of mise-en-scène.22 The essays and monographs of Robin Wood were perhaps the most influential work in this vein, although the earliest in-depth study by a Movie critic would seem to be Ian Cameron’s probing 1962

12

STAGING AND STYLE

essay on Antonioni.23 In Film as Film (1972), Movie’s V. F. Perkins put mise-en-scène firmly at the center of cinematic artistry: “The director’s most significant area of control is over what happens within the image.”24 The tastes of mise-en-scène critics varied somewhat. Rohmer found in Murnau the perfect model of a style that blended a respect for natural appearances with the sort of idealization found in classical art. Rivette championed Hawks and Rossellini, who put the serenely poised body at the center of the mise-en-scène. The MacMahonists created a very diverse canon, ranging alongside Lang, Preminger, and Mizoguchi Kenji directors like Joseph Losey, Cecil B. DeMille, Ida Lupino, and Vittorio Cottafavi (perhaps best known for Hercules Conquers Atlantis, 1961). The Movie critics, who could wring a dense page of analysis out of the first shot of Marnie, sought films that created the sort of intricate implication that literary critics found in John Donne or Wallace Stevens. Sarris was more mystical, finally at a loss to define mise-en-scène except through a hallucinatory rapture. Watching a little-known Mizoguchi film without subtitles at the Museum of Modern Art, he was swept away: Then at the end the beleaguered heroine walks to a restaurant on a hillside overlooking the sea, and she orders something from a waiter in white, and the camera is high overhead, and the morning mists are bubbling all around, and the camera follows the waiter as he walks across the terrace to the restaurant and then follows him back to the heroine’s table now magically, mystically empty. It is as if death had intervened in the interval of two camera movements, to and fro, and the bubbling mists and the puzzled waiter provide the Orphic overtones of the most magical mise-en-scène since the last deathly images of Murnau’s Tabu.25

Despite diªerences in taste, however, most critics treated mise-en-scène as a knowing yet natural deployment of figures before the camera. Many still do. Thierry Jousse reminded 1994 readers that all the recent talk about crafting tight screenplays should not obscure the deciding factor: “Parallel to the dramatic construction of the plot, which can be written down, there is a whole technique, or rather an art, of impregnation, of the directing of bodies, of choreography, of the occupation of space, of natural movement which remains irreducible to the script.”26 And in 2000 a brigade of French reviewers, when asked to explain what mise-en-scène meant to them, responded that it centered on how the human figure is displayed across time and space. Mise-en-scène, says one, is a demiurgic will to control the entire space of a film, the quest for a perfectly adequate image.27 The concept of cinematic staging, as I’ll be using it in the pages that follow, has undeniable links to the idea of mise-en-scène, and the critical tradition I’ve just reviewed has informed my thinking in ways that should be apparent. Several of my points about Hong Sang-soo’s drinking scene owe a lot to Bazin, especially his emphasis on the dramatic advantages of presenting two pieces of information to us simultaneously.28 Nonetheless, it’s worth briefly signaling some distinctions I’ll be drawing.

STAGING AND STYLE

13

MISE-EN-SCÈNE AND CLASSICAL CONTINUITY CINEMA During the 1920s and 1930s, Hollywood filmmakers blended continuity editing tactics with prolonged shots in which bodies moved gracefully around the frame. Longish takes allowed actors to deliver fairly lengthy lines to others, who could react not only through facial expression but also through gesture and posture. A sophisticated example is the choreography in Howard Hawks’s His Girl Friday (1940). The first extended scene is a quietly virtuosic lesson in how to build a conversation around body language: stance, eyeline, orientation, and gesture. Consider only one passage out of a virtually seamless whole. Hildy has come to tell Walter she’s getting remarried and quitting the newspaper. Walter has just ordered his o‹ce manager to contact the governor and pressure him to pardon the condemned murderer Earl Williams. As Walter launches his eªorts to win Hildy back through both seduction and intimidation, Hawks asks his players to use their hands, knees, laps, and angle of head to plot the moment-by-moment flow of comedy, while his shrewd compositions create and resolve spatial tensions (Figs. 1.20–1.26). Although the Cahiers critics never analyzed scenes in this sort of detail, they intuitively registered a subtlety of which the audience was unaware, and they celebrated that “transparency” as a central quality of “classical” mise-en-scène.

1.20 His Girl Friday (1940): Walter has ordered his cronies out of his o‹ce, and the camera dollies in to frame Walter and Hildy at opposite edges. She stands warily alert; he stands, feet outspread, hands thrust in pockets, sizing her up.

14

STAGING AND STYLE

1.21 Walter strides forward as she teases him— “Well, Walter, I see you’re still at it”—and he sits down, remarking that it’s the first time he’s double-crossed a governor. The diagonal running from upper left to lower right will soon be reversed.

1.22 Cut in to a medium shot of Hildy, singling out her reaction and marking a new phase of the scene: “Would you mind if I sat down?”

1.23 Cut back to the master shot, with Walter still in the lower right. “There’s been a lamp burning in the window for you, honey.” He lifts his knee above the edge of the desk and beckons to her.

1.24 She coolly spurns his invitation, walks to the table, and perches on it. Refusing to acknowledge her change of position, Walter remains facing away from her. He starts to light a cigarette, pointedly not oªering her one and forcing her to ask for it.

1.25 She asks for a match, and he hands her a box of matches. In a kind of spatial suspense, the unbalanced composition holds while Hildy lights up . . .

1.26 . . . before Walter deigns to recognize her, rotating his chair to rebalance the shot. (Compare Fig. 1.21.) He asks the loaded question: “How long is it?”

STAGING AND STYLE

15

Proponents of mise-en-scène aesthetics seldom distinguish among aspects of it, but we can. I take the core technical sense of the term to denote setting, lighting, costume, makeup, and performance within the shot. Some critics would also include camera movement as an element of mise-en-scène, but I think that it’s best to keep it as an independent variable; it’s really a feature of cinematography, the manipulation of the camera, rather than a feature of what is filmed. Three of my directors coordinated camera movement with “mise-en-shot.” Likewise, the prototype of a mise-en-scène image is usually assumed to be a longtake shot exhibiting great depth of field. To be sure, many of my analyses will concentrate on just such images, but we capture the complexities of filmmaking practice most fully by teasing out these techniques along distinct and graduated dimensions. So, for instance, instead of counterposing long takes to brief shots, we should keep in mind that shots can be presented in any duration, from one twenty-fourth of a second to a dozen minutes or more. More pertinent are the shot-length norms relevant for any particular period or group style. Along another dimension, the framing can be close or distant, and there’s no reason to assume that long shots must run long and close-ups must be cut fast. Ingmar Bergman’s films often feature prolonged close-ups (Fig. 1.27). Depth compositions can be found in short shots or long takes, distant framings or medium shots (as in The Virgin Stripped Bare . . . ) or even close framings (Fig. 1.28). Rohmer claims that if actors aren’t on two or more planes, “there is no real mise en scène, no deeper meaning,”29 but we should acknowledge the artistic resources of lateral staging, even if it displays minimal depth (Fig. 1.29). Further, our impression of depth is built out of many spatial features, and these can vary independently, creating significantly diªerent patterns of staging. Deep or shallow, cinematic staging relies on a perspectival projection of space. That projection is most evident in photographic media, but it’s also attainable through nonphotographic means—in classical painting and, most obviously for us today, in digital image-making. All the artistic resources of cinematic staging are in principle attainable in animation or computer-generated imagery, as long as the process retains the perspectival projection characteristic of camera lenses (Fig. 1.30). Here I part company with Bazin and the Cahiers critics: the cinematic image yields a perspectival projection, but that projection need not record a preexisting world. Once a film’s staging no longer need capture an independent space and time, we can’t judge artistic value on grounds of faithfulness to reality. In some film movements, as in the New Taiwanese Cinema of the 1980s, the long take was indeed felt to be a blow struck for realism, but even in that school it yielded other advantages. In the chapters to come I’ll touch on several of these: the ways in which staging can guide our attention across a complex visual field, play hide-and-seek with our expectations, summon up expressive qualities like delicacy or dynamism, and participate in a broader narrative patterning. Writers who admire mise-en-scène tend to look on editing as intrusive and heavy-handed, but

16

STAGING AND STYLE

1.27 Close-up and long take in Bergman’s Persona (1966).

1.28 Layers of depth in close-up in Mikhail Kalatozov’s Letter That Was Never Sent (1959).

1.30 A “deep-focus” action scene in Otomo Katsuhiro’s Akira (1988). 1.29 Relatively shallow space creating dynamic textural and compositional patterns (all the way to the corner of the frame): Kriemhild points to her murdered husband in Lang’s Kriemhild’s Revenge (1924).

I’ll often have occasion to show how staging and editing cooperate. In a context dominated by long takes, cuts take on great importance, and many a depth-oriented director has used editing to maintain or transform spatial relations in the course of a scene. In this respect my approach is close to that articulated by Sergei Eisenstein in his writings and teachings of the 1930s. Although Eisenstein will always be associated with concepts of montage, he also studied staging in unprecedented detail. He distinguished among mise-en-scène, the arrangement of the action as if on a theater stage; mise-en-cadre (“mise-en-shot”), the staging within the frame of the image; and montage, the sequencing of shots. For Eisenstein all worked together to communicate the story situation and to heighten its expressive impact. Of particular interest to us is Eisenstein’s 1933 discussion of staging in depth. He required his pupils at the USSR Film School to stage the

STAGING AND STYLE

17

1.31 Eisenstein’s 1933 Crime and Punishment exercise: depth staging for Raskolnikov’s approach.

1.32 The attack on the pawnbroker becomes a jolting thrust toward the viewer.

scene from Crime and Punishment in which Raskolnikov murders the old pawnbroker. He added a crucial proviso: it had to be filmed in one shot and from a fixed camera position. Ten years before Bazin celebrated Citizen Kane’s profondeur de champ as “a dialectical step forward in the history of film language,”30 Eisenstein was working out an intricate “Wellesian” approach to staging—from the standpoint not of photographic realism but of dynamic storytelling and spectatorial eªect. With his students Eisenstein decided to bring the action steadily toward the viewer. Raskolnikov erratically follows his victim forward (Fig. 1.31), and when he strikes her, knocking her out of the frame, her hands fly up in the foreground (Fig. 1.32).31 Perhaps Eisenstein’s speculations on mise-en-cadre were influenced by an equally provocative proposal from his contemporary and teacher Lev Kuleshov. In 1929 Kuleshov pointed out that because the film image results from a perspectival projection, the playing space constituted a pyramid tipped on its side, the point resting at the camera lens. In the engineering spirit that characterized his thinking Kuleshov proposed that this pyramid could be treated as a “metrical spatial web,” marked oª at intervals that could specify any point in the height, width, and depth of the space. An actor might move her hand from, say, point 62/44/32 to point 82/47/38. It’s a startling anticipation of the 3-D coordinates used to plot volumes and positions in computer-generated imagery, and like Eisenstein, Kuleshov contributed to a vein of thinking about mise-en-scène not as photographic recording but as purely spatial manipulation within a projective geometry.32 We can trace the problem one more step back. Before the Bolshevik Revolution turned Kuleshov into a Soviet filmmaker, he worked as a set designer and actor with Yevgenii Bauer, one of the greatest Tsarist directors. Bauer would have been acutely aware of that visual pyramid that Kuleshov imagined mapping so meticulously. In the 1910s, before Hollywood’s continuity editing came to dominate European cinema, most directors staged their scenes in long, fairly distant takes. Directors were obliged to explore, however intuitively, the possibilities inherent in the unedited image, and that meant understanding

18

STAGING AND STYLE

1.33 The Revolutionary (1917).

1.34 The Revolutionary.

1.35 The Revolutionary.

1.36 The Revolutionary.

the perspectival constraints of the camera eye. Take yet another dinner-table scene, this one from Bauer’s last film, The Revolutionary (1917). At breakfast the old father who has suªered under Tsarist repression is sitting on the far right, with his brother turned from us and his daughter facing us in the center (Fig. 1.33). The son enters and kisses his father (Fig. 1.34) before kissing his uncle (Fig. 1.35) and sitting down on frame left, opposite the old man (Fig. 1.36). The father, who has just read about a Russian victory in the world war, passes the newspaper to his son through the shot’s central zone (Fig. 1.37). As the son reads the news, the daughter leans aside to read with him, blocked out by the uncle and the samovar on the table (Fig. 1.38). Now there is nothing to distract us from studying the son’s reaction. He folds up the paper and passes it back (Fig. 1.39), and a dialogue title gives us his response: “But our party is against continuing the war.” The father looks at the uncle, who turns in a three-quarter view like that of an “over-the-shoulder” shot in conventional editing (Fig. 1.40). The uncle, dismissing the boy’s naivete with a wave of his hand, rotates fully away from us, and again the daughter is blocked (Fig. 1.41). The father rises and kisses his daughter (Fig. 1.42),

STAGING AND STYLE

19

1.37 The Revolutionary.

1.38 The Revolutionary.

1.39 The Revolutionary.

1.40 The Revolutionary.

1.41 The Revolutionary.

1.42 The Revolutionary.

20

STAGING AND STYLE

1.43 The Revolutionary.

1.44 The Revolutionary.

1.45 The Revolutionary.

then goes over to collect his son (Fig. 1.43), who rises and goes out with him (Fig. 1.44). The tension has abated; the diªerence of political opinion isn’t a problem—a fact signaled by the daughter, still in the center, who moves out from behind the samovar to smile over her teacup (Fig. 1.45). Kuleshov completed The Revolutionary after Bauer’s death, so we cannot be sure who directed this sequence. Still, it oªers a vivid example of the practical resources yielded by the camera’s monocular vision. The gradation of emphasis is clear. The staging presents the two principal characters as potential antagonists, facing each other at opposite ends of the table, with the daughter as a secondary observer and the turned-away uncle as of least importance. At crucial moments the uncle and the samovar block the daughter in frame center so that we may concentrate on the two major players at either end of the table. Everyone’s position and movements could be plotted on a “metric spatial web,” but the practical director need know only that in the absence of editing the camera-eye’s limitations can also become strengths. This is why our study of staging can profitably start with early film and concentrate on Feuillade and other ingenious 1910s

STAGING AND STYLE

21

directors. It was in those years that filmmakers realized, however intuitively, that the power to see everything is meaningless unless at certain moments we notice only the most important thing.

I’d love to see directors not shoot so many close-ups. I’d love to see directors start trusting the audience to be the film editor with their eyes, the way you are sometimes with a stage play, where the audience selects who they would choose to look at while a scene is being played, with two characters, four characters, six characters. There’s so much cutting and so many closeups being shot today. . . . It’s too easy for filmmakers. It’s very easy to put somebody up against a wall and shoot a close-up, and they say the words and you go on to the next shot.

S T E V E N S P I E L B E R G 33

In the 1950s Bazin and his young colleagues could talk of American cinema as one of the strongholds of subtle cinematic staging. By and large that’s no longer the case. The technique on display in today’s Hollywood film is so at odds with that of Bauer or Hong Sang-soo or even Hawks that these filmmakers look eccentric or old-fashioned. The stylistic strategies on which this book concentrates quietly challenge today’s most widespread habits of making and watching movies. Despite critics’ complaint that movies are packed with chases, explosions, and gun battles, the standard scene remains a conversation. These scenes are dominated by two staging norms, what practitioners call “stand-and-deliver” and “walk-and-talk.” Both flourished in classical studio cinema as well, but between these extremes there was also the choreography found in His Girl Friday (see box, pp. 14–15). Today this midrange blend has almost completely vanished. Instead, filmmakers have developed tactics of editing and camerawork that enliven static dialogue or conversations on the go. In the stand-and-deliver approach, which may turn into a sit-and-deliver approach, the characters are shown taking up positions, usually in a master shot. An axis of action governs the actors’ orientations and eyelines, and the shots, however varied in angle, are taken from one side of the axis. Actors’ movements are matched across cuts. As the scene develops, the shots tend to get closer to the performers, carrying us to the heart of the drama. Analytical editing presents shot/reverse shots, over-the-shoulder angles, and singles—all the coverage required by long-standing premises of classical continuity editing. When characters change position within the space, a new establishing shot gives us the information. Whatever variants or alternatives we might find, these schemas remain default values for most filmmakers. “If all else fails,” Ron Howard tells a neophyte director, “shoot a master and shoot overs [over-the-shoulder shots] and singles.”34 The same advice could have been given in 1925. The crucial diªerence is that by the 1930s a scene like Hawks’s

22

STAGING AND STYLE

from His Girl Friday could integrate analytical editing with intricate choreography. In contemporary cinema the stand-and-deliver strategy dominates. Having fixed their players in place, filmmakers dynamize the dialogue through specific devices of cutting and cinematography. We can call this new style, which emerged in the early 1960s and became dominant by the 1980s, intensified continuity.35 Early in Jerry Maguire (1996) the hero and heroine “meet cute” at an airport luggage terminal. Dorothy is with her son, Ray, and they’ve traveled on the same flight with Jerry. Dorothy is an accountant in the sports agency where Jerry works, and she’s been moved by his “mission statement” proposing that the agency start to give clients more personal attention. The scene opens with a distraught Dorothy calling for Ray, who has run oª. Jerry oªers to help and spots Ray riding on the luggage conveyor belt. After Dorothy has retrieved her son, she tells Jerry how inspiring she found his memo. His ego, and perhaps his libido, aroused, Jerry stays to talk with her. He’s about to oªer her a ride home when her sister arrives and Jerry departs. The scene brings the couple face-to-face for the first time and lays down the terms of their relationship. Jerry is vain, self-absorbed, and irresolute; although he likes taking credit for his mission statement on sports agenting, he’s prepared to disavow it. Earlier, after writing it, he had tried to withdraw it for fear that it would damage his career. When he’s fired and suªers the hard knocks of freelancing, he’ll be tempted to abandon the ideals he championed. Dorothy believes more unswervingly in those ideals than Jerry does, and she’s attracted to him, as she’ll admit later, partly because of “the man he almost is.” Moreover, in the film’s opening Jerry has been presented as having a special fondness for children. As he and Dorothy talk in the airport, each clasps one of Ray’s hands, and the boy swings carelessly between them, prefiguring the role he will play in drawing Jerry into marrying Dorothy; in the film’s epilogue, as the couple return from the zoo with Ray, he swings on their arms between them. The airport scene is edited classically. After an establishing shot (Fig. 1.46), the first phase deals with Dorothy’s finding Ray with Jerry’s help. In this portion of the scene the couple’s dialogue is handled in over-the-shoulder medium shots (Figs. 1.47–1.48). Just as Jerry tells her where Ray is (riding the luggage conveyor), the conversation is underscored by a shift to tight close-ups (Figs. 1.49–1.50). This passage establishes their bonding around the boy, the force that will later hold their marriage together. Dorothy then scolds Ray for running oª (“You scared me!”), an act that reconfirms her as a concerned but harried mother trying to enforce discipline (Figs. 1.51–1.52). Jerry says good-bye, but Dorothy’s calling, “I loved your memo!” after him brings him back to take some more credit (Figs. 1.53–1.54). The next phase of the scene presents Dorothy’s admiration and Jerry’s self-absorption, played out in orthodox eyeline-matched reverse shots (Figs. 1.55–1.56). These framings are about as close as those seen earlier, but there are many more of them here, signaling the development of the romance. Often in classical scenes, as the drama proceeds to its

STAGING AND STYLE

23

1.46 Jerry Maguire (1996): Dorothy walks into the establishing shot looking for her son, Ray.

1.47 Long lenses capture classical over-the-shoulder framings on Dorothy . . .

1.48 . . . and Jerry.

1.49 An abrupt cut to a closer view underscores Dorothy’s sudden awareness that Jerry knows where Ray has gone.

1.50 A matching framing on Jerry as he calms her.

1.51 Wide-angle close-ups present Dorothy chastising Ray . . .

1.52 . . . and his reaction.

24

STAGING AND STYLE

1.53 Another pair of long-lens framings present the eyeline matching of Dorothy . . .

1.54 . . . and Jerry as he pauses to listen to her praise his mission statement.

1.55 As Dorothy bursts out with admiration, Jerry starts to size her up, in a series of very tight close-ups.

1.56 A comparably tight close-up on Dorothy.

1.57 One of the few knees-up shots, here presenting Ray swinging on the adults’ arms.

1.58 Jerry departs, in a still longer-lens framing.

1.59 The reverse angle in another telephoto shot, as Dorothy and Ray watch him leave.

STAGING AND STYLE

25

core issues, the characters become more prominent. At one point a broader three-shot is introduced, but that’s to accommodate Ray’s swinging between the couple (Fig. 1.57); soon enough we’re back to dialogue in tight close-ups. When Dorothy’s sister, Eileen, arrives, oªscreen at first, Jerry leaves, followed in a long-lens medium shot (Fig. 1.58). The scene ends with Dorothy and Ray embracing and looking oª (Fig. 1.59), leaving us to expect further romantic developments.36 This sequence isn’t as galvanizing as a full-tilt car chase—supposedly the prototype of today’s Hollywood—but it’s more typical. Most scenes in commercial movies (and TV shows) are staged, shot, and cut as this conversation is. In stand-and-deliver passages like this, the visual interest springs not from complex staging but from techniques of editing and camerawork. The scene moves through sixty-seven shots in three and a half minutes, yielding an average shot length of only 3.2 seconds. This figure points up a notable shift in the classical style. From the coming of sound to about 1960, most Hollywood feature films contained between three hundred and seven hundred shots, yielding a range of average shot lengths (hereafter ASLs) falling between eight and eleven seconds.37 Since the 1960s, cutting rates have steadily accelerated, with typical 1970s ASLs running five to nine seconds and 1980s ASLs running three to eight seconds. The 1990s set a still faster pace, with several films containing more than two thousand shots and ASLs ranging from two to eight seconds. As one would expect, action films led the way, but at the century’s end the typical film in any genre, even romantic comedies like Shakespeare in Love (1998) and Runaway Bride (1999), had an ASL of between four and six seconds. Jerry Maguire, with its nearly two thousand cuts, averages 4.1 seconds per shot. Faster cutting, however, didn’t lead Hollywood films toward the discontinuities of Soviet silent montage. By and large, as in our Jerry Maguire example, editing simply accelerates the pace of scenes presented in ordinary ways. Instead of three or four reverse shots we might get ten or twelve, with each line or facial reaction assigned a separate shot. The premises of spatial continuity still govern the way the scene is staged, shot, and edited. Indeed, the greater number of shots strengthens the reliance on classical continuity principles; because each shot is so brief, it needs to be more redundant in indicating who is where, who is speaking to whom, who has changed position, and so on. Reciprocally, stand-and-deliver staging allowed directors to cut faster, since positions could be taken for granted. Directors also intensified continuity principles through more close framings and singles. The sequence from Jerry Maguire is characteristic. It contains only four long shots, and two of those become closer views when a character moves into them. The opening shot establishes the luggage terminal very briefly before Dorothy steps into the foreground and the camera starts to follow her search for Ray (Fig. 1.46). The scene relies on singles: two-thirds of the shots show only one of the three characters. In addition, the long lenses tend to turn even somewhat distant framings into singles (Fig. 1.53). As a result, there are twenty-two waist-up or chest-up medium shots (for example, Fig. 1.48).

26

STAGING AND STYLE

Even more surprising, more than half of the scene’s shots are close-ups of Jerry, Dorothy, or Ray. The medium shot, once the workhorse of studio filmmaking, has been replaced by various degrees of close framings. Since our orienting view in the early phase of the scene is the over-the-shoulder medium shot (Figs. 1.47–1.48), the only way to heighten emphasis is to move toward tight singles: of Jerry and Dorothy (Figs. 1.49–1.50), of Dorothy and Ray (Figs. 1.51–1.52), and of Dorothy and Jerry when they start warming to each other (Figs. 1.55–1.56). Even the singles in His Girl Friday (for example, Fig. 1.22) are nothing like this. Continuity cutting has been rescaled and amped up, and the drama has been squeezed down to faces—particularly eyes and mouths. Only Ray’s play at the grownups’ knees makes use of the lower parts of the couple’s bodies. No wonder that most contemporary acting manuals assume that the performance is almost wholly a matter of voice and face.38 Such tight facial shots are typical of Hollywood cinema since the 1960s. In the 1930s and 1940s the baseline framing, what the French called the plan-américain (“American shot”), tended to cut the figure oª at the knees or midthigh, as in our His Girl Friday scene (Figs. 1.20, 1.26). Many directors built their scenes out of prolonged two-shots in a midrange framing, a scale well-suited to the squarish 1.33:1 aspect ratio (Fig. 1.60). In the late 1950s and early 1960s filmmakers began to rely more on closer views, despite the fact that virtually every movie was shown in a widescreen format. Filmmakers working in the somewhat wide ratio of 1.85:1 had little trouble fitting close-ups into the shot; in eªect many shots in the new ratio simply masked oª the top and bottom of shots framed in the 1.33 format, leaving less headroom and occasionally cropping out foreheads and chins. The wider anamorphic formats like CinemaScope (2.35:1) at first seemed to favor more distant framings, and many directors dotted the frame with bodies and decor—sometimes awkwardly, like clothes hanging on a line (Fig. 1.61). Many believed that 2.35:1 format rendered the close-up obsolete; 1950s anamorphic lenses often distorted facial shots anyhow (creating “CinemaScope mumps”). By the 1960s, though, thanks largely to innovations introduced by the Panavision company, close-ups and even extreme close-ups became feasible in anamorphic and wide-film formats (Fig. 1.62).39 Surprisingly, the wide image could incline a director toward getting closer to the actors. Close views did not slice faces free of their surroundings, since directors could now include a lot of setting even in fairly tight framings (Fig. 1.63). Moreover, the 2.35 aspect ratio posed a problem with full-length shots; if you wanted to show the actor’s body you’d have to get back quite far or leave large empty areas. “What pulled me into shooting close-ups,” Steven Spielberg remarks, “was when I shifted to the widescreen format.”40 Medium shots and close-ups of individual actors were common in silent cinema, but in the sound era, when two-shots carried the burden of many scenes, a cut-in to a single could carry considerable force. (See Figs. 1.21–1.22.) As rapid cutting developed in the 1960s, filmmakers broke conversation scenes into many close views, and today these

STAGING AND STYLE

27

1.60 Pat and Mike (1952): a lengthy dialogue in medium two-shot; Cukor’s framing allows Tracy and Hepburn to convey casual intimacy through posture and arm positions. Compare the staging in Figs. 1.47–1.48.

scenes are built out of singles. Singles allow the editor to vary the scene’s pace and to pick the best bits of each actor’s performance.41 And since singles constrain the directors’ choice about framing, we shouldn’t be surprised to find filmmakers working along that narrower scale from medium shot to extreme close-up. The filmmaker doesn’t foreswear long shots altogether, of course, but they can be fewer and briefer, as in our Jerry Maguire scene. Now a long shot can serve as punctuation; it can demarcate phases of the action or provide that striking visual beat that close-ups no longer muster. The airport meet-cute in Jerry Maguire carries yet another lesson in intensified continuity. Nearly every shot is filmed with a long lens—even the over-the-shoulder shots (Figs. 1.47–1.48). Those shots not filmed with a long lens are given a slightly grotesque bulge with a wide-angle treatment (Figs. 1.51–1.52). This polar opposition between long and short focal lengths became marked in the 1960s, when filmmakers wanted to retain the expressive resources of Hollywood’s wide-angle tradition (associated with Citizen Kane) but also wanted to exploit the telephoto eªects seen in European cinema. Studio-era filmmakers had reserved long lenses (75 mm and up) for making facial close-ups or following animals at a distance. By the 1970s, however, filmmakers were staging entire scenes for telephoto pickup, with rack focusing (shifting focus from one plane to another) becoming a common way to guide the viewer’s eye. Our sequence from Jerry Maguire doesn’t exemplify every aspect of today’s Hollywood style. It doesn’t, for instance, display what we might call the free-ranging camera. Across most of the history of cinema, directors have had occasion to move the camera independently of the action, but today this technique has come to the fore to an unprecedented degree. The camera often tracks in or out from a figure, even in a shot/reverse-shot exchange (push in on A/push in on B/push in still closer on A/push in still closer on B, etc.). It spirals around the actors, as in the Reservoir Dogs scene (Figs. 1.9–1.10). A sequence

28

STAGING AND STYLE

1.61 A CinemaScope framing gathers the three couples at a soda fountain, and the ne’er-do-well reveals that he’s actually rich (How to Marry a Millionaire, 1953).

1.62 Panavision permits a tight close-up, if a bit distorted by the wide-angle lens (You Only Live Twice, 1967).

1.63 A typical long-lens Panavision shot from the 1970s, allowing a medium close-up of the figure and still including a good deal of setting (Klute, 1971).

may start with an arcing or sidling movement past a foreground element, or the camera may crane down to meet the players. Stand-and-deliver is the most common staging option, but a second approach provides more dynamic figure movement. In the walk-and-talk option the characters are striding down the street, through an o‹ce, or along a corridor, usually toward the camera in a lengthy tracking shot. This was a stylistic signature of Max Ophuls and later Stanley Kubrick, but it was also common in Hollywood, especially for a film’s opening sequence (for example, Scarface, 1932; Ride the Pink Horse, 1947; Touch of Evil, 1958). By the 1980s, thanks to auteurist film criticism and increasingly lightweight cameras and handheld supports, the prolonged following shot became de rigueur for nearly every film. Such shots, in which overcoming logistical tangles becomes an end in itself, may have been an eªort to counterbalance the demands for fast-cut stand-and-deliver. It allowed cinephiliac filmmakers to flaunt a pluralistic style, a knowing synthesis of both “montage” and “mise-en-scène.” In either stand-and-deliver or walk-and-talk, long takes typically demand camera movement. When a scene is based on a “sustained master,” as in Woody Allen’s films or Spike Lee’s Do the Right Thing (Figs. 1.64–1.65), it’s usually a moving shot. Today’s Hollywood style is always on the move—if not through cutting, then through camerawork. What has gone almost completely unexplored is the unbudging long take. It can occasionally be found in 1960s and 1970s films (for example, Nashville, 1975), but it became much rarer in the 1980s. Today perhaps the most radical thing you can do in Hollywood is put your camera on a tripod, set it a fair distance from the action, and let the whole scene play out.

STAGING AND STYLE

29

1.64 The “moving master” in Do the Right Thing (1989), as Sal’s son complains about working in a black neighborhood . . .

1.65 . . . and the camera slowly comes in to capture Smiley at the window.

Where did this change in the continuity style come from? No one has yet traced its emergence in detail, but prime influences would surely include television techniques (which since the 1960s have relied on fast cutting and camera movement), the need to make a scene read well on TV monitors (so lots of close-ups), the influence of showy directors like Hitchcock and Leone, new technology (like the Steadicam and computer-based editing), and changes in production routines (for example, the use of multiple cameras armed with long lenses). In all these respects, and probably others, responses to production demands created standardized practices, and those in turn supported stylistic trends. It is here, within the film community—where concrete problems must be solved and successful solutions tend to become models—that we often find the most precise explanations for changes in visual style. No style is without advantages, and the schemas of intensified continuity have been brilliantly exploited by Michael Mann, David Lynch, and others. The style has also had enormous influence across the world, particularly in the popular cinemas of Europe and East Asia. In Hong Kong directors like John Woo, Tsui Hark, Ringo Lam, and Johnnie To embellished the style by putting it at the service of local traditions of percussive movement and sensuous display. A Thai film like Nang Nak, a Korean film like Shiri or Tell Me Something, and a Japanese film like Monday (all 1999 releases) indicate that intensified continuity has become the default value for international mass-market cinema. Aspects of the style are on display in more prestigious products too; look, for instance, at the choppy telephoto close-ups and endlessly drifting camera of Jane Campion’s Portrait of a Lady (1996). True, a few American independents like Woody Allen have favored longer shot lengths,42 but most directors have assimilated other aspects of intensified continuity, as can be seen in such disparate products as sex, lies, and videotape (1989) and Happiness (1998).43 More creatively, some independent filmmakers have subjected conventions of intensified continuity to unusual variations. Joel and Ethan Coen and Sam Raimi have exploited the style for grotesque eªects, but others have made it serve storytelling in more austere ways. For example, Hal Hartley’s scenes are shot predominantly in medium shot and close-up (thus saving money on sets and lighting), but the staging avoids the overcovered stand-

30

STAGING AND STYLE

1.66 Simple Men (1992): at the window, as Bill and Mary talk, the camera keeps tightly framed on them.

1.67 They move to the counter to order, still not looking at each other, questioned by an unseen waitress.

1.68 Back at the window, in another tight framing, they continue to avoid each other’s eyes.

1.69 Later, after Bill has met Kate, they kiss cautiously, their bodies turned away from one another.

and-deliver approach favored by Hollywood, moving closer to the elliptical close-ups in late Godard films like Détective (1985). In Simple Men (1992) Bill, a hardened thief on the run, meets his ex-wife in a coªee shop and gives her money for their child. Hartley supplies no establishing shot, instead working with four shots (one very brief ) in which characters’ faces move in and out of the frame and secondary figures are put oªscreen (Figs. 1.66–1.68). The result is a careful choreography—not of full bodies, as in His Girl Friday, but of faces and shoulders, which register changes in the drama through slight turns and deflected looks. In Jerry Maguire the couple’s eyes lock, but Hartley’s scene is built out of partners turning away or glancing aside. These movements are recalled much later in the film, when Bill is drawn to another woman. Hartley now sets the faces in profile but twists the bodies away from one another, suggesting only a half-trusting intimacy (Fig. 1.69). Most proponents of intensified continuity, however, aren’t interested in such austerity. By treating ordinary scenes with the sort of inflated technique once reserved for suspenseful climaxes and courtroom revelations, today’s director not only risks bombast but

STAGING AND STYLE

31

is obliged to raise the stakes. Now the real climaxes must be treated with ever more outré eªects—slow-motion, smeared color, digital doodlings, crashing electronic chords. Too often intensified continuity turns style into Style (or STYLE). We can’t say we weren’t warned. Back in the 1960s, American critics spotted strident technique in Kubrick, Kazan, Frankenheimer, and the two Sidneys, Lumet and J. Furie. Dwight MacDonald objected to movies that were “overdirected,” claiming that Tony Richardson was “unable to compose a scene clearly in front of the camera and so [was] forced to overuse the close-up.”44 Andrew Sarris complained of Mike Nichols’s “facile shock tactics” and Richard Lester’s and John Schlesinger’s “weakness for the immediate click-eªect at the expense of an overall conception.”45 Pauline Kael proved prophetic: The shifting camera styles, the movement, and the fast cutting of a film like Finian’s Rainbow—one of the better big productions—are like the “visuals” of TV commercials, a disguise for static material, expressive of nothing so much as the need to keep you from getting bored and leaving. Men are now beginning their careers as directors by working on commercials—which, if one cares to speculate on it, may be almost a onesentence resumé of the future of American motion pictures.46

Nearly forty years later the pumped-up technique these critics protested against has become the norm, and like all norms it excludes some fruitful alternatives. One goal of this book is to remind us of those alternatives and to coax young filmmakers into exploring them.

You may think I’ve dodged some obvious questions from the outset. Why should we inquire into style at all? Isn’t content—the story, or the film’s larger significance—finally more important? Why plunge into the formalistic niceties of lenses and shot lengths? And even if style is worth studying, is concentrating on the concrete and practical decisions made by filmmakers the most fruitful way to do so? How, today’s film scholars might demand, can we “theorize” the study of style? Film style matters because what people call content comes to us in and through the patterned use of the medium’s techniques. Without performance and framing, lens length and lighting, composition and cutting, dialogue and music, we could not grasp the world of the story. Style is the tangible texture of the film, the perceptual surface we encounter as we watch and listen, and that surface is our point of departure in moving to plot, theme, feeling—everything else that matters to us. And since filmmakers devote painstaking care to fine points of style, we must dig into details. A comprehensive discussion of any film can’t stop only with style, but style should claim a lot of our attention. Yet most film scholars don’t analyze style, particularly visual style.47 Ironically, as films have become more available for close analysis than ever, interest in stylistics has waned. Why? Partly because film studies has for some time attracted scholars of a literary turn

32

STAGING AND STYLE

of mind, more comfortable with hermeneutics than stylistics (which remains a minor discipline in literary studies). In addition, even scholars have di‹culty attending to the minutiae of technique. As we watch a film, we absorb its images but seldom notice how they’re lit or composed. So critics and scholars find it more natural to talk about characters’ psychological development, about how the plot resolves its conflicts and problems, or about the film’s philosophical or cultural or political significance. Our neglect of images is probably tied to our habits of taking in the world. We move through life cheerfully unaware of all the complicated ways our perceptual systems sift and elaborate the information pouring in.48 The acts of seeing and hearing are, as psychologists say, “informationally encapsulated.” Life has no pause or rewind buttons, and we can’t stop to examine the mechanisms in our heads that deliver a three-dimensional world of objects and activities to us. For normal purposes this is all to the good. We evolved so as to act on the outputs of those mechanisms, and being aware of them could only slow us down in a world where seconds matter. Yet the habits of ordinary perception, easy and quick, oªer little help when we want to understand how cinema works. Moreover, only a few disciplines, such as art history, have developed a vocabulary for analyzing how images create their eªects. This vocabulary, inevitably a technical one, comes far less spontaneously to us than the language we use to describe characters and plot. Of agents and activities we know; our commonsense psychology has given us a plentiful repertory of ways of describing human action. Of composition and color and patterns of movement we can say much less. In particular, to study cinematic staging is to focus on issues that, for historical reasons, have received little attention even from students of cinema. A comprehensive and detailed theory of film style isn’t in the cards here, especially in the dwindling pages of this introduction. But I should oªer a preview of this book’s treatment of two fundamental issues: How can we usefully describe and analyze film style? And how can we fruitfully explain its historical causes and consequences? At almost every turn my answers can be challenged, and on some points they already have been. Rather than taking on criticisms in advance here, the final chapter reviews some major objections to this sort of analysis and oªers some rebuttals. The reader is in a better position to appraise the criticisms, I think, if she or he has already seen what insights flow from pursuing this line of inquiry. This is primarily a book of film criticism, oªering some tools for stylistic analysis. But no criticism works without some theoretical presuppositions, so I’ll briefly lay out the ones most important for this project. I will assume that style has four broad functions. First, it serves to denote a fictional or nonfictional realm of actions, agents, and circumstances. Just as in literature, the denotative function of film style governs a great deal— the description of settings and characters, the account given of their motives, the presentation of dialogue and movement. In storytelling cinema, the denotative functions of style are everywhere evident. Each shot presents a slice of space and a segment of time, a set of persons and places that we are to take as part of a fictional or nonfictional world.

STAGING AND STYLE

33

Jerry Maguire and Dorothy, Walter Burns and Hildy, Su-jeong and her friends, the breakfasts in The Revolutionary and in Simple Men—all come to us courtesy of stylistic processes working in the denotative dimension. This quality is so basic to filmmaking that we’re inclined to take it for granted, but shortly I’ll argue that it should be central to any account of cinema stylistics. Style need not only denote concrete objects and persons; it can display expressive qualities too. A great deal of musical style is devoted to representing emotional states, such as majesty, sprightliness, or menace. Abstract Expressionist paintings are often taken to represent turbulence or anxiety. In most films expressive qualities can be carried by light, color, performances, music, and certain camera movements, such as the blurry swirl that can express vertigo. There are many philosophical arguments about how expression works in the arts, but for my purposes one further distinction is helpful. We can distinguish between style presenting feelingful qualities (“The shot exudes sadness”) and causing feelings in the perceiver (“The shot makes me sad”). For the most part I will confine myself to the first dimension of expression, but there is no overlooking the power of style to “infect” the spectator with strong feelings.49 Style can yield more abstract, conceptual meanings as well; style, that is, can be symbolic. In a painting, a (denoted) tree might also be a symbol of fecundity; the squished skull tipping into the foreground in Holbein’s The Ambassadors may be a memento mori. Symbols in literature, like the sign of the staring eyeglasses in The Great Gatsby, and in cinema, like the copulating jet planes at the beginning of Dr. Strangelove (1964), work by the same principle, of suggesting more general or abstract meanings. Films may evoke symbolic implications through color schemes, lighting design, setting, and musical associations. Finally, in any medium style can work somewhat on its own. It creates discrete moments and broader patterns that engage us for their own sake, coaxing us to discover order or notice fine diªerences. Granting the term’s inadequacies, we might say that here style functions decoratively. Decoration, in this sense, asks us to apprehend the sheer pattern-making possibilities of the medium.50 In addition, we think of decoration as ornamentation, as an application grafted onto some support. We decorate an apartment’s walls, or we decorate our bodies with tattoos. In storytelling films, style can be decorative in just this sense: the pattern making operates alongside or “on top of ” other stylistic functions. In purely nonrepresentational films, style is sheer decoration, creating wholly abstract patterns. But since most films are representational, we may find decorative qualities added on, as when the filmmaker creates pure patterns while also denoting a world or conveying expressive states. We might consider decoration a matter of one-oª moments, as when today’s crane shot weaves arabesques around a minor action. In this sense decoration seems to be a sign of the mannerism that some commentators celebrate in such directors as von Sternberg, Busby Berkeley, and Brian De Palma and that Kael, Sarris, and MacDonald faulted in 1960s Hollywood releases. I’d suggest, though, that decorative uses of style can also work in a more pervasive, systematic way.51 What I’ve elsewhere called “parametric nar-

34

STAGING AND STYLE

ration” can be thought of as a highly organized decorative stylistic execution.52 Other arts can start from pure patterning and add other dimensions: the composer of a fugue will try to find an expressive melody, and Mondrian gave a denotative-expressive spin to one abstract painting by calling it Broadway Boogie-Woogie. By contrast, systematic use of decoration is pretty rare in cinema, partly because this art form is historically so tied to denotation. We ought, for this reason, to resort to decorative explanations of film style only after fully considering other functions. I’ve separated out these various functions for analytical purposes; in actual cases any particular technique can serve several functions at once. New Hollywood’s intensified continuity provides many instances. Intercutting the camera movements tracking in toward two stationary characters can be at once denotative (magnifying facial reactions), expressive (signaling a growing tension in the scene), and mildly decorative (creating a parallel repetition of the stylistic device). A passage of rapid editing can inform us that the pursuers are closing in on the hero and simultaneously express the excitement of the situation. The last examples suggest that any function of style can also yield a payoª in the viewer’s response. Naturally, expressing an emotion, through whatever technique, often triggers an appropriate emotional response in the viewer. If both picture and sound convey a character’s grief, we may be moved to pity. The combination of nearly hysterical performances, relentless forward tracks, and pulsating, threnodic musical phrases give Magnolia (1999) an insistently anxious quality that seeks to work on the spectator quite viscerally. Less showily, a denotative use of style, simply presenting the twists and turns of the story, can arouse a range of reactions, from curiosity and suspense to melancholy and elation. Even the abstract pattern-making of decorative processes can build up frustration (why is this technique repeated when it doesn’t seem to have a representational purpose?) and satisfaction (now I discern the overarching design). Style informs our experience of a film at many levels.

Generally speaking, in blocking and framing a shot, the most important thing is to make sure the audience is looking where you want them to look. You have to make sure they clearly understand the story point that is taking place, and that they grasp all of the pertinent information while sort of ignoring everything that’s not important. Short of force-feeding the issue by blacking out everything else in the frame or providing blinking arrows, this is primarily accomplished through shot-composition and careful blocking.

R O B E R T Z E M E C K I S 53

All the foregoing provides only a rough characterization of style’s functions in cinema, but it allows us to begin analyzing cinematic staging. We can use the distinctions I’ve just outlined to suggest some priorities. First, denotation is the central function that style fulfills in virtually any representa-

STAGING AND STYLE

35

tional film. This fact has probably made denotation hard to notice: critics have largely overrated the symbolic dimension and underrated the denotative one. In analyzing a film’s style, then, I propose that our customary point of departure should be the denotative functions that style fulfills. Moreover, in analyzing a narrative film we should hypothesize that for the most part stylistic denotation serves to present the information most relevant to the ongoing story. All other things being equal, film style typically operates as part of narration, the process of cueing and guiding us to construct a story out of what is presented on the screen and on the sound track.54 Of course, that construction may be creatively blocked through ellipses, concealment, misdirection, and the like (themselves all narrational maneuvers manifested through style). It would be more exact to say that the film encourages us to construct the story along certain preferred pathways. And that construction will take place in and through stylistic patterning. Not every technique in a film will have primarily denotative functions. The premises serve as default values and working hypotheses to guide our inquiry; being alert to them doesn’t keep us from noticing other matters. In many cases, once we have determined the denotative role of a stylistic device, we may not need to go farther to account for its traits. But if we can’t find a denotative role for the device, we will be all the more encouraged to seek others. Since a device may serve several functions, we can always continue to consider expressive, symbolic, and other purposes of the technique. Using narrative denotation as a default value oªers us both coverage and economy. This hypothesis allows us to acknowledge that a film’s stylistic texture is pervasive, uninterrupted from first moment to last. By contrast, symbolic functions don’t saturate the film’s style to the same degree. It’s unlikely that every bit of lighting, every camera movement, every shift of a figure in the frame, every cut conveys an abstract meaning; indeed, those moments when we spot Christ imagery or a highlighting of gender identity stand out from the background of more ordinary shots and cuts. That constant background has a more primary purpose—what we pretheoretically call “telling the story.” The swiftly cut singles of Jerry Maguire’s airport scene underscore certain lines and facial reactions. Within a wider image, as we’ve seen, a character’s slightest movement can channel the flow of dramatic information: a man’s rising to leave a drinking bout becomes a pivotal action, and even a samovar blocking a face plays a role by obliging us to look for story developments elsewhere. Moment by moment, whatever else it may do, style usually helps tell the story in a particular way. Starting from narrative denotation also yields some explanatory economy: we can gather many diªerent techniques under a few general principles. One of the most important, it seems evident, is the principle of directing attention. Typically, a storyteller in any medium seeks to call certain items to our notice and let others recede or drop out altogether. In literature no description of a character or setting constitutes a complete inventory; the writer selects the aspects relevant to the story and the eªects to be achieved. Watson doesn’t mention that Holmes has lungs, but they’d become quite relevant if Holmes were to contract cancer from his shag tobacco. In photography or cinema the frame excludes much

36

STAGING AND STYLE

more than it includes, and within the frame, composition, color, and all the techniques of each medium work to highlight the most important qualities on display. Artists are practical psychologists, trusting that our perceptual proclivities, developed through both natural endowment and cultural training, will make some aspects of the story more salient than others. Storytellers further trust that our attention tends to fasten on aspects that attract us in the world outside art. Take the body. Primates’ bodies emit a constant stream of signals about their emotional states and their relations to their conspecifics, and we have evolved to be highly sensitive to the postures, gestures, and movements of those around us. The distance and angle at which two people stand can instantly convey intimacy or dominance, flirtation or confrontation; in all cultures an erect bearing tends to suggest higher status than does a cringe. Gestures punctuate and emphasize speech, and some do duty for words, most through culturally specific signs (such as the cuckold’s horns), a few through universal ones (applauding, shrugging). Many nonverbal cues, such as the yawn or the clenched fist, may aªect us more powerfully than a remark does, since they are hard to suppress and thus more likely to be genuine.55 Above all, we are attuned to the face, the most informative part of the body. There is reason to believe that the face evolved partly under pressures to signal socially relevant information, and it now serves as a constantly changing theater of moods. At least six basic facial expressions (happiness, surprise, fear, sadness, anger, disgust or contempt) appear to be recognized across a great many cultures. Other expressions, such as the raised eyebrow for skepticism or hauteur, seem more localized, and cultures also vary in what Paul Ekman calls “display rules” for the circumstances under which expressions are appropriate.56 Film acting relies heavily on widely recognized facial expressions.57 Within the face the most informative zones are the mouth and the eyes, the two areas on which conversing people focus. The eyes are particularly important, since gaze is associated with facial expressions (an aggressive expression and a direct gaze go together in chimps and other primates like us), and we are preternaturally sensitive to the particular angle of the gaze. In a conversation the listener is likely to look more steadily at the speaker than the speaker looks at the listener; the speaker glances away at intervals, using the gaze to punctuate the talk and to check for reactions. The whole process is carefully coordinated so that when the listener speaks, the roles switch. The eyebrows add their own commentary on both the words and the pattern of gazing.58 Since in most films human (or humanlike) creatures are the center of interest, we ought to expect that our perceptual system will home in on bodies and faces. Perhaps this is one reason why the classical schemas of shot/reverse shot and eyeline matching communicate so easily across cultures. They build on the rhythm of turn-taking and the pickup of facial and bodily cues in typical human interactions.59 Yet classical schemas don’t simply replicate what happens in ordinary interchanges. For the sake of comprehension the schemas simplify the process. They stress faces through medium shots and close-ups, while the shot scale and lighting make eyelines absolutely evident. The actors look steadily at each other far more often than people would in every-

STAGING AND STYLE

37

day life, so in movies an averted gaze becomes not a normal maneuver but an expression of distraction, guilt, shyness, or some other relevant state. Similarly, filmmakers reduce the normal blink rate. We typically blink twelve to fifteen times per minute, but movie characters blink far less frequently.60 In our Jerry Maguire passage there are fewer than thirty blinks among three characters in a passage of nearly three minutes, and many of these make an expressive point (such as Jerry’s prolonged I-hold-the-floor blinks). If classical continuity streamlines the most salient interpersonal cues, intensified continuity exaggerates and amplifies them. Simple Men, despite its longer takes and more complex staging, obeys such a strategy. In order to express coldness or wariness, Hartley has his players make eye contact significantly less than a conversing couple would in ordinary life (Figs. 1.66–1.68). The result is as artificial in its own terms as the rapt gaze binding Dorothy and Jerry at the luggage conveyor. Even styles less tied to close-ups make expressions and eyelines salient by means of composition, lighting, and movement. As we’ll see, Feuillade, Mizoguchi, Angelopoulos, and Hou Hsiao-hsien could not do without the information available in faces. It’s just that in their traditions a range of other nonverbal cues, from posture and gesture to landscape and architecture, provide alternative centers of attention. Faces still matter, but so do the drinking gestures of the foreground figures in The Virgin Stripped Bare by Her Bachelors and the angular tilting to and fro of the uncle at the breakfast table in The Revolutionary. As the chapters proceed, we will see how our directors creatively conceal facial expression, steering us to the denotative, expressive, and decorative possibilities of other aspects of the image. A large part of cinematic storytelling crosses cultural borders so easily, and one plausible reason is that it rides on widespread perceptual tendencies. The styles that came to dominate filmmaking seem to have tuned in, through trial and error, to some perceptual and cognitive universals.61 Long before cinema came on the scene there arose certain commonalities in the way all humans saw, heard, and understood the world’s regularities.62 Those commonalities were most probably produced by both biological evolution and converging cultural practices.63 Viewers come to movies dragging their perceptual and cognitive proclivities with them, so, seeking to aªect viewers, filmmakers have developed techniques that piggyback on cross-cultural a‹nities. All film styles are arbitrary in the sense that they could have been otherwise, but certain styles are more eªective for certain purposes, such as conveying a story, than rival styles are. If you want your viewer to focus on character reactions, shooting from an angle displaying the faces is more e‹cacious than shooting, say, from an angle that shows only shoes. Of course, if the filmmaker has other purposes, such as creating a film lyric or an experimental narrative, then uncommon stylistic choices become more appropriate; but even those will work with (or against) the perceptual and cognitive proclivities of potential spectators. If, as seems likely, directing attention is one crucial aim of narrative filmmaking, we can usefully invoke another cross-cultural regularity: the fact that all humans use their eyes to search their environment. Because only a narrow region of our eye’s anatomy, the

38

STAGING AND STYLE

fovea, possesses critical focus, the eyes move to let the fovea attend to items of interest. Sometimes the eyes track slowly moving objects via smooth pursuit movements; more often, three or four times per second, our eyes jump from spot to spot in what are called saccades. Saccades sample the environment, bringing features into sharp focus for only about a quarter of a second. If the item is worth studying, microsaccades or flicks shift the fovea slightly over the target.64 The process of visual search is active, fast, and indebted to our biological heritage. Scanning the environment and monitoring changes in it by fixating on crucial aspects yield robust evolutionary advantages to mammals like us. Unsurprisingly, newborn babies’ gazes linger on faces, particularly eyes and lips. Both saccades and pursuit movements are purposive, prompted by information arriving at other parts of the optical system (chiefly peripheral vision) and other perceptual systems, such as hearing. Psychologists speak of a “covert attentional system” guiding our eye movements.65 Since saccades bring the most informative regions to the fovea, it seems likely that the mind directs saccades by generating hypotheses about what to look for—in eªect, a cognitive map. Nonfoveal vision sketches a rough spatial layout that gets filled in through closer inspection of the field.66 At an intuitive level artists have long been aware of visual search. During the late seventeenth century, members of the French Academy, who oªered the first detailed criticism of works of painting and sculpture, based their theory of pictorial order on guiding the spectator’s scanning over time.67 Hogarth declared that a sinuous line “leads the eye a wanton kind of chase.”68 Centuries later Katharine Hepburn, according to the cinematographer for On Golden Pond (1981), tried to upstage her colleagues by wearing a scarlet cardigan because, as with the red accent in an impressionist painting, “your eye goes straight to it.”69 The contemporary intensified-continuity style would seem to require leading the eye rather brusquely, in the vein of Kuleshov’s demand for legibility. “When a shot is only going to be on screen for three seconds,” notes cinematographer Vilmos Zsigmond, “ . . . there’s no time to decide what is important, so you have to direct their eye, force it.”70 Guiding attention has a well-established biological and psychological basis, and it is central to the denotative aims of staging. By appealing to the principle of directing attention we can economically explain many tactics at work in our sample scenes. Techniques of lighting, composition, and performance reinforce our natural drift toward bodies, gazes, and facial expressions. And the filmmaker who packs the shot with items competing for our attention will often steer us carefully to the most important regions. Many tactics of staging can therefore be seen as ways of managing the viewer’s arc of visual search—not necessarily “forcing it,” in Zsigmond’s strong sense, but guiding it, or oªering a line of least resistance that, given the time pressures set by the viewing situation, will pull most spectators along.71 Of course, style analysis doesn’t demand that we earn degrees in perceptual psychophysics, but in the chapters that follow, while being alert to the expressive, symbolic, and decorative qualities of visual style, I’ll often take the shaping of attention as a default value. Far more often than we usually allow, stylistic choices

STAGING AND STYLE

39

assist narrative denotation principally by suggesting where and where not to look—always with the understanding that the spectator’s interest is already subject to the gravitational pull of the bodies we see traced in the light of the screen.

Given the various functions of film style, how can we fruitfully explain a stylistic trend’s historical causes and consequences? Another hypothesis can guide our investigations. Since my concern in this book is with fictional narratives, we should assume that the stylistic organization we detect is the result of filmmakers’ choices among the alternatives available to them. These choices may have been planned before shooting, may have emerged spontaneously in the course of shooting, or may have obtruded themselves in postproduction. To make an oversimplified distinction, they may be “free choices” exactly fulfilling filmmakers’ intentions, or they may be “forced choices” born from external constraints such as time, budget, or lack of authority. From this perspective in order to explain continuity and change within film styles, we ought first to examine the circumstances that impinge most proximately on filmmaking—the mode of film production, the technology employed, the traditions, and the craft routines favored by individual agents. More “distant” factors, such as broad cultural pressures or political demands, can manifest themselves only through those proximate circumstances, in the activities of those historical agents who create a film. The zeitgeist can’t switch on the camera. The argument I’ll be pursuing runs counter to explanatory frameworks that have dominated film historiography. These have been macroscopic accounts, tracing film style (or other aspects of films) to a spirit of an age, an ideology, national culture, or epochal conditions such as “modernity” or “postmodernity.” I’ll try to show in the last chapter that for the most part these frameworks are vague; their proponents have not spelled out the causal mechanisms by which such elusive factors could shape the fine-grained qualities we can find in the films; and those proponents have proven reluctant to test their claims through counterexamples. In sum, our propelling question—What are the aesthetic possibilities and historical ramifications of the staging-based traditions?—isn’t adequately addressed by such broad-based accounts. For example, we’re tempted to invoke broad cultural circumstances to explain the cutting pace of intensified continuity. In a media-saturated age, some would say, our sensibilities are attuned to rapid-fire imagery. At the limit perhaps our perceptual threshold has been raised; we may be “visually literate” in a way that previous generations were not, and we can take in pictorial information at a faster clip. Yet although we may have indeed developed a taste for fast-paced imagery, it seems unlikely that we’ve somehow expanded our powers to assimilate it. After all, audiences have seen fast cutting before. In the American cinema of the 1920s shots averaged four to six seconds each, not so far from the norm of the 1980s and early 1990s. Moreover, although most US films of the 1930–1960 period used longish takes, the many that did employ fast cutting weren’t illegible. Audiences are versatile: today’s films favor one of many cutting rhythms that viewers have

40

STAGING AND STYLE

long been able to assimilate. Through a mutual process of feedback and correction, filmmakers have shifted toward new norms, and those norms most plausibly arise from the historical dynamic I’ve tried to specify, such as the demands that filmmakers believed television to have placed on them. Fast cutting would seem easy meat for an epochal explanation; the task gets a lot harder when we ask how a cultural shift in perception could explain the rise of other features of intensified continuity, such as the reliance on stand-and-deliver staging or long lenses or arcing camera movements. In studying intensified continuity we find that one advantage of teasing out quite specific stylistic traits is that we force ourselves to propose fairly precise local causes for them. Similarly, in considering staging I’ll be pointing to some pictorial strategies of extraordinary delicacy, and these resist being subsumed into the broad, quasi-Hegelian frameworks (modernity, postmodernity) on oªer today. The next four chapters proceed inductively, working from particular stylistic features toward the most pertinent and proximate functions and causes we can detect. The final chapter argues that the sort of stylistic features brought to light are not crisply explained by national culture or epochal conditions. As a first approximation I suggest that the most salient way to assign functions and causes is to treat stylistic strategies as solutions to problems. Indeed, says David Zucker, director of Airplane! and Ghost, “Directing is mostly about solving problems.”72 What kinds of problems? In narrative cinema many problems revolve around telling a story in a specific way. How to create, through the actor, the character’s reaction and then make sure the audience sees it at the proper moment? How to move actors around the frame to indicate their relationships, their emotional states, their responses to the ongoing drama? When do you cut to a reverse shot of B—a bit before A finishes the line, just after A finishes, or after a pause holding on A? These are, at a primary level, problems of narrative denotation, expressivity, and audience uptake. Fortunately, however, the filmmaker doesn’t have to solve each problem from scratch. The tradition in which she or he works furnishes a repertory of schemas, those norms of style that experts can recycle, modify, or reject. When Ron Howard advises the novice to shoot masters, then OTSs, and then singles, he’s pointing to a cluster of schemas that have over the years reliably solved the problem of conveying a conversation with rising emphasis. And schemas, like all constraints, always oªer further zones of choice. Even within the analytical continuity tradition Howard mentions, one still has a wide latitude in choosing shot scale, composition, timing of the shots and cuts, and so on. You might shoot the master as a bird’s-eye view, as De Palma sometimes does, or frame the reverseshot singles with the characters looking directly at the camera, as Jonathan Demme does. Here is why I started out by concentrating on particular options historically open to filmmakers. Filming a dinner-table conversation is a problem many directors face, and there are canonized ways to solve it, each with advantages and drawbacks. The filmmaker selects from the repertory available, or modifies the norms she knows, and thereby steers the moment-by-moment flow of the scene in a particular direction, with concomitant ap-

STAGING AND STYLE

41

peals to spectators. Originality emerges when a filmmaker creatively fits a new means to a recognized end, or decides on new goals that rework an old means. The problem-solving perspective entails that every problem has more than one solution, yielding varying rewards. Some solutions are favored by tradition. The artist inherits a body of craft practices, and those practices settle into stylistic standards. In cinema the emergence of Hollywood’s powerfully pervasive international style yielded a solid basis for both conservative and experimental filmmaking. Dalí and Buñuel’s Un chien andalou (1928), a prototype of avant-garde cinema, achieves its force by presenting repulsive or amusing absurdities while apparently following canons of classical continuity. In explaining why a group style changes or maintains stability, or in analyzing how a filmmaker achieves a distinctive style in a single movie or across her career, we arrive at the most concrete results when we reconstruct a historically plausible choice situation. In order to do this, we must study other films both inside and outside the tradition; search for evidence of the filmmakers’ purposes; examine the technology available; and reconstruct relevant institutional constraints. For example, we can understand the emergence of intensified continuity by considering how television shaped both production and exhibition, how new technologies like the Steadicam favored certain kinds of shots (and encouraged filmmakers to expand their ambitions), how demands for more e‹cient production and more choices in the editing phase encouraged the use of multiple cameras and shorter shots. Such is the milieu in which filmmakers must survive, and they are always alert for ways in which such tangible forces can expand their range of choice, solve standing problems, or pose new ones that can be creatively overcome. With this framework in mind we can ask, What are the aesthetic resources of stylistic traditions grounded in cinematic staging? How have those resources been explored by four major filmmakers across history? In seeking answers, we’ll often find that although directors draw strength from their social milieus, they also belong to the international community of filmmakers, a community with a craft tradition of shared problems and solutions, norms and schemas, and bound by the common aim of arousing audiences through the power of moving pictures.

42

STAGING AND STYLE

2 FEUILLADE, OR STORYTELLING

Louis Feuillade is probably the most prolific of the great filmmakers. Between 1906 and 1925 he made at least five hundred films for the Gaumont company—some only a few minutes long, others running more than seven hours. He worked in every genre— comedies, costume pictures, contemporary dramas, mystery films, sentimental melodramas, patriotic epics, even westerns shot in Parisian parks. He cast his lot with the mass audience. Brought up in the southern town of Lunel, he was a passionate cyclist, athlete, and bullfight fan, and it is not surprising that a lover of the corrida, with its unremitting physicality and frenzied crowds, should find in cinema “the popular art of our time.”1 His films, he said, were aimed at the ticket punchers in the Métro, not the elite wooed by the avant-garde. For the public “the only thing that counts is to know if, in [a film’s] twenty-six reels, there lies a sleeping princess whom a magician will awake with the beam of his marvelous lamp—I mean, a good story. That is the sole point: the story, the tale, the fiction, the dream.”2 Feuillade unabashedly put on display all the clichés of popular narrative. There are dying mothers, orphaned children, saucy maids, benevolent American millionaires, henpecked husbands, seductive adventuresses, and cunningly disguised villains armed with poison, bombs, and the power to hypnotize innocents. If you want your movie to sell, he told one director, hire a child or a dog.3 Hundreds of his films survive only as negatives, locked away in the Gaumont vaults, but by hopscotching among archives and revivals, one can study many of the works on which his reputation rests—very long movies made to be shown in installments. Feuillade’s first, and in many ways best, installment film was Fantômas (1913–1914), based on

FEUILLADE, OR STORYTELLING

43

the popular crime novels in which Inspector Juve and the journalist Fandor pursue the archcriminal Fantômas and his sinister mistress, Lady Beltham. Its successor was Les Vampires (1915–1916), “grand film de mystère et d’aventures en plusieurs épisodes,” and thanks largely to the surrealists, it became Feuillade’s most famous work. Heavily improvised because of the constraints of wartime shooting, this delirious saga shows a criminal organization ransacking and terrorizing every level of French society. The investigators are the journalist Philippe Guérande and his sidekick Mazamette (owlish stage comedian Marcel Levesque), but the most charismatic figure is the anagrammatically named Irma Vep. Played by the kohl-eyed Musidora, Irma is equal to any man in treachery and resourcefulness. As the installments unfold, each new Vampire chieftain is captured or killed (usually because the actors were called to military duty), but Irma survives to the climax. Fantômas and Les Vampires were called films aux épisodes, series in which the continuing characters passed through one self-contained adventure per installment, as in a TV series today. Then, imitating Pathé’s Perils of Pauline (1914), Feuillade turned to serials proper, sometimes called ciné-romans (“film novels”). These presented a long tale broken into segments, often accompanied by a serialized feuilleton in a daily newspaper.4 Feuillade’s first serial was the most popular long film of his career, Judex (1917). Bowing to the public criticism that Les Vampires glorified crime, he gave his public a darkly handsome superhero. In sweeping cape and floppy hat, René Cresté played the mysterious avenger who metes out justice with some help from the bumbling detective Cocantin (Levesque once more). Feuillade immediately made a sequel, La Nouvelle Mission de Judex (1917), and two more adventure serials, Tih Minh (1919) and Barrabas (1920), as well as the wartime drama Vendémiaire (1919). The vogue for crime films having passed, he turned to sentimental melodramas, which he also serialized on a grand scale.5 He was a tall, stocky man with large, heavy-lidded eyes and a close-cropped mustache stained by endless cigarettes. His duty, he said again and again, was to make the company profitable. He had a stake in those profits because Gaumont pegged part of Feuillade’s salary to the amount of finished footage he turned out, and he got a bonus if a film proved popular. So he enforced a ruthless e‹ciency. The former cavalry rider ran his team as a regiment, bellowing orders in a deep voice and using a drillmaster’s whistle to synchronize players’ gestures. He kept a cane handy, not to support himself but to snap during fits of rage. He forgave no infraction. When Jean Ayme, playing the leader of the Vampires, showed up late because of illness, Feuillade filmed Musidora blasting Ayme with a pistol. Getting up, Ayme asked what he should do next. “Nothing,” Feuillade told him. “You’re dead. You’ve got no more to do in Les Vampires. Go to the cashier and collect your wages.” Feuillade wrote a new chief Vampire into the next episode and never hired Ayme again.6 No time to spare for a sick actor; during the first several years, no time to work out scenes in advance. His players had to share his swift, intuitive craft and be ready to improvise. He usually couldn’t explain exactly what he wanted, so he would act it out, and soon he built up a stock company who learned what pleased him. His trusted players could sustain a shot for four minutes, and they would seldom need more than one take. Thanks

44

FEUILLADE, OR STORYTELLING

to his unsparing discipline he could knock oª a short film in an afternoon and a fortyminute episode in three days.7 At a pinch, footage shot for one purpose could be squeezed into another film. Why, in the midst of Les Vampires, is there suddenly an inserted tale featuring a horseman galloping after a bull? Feuillade had shot the chase in Spain months before and wanted to use it somewhere. Incredibly, he also found time to run the studio. He was “Artistic Director” of the company, meeting with Léon Gaumont six mornings a week at 7:30. Filming started punctually an hour later. Feuillade bought stories, hired actors, and supervised other directors’ work. At night he read scripts. He seldom took vacations. His frail wife lived in Montpellier until her death in 1911; his closest family, it seems, was his troupe. One of his actresses became his second wife, and his cameraman, Maurice Champreux, married his daughter. Loyal members of the team were guaranteed steady work for years. During the war, when the Paris studio proved di‹cult to maintain, Feuillade shot in Marseille, and in 1917 he relocated to the Gaumont facility in Nice. There, among roomy sets, picturesque estates, dazzling light, and rugged seashores and mountains, he could run everything to his liking. Every morning his assistant came to his courtyard, and from a window Feuillade would shout down the day’s requisites: a yacht, an Algerian birth certificate, a metallic spiral staircase one story high.8 Filming became a family outing. The actors would arrive at the studio for a seven o’clock breakfast, and over roasts, fish, and vegetables Feuillade would outline the day’s work. Around eight the cars would arrive to transport everyone to the location, where they filmed until nightfall. Then there was dinner at an inn, with singing.9 In Nice he reigned paternally over his troupe, turning out at least one ciné-roman a year: the crime serials following Judex, then the melodramas Les Deux gamines (1921), L’Orphéline (1921), Parisette (1922), Le Fils du Flibustier (1922), Vindicta (1923), and L’Orphéline de Paris (1924). At a period when Gaumont had few hit films and the US invasion of French screens was in full force, Feuillade could dictate terms. If Léon Gaumont wanted to see the latest footage, he had to come to Nice.10 Feuillade died of peritonitis in 1925, aged only fifty-two; his last film, Le Stigmate, was completed by Champreux. Feuillade flourished in a time of transition. When he started directing in 1906, films ran a reel (fifteen minutes) or less and were sold outright to exhibitors. The pictures looked very theatrical, with action staged far from the camera and performers waving and jumping about (Fig. 2.1). But films were starting to be shown in movie houses rather than vaudeville theaters and cabarets, and the audience was expanding. Soon movies grew to several reels, running from thirty to fifty minutes, and production companies realized that it was more profitable to rent prints than to sell them. The presentations became more upscale, providing customers printed programs like those in legitimate theater.11 Technique was also becoming more elaborate. American filmmakers were blending a variety of editing devices into a fast-paced, carefully calibrated style. A flurry of shots would trace characters bustling through doorways and rushing along sidewalks. Crosscutting might flash between simultaneous actions taking place in two locales. Some di-

FEUILLADE, OR STORYTELLING

45

2.1 Une Nuit agitée (1908): a harried landlord awakened by bohemian tenants.

2.2 A scene in The On-the-Square Girl (1917) begins with an establishing shot backstage at a fashion show. Anne enters.

2.3 In a reverse-angled medium shot René responds haughtily.

2.4 Another reverse angle reveals Anne, in medium close-up. Most US features from 1917 onward include such passages of analytical cutting.

rectors broke up a scene into medium shots and close-ups. The new style did not emerge in any steady, linear way, and there were false steps and blind alleys, but around 1917 the Americans consolidated a powerful storytelling strategy that involved building a scene out of several shots. A long shot, usually brief and at the start of the scene, would establish the characters’ positions. Then the bulk of the action was played out in medium shots and close-ups, usually taken from a variety of angles, so that characters’ expressions and gestures were magnified for clarity and impact. (See Figs. 2.2–2.4.) This quickly became known as the “continuity” style, and it insured that at any given moment the most important bit of story information would be isolated for the viewer.12 European directors used such devices more sparingly. Exterior scenes would include some cutting, particularly in situations of spying or pursuit, but usually the action wouldn’t consume as many shots as the Americans would use. In interior scenes, editing would

46

FEUILLADE, OR STORYTELLING

2.5 A ballet scene in Les Vampires (1915–1916) . . .

2.6 . . . uses an axial cut-in to enlarge the dancers.

be even less important. Most of the action would be played out in a single orienting view, called in France the tableau. The nearest characters would be framed from the thighs or knees up. The director might cut in a close-up to show a letter or newspaper article or visiting card, since such items were too small to read in a master view. At a point of intense drama there might be a medium shot of a character’s reaction. Such a cut-in would probably be axial—that is, taken right along the lens axis of the tableau framing (Figs. 2.5–2.6) rather than from the oblique angles on display in many American scenes (Fig. 2.3). The axial cut-in is the most conservative strategy for breaking down the scene, since it is least likely to disorient the audience. The Europeans’ emphasis on the tableau led them to minimize crosscutting as well. In the last episode of Les Vampires, “The Bloody Wedding” (June 1916), the journalist Philippe Guérande summons the police to the gang’s hideout, where they are celebrating their triumphs over staid society. Instead of swiftly alternating shots of Philippe waiting, the police racing to the rescue, and the Vampires partying, Feuillade gives us a thirtythree-second shot of Philippe standing in the street, with the police van emerging in the very far distance and eventually drawing up to him. Then come six shots of cops piling out and climbing over the estate’s wall (consuming fifty-two seconds). Only then does Feuillade cut back to the party, when the Vampires launch a dance that lasts a minute. Cut back outside, to two shots of the raiders approaching the house and entering through a rear door (forty-five seconds). Cut back to the party, where the chief Vampire amuses his crew by drawing a caricature of Mazamette and then firing his pistol at it, a gag that employs axial cut-ins and takes another seventy-five seconds. And then, as the Vampires strike up a frenzied dance in the full tableau, the police burst in at the rear.13 The nineteen shots in this sequence, the climax of Les Vampires, average 15.5 seconds each (a characteristic ASL for any episode of the serial). By contrast, several American films from 1915 to 1916 average four to ten seconds per shot across their entire running time,14 so a comparable climax in one of them would have a much lower ASL than Feuil-

FEUILLADE, OR STORYTELLING

47

lade’s. Gri‹th and some of his contemporaries would have split the police raid into many more shots and crosscut more often among the converging dramatic forces. Consider Gri‹th’s Intolerance, released only a few months after Les Vampires’ “Bloody Wedding” installment. The final phase of the rescue of the Boy from the gallows covers a mere 102 seconds in thirty-six shots, yielding an ASL of only 2.8 seconds.15 Just as important, Gri‹th repeatedly switches between the rescue party and scenes at the prison, returning to the Boy at the gallows six times. Feuillade makes do with a minimum of alternation: Philippe and the police (7 shots) / the Vampires (2 shots) / Philippe and the police (2 shots) / the Vampires interrupted by the police (8 shots).16 He has not absorbed Gri‹th’s lesson that each line of action can be reduced to one brief, telling image (often a close-up), thereby enabling many more switches between plotlines.17 We shouldn’t exaggerate the diªerences: European filmmakers relied on a rudimentary form of continuity, and many American films employed virtuosic staging within the frame.18 Still, the broad contrast provides a useful point of departure. Well into the late 1910s most directors outside the United States tended to favor ensemble performance within the sustained single shot.19 Long before Alexandre Astruc asked, “What is miseen-scène?” European directors were providing brilliant examples.

What did these filmmakers gain by playing down editing? Feuillade can help us find an answer. He stands between the first generation of filmmakers (Lumière, Méliès) and that younger generation (Gance, L’Herbier, Kuleshov, Dreyer, Murnau, Lang) who took up the new American style. His pressurized production routines seem to have favored a technique far more straightforward than that developed by his contemporaries; he is, we might say, the Howard Hawks of the 1910s, exploiting current conventions of staging with an almost diagrammatic sobriety. His straight lines, uniform illumination, and sparsely decorated sets make his assured choreography of bodies in space all the more vivid. Start with one of the crucial representational tasks that every filmmaker faces: directing the viewer’s attention. The idea behind American-style cutting was that changing the shot scale and angle would highlight what is dramatically significant at each moment. Some Soviet directors, admiring the American methods, saw this guidance as a central virtue. A protracted full shot permitted the viewer’s attention to stray to a minor player or an irrelevant bit of setting, but a close-up drove the eye to the essential detail. What continuity advocates tended to ignore, however, was that before cutting intervenes, we are already scanning the frame for important elements, and many factors cooperate to guide us. As humans, we are likely to pay attention to other humans, particularly to the most informative features of their bodies: eyes, mouths, and hands. Fortunately, those are features that tend to carry a great deal of dramatic weight as well. In following the story we are encouraged to concentrate on characters—their carriage and expressions, their gestures and the props they handle. And since movies consist of images, the filmmaker must use pictorial cues to guide us too. The director can manipulate movement, the size of the figures in the frame,

48

FEUILLADE, OR STORYTELLING

2.7 Tih Minh (1918): in a breathtaking shot during the final chase all five of our heroes clamber into a frighteningly tiny cable car.

2.8 The camera pans slightly as they slide oª into the distance, dangling far above a vast river; as their car dwindles, the cables converge to make it the vanishing point.

diªerences of tonality, and the compositional design as a whole. Such cues can reinforce the elements toward which our eye would gravitate anyhow, for example by centering the protagonist within the frame. Alternatively, pictorial cues can provide a gradation of emphasis whereby figures on whom we might spontaneously concentrate are given less weight than other components of the scene. A protagonist in the foreground, for instance, can turn away from us in order to draw our attention to someone in the distance.20 Moreover, in Feuillade’s day it was not enough to direct the spectator’s eye; one had to do so gracefully. Directors were expected to display a skill in pictorial composition—not least in order to help raise the cultural prestige of cinema. Feuillade wrote in 1910: “In reality, the cinema proceeds as much, if not more, from the art of the painter than from that of the stage director, since cinema addresses itself to our eyes by combinations of light and changing tonalities, and by qualities of composition. Why have films not given us sensations of beauty comparable to those aroused by a painting by Millet or a fresco by Puvis de Chavannes?”21 By 1910 a director could conceive the film image as requiring the kind of organization found in late-nineteenth-century academic painting.22 But how to achieve it? The most obvious way to highlight a piece of information is to put it smack in the middle of the picture format, as Feuillade does in Une nuit agitée (Fig. 2.1). In several of his 1910s films Feuillade uses a stark geometrical centering to heighten the drama (Figs. 2.7–2.8). Often Feuillade utilizes the center to introduce new material, as when in Les Vampires the supposedly dead bank courier reappears in the o‹ce and flabbergasts Irma Vep (Figs. 2.9–2.10). Yet this moment also indicates how even a little variance from geometric centering becomes quite compelling. No great sweeps of our eyes are required to move oª center from the resurrected “Spectre” to Irma, or, in Figure 2.11, from Fantômas to the mysterious face in the train car above him and to the right. Such instances remind us that it is often useful to think of an image, in painting or in

FEUILLADE, OR STORYTELLING

49

2.9 Les Vampires: Irma Vep discovers . . .

2.10 . . . the bank courier she has supposedly killed oª.

2.11 Fantômas (1913–1914).

cinema, as having several diªerent centers of interest. The challenge for the skillful director is to guide our attention among them. In another shot from Fantômas Feuillade tucks a doorbell into the right upper corner of the image, thereby setting himself the problem of directing the audience’s eyes to the ringing bell and back to Moche. Feuillade solves the problem by the straightforward expedient of obliging Moche to look at the bell (Fig. 2.12) and then turn to us, thoughtful (Fig. 2.13). In the 1910s, as today, an image need not be strictly centered if its design harmonizes elements within the frame. Such harmony usually involves one of two basic principles, both of which Feuillade handily delineates in his diverting 1908 short Les agents tels qu’on nous les représente / Les agents tels qu’ils sont (Policemen as We Portray Them / Policemen as They Are). The first part, consisting of a single shot, shows cops sneaking a smoke while an upright citizen is attacked and abducted by thieving apaches (Fig. 2.14). The second part employs two shots to present burglars spotted and pursued by tenacious policemen (Figs. 2.15–2.16). The thieves are captured, but one cop is wounded in the fray, and a final shot shows him recovering at a sanitarium, where he is awarded a medal (Fig. 2.17). Each episode exploits

50

FEUILLADE, OR STORYTELLING

2.12 In Moche’s o‹ce the doorbell rings in the upper right corner of the shot.

2.13 Then he turns toward the camera thoughtfully.

2.14 Les agents tels qu’on nous les représente (1908).

2.15 Les agents tels qu’ils sont.

2.16 Les agents tels qu’ils sont.

2.17 Les agents tels qu’ils sont.

FEUILLADE, OR STORYTELLING

51

a diªerent staging schema. In the first the cops and robbers are stretched out laterally along a fence. In the second the action moves diagonally, as was common in chase films of the time, drawing the action from the distance into the foreground. (Interestingly, the satire on cinematic stereotypes utilizes the horizontal arrangement, whereas the portrayal of heroic cops “as they are” relies on deep diagonals, as if depth signaled greater realism.) When staging horizontally—that is, setting two or more figures on the same plane— Feuillade sometimes leaves the center of the image vacant. This obliges him to organize the frame by balancing competing centers of attention, as in the shot of Moche’s doorbell (Figs. 2.12–2.13). Usually, however, such balance is not achieved by planting the figures in one place for the duration of the shot. Much of the appeal of Feuillade’s films, and of the cinema of his epoch, stems from the fact that cinematic composition creates not a snapshot but a dynamic flux whose stabilities and instabilities unfold over time. In the first part of Les agents . . . , Feuillade carefully alternates centered action, such as the robbery of the passerby, with two centers of interest, spread symmetrically across the frame (Fig. 2.14). Feuillade and his contemporaries show that in the cinema centering, balance, and other pictorial strategies find their most subtle fulfillment instant by instant. An element may move out of the geometrical center, arousing a spatial tension that is relieved or developed by a new element’s arriving in the vacant spot. At such moments decentering may be only a prelude to recentering; directing the audience’s attention to one element often prepares for shifting attention to another item. Our study of Feuillade’s unfolding horizontal arrangements can start with a comparatively simple example from Bébé apache (1910), one of his many comedies starring the child star Bébé. With his girl companion Bébé arrives at an outdoor party held by apaches. The first phase of the action seesaws from the extreme left of the frame (Figs. 2.18–2.20) to the right (Fig. 2.21), and the bulk of the scene develops in the geometrical center of the frame (Fig. 2.22). There is comparatively little sense of diagonal depth, and the main plane of action is quite distant from us. In Fantômas, three years later, Feuillade sets the action closer to the camera and creates a more spare and dramatic lateral arrangement. Lady Beltham has lured the actor Valgrand to her home while Gurn (a.k.a. Fantômas) is lurking behind the curtain. The plan is to call the police and induce them to mistake the actor for the escaped convict he has portrayed onstage. The action starts with Lady Beltham seated on the left and Valgrand on the right (Fig. 2.23). He reacts to the drug she has put in his drink, and she walks to the right side of the frame. As she passes the center of the frame, the curtains part and Gurn starts to peek out (Fig. 2.24). Once Lady Beltham arrives at Valgrand’s side, Feuillade has her look leftward, directing our eye back to Gurn, who obligingly widens the curtain’s gap (Fig. 2.25). The revelation of Gurn would not have been so vivid if Lady Beltham had not cleared his zone of the shot by going to frame right. The frame dynamic is soon inverted. Lady Beltham strides leftward, turning to

52

FEUILLADE, OR STORYTELLING

2.18 Bébé apache (1910). We are given time to see the apaches in the middle ground before the group freezes and turns our attention to an area at the extreme left.

2.19 A couple left of frame center dodges aside to the right . . .

2.20 . . . creating a path for Bébé and his girl companion, who trot in on the extreme left.

2.21 They seat themselves in the empty chairs on center right.

2.22 Soon the zone between the tables becomes the main site of action—including a knife fight between Bébé and an apache.

FEUILLADE, OR STORYTELLING

53

2.23 Fantômas.

2.24 Fantômas.

2.25 Fantômas.

2.26 Fantômas.

2.27 Fantômas.

54

FEUILLADE, OR STORYTELLING

Valgrand at about the center of the frame. This clears a space on frame right for the prison guards, who enter almost immediately. Gurn closes the curtain (Fig. 2.26). After the guards have arrested the groggy actor and departed, Lady Beltham sees them out on the right while Gurn emerges, creating a neatly balanced composition (Fig. 2.27). Throughout the scene, decentering serves only to resettle the shot and to direct attention to key elements. This example is instructive in another way. It reminds us that directors of the 1910s do not have to assume that every viewer must pick up on every item at the same instant. Rather, without seeming to be merely marking time, the director must stretch out an action long enough for all viewers to notice. For instance, perhaps some viewers spot Gurn at the moment Lady Beltham starts her rightward stroll across the frame (Fig. 2.24). But by the time she turns to look at him as he peers out of the curtain (Fig. 2.25), most viewers have registered his presence. Feuillade can only prompt, not force, the spectator’s attention; but he can prompt it through a cascade of cues. We usually think that a deep pictorial space demands diagonal compositions, but lateral staging can achieve considerable depth by setting rows of figures perpendicular to the camera. Near the end of the same episode Juve believes that Fantômas is disguised as Gurn, so he brings Gurn into the jail, only to discover that he is actually the actor Valgrand (with whom Fantômas/Gurn had switched identities). Instead of the single horizontal plane of Les agents . . . , this scene is conceived in depth but with depth itself consisting of several parallel layers (Figs. 2.28–2.34). The main alternative to lining actors up, either in one layer or several, is diagonal staging. Staging of this sort was common in exterior shots since the beginning of cinema, and within interiors it became more frequent between 1906 and 1913. Kristin Thompson has traced how American directors worked to fill the empty foreground area of interiors by designing sets with rear entryways and by placing furniture downstage so that figures were motivated to approach the camera.23 Feuillade’s films make robust use of diagonal depth, usually with all planes in focus. Gaumont took pride in his films’ crystalline photography, achieved largely by shooting in daylight, both in exteriors and in sets built in the glass-walled studio. Sunlight allowed Feuillade’s cameramen to keep both foregrounds and backgrounds sharp, and moving to the brilliant climate of Nice only increased the prospect of extreme depth of field, an eªect apparent in La Nouvelle Mission de Judex and made altogether splendid in Tih Minh (Figs. 2.7–2.8). Striking though diagonal depth is, it also poses compositional problems. Given players of comparable height, to put one closer to the camera is to unbalance the shot and give him or her more prominence. Three solutions suggest themselves. The director can push the more distant figure to the geometrical center of the format, compensating for its lesser size by its prime location. In some scenes we’ve already examined, a background figure gains saliency simply because it is more centered than a foreground one (Fig. 2.9). A second solution relies on the fact that in cinema as in painting, frontality is an eyecatcher. So a foreground figure can compensate for its large size by turning its (and our)

FEUILLADE, OR STORYTELLING

55

2.28 Another scene in Fantômas makes Juve compositionally prominent as he waits for Gurn to be brought in.

2.29 The police o‹cials lead Gurn in and plant him on a stool in the foreground. But now, as so often, Feuillade makes things hard for himself.

2.30 In order to register Gurn’s arrival fully, we ought not to be distracted by keen-eyed Juve, so Feuillade quietly insures that Juve is swallowed up in the crowd.

2.31 How to retrieve him? The o‹cials shift away slightly, revealing Juve—a squat, gray figure among all the men in black—in a fresh spot, left of the center.

attention to a more frontal figure in the distance. This is what happens when the bank manager swivels to call attention to the disguised courier in Fig. 2.10. Finally, the depth arrangement can exploit movement. We can add visual weight to the smaller figure by letting it change position while the foreground actor stays comparatively still. This tactic operates in Figs. 2.31–2.32.24 All three tactics allow depth staging to achieve compositional balance. As with lateral staging, the most dynamic depth compositions are constantly being reconfigured, so very often the person in the rear creates compositional balance in yet another way—by moving to the foreground. From the very beginning of cinema, directors displayed a strong urge, at least in exteriors, to pull distant figures to meet their counterparts in the front of the shot.25 This tactic not only permits performers to interact on

56

FEUILLADE, OR STORYTELLING

2.32 As Gurn is questioned by the o‹cial on the right, Juve advances on the left, completing a row of men lined up perpendicular to the camera.

2.33 With a flourish Juve tears oª the man’s mustache, revealing him to be not Fantômas but Valgrand, and a few men huddle behind the prisoner.

2.34 When the o‹cials take the hapless actor away, Juve sighs and collapses onto the stool in defeat. The shot makes a comfortable parallel to Fig. 2.28.

the same plane; it also promotes compositional stability by spreading roughly equal masses across the frame. Thus diagonal interaction often becomes only one moment in a suite of activities that culminates in horizontal layout. And, since the foreground area is by optical necessity narrower than that of the background, characters coming forward are squeezed closer together. With its eªect of enlarging the figures, bringing them to the same plane, and narrowing the gap between them, the movement from background to foreground became the norm for initiating a scene’s action or developing the drama to a higher pitch. Feuillade pursues exactly these implications of forward-thrusting action, as a simple example from Fantômas will illustrate. The crooked prison guard Nibet calls on Lady Beltham, preparatory to the substitution of Valgrand for Gurn. He arrives far back in the

FEUILLADE, OR STORYTELLING

57

2.35 Fantômas.

2.36 Fantômas.

shot, while she stands downstage. But by turning away from the camera, she grants Nibet some saliency (Fig. 2.35). Moreover, instead of having him simply walk to the vacant foreground chair on the right, Feuillade has Lady Beltham cross to sit in the chair while he strides diagonally forward to assume her previous position (Fig. 2.36). By passing through the central zone of the shot, Nibet receives momentary emphasis before he sits on the love seat on the left, rebalancing the shot. Again, once Feuillade has mastered this principle, he sets himself fresh di‹culties. During one scene of La Nouvelle Mission de Judex, the foreground figures pile up so rapidly that Feuillade must juggle them very carefully. He draws two, then three, then four men from the distance into a crowded middle ground (Figs. 2.37–2.41). To make things interesting, he adds a woman. Her arrival motivates a judicious cut-in (Fig. 2.42), after which the front-row ensemble has been reconfigured to make room for her (Fig. 2.43–2.44). Feuillade solves his visual problems by elegant rules of thumb: line the figures up in planes; given an even-numbered group, balance them around the frame’s vertical axis; with an odd-numbered grouping keep one figure in the center. And, we might add, block and clear a central avenue for upcoming developments. At intervals Roger conceals the doorway (Figs. 2.39, 2.41), which is revealed at just the right moments when he swings aside (Figs. 2.40, 2.43). More sustained is the virtuosic play with visibility in the opening scene of Les Vampires. The journalist Philippe Guérande comes in to his o‹ce and finds his dossier on the Vampires missing; he accuses the clerk, Mazamette, who confesses. Apart from some titles and inserts of written matter and a photograph, Feuillade plays out the four-minute scene within a single camera setup. Philippe enters through the rear door—an aperture that, throughout the scene, will frame foreground elements and provide a channel for entrances and exits (Fig. 2.45). In the frontmost plane Philippe bends to investigate the drawer, and his body blocks the central journalist (Fig. 2.46). This allows our attention to shift to Mazamette’s worried response, given in a fairly frontal view: we may already start to suspect that he is the culprit.

58

FEUILLADE, OR STORYTELLING

2.37 La Nouvelle Mission de Judex (1917): at his desk Jacques (a.k.a. Judex) tells Cocantin that Primerose and his son have vanished.

2.38 From the doorway in the rear Roger bursts in . . .

2.39 . . . and rushes to the foreground to announce that the dogs have been poisoned. He is given a central position, framed by the doorway.

2.40 Immediately, Dr. Howey enters from the same doorway, and this poses a problem for Feuillade: where to put him? The solution is to let Cocantin look behind Roger, who then moves aside to let us see the doctor’s approach . . .

2.41 . . . before creating a new, fairly balanced composition of the four men.

2.42 But now Jacques’ wife enters from the same doorway! Just as Jacques swears the men to secrecy, Feuillade cuts in to a medium shot of her.

FEUILLADE, OR STORYTELLING

59

2.43 We then return to the original setup, which now puts her at the center, two men squarely on each side like bookends, each pair turned and looking in somewhat symmetrical directions.

2.44 Naturally, the wife comes forward. The foreground is now full, with the flanking characters turning away from us so as to heighten the conversation between the central figures.

Philippe rises as he discovers the robbery, and Feuillade further calls our attention to Mazamette by having him rise in synchronization on frame left (Fig. 2.47). Philippe interrogates his colleagues, and Mazamette tries to sneak out in the distance. Philippe’s body blocks our view of him, accentuating the moment when, clearly if distantly in view on the threshold, Mazamette is halted by Philippe’s question (Fig. 2.48). Thereafter, as Philippe interrogates Mazamette in the foreground, the pair alternately block the other journalists and reveal their reactions (Fig. 2.49). Note that this sequence could not be as tightly choreographed on the stage. For those spectators at the center of the auditorium, Philippe’s entry (Fig. 2.45) would be visible in the gap between his two colleagues. But for the theatergoers on the side Philippe would be partly blocked in the course of his entrance. For viewers in the balcony his face would not be centered between them. Similarly, the moment when Mazamette halts in the doorway (Fig. 2.46) could not be seen by everyone in a theater situation; for many viewers Philippe would be standing in the way. And the minute blocking and revealing of the reporters simply could not be achieved on the stage, since only a few spectators would have a view corresponding to that given in our image. Theater staging, working within a wide and shallow rectangle, tends to be spacious and lateral, spreading the figures out to accommodate many sightlines. By contrast, thanks to the laws of optics, the film camera captures a pyramidal chunk of space, with the tip of the pyramid at the lens and the playing space radiating out from there (See box, Fig. 2.50). Since the camera views a unique configuration of bodies in space, you can not only pack the shot with many figures and objects. You can move each one as slightly as you like, blocking or exposing whatever area is necessary, all in the confidence that, in contrast to the live-theater situation, every viewer can easily notice the changes. This is what

60

FEUILLADE, OR STORYTELLING

2.45 Les Vampires.

2.46 Les Vampires.

2.47 Les Vampires.

2.48 Les Vampires.

2.49 Les Vampires.

FEUILLADE, OR STORYTELLING

61

THE CINEMATIC PLAYING SPACE In Feuillade’s day filmmakers were quite aware of a crucial diªerence between film and theater. On the proscenium stage the playing space is broad and comparatively shallow, so directors tend to spread actors across the horizontal expanse for the benefit of many sightlines in the auditorium. But in cinema only one sightline matters—that of the camera lens. Thanks to projective optics, the playing space of cinema constitutes a horizontally tipped pyramid, with light rays gathering at the lens. This yields a very narrow but rather deep area of visibility. (See Fig. 2.50.) When an actor approaches the camera, she fills up more of the foreground and blots out things behind her. Approaching the camera and concealing or revealing items in depth thus become powerful devices for guiding the viewer to salient information in the shot. In practice the eªective playing area is not pyramidal but trapezoidal. Players too close to the lens will be out of focus, so they must be held to a well-focused foreground plane— what writers of the 1910s called the “front line.” European filmmakers tended to keep the front line a fair distance from the camera, ten to twelve feet or even more. This yielded the so-called “French foreground,” with the nearest players presented in full-figure or cut oª around the shins. Several American filmmakers, particularly those working at the Vitagraph studio, favored a closer framing, with the camera set about nine feet away and cutting the actors midway at the knees or even around the hips. These “American foregrounds,” filling the frame with large figures, aroused some objections. “The actors cannot circulate,” complained one French observer; “they obscure one another. When a scene represents a room, we never see more than a part of it. . . . When it is a question of a salon or reception rooms, the director is obliged to project [the action] in depth; and the spectator sees no more than a corridor whose width in close shot corresponds at most to the space necessary for three or four characters.”26 Filmmakers commonly marked the trapezoidal playing area with chalk lines, string, or planks. A 1917 British manual advises cinematographers to start by selecting “a suitable depth for the front stage [front line]” and then calculate the “back line.” With the standard two-inch lens, the author recommends a front line six feet across and a rear plane twelve feet wide.27 Once this area was decided, the cameraman could set focus and determine exactly what would be onscreen. Even when the playing space was not so explicitly traced out, actors could be told of the scene’s “limits”: “Jill, Alan, and Jellicott are in this scene. Your limits are here on the right—by the table—and on the left here—by the harmonium. Don’t advance farther forward than the footstool.”28 The trapezoidal playing space of cinema is crucial to all cinematic staging, and the 1910s artisans exploited its resources with full awareness of its powers.29

62

FEUILLADE, OR STORYTELLING

2.50 Cinema’s pyramidal playing space, mapped onto a characteristic US set of the period, with the camera pointing into the corner of a ranch house. Note how the playing space tapers toward the lens. The front line, called here the “working line,” is set about fourteen feet from the camera and aªords a playing space only a little wider than four feet. Further back the width increases proportionately. The shaded areas show portions of furniture that will appear onscreen. The lens is a 50 mm one, standard for the period. With the lens focused at the front line and a diaphragm opening of f/5.6, a common setting, the shot would display reasonably sharp focus from about ten feet to twenty-six feet; that is, all planes in the scene would be in acceptable focus. The diagram is adapted from J. Berg Esenwein and Arthur Leeds, Writing the Photoplay (Springfield, MA: Home Correspondence School, 1913), 160.

FEUILLADE, OR STORYTELLING

63

allows the samovar and the uncle’s head (Figs. 1.33–1.41) to block the revolutionary’s daughter so exactly. The tactic of blockage and revelation thus becomes one of the most useful tools in the metteur en scène’s kit. Blocking and revealing can guide the spectator’s attention quite delicately across layered arrangements, as we’ve seen when the crowd of top-hatted o‹cials momentarily blots out Juve and allows us to concentrate on Valgrand (Fig. 2.30). The tactic can work in diagonal layouts, too, as the opening sequence of Les Vampires shows: thanks to the optical pyramid, tiny slices of space slip in and out of view when actors shift position. One schema for unobtrusively guiding the viewer’s attention lets a downstage character occupy a major zone before moving away to reveal another figure that is en route to the foreground. We’ve seen a simple example already in Bébé apache, where the apaches in the front slide oª to let us see the children’s arrival (Figs. 2.18–2.20); their movement prepares us to watch the area they vacate. A scene in La Nouvelle Mission de Judex provides a more graceful integration of frame balancing with this blocking-and-clearing maneuver. Judex traps two crooks in a cellar and materializes behind them. The most rudimentary diagonal staging would have put the two men in a foreground corner, with Judex emerging from the opposite side of the background. And indeed, the shot begins with the two men in the far right foreground (Fig. 2.51). But when they hear Judex shout, “You’re prisoners! Throw down your weapons!” they move forward and leftward as they drop their guns and raise their hands (Fig. 2.52). There’s now a problem of visibility: the man with the cap blocks the doorway. So while his partner halts, he continues leftward just as the door opens (Fig. 2.53), and Judex is revealed (Fig. 2.54). The crook continues his westerly drift until Judex stretches his arm commandingly across the doorway (Fig. 2.55). The man’s foreground movement wiping across the doorway draws our attention to the center while also delaying and underscoring the revelation of Judex. The crook’s final freezing in place gives the climax of the shot a nicely triangulated stability. What is striking in nearly every instance I’ve mentioned is the timing of a coordinated ensemble. Despite the haste with which these films were made, the synchronization of movements and poses is uncannily precise. Exactly as the Spectre turns away to be led to the background, Irma Vep swings back to us so that we can register her shock (Fig. 2.10). In Fantômas, when Fandor follows Nibet and his accomplice through an underground passage over a canal, Feuillade brilliantly orchestrates the characters’ silhouettes leaping from side to side, pursuing each other in a gentle oscillation across the central axis, blending moments of poise with symmetrical arrangements (Figs. 2.56–2.58). Could any director today execute the opening of Les Vampires, that ballet of smooth changes of position and short, sharp instants of stasis involving so many figures in a stationary shot? Our practitioners of intensified continuity don’t display much mastery of flexible centering, the use of doors and windows to mark out zones of action, the virtually unnoticeable balancing and rebalancing of the frame, the subtle pivoting of characters to

64

FEUILLADE, OR STORYTELLING

2.51 La Nouvelle Mission de Judex.

2.52 La Nouvelle Mission de Judex.

2.53 La Nouvelle Mission de Judex.

2.54 La Nouvelle Mission de Judex.

2.55 La Nouvelle Mission de Judex.

FEUILLADE, OR STORYTELLING

65

2.56 Fantômas: silhouettes in rhythmic pursuit through a tunnel.

2.57 Fantômas.

2.58 Fantômas.

and from the camera, the power of lateral staging, the rush to the foreground and the clearing of the front line, the slight shifts that conceal and reveal background action. Feuillade’s films of the 1910s oªer inexhaustible lessons in a craft nearly forgotten. They show us how the simple choreography of bodies in space can create a dynamic, intelligible presentation of a scene’s action. The same tactics were employed by Feuillade’s peers too, but he gave them a trim e‹cacy that suited the swiftly moving stories he had to tell.

While Feuillade and his European contemporaries were refining techniques of ensemble staging, American filmmakers were developing editing tactics that crystallized into the Hollywood continuity system. Many of these devices appeared sporadically in earlier European and Russian films. Directors felt free to cut to action in adjacent rooms or hall-

66

FEUILLADE, OR STORYTELLING

ways or nearby areas outdoors. The “insert” of printed matter or a photograph was very common. Well before 1913 Feuillade occasionally used closer views, changes of angle, and point-of-view shots,30 and the films I’m considering here display a gradual adoption of continuity devices. The reciprocities between staging and editing ought not to surprise us, since both are tactics for guiding our attention. Most European filmmakers (and some of their US counterparts) treat cutting not as an absolute alternative to staging in depth but rather as a complement to it, often an instrument of it. Broadly speaking, American découpage after 1915 subordinates staging to editing. The master shot establishes and orients; the space will be articulated primarily through closer views, matches on vision or movement, cuts to adjacent spaces, and the like. These carry us to the heart of the action (Figs. 2.2–2.4). Here cutting back to the long shot serves as a transition, usually to show the characters moving into a new arrangement, which will then be dissected by a new passage of analytical editing. Sometimes the cut back to an establishing shot will provide an expressive accent: in today’s cinema, a cut back to a master shot underscores a moment of tension or a pregnant silence. Still, in this tradition, from The On-the-Square Girl (Figs. 2.2–2.4) to Jerry Maguire (Figs. 1.46–1.59), the primary burden of a drama within a single locale is carried by analytical cutting and shot/reverse shots. To this we can counterpose an alternative that was prominent in Europe during the 1910s—a minimal continuity editing in which the tableau carries the dramatic weight. The few closer views tend to be straight cut-ins, serving usually to enlarge small items like letters and newspaper articles. Less often, the cut-in creates an accent, even a slight shift in emotional register, usually by enlarging a character’s expression at a key moment. This minimal or attenuated continuity editing, subordinated to staging, can highlight a region of the tableau for our scrutiny—as in one moment of Judex, when a single closeup of a photograph assures that we will make it a target in the dynamic of glances played out in the full shot (Figs. 2.59–2.62). Throughout the scene the photograph of Jacqueline Favraux sits quite visibly in the foreground right of Judex’s desk, and we have plenty of time to become aware of it. Sometimes, however, composition or movement in the tableau cannot su‹ciently stress dramatic factors, and then cutting can supplement staging. Sometimes this leads to remarkably daring découpage. In Tih Minh the Marquise Dolores is stealing Sir Francis Grey’s wallet, but she hides herself when he enters. Grey bends over his drugged valet, alongside a full-length mirror (Fig. 2.63). A sharp-eyed viewer may spot, in the bottom of the mirror’s reflection, Dolores’s face peering out from under the bed. Feuillade could have seized our attention by letting Dolores make a big movement, but instead he cuts in to the reflection at the moment she pokes a revolver out (Fig. 2.64). More boldly, a return to the initial framing shows the pistol firing (Fig. 2.65) and Grey ducking aside (Fig. 2.66) before he hurls himself onto the oªscreen bed—an action revealed in the upper right of the reflection (Fig. 2.67). The close-up has encouraged the viewer to attend to a

FEUILLADE, OR STORYTELLING

67

2.59 Judex (1916): while old Kerjean waits for Judex and Roger to return from their secret prison, he picks up a mounted photo from the desk.

2.60 An inserted close-up reveals that it shows Favraux’s daughter. Kerjean infers that Judex has fallen in love with Jacqueline, and soon we must remember the photo and watch how other characters interact with it.

2.61 When Judex and Roger enter from the rear, Kerjean gestures toward the photo.

2.62 After the old man leaves, Judex broods longingly over the picture while Roger averts his eyes tactfully, as he comes to Kerjean’s conclusion as well.

slice of space available in the mirror, highlighting the narrative action while also leading us to notice activity around the bed, an adjacent region of reflected depth. In such ways the principle of enlarging a detail can be extended to a piece of action, even when the staging makes that action di‹cult to see. Our example of characters rushing toward the foreground desk in La Nouvelle Mission de Judex has shown that the blocking-and-concealment strategy can be combined with editing: when Jacqueline comes into the room, Feuillade cuts in to her (Fig. 2.42) before returning to the master shot (Fig. 2.43). Here Feuillade’s cut stresses the character whose reaction will be most important. (Judex has just insisted that she not be told of the danger facing her family.) But the more general principle seems to be that when a foreground figure blocks something in depth, that may motivate a cut to whatever is momentarily concealed (Figs. 2.68–2.70).

68

FEUILLADE, OR STORYTELLING

2.63 Tih Minh.

2.64 Tih Minh.

2.65 Tih Minh.

2.66 Tih Minh.

2.67 Tih Minh.

FEUILLADE, OR STORYTELLING

69

2.68 Early in Tih Minh, when Jeanne comes forward to a desk, she is followed by her husband, Jacques.

2.69 Feuillade cuts in and she turns, but instead of revealing him (as in Fig. 2.40), the foreground figure obscures him.

2.70 Feuillade then supplies a closer view of Jacques before cutting back to the master shot. Compare Figs. 2.2–2.4.

In our early example from The On-the-Square Girl the figure entering in the background is perfectly visible in the establishing shot (Fig. 2.2), and the cut drastically shifts camera position to show René’s disdainful reaction (2.3) before enlarging Anne’s face (Fig. 2.4). Feuillade, however, remains tied to the camera axis; in our Tih Minh case (Figs. 2.68–2.70) he cuts directly “through” Jeanne to Jacques and provides no reverse shot of her. Curious though this practice may seem today, it suggests that Feuillade feels obliged to provide a closer view only when something cannot be made fully visible in the long shot because it is tiny (printed matter, a photo), peripheral (a hand under a bed), or blocked from view. (Had Feuillade been directing the On-the-Square scene, he would presumably have made René’s expression visible by having him turn to face the camera, as we’ve seen in Figs. 2.10, 2.43, and 2.47.) In this respect, as others, Feuillade’s method seems typical of the European approach. Directors staging in depth tended to push their luck, filling the frame

70

FEUILLADE, OR STORYTELLING

2.71 Fantômas.

2.72 Fantômas.

2.73 Fantômas.

with so many blocked and revealed figures that some actors were obscured. An axial cut could bring those souls back into the action, even if the positions did not match perfectly (as often happens in Feuillade). So the 1910s director could choose between full-fledged découpage in the American manner, which emphasized the string of closer views from varied angles, and a kind of sparse or attenuated continuity, which emphasized the full shot and supplemented it with a cut-in or two. Alternatively, the director could treat the larger views and the cut-ins as roughly equal in weight. This usually occurs when the cut-ins are not tight close-ups but medium shots, and fairly densely composed ones at that. Interestingly, they don’t fulfill the Soviets’ dictum that an insert should carry only one piece of information. Feuillade and his contemporaries show that cutting together moderately close views can creatively rearrange data given in the full view.

FEUILLADE, OR STORYTELLING

71

2.74 Judex: after a master shot Favraux’s banquet is broken into closer views . . .

2.75 . . . with angles only slightly varied from the camera axis. This strategy allows Feuillade to continually reshift compositional elements . . .

2.76 . . . and introduce new foreground characters (here, the mysterious bearded assistant) while maintaining spatial reference points (the banker Favraux in the distance).

For instance, during the famous gun battle among the barrels in Fantômas, changing camera positions yield a variety of foreground/background configurations, all anchored in adjacent space by the skiª moored in the distance (Figs. 2.71–2.73). No less remarkable, though more dramatically motivated, is the banquet scene in Judex. As the guests at banker Favraux’s home await the clock’s striking ten, Feuillade’s analytical editing permutes the elements of the scene, preserving and refreshing the depth relations (Figs. 2.74–2.76). Arguably, this is a more inventive use of closer views than the fairly naked medium shots of The On-the-Square Girl (Figs. 2.3–2.4). Historians influenced by Kuleshov and Pudovkin have too often presupposed that cutting to closer views necessarily reduces and simplifies the pictorial field. Feuillade’s jammed but cogently composed midrange shots remind us that guiding the audience’s attention is still necessary when the camera is fairly close to its subjects.

72

FEUILLADE, OR STORYTELLING

2.77 Les Deux gamines (1920): Feuillade employs eyeline-match cutting from one medium shot . . .

2.78 . . . to another. These spare views contrast with the dense medium shots of Figs. 2.74–2.75, coming somewhat closer to the US norm (Figs. 2.3–2.4).

Feuillade was not entirely conservative; he adopted continuity devices selectively and gradually. Judex (1917) sometimes opens scenes with medium shots of characters, and La Nouvelle Mission (1917) includes occasional shot/reverse shots and matches on action, although usually without distinctly changing the angle or penetrating the scenic space in the US manner. Across the 1910s his cutting pace accelerates, rising from 17.1 seconds per shot for Fantômas (1913) to 7.6 seconds for Barrabas (1919). By the time he made Les Deux gamines, Feuillade was exploring most of the découpage alternatives I’ve reviewed. Some scenes are built almost completely out of stripped-down American-style medium shots (Figs. 2.77–2.78). Others revert to tightly timed depth staging, with cut-ins to accentuate a scenic element. In one clever scene the girl Ginette goes with her benefactor Bersange to see her father, now working in a café (Figs. 2.79–2.84). Such moments show that complex staging was not an absolute alternative to editing but could often be supplemented and strengthened by it. In his last years Feuillade converted to something close to découpage in the American manner. All eight of the 1920s films that I’ve seen have ASLs between four and seven seconds. Sometimes scenes of physical action are cut very fast; L’Orphéline (1921) contains shots shaved down to five or six frames. In these films Feuillade’s staging serves principally to arrange figures in the establishing shot and reestablishing shots. Ironically, now his long shots often avoid depth in favor of simple lateral layouts. He seldom attempts shot/reverse-shot cutting in these films (and when he does, he sometimes botches it). Instead, his ciné-romans rely on a characteristically cheap and fast solution, one he had explored in the 1917 serials: the medium two-shot showing people in lengthy conversation (Fig. 2.85), broken by many dialogue titles. (In some of his late films, titles constitute a fifth of all shots.) As prudent a strategy as any he adopted earlier, the sustained two-shot minimized changes of camera position. This proto-talking-cinema style doubtless allowed him to turn out footage quickly. By the early 1920s he had added a

FEUILLADE, OR STORYTELLING

73

2.79 Les Deux gamines: in a long shot Bersange brings Ginette into the bar, where her father is clearing a table on the left and the café owner is tidying up a room in the background.

2.80 Cut to a three-quarter medium shot of Ginette smiling, her father visible in reflection in the mirror; this, the scene’s one cut-in, underscores her moment of recognition.

2.81 In a closer long shot, rather more a tableau, the father glances apprehensively to the rear, not wishing to attract his boss’s attention . . .

2.82 . . . and Bersange turns just as the woman becomes visible for us.

2.83 He helps Ginette toward her father and then steps to the center and turns away from us, hiking up his arms “naturally” so that the two can share a furtive embrace on the left. By the logic of the optical pyramid, blocking our view of the woman equals blocking her view of father and daughter!

2.84 Ginette then hurries back to the right, and as Bersange swivels, the old woman is revealed and comes forward in a new, harmonious composition.

74

FEUILLADE, OR STORYTELLING

2.85 A sustained medium two-shot from Barrabas (1920).

second cameraman, who would film details while the principal camera concentrated on the whole scene.31 Moments of clever staging crop up in Feuillade’s late films, but by and large he had abandoned complex choreography.

Feuillade’s staging principles grow out of two decades of explorations conducted by filmmakers throughout Europe. We can trace the principles back to the Lumière films, although a more proximate predecessor would be the French “Film d’Art” cycle, an eªort to upgrade the new medium by importing plots and players from the legitimate stage. The most famous Film d’Art production, L’Assassinat du Duc de Guise (1908), was long deprecated by historians who saw editing as the essence of cinema. Although granting that the film brought a new realism of performance to cinema, Georges Sadoul claimed that L’Assassinat “retains, ten years after L’Aªaire Dreyfus, the technique which Méliès created in his Montreuil studio. The film didn’t know, didn’t even suspect, the existence of editing [montage], although several contemporaneous films had begun to make systematic use of it. The work is merely a succession of tableaux vivants, each scene taken from a vantage point centered in the orchestra seats. In this respect, the Duc de Guise is a piece of filmed theatre.”32 Eventually, historians would praise L’Assassinat for its smooth contiguity cutting, which carries its hapless title character through a series of adjacent rooms.33 Now we can see that the film also constitutes a model of the techniques of balance, aperture framing, and precisely timed rearrangement that would be so important for the next dozen years. The scene leading up to the assassination opens with the duke moving toward the camera; it closes with his exit via the same doorway. Across the scene the actors play out a ballet of blocking and revealing, concentrated in a few zones of spatial and dramatic tension (Figs. 2.86–2.87). As early as 1908, then, filmmakers were creatively exploiting cinema’s trapezoidal playing space to create precise compositions. Our early examples like Bébé apache suggest that Feuillade mastered the staging tactics on display in L’Assassinat and proceeded to bend them to his own ends.

FEUILLADE, OR STORYTELLING

75

2.86 From the fifth tableau of L’Assassinat du Duc de Guise: Duc Henri stops in the foreground, becoming the most salient figure—centered, frontal, comparatively close, and the object of others’ attention.

2.87 Henri swivels, revealing the guard in the background, now lounging on a chair and blocking his escape.

In studying those ends I’ve concentrated on certain denotative qualities, particularly the ways in which staging steers our attention to narratively salient information. Instant by instant Feuillade draws our eye to the most important zones of the frame without our being aware of the process. This is a primary goal of any plot-driven cinema. Only by tracking the flow of the drama can we respond with curiosity, suspense, or surprise. In order to enjoy the stages through which Valgrand falls into the trap laid by Fantômas and Lady Beltham, we must follow Feuillade’s lead and pick up on the salient shifts of figures (Figs. 2.23–2.27). With a lovely inevitability each image provides a wholly adequate viewpoint on the action. At bottom, Feuillade’s staging techniques function as crisp, unobtrusive cinematic narration. His work reveals a kind of optimal design within the 1910s long-take tradition. This is not to say that his style lacks expressive qualities. The clean lines of his staging permit a relatively understated acting. “Feuillade asked us to suppress the useless gestures of pantomime,” the actor René Navarre remarked years later. “He all but forced me to keep my hands in my pockets and to act just with my eyes and my body.”34 In the crime serials the sobriety of staging and performance throw into relief the outlandish plot twists. A python slithering out of an air duct or a cannon assembled in a hotel room are treated as laconically as a man slitting open a telegram at the breakfast table. Moreover, although the pithiness of the staging doesn’t invite us to call it “ornamental,” it displays a strong impulse toward symmetry, echo eªects, visual tensions sustained until they subtly resolve—all of which can be savored for their own sake. Throughout, Feuillade sets himself problems that he overcomes, often in elegant ways. Feuillade eventually won a measure of fame outside specialist circles; at intervals Les Vampires is rediscovered as a great classic that is also good fun. (Perhaps some day the same recognition will be granted to the superb Tih Minh and Feuillade’s sparkling short comedies.) Yet the subtleties of his art remain little appreciated. Even less celebrated are his gifted contemporaries who worked in the same tradition. First among them, it now

76

FEUILLADE, OR STORYTELLING

2.88 Georg af Klercker’s Mysterious Night of the 25th (1917): an actual club room, filmed with a wide-angle lens, yields many areas of activity; compare Feuillade’s more sparsely furnished interior sets.

2.89 Robert Reinert’s Nerven (1919): as the neurotic protagonist Johannes lurches to the camera with a bottle of poison, the face of his blind sister peeps out in the gap between his arm and the cupboard door.

seems, was Victor Sjöström, remembered more for The Phantom Carriage (1921) and The Wind (1928) than for his splendid films of the early 1910s, particularly the masterpiece Ingeborg Holm (1913).35 There are other masters well worth studying, such as the Swedish director Georg af Klercker, who delighted in cramming his interiors with tables, lamps, armchairs, and hanging curtains, creating a maze of furnishings that the actors would negotiate unflappably. In eªect, he plays Visconti to Feuillade’s Hawks (Fig. 2.88). There is also Robert Reinert, a German filmmaker who also presents a well-filled frame, but thanks to the narrowing tip of the optical pyramid, he stages many scenes in claustrophobic proximity to us (Fig. 2.89). Eisenstein might say that Reinert employs montage within the shot, but we can also see his tactic as one European director’s way of finding functional equivalents of cut-ins, a solution that yields a bonus—an expressively asymmetrical composition that heightens the intensity of the developing action.36 Both Klercker and Reinert give us quite florid elaborations of the premises that governed Feuillade’s staging. Wilder still are the baroque inventions oªered in Yevgenii Bauer’s films. He, too, found ways of packing his sets, particularly with massive columns, which divide the space into zones and provide opportunities for character hide-and-seek. One of Bauer’s most memorable scenes, however, uses yawning vertical spaces surrounding his characters in order to disorient us again and again. The intertitles for The King of Paris (1917) have been lost, so it’s di‹cult to reconstruct the exact purport of the scene taking place on the staircase. Yet that hasn’t kept it from becoming a favorite of researchers into 1910s film.37 The rapacious Marquis de Prédalgonde, the “king” of the title, flirts with a young woman in a large foyer. After she leaves, he talks with his mentor, Rascol. Nothing complicated about this, except for a simple addition to the decor: a stupendous mirror. It stands across from a large open staircase, and it has an extraordinary power to confound our sense of the locale. At the start of the shot, because we cannot see the edges of the mirror, we are inclined to assume that it starts on

FEUILLADE, OR STORYTELLING

77

2.90 The King of Paris (1917).

2.91 The King of Paris.

2.92 The King of Paris.

2.93 The King of Paris.

the left and ends midway into the image, with the staircase extending rightward past the central column and oª into the background. Bauer confirms this assumption at the outset of the shot by showing the woman touching up her hair, complete with reflection, while other women go up the stairs on the right side (Fig. 2.90). We cannot know that they too are reflections in the same mirror. The shot unfolds as a string of weird incompatibilities and delays. At one point, in reflection the marquis seems to be moving quite close to the woman (Fig. 2.91), but in reality he is far away, indeed oª-frame right (Fig. 2.92); it’s as if there are actually two couples, mimicking each other slightly out of sync. Similarly, when the marquis chats with the woman, the spatial cues force us to assume that Rascol is descending the stairs behind them, in a classic eavesdropper situation (Fig. 2.93). Yet when the woman bolts out frame right (Fig. 2.94), she suddenly reappears (Fig. 2.95) in a reflection on the very stairs we thought we were observing directly, revealing as well that Rascol has been standing not behind the couple but across the foyer facing them. This challenge to our sense of the layout is reinforced after a pause: Rascol descends the stair (Fig. 2.96) and enters

78

FEUILLADE, OR STORYTELLING

2.94 The King of Paris.

2.95 The King of Paris.

2.96 The King of Paris.

2.97 The King of Paris.

not from the background but from the right, coming to stand by the marquis (Fig. 2.97). So robust is the illusion that we have to struggle constantly to remember that the staircase “behind” the characters is actually oªscreen. Bauer cunningly gives us a recognizable mirror in the left half of the frame and a “hidden” mirror in the right half. Whereas Feuillade strives for a maximally informative spatial layout, Bauer exploits our perceptual inclinations in order to disorient us (in a manner reminiscent of Manet’s famous painting Bar at the Folies Bergère). Perhaps the original dialogue in The King of Paris gave these treacherous reflections a thematic significance, but in any event the virtuosic play with illusion seeks to frustrate our eªorts to grasp the action, while also indulging in flagrant pattern making—a swerve toward a decorative use of style.

How might we explain the emergence of such a rich tradition of staging in the period 1908–1920? If we reflect on the tasks and problems faced by filmmakers, we can sketch out some answers. E. H. Gombrich has suggested that many changes in the history of vi-

FEUILLADE, OR STORYTELLING

79

sual representation can be understood as eªorts to adapt inherited visual forms (the “schemas” I’ve invoked) to new purposes or functions. One example is the growth of realistic painting in Europe around the thirteenth century. The Christian Church, Gombrich argues, sought to dramatize sacred stories for the unlettered faithful. In order to fulfill this purpose, artists experimented with ways to depict figures and locales with greater expressiveness and spatial realism. Painters were obliged to adapt the schemas available to them, drawing on the narrative painting of the Greeks and on the Hellenistic traits that persisted in Byzantine art. Once artists had discovered new means of realistic and expressive depiction, those very means became ends in themselves, and artists competed to find ever-more evocative and virtuosic ways to render the appearance of the visible world.38 Imagine a similar process at work in film around 1909. From the start films had tended to grow longer. One-shot films became multiple-shot films; multiple-shot films became one-reelers; films expanded to several reels and then to feature length—all in response to industrial and cultural demands. As films got longer, they were obliged to tell more complicated tales.39 So filmmakers linked episodes, embedded stories within stories, multiplied lines of action, added twists and complications, and created more complex characters. At the level of style directors were faced with several tasks, notably that of guiding the audience’s attention to the most important elements of these increasingly complicated plots. And as the film became longer, the director would have to find ways to intensify the audience’s emotional involvement—to make the extra time spent on the film a satisfying investment. One cluster of solutions was pursued by the Americans: let editing pick out a scene’s salient details, switch from one plotline to another, and arouse the viewer through delay and suspense. Another option was to direct the audience’s attention by means of more intricate staging. This alternative encouraged directors to expand their compositional repertoire so that the audience could follow a more elaborate intrigue and enjoy a stronger emotional response. From this perspective Philippe’s discovery of the rifled dossier in Les Vampires (Figs. 2.45–2.49) and the piling up of figures in the foreground desk of La Nouvelle Mission de Judex (Figs. 2.37–2.44) constitute elegant solutions to the problem of conveying a rapid flow of story events while also maximizing curiosity and excitement. Another goal may have governed filmmakers’ eªorts. As films became longer and moved into dedicated venues, apologists for the new medium sought to justify film’s cultural credentials. The prestige of literature and theater could legitimize cinema, so directors drew plots, players, and techniques from the stage.40 Significantly, as we’ve seen, Feuillade broached the idea that artistic cinema would borrow from tasteful salon painting. His eªorts to balance the frame smoothly, to create an overall visual dynamic that lets each narrative element stand out cleanly at the proper moment, may constitute another means of diªerentiating a more aesthetically dignified cinema from its comparatively crude predecessors. Any new means that Feuillade and his contemporaries might devise could build on deep-seated visual predispositions: the eye is attracted by contrast, frontality, movement,

80

FEUILLADE, OR STORYTELLING

2.98 William Frederick Yeames, Defendant and Counsel (1895). Oil on canvas, 53 1⁄2" × 78". Bristol City Museum and Art Gallery.

and specific items, such as faces and gazes. At the same time, filmmakers could borrow and modify certain pictorial devices already available in their milieu. Many tactics, such as turning from the viewer, letting secondary characters look at key figures, and balancing figures laterally, probably came from the theater. Comparable schemas were doubtless available from academic painting, particularly realist and narrative traditions of the late nineteenth century. Consider, as one example, William Frederick Yeames’s Defendant and Counsel (1895; Fig. 2.98). The dark-robed lawyers balance the brightly lit defendant, and the men’s wigs and profiles form a wedge pointing to the pale, frontally positioned woman on the right. One task that Feuillade and his peers set themselves was to shift actors into and out of such harmonious compositions, extending to the moving image principles of visual design already canonized in academic art.41 Gombrich also points out: “Each gain or progress in one direction entails a loss in another.”42 The Renaissance painter had to reconcile the commitment to a new realism of detail with the demands of clear and harmonious composition.43 Similarly, Kristin Thompson has suggested that editing was a mixed blessing for early film, with the cut promising a freedom of time and space but also threatening a loss of narrative clarity.44 From this standpoint the mise-en-scène of the European tradition can be seen as a conservative, low-risk strategy. At a time of stylistic change directors who had already mastered ways of guiding the viewer within the sustained shot may have been reluctant to tackle the new demands of continuity editing. Economic factors may also have played a role. A cinema dominated by editing required

FEUILLADE, OR STORYTELLING

81

fairly careful planning, if not full-blown shooting scripts. (As Feuillade cut more, he began to write out his scripts in more detail.) By contrast, staging-based production can get by with less blueprinting. Cutting-based shooting also requires new production roles, such as that of the American “script girl,” who kept track of the many camera setups. If the director has authority in the studio and works with the same personnel from film to film, as Feuillade did, a style based on long takes can be very e‹cient, turning out many minutes of footage per day. Feuillade was paid by the meter, so any time spent planning rather than filming would eventually have to justify itself in greater output. Perhaps directors in Europe, Russia, and the Nordic countries relied on master scenes for so long partly because they had no incentives to embrace the pre- and postproduction control aªorded by editing or the constraining routine of shooting many setups. Needless to say, a reliance on long takes triggered problems—problems of pacing, of maintaining interest, and of keeping everything in proportionate significance as the action unfolded. Precision staging in depth oªered a reliable solution. A frontal figure in the foreground is an eye-catcher, a point of orientation for the rest of the action. By presenting other figures as competing centers of interest, or as engaged in more vigorous movement, the director could balance the shot and encourage a scanning that would pick out the salient elements. And, as we’ve seen, the powerful advantages of carrying the players downstage coordinate naturally with the rising tension of a scene’s action. To adopt precision staging, however, initiates a fresh crop of problems. Once characters move closer to the camera, they occupy more frame space and may mask important action behind them. Hence the need to alter the arrangement in the course of the scene. Diagonal layouts of figures become lateral ones in the foreground (Figs. 2.35–2.36), lateral arrangements rotate or oscillate around the central axis (Figs. 2.23–2.27), impulses thrusting directly to the camera hit the foreground and spread out (Figs. 2.38–2.44). The admirable choreography developed by Feuillade and his contemporaries constitutes a sustained eªort to shape the viewer’s attention by ever-more refined control of composition and depth relations. Once the incentive to innovate was in place, artists had a formal problem: to shape existing schemas to the new demands of narrative presentation. Feuillade and his peers discovered staging techniques adequate to films displaying more twists of action and greater emotional complexities than the one-reelers of earlier years. As filmmakers’ mastery increased, many also explored the expressive resources of staging. They set themselves new di‹culties, often sheerly for the sake of overcoming them. In the process they created a cinema of subtle beauty.

82

FEUILLADE, OR STORYTELLING

3 MIZOGUCHI, OR MODULATION

American films swept the globe in the wake of World War I, and the découpage methods they employed soon became a cinematic lingua franca.1 Feuillade and his contemporaries adjusted to the new norms, and a younger generation—Abel Gance, Jean Epstein, Carl Dreyer, Fritz Lang, F. W. Murnau, and many others—embraced them eagerly. The few Asian films surviving from before 1930 suggest that Japanese filmmakers also developed assured, idiosyncratic variants of the emerging style. Fragments from the 1910s indicate that, like Western filmmakers, Japanese directors used both lateral staging, with figures strung out perpendicular to the camera, and a more diagonal arrangement (Figs. 3.1–3.2). Soon came demands to modernize Japanese cinema. The “Pure Film” movement urged directors to adopt close-ups and crosscutting, as well as Hollywood script formats.2 Magazines published explanations of American continuity practice, and one column habitually compared the number of shots in foreign films to the cutting rate in Japanese films.3 Some advocates hoped that “Americanized” filmmaking would force directors to abandon outworn theatrical traditions; other writers argued that only by adopting foreign techniques could companies find export markets. New studios, most notably Shochiku, were formed with the avowed purpose of producing modern, world-class films. Japanese directors were thus pressured to master continuity editing, and by the mid1920s most had.4 A climactic sequence of The Cuckoo (Ikeda Yoshinobu, 1922), for example, displays notable subtleties of cutting and camera angle (Figs. 3.3–3.4). Apart from the unusual variety of angles, the scene illustrates a characteristically Japanese interest

MIZOGUCHI, OR MODULATION

83

3.1 Some scenes in Makino Shozo’s Chushingura (1913 or 1917) recall European cinema of the period in staging their action in parallel horizontal rows; the kneeling vassals bring the two disputants into prominence. (Compare Figs. 2.1 and 2.14.)

3.2 Makino also stages actions in marked depth, as when Lord Kira leaves the corridor with his attendants. (Compare Figs. 2.17 and 2.87.)

3.3 The Cuckoo: several close views show Namiko’s family gathered around her sickbed.

3.4 In a striking aperture framing, one woman’s face is enclosed by the curving bedstead.

in the decorative dimension of the image. The curving iron bedstead, partly hiding the family’s faces, becomes a vivid graphic component, adding a purely pictorial dimension to the mournful tenor of the scene. The modernization of Japanese cinema was accelerated by a 1923 earthquake that demolished large stretches of Tokyo. There came a call to rebuild Japan on a scale commensurate with the glittering metropolitan West. The major cities began to sport o‹ce buildings, coªee shops, comic books, jazz, flapper styles, and other marks of urban sophistication.5 Film companies eliminated remnants of theatrical tradition, such as the oyama (female impersonator). Production accelerated to an astonishing six hundred to eight hundred films per year, making Japan’s industry the world’s most prolific. Into the studios came a new generation of men who admired foreign films and who were aware of contempo-

84

MIZOGUCHI, OR MODULATION

3.5 A triumph of the pictorialist trend: servants gather outside while the Abe brothers discuss their future after their father’s death (The Abe Clan, 1938). Compare Fig. 3.1.

rary trends in theater, art, and politics. Kinugasa Teinosuke, Ito Daisuke, Makino Masahiro, Mizoguchi Kenji, Ozu Yasujiro, Naruse Mikio, Gosho Heinosuke, Uchida Tomu, Shimazu Yasujiro, Shimizu Hiroshi, and others forged modern Japanese film. Beginning their careers in the 1920s and early 1930s, they defined their country’s cinema for thirty years.6 Directors found varied solutions to the problem of catching up with Hollywood and Europe. Many films surviving from the period straightforwardly obey the premises of analytical continuity. Beyond this baseline three distinct tendencies emerged. Most arresting is what we can call a calligraphic style, showcased in the swordplay films (chanbara). Here swordsmen hurl themselves into combat at hyperaccelerated speed, popping in and out of the frame unexpectedly, glimpsed in swift, often bumpy panning and tracking movements. Cutting is very rapid, often flamboyantly disjunctive. A much calmer technique is on display in what we can call “piecemeal” découpage. Here each scene is broken into many distinct medium shots and close-ups, not only of faces but of hands, props, and details of setting. The Shochiku studio cultivated this American-Lubitsch approach, and it is on display in the silent films of Ozu, Naruse, and Gosho. A third style is predicated on distant shots, meticulous compositions, and graceful camera movements. Mizoguchi may have helped solidify this “pictorialist” approach, which became a marker of prestige in the late 1930s (Fig. 3.5).7 Although all three approaches derived from classical continuity, each fostered experimental impulses too. The calligraphic style encouraged camera tricks and hectic cutting, sometimes with little concern for anything but visceral arousal. The result is a frenzied pictorial style that has few counterparts in Western cinema until the 1960s. The piecemeal approach allowed Ozu to expand Hollywood’s 180-degree dramatic space into a circular arena of action and then cultivate a playful approach to framing, setting, and shot design.8 The pictorialist tendency, as we will see in Mizoguchi, permitted the filmmaker

MIZOGUCHI, OR MODULATION

85

to cultivate exceptionally delicate staging tactics. Overall, the urge to decorate the stylistic surface, to embellish even a straightforward scene with momentary flourishes, remained a hallmark of Japanese films. At the climax of Wife, Be like a Rose! (1935), for example, Naruse constantly tracks the camera up to and away from his players, cutting these sweeping movements together in overtly parallel patterns. Throughout the world the coming of sound modified the continuity framework. Shots became lengthier; whereas the typical US film of the 1920s averaged around five seconds per shot, that duration doubled in the sound era. Several factors slowed cutting pace. Shots could not be broken up by intertitles, and it was di‹cult to synchronize dialogue during editing. Because moving the heavy sound cameras to various setups was time-consuming, directors tended to sustain their shots. However, once mobile supports were devised to shift the camera around the set, directors began to utilize more frequent tracking shots than they had in the silent era, if only to add the visual variety lost through a slower cutting pace. Likewise, depth staging and the deep-focus image were integral to early silent cinema, but filmmakers from Europe, Russia, the United States, and elsewhere found that these techniques could also fulfill the demands of dialogue-based production. Many directors who achieved prominence in the mature sound cinema, from Jean Renoir and Max Ophuls to William Wyler, George Cukor, and John Stahl, made long takes and staging in depth central to their aesthetic repertoire. Still, their innovations operated within the framework of analytical cutting, shot/reverse shot, and judiciously placed close-ups highlighting the main turns in the action. Japan’s changeover to sound was slowed by indecision about technology and the resistance of the powerful benshi—the commentators who accompanied film screenings by explaining the story and taking on the voices of the characters. Once sound filmmaking arrived, however, stylistic change paralleled the Western path. The calligraphic trend died out. Piecemeal découpage survived, as did the pictorialist approach, which now often favored stately, distant framings and rather long takes. And the baseline remained a more sober, quasi-Hollywood technique. This last became dominant during the Pacific war period, stylistically the most conservative years of Japanese filmmaking. After the war the most ambitious young directors, Kurosawa Akira and Ichikawa Kon, sustained the vigorously flamboyant approach to filmmaking pioneered in the 1920s. Older masters like Ozu and Shimizu continued to refine their unique styles. Soon, however, the audience became younger, the genres of juvenile delinquency and yakuza (gambler) movies arrived, and movie attendance went into a tailspin. The arrival of directors like Oshima Nagisa and Imamura Shohei in the late 1950s definitively marked the end of the golden age. Still the anarchic, abrasive power of the young men’s films and their rebellion against their fathers’ cinema should not make us forget a simple fact: on the foundation of classical Hollywood découpage the interwar generation of Japanese directors built perhaps the most subtle and variegated national film style the world has yet known.9 In this process Mizoguchi Kenji—arguably film history’s greatest exponent of staging— played a central role.

86

MIZOGUCHI, OR MODULATION

The eye wants to overcome di‹culties. One should of course set it tasks capable of solution; but the whole history of art is proof that this morning’s “clarity” is now boring and that visual art can aªord to dispense with partial obscurity, momentary visual puzzlement, as little as music can dispense with dissonances, false endings, etc.

H E I N R I C H W Ö L F F L I N 10

Born in 1898, Mizoguchi belonged to what Japanese historians call the “generation of 1900.”11 As these young people were growing up, their nation proved itself a military power, seized Korea, launched its first experiment with parliamentary democracy, and opened its doors to Western culture, both high and low. Although the Communist Party was weak, Marxism had a wide influence on intellectuals and artists. So too did a stream of translations, foreign films, and imported visual art. In the midst of this ferment Mizoguchi, the son of a Tokyo carpenter, discovered a love of art. As a boy he apprenticed to a design house, and in his teens he attended a school teaching Western-style oil painting. All the while he was reading widely and attending the theater. In Kobe he worked in a publicity firm while writing poetry, flirting with Christian socialism, and founding an amateur theater troupe. In 1921 an actor friend found him a job as a player and assistant at the Nikkatsu studio, and he began as a director in 1922. Other aspects of his life might have furnished the plot for one of his own melodramas. At age seven Mizoguchi saw his father go bankrupt and sell his sister as a geisha. When his mother died ten years later, he quarreled with his father and went to live with his sister, who by then had become an aristocrat’s mistress. Soon after Mizoguchi started directing, he began an aªair with a geisha who was also involved with a gangster. One night the woman slashed his back with a razor, an incident that won Mizoguchi notoriety and a three-month suspension from Nikkatsu. (He was fond of displaying the scar and remarking, “You cannot portray women unless you have been stabbed like me!”) Shortly afterward, he met a gangster’s wife at a dance hall and persuaded her to leave her husband. Once they were married, the couple quarreled bitterly, and from 1941 Mizoguchi’s wife was confined to a mental hospital because of syphilis.12 Despite his sensational private life, he proved himself an e‹cient and accomplished director. After the 1923 earthquake he moved to Kyoto, and there, amid the old Japan of temples and licensed pleasure quarters, he would reside the rest of his life. His earliest films appear to have explored a variety of modes: lyrical naturalism, aesthetic experiment, military propaganda, and a quasi-Marxist proletarian realism. Mizoguchi made some commercial hits, such as Tokyo March (1929), and his work began to place in the annual “Best Ten” lists compiled by the film journal Kinema Jumpo. With Taki no Shiraito (a.k.a. White Threads of the Waterfall, 1933), which placed second in the Kinema Jumpo poll, he joined the front rank of directors. One mark of his new renown was that whereas most directors worked steadily for a single company, he was able to break with Nikkatsu and migrate among companies from 1933 onward. By then he was known as a specialist in feminine melodrama. Because Nikkatsu’s fore-

MIZOGUCHI, OR MODULATION

87

most director, Murata Minoru, preferred making films centering on men, the company had forced Mizoguchi to take up women’s genres, but eventually he came to like the idea. Literary and theatrical traditions provide any Japanese artist with a storehouse of situations based on victimized women. Merchants or clerks fall in love with prostitutes or aristocratic ladies, maternal prostitutes sacrifice themselves for men’s careers, girls are sold into bullying marriages.13 All these pathetic situations, derived from kabuki and puppet theater, furnished plots for popular novels serialized in newspapers and an emerging form of theater called shinpa. Shinpa was a romanticized, quasi-modern version of kabuki that became hugely popular at the turn of the century. Many plays highlighted independent women, often in the sex trade, crushed by social obligations and unfeeling men. When Mizoguchi entered the film industry, any tear-jerking film set in modern times was considered an oªshoot of shinpa. He adapted some shinpa plays, and once he went oª on his own, he clung to the shinpa aesthetic for the rest of his career. Mizoguchi found his greatest degree of independence in working at Dai-Ichi Eiga, a company he founded in 1934 with the ambitious producer Nagata Masaichi. He made several important films for Dai-Ichi Eiga, and two 1936 projects, both written by Yoda Yoshikata, proved to be groundbreaking. “I didn’t begin portraying humanity accurately until Naniwa Elegy and Sisters of Gion.”14 In the Kinema Jumpo poll Naniwa won third place, and Sisters won first. When Dai-Ichi Eiga dissolved, Mizoguchi moved with Nagata to the Shinko company and then to Shochiku, which placed great resources at his disposal. His film about a kabuki actor in the Meiji era, The Story of the Last Chrysanthemum (1939), won a government award, and he went on to create a cluster of films celebrating traditional Japanese arts. As the nation mobilized for the Pacific war, he was assigned to direct a colossal version of the classic 47 ronin tale Genroku Chushingura (1941–1942), the most expensive Japanese film yet made. By now he was motion picture consultant to the cabinet and head of the Japanese Filmmaking League. One critic called him “the master among masters in the Japanese film world. . . . To talk about this man is at the same time to talk about the path upon which Japanese film has progressed.”15 Starting out in the 1920s had toughened him; directors often confronted lazy crews protected by yakuza gangs, and shooting might be disrupted by brawls. Mizoguchi adopted a ferocious demeanor. The “three-clawed fiend,” as Yoda called him, was capable of demanding that telegraph poles be cut down because they spoiled a composition. He insisted on lengthy rehearsals before filming, but he refused to instruct the actors; when their performances fell short, he screamed at them, with furious spasms shaking his right shoulder. “He drank a lot and had a very foul mouth on the set,” one recalled.16 Coming in each morning, actors would confront a blackboard bearing entirely new dialogue for the scene. Mizoguchi sometimes worked his staª through days and nights, keeping a chamber pot nearby so that he would never have to leave the set. Yoda’s memoirs are a catalog of caprices and indignities. On the day of Yoda’s marriage Mizoguchi badgered him about a new script.17 Yet after a drinking bout the fiend could suddenly start pounding his head on the tatami mat, shouting, “This man they call Mizoguchi is an idiot!”18

88

MIZOGUCHI, OR MODULATION

He played out political contradictions too, as did many men of his generation. During the late 1930s their youthful attraction to Western culture and Marxist politics gave way to patriotism and a rededication to “national purity.”19 Mizoguchi made several military propaganda films and put his name to ghostwritten declarations of faith in Japanese conquest. (“I believe in the film of war.”)20 After Japan surrendered and US forces occupied the country, he executed another about-face. Conforming to Occupation demands, he criticized feudal traditions and pleaded for women’s rights in a string of films that often showcased the brilliant actress Tanaka Kinuyo. He also mounted lush melodramas, usually set in upper-class surroundings. In 1953 Mizoguchi joined Nagata Masaichi at his revived studio, now called Daiei, and he alternated contemporary dramas with historical pieces. Most of his postwar films were received coldly by both the press and the public, and some critics complained about his old-fashioned style and shinpa romanticism, but he continued to occupy the spotlight. He served as adviser to the ministry of education, president of the production labor union, a member of Daiei’s board of directors, and president of the Film Directors Association from 1949 until his death. Many of his postwar projects also exemplify that “return to Japan” (nihon kaiki) that is said to be characteristic of Japanese artists who cast aside their youthful fascination with the West.21 Early in Mizoguchi’s career he adapted European and American literary works, and he professed an interest in the films of von Sternberg and Lubitsch and in the painting of Picasso, Matisse, and de Chirico. His wartime films were required to project a spiritualized image of Japanese culture, but he did not swing back to an a‹rmation of Western values when peace arrived. Instead, he intensified his interest in traditional Chinese and Japanese art. Accordingly, the Occupation-era project Utamaro and His Five Women (1946) paid tribute to the great eighteenth-century painter. Mizoguchi converted to Buddhism and began rereading the Japanese classics. Along with the faithful Yoda he adapted a play by the great Chikamatsu (Chikamatsu Monogatari; a.k.a. The Crucified Lovers, 1954), drew The Life of Oharu (1952) from a novel by the satirist Saikaku, and oªered his treatment of Japan’s greatest historical epic in Shin Heike Monogatari (a.k.a. New Tales of the Taira Clan, 1956). He adapted several other films from well-respected modern novels. Thanks largely to Nagata’s shrewd export strategies, Mizoguchi startled the film world by winning the top prize at Venice three years in a row, for the historical dramas The Life of Oharu (1952), Ugetsu Monogatari (1953), and Sansho the Bailiª (1954). By the time he died of leukemia in 1956, he was second only to Kurosawa as an emblem of Japanese cinema. The French, as so often, led the way to his canonization, with Cahiers du cinéma running reviews and translated pieces, including Yoda’s memoirs. Monographs appeared.22 In a period in which Ozu, Naruse, and others remained virtually unknown in the West, Mizoguchi constituted the only major alternative to the rugged, flashy Kurosawa. The praise ran to extremes. “Kenji Mizoguchi is to the cinema,” wrote Jean Douchet, “what Bach is to music, Cervantes is to literature, Shakespeare is to the theatre, Titian is to painting: the very greatest.”23 His languid pace and burnished imagery embodied that mystique of mise-en-scène

MIZOGUCHI, OR MODULATION

89

3.6 The Life of Oharu (1952).

3.7 The Life of Oharu.

that was central to the Cahiers aesthetic. The ecstasy of Sarris before Madam Yuki’s shimmering lake vista, unspoiled by subtitles, belongs to this tradition. Mizoguchi aimed, wrote Godard, simply “to leave things to present themselves, with the mind intervening solely to eªace its own traces.”24 Jacques Rivette noted that these films, however distant culturally, “speak to us in a very familiar language. Which one? The only one to which a director must aspire: that of mise en scène.”25 Thanks partly to the eloquence of the Cahiers writers, Mizoguchi has remained known chiefly through his richly realized costume pictures. To understand his achievement more exactly, however, we must attend to details. And then we discover a practice of staging whose visual ambitions are almost without peer in film history. For Feuillade, the image was primarily a vehicle for storytelling; pictorial invention, however important, flowed from that purpose. For Mizoguchi cinematic staging was above all a pictorial art—one that, he believed, paradoxically came into its own only with sound filming. His task was to integrate bold and intricate visual design with the demands of narration and emotional expression. A concrete case aªords us a good starting point. Early in The Life of Oharu (1952), the heroine has been sent home from court after a scandalous aªair with a low-ranking samurai. She and her family have been exiled. Before her lover, Katsunosuke, is beheaded, he sends her a message. He asks her to marry a good man for whom she feels true love, adding defiantly that he hopes for a day when no love is forbidden by social rank. On reading Katsunosuke’s letter Oharu grabs a knife and rushes outside to kill herself; her mother pursues her into the woods and disarms her, leaving both women sobbing on the ground. The pivotal portion of the scene occurs when Oharu’s mother delivers Katsunosuke’s message. The shot begins normally enough, with the two women seated (Fig. 3.6) in a fairly distant framing. Warning that the father must not learn of this, Oharu’s mother turns from us and rises to keep watch for him. Oharu slides behind a hanging kimono to read the letter (Fig. 3.7). Strikingly, the kimono conceals her face; but we do hear a sob. Suddenly Oharu scrabbles toward the lower right of the frame, visible only in a squar-

90

MIZOGUCHI, OR MODULATION

3.8 The Life of Oharu.

3.9 The Life of Oharu.

ish gap left by another kimono. As the mother turns and lunges behind the kimonos, a thrashing movement can be sensed in the gap, broken by an intermittent glint (Fig. 3.8). The central kimono is yanked down as the women struggle (Fig. 3.9). Oharu rushes out of the frame, and now a knife blade quickly flashes past us (Fig. 3.10). Her mother runs after her. Outside, a high-angle camera movement catches Oharu racing out of the house (Fig. 3.11). The camera follows her as her mother wrests the knife from her (Fig. 3.12). The two run diagonally into the depth of a bamboo grove (Fig. 3.13) before collapsing in extreme long shot, sobbing (Fig. 3.14). It is a commonplace of Mizoguchi criticism that he favors deep space and long takes (here the first shot runs about forty seconds, the second about seventy). But it is his staging principles that give the scene its distinct quality. For a start, Oharu’s distraught reaction to the message is not presented via close views of the heroine, music underscoring her tearful face, or an emphasis on the knife she snatches up. Instead the camera stays at a considerable distance and lets key actions remain almost entirely hidden. Mizoguchi has also divided our attention almost perversely. The vigilant mother and the prominent doorway stir some expectation that the father may discover them (Fig. 3.7), but he doesn’t. Instead, the drama develops through Oharu’s reaction, which is concealed from us by the hanging kimono. Her grief is signaled on the sound track, which yields a sob and then the frantic rustle of movement. Her pain becomes resolutely private, as if the narration were reluctant to intrude on the purity of her relation to her lover. The knife itself is scarcely visible, squeezed into a sector of the frame we normally ignore (Fig. 3.8, on extreme lower right). The mother’s cries of distress oªer some clues as to Oharu’s overwrought state, but the daughter’s intentions remain uncertain until the knife flashes through the lower right foreground (Fig. 3.10), just long enough for us to see it. As Mizoguchi had confined the knife to an aperture, so now he lets us glimpse Oharu fleeing the house through a tunnel of bamboo trees (Fig. 3.11). He now reverses his scaling strategy, bringing Oharu to the foreground as she gives up the knife to her mother

MIZOGUCHI, OR MODULATION

91

3.10 The Life of Oharu.

3.11 The Life of Oharu.

3.12 The Life of Oharu.

(Fig. 3.12). Then Oharu shrieks, “I’ll go to Katsunosuke!” announcing that she’ll find another way to join him in death and running oª to the distance. The mother pursues her in a zigzag course and grapples with her in extreme long shot before both drop to the ground (Fig. 3.13). The urgency of the first shot tapers into something more lyrical as the women plunge into the grove. Once more Mizoguchi refuses us proximity and facial expression. It is in the writhing and collapsing of Oharu’s body that we grasp her passage from wild abandon to resignation (Fig. 3.14). Her desolation is made all the more poignant by the stillness of the grove (crows call on the sound track) and the sedateness of the camera movement. Mizoguchi’s oblique narration—suppressing Oharu’s response, diverting us with the pretext that the father might catch her—creates a pictorial field that he can charge with emotion. Starting as a straightforward presentation of the action, the first shot transmogrifies into an exploration of the design possibilities of the two hanging kimonos. They begin as a realistic backdrop; then one becomes a convenient screen for masking Oharu’s reaction; then the other yields a tantalizing aperture within which she seizes the knife.

92

MIZOGUCHI, OR MODULATION

3.13 The Life of Oharu.

3.14 The Life of Oharu.

Before our eyes, several zones of space—the doorway, the two kimonos—bristle with the promise of action and expression. The picture, richly composed without looking stuªed, bursts into violence thanks to the lures and traps built into it. In the bamboo grove the placid landscape absorbs Oharu’s despair, leaving her with only acceptance; but it also continues the game with visibility by letting her and her mother veer from the camera and dodge through the bamboo stems. Mizoguchi has brought the denotative, expressive, and decorative aspects of the image into remarkable harmony. We could find scores of such dazzling scenes in this director’s oeuvre, but broader questions nag us. We are very far from the trim, lucid world of Feuillade. What historical circumstances led Mizoguchi to pose such staging problems and solve them so inventively? How did the problems and solutions change, and why? Any answers to these questions must be tentative, chiefly because we cannot chart his overall development with much confidence. Mizoguchi started directing in 1923, and by the time he started his first talkie, Hometown (1930), he had made no fewer than fortythree films. All of them are lost save one (Song of Home, 1925) and one condensation (Tokyo March, 1929). We will probably never see his Arsène Lupin vehicle (813, 1923), his adaptation of Anna Christie (Foggy Harbor, 1923), his Expressionist film (Blood and Soul, 1923), his first critically praised melodrama (A Paper Doll’s Whisper of Spring, 1926), and his version of a best-selling novel about Townsend Harris’s mission to Japan (Okichi, Mistress of a Foreigner, 1930). Several works from the next decade have gone missing as well. Even his years of fame are riddled with lost films, at least two of which—Woman of Osaka (1940) and Life of an Actor (1941)—are likely to have been masterpieces. In all, of the eighty-three feature films he completed, only thirty-one survive more or less intact,26 and seventeen of these come from the last decade of his career. We rightly regret the loss of such treasures as von Sternberg’s Sea Gull (1926) and Murnau’s Four Devils (1928), but faced with deprivation on the Mizoguchian scale, one is tempted to despair. Still, the surviving films are unambiguous on one score: the man liked long takes. He

MIZOGUCHI, OR MODULATION

93

named his method “one scene–one cut.” (Like US filmmakers of the same period, Japanese often called a shot a cut, as in a cut of meat.) From 1935 to the end of his career, the average shot lengths of Mizoguchi’s films range from fifteen seconds to ninety seconds, with most ranging between twenty-five and forty seconds. Although his 1930s contemporaries favored somewhat long takes (an average of ten to twelve seconds for a talkie),27 his shots are unusually extended for any period or place. Feuillade chose the long take as the default option, simply the received way of shooting scenes. Mizoguchi, starting his career at the moment in which continuity cutting triumphed internationally, clung to the long take against the pull of current practice. His stubborn adherence to it added to his production di‹culties. In sound filming, with a much heavier camera and all the vagaries of recording dialogue, very protracted shots were more demanding than in the silent era. Moreover, Japanese studios demanded that film stock be conserved, so most directors could shoot few retakes. Mizoguchi’s collaborators recalled that an almost suªocating anxiety gripped everyone as the camera was turned on. The slightest mistake would spoil the shot; time and film would be wasted; worst of all, cast and crew would face the demon’s wrath. “He asked the most of everyone, and all at the same second,” recalled his cameraman Miyagawa Kazuo. “If there was a characteristic quality of Mizoguchi, it was this tension and its prolongation.”28 When producers complained that the long takes delayed the shooting schedule, Mizoguchi replied that the actors weren’t good enough. Why cling to the long take in the face of such obstacles? We might be tempted to treat the device as an end in itself: the longer the take, the more daring (or “radical”) the shot. Yet our Oharu scene prompts us to consider the long take as a by-product of other concerns. The most salient one, it seems to me, is also one of the striking features of Japanese cinema of the late 1920s and the 1930s: an extraordinary passion for densely composed images. Consider a moment from Naruse Mikio’s The Whole Family Works (1939). A factory worker has fathered nine children, and several of his sons want to continue in school. One morning he talks with his eldest son about whether to consult the teacher of Eisaku, his fourth son. By the standards of European and US filmmaking, Fig. 3.15 is an overwrought establishing shot. The architecture splits the image into several pigeonholes, and a foreground figure (Eisaku, with his back to us) serves to further articulate lateral zones of action. Lighting helps by establishing a gradation of emphasis. The two brothers caught in the lattice of the left wall, being somewhat grayer, become less important than the father and eldest son conferring in the better-lit distance on the right. Eisaku, the object of the decision, is given prominence through his silhouette. The shot presents story information by establishing several brothers watching the father’s deliberations and by highlighting Eisaku (centered, largest in the frame). But Naruse adds a decorative surplus. Always adept in composition, he sets himself the challenge of arranging seven figures in three areas of the room, achieving both narrative clarity and an elaborate visual pattern.

94

MIZOGUCHI, OR MODULATION

3.15 The Whole Family Works (1939). In the distance right of center, the father considers letting his son continue school.

Mizoguchi, trained as a painter, was probably quite sensitive to such compositions, and he seems to have realized that his peers didn’t grasp all their implications. “Filmmakers must study the film image and its potential for expression,” he once remarked. “This is our primary responsibility.”29 He surely noticed that his contemporaries did not sustain their ornate images for very long, being content to enlist them as establishing shots or momentary flourishes. (Naruse uses nineteen shots in the scene we’ve considered, following the dense master shot in Fig. 3.15 with singles and two-shots linked by eyeline matches.) By the early 1930s Mizoguchi saw that he could sustain the dense image for some time if he could make it unfold in compelling ways. Citing a psychology professor who claimed that film images became boring after five seconds, he began to pursue a “one scene–one cut” method that would absorb the audience for minutes on end.30 How to accomplish this? On the evidence of his surviving films, his chief strategy was to oªer a richly modulated image. His peers had developed schemas of elaborate shot design; he would take such schemas through unpredictable twists and turns. Other directors had made each character’s placement matter visually; he would hold the shot so that a slight shift in position would gather great power. They created nuanced compositions; he would elaborate such nuances for minutes on end. In sum, by means of the long take he could prolong and probe the pictorially arresting norms that emerged in this period. Such dense long takes seemed, in Mizoguchi’s judgment, especially appropriate to talking pictures.31 Whereas fast cutting suited silent film, dramas of his sort, especially in the sound era, favored a long-take style. Editing breaks up “the psychological weight or density that the audience experiences. . . . If you use cross-cuts, there are invariably a few cuts [shots] that shouldn’t be included. And it’s a big mistake to rationalize this by saying they’re short. The hypnotic power has been impaired.”32 More paradoxically: “Cinema with short cuts is too cinematic.”33 His reference to the “hypnotic power” of cinema indicates yet another goal that the long take fulfilled. Mizoguchi evidently wanted to hold

MIZOGUCHI, OR MODULATION

95

the spectator in the same sort of rapt apprehension that enfolded his staª on the sound stage. Late in his career he explained: “During the course of filming a scene, if an increasing psychological sympathy begins to develop, I cannot cut into this without regret. I try rather to intensify and prolong the scene as long as possible.”34 Certainly this rapt attention involves feeling for or with the characters, but the design of the shots suggests that even more is at stake. The image invites the spectator to explore every cranny of the space, to scrutinize even small movements. The long take, Mizoguchi added, “allows me to work all the spectator’s perceptual capacities to the utmost.”35 Mizoguchi seeks to entrance the viewer on several registers at once, yet his images oªer themselves almost haughtily, indiªerent to the viewer’s inclinations. The staging narrates, but it often unfolds the story in an opaque or ambiguous fashion. It oªers decorative resonance but with a subtlety that demands that the viewer explore the frame. And the staging becomes expressive but mutedly so. As the Oharu scene suggests, the interplay of visual pattern, narrational display, and expressive qualities can become quite intricate. When Feuillade blocks something and then reveals it, he usually does so unobtrusively, just before we need to see it: an art of clarity. Mizoguchi blocks and only partly reveals, in a teasing visual tremolo; or he creates an opacity that registers as a refusal to specify the story point; or he presents the very intermittence of vision as itself a refined eªect worth savoring. In his narrational games of vision Mizoguchi often relies on characters’ dialogue and exclamations to express the scene’s feelings, as with Oharu’s unseen sob. In addition, he seeks to make his pictures expressive in themselves. Herein lies one of his great contributions to the history of cinematic staging. By the 1920s the problem of telling a story concisely and suspensefully had been solved, both by the European long-take directors and by their Hollywood peers. Directors faced new tasks, notably that of telling a story with maximal expressiveness, using all available film techniques. Some directors, notably Sjöström, had found a way to heighten emotion within the tableau tradition; others, such as Gance, Epstein, and the Soviet montage directors, achieved emotional expression by pushing editing to extremes.36 By the mid-1930s filmmakers around the world knew how to quicken the emotional pace through rapid cutting, presenting a train rushing through a landscape, a ride on a merry-go-round, a suspenseful exchange of glances. Hitchcock’s late silent films are typical of the ways in which the montage of close-ups had become a standardized visual rhetoric. Synchronized sound, for reasons of technology as much as taste, made cinema less overtly “avant-garde,” as André Bazin was among the first to recognize.37 Classical découpage, accentuated by camera movements and taken at a slower pace, remained the norm from Los Angeles and Berlin to Moscow and Shanghai. Cutting and camera angle usually served to present the action neutrally; the expressive dimension of the story would be rendered through the script, the music, and the mise-en-scène—above all, the actors’ performances. Through his pictorialism, Mizoguchi sought to make staging more subtly expressive than most directors of the 1930s did. The plot traditions that he inherited

96

MIZOGUCHI, OR MODULATION

emphasized the slow accumulation of emotion, and in order to mesmerize the spectator, his scenes of suªering grew into long takes that mapped the fluctuations of feeling onto the movement and positions of bodies in a thickly furnished space. The obliqueness of his visual narration could foster emotional suggestion. His genre proved helpful too. Because melodrama tends to endow the spectator with an omniscient range of knowledge, we are able to anticipate a character’s reaction to news we already know. Mizoguchi can play on this foreknowledge in two ways. Sometimes, his roundabout presentation of a situation can be taken both as tact and as a creative treatment of a stock situation. We have already witnessed the final moments of Oharu’s lover, so we expect her to feel grief when she gets his message. Our expectation is confirmed even though we don’t exactly see it (the mother distracts us, the kimono hides Oharu), but this in turn makes the image register exactly as a sensitive handling of the moment. Alternatively, Mizoguchi can transmit the story situation with little impedance, as we will see in The Story of the Last Chrysanthemum and Utamaro and His Five Women. Such occasions allow him to enrich the conventional moment through microactions and barely noticeable diªerences. A lover’s farewell or a woman’s confession, presented straightforwardly, can play on minute changes of position, glance, shadow. We can take the basic action for granted and scrutinize pictorial nuances that other directors never summon up. “An art of modulation,” Jacques Rivette calls it, and Philippe Demonsablon adds a gloss: “He emits a note so pure that the slightest variation becomes expressive.”38 The long take, then, is not an end in itself but a means of riveting the spectator to the exquisite image. The result, in Mizoguchi’s hands, yields something Westerners have long considered “purely Japanese.” When we point out kindred examples, such as the shot from The Whole Family Works, some may want to say that this, too, manifests a typically Japanese use of space. In what follows I’ll be avoiding this inference—or, rather, reformulating it. I’m skeptical of a view that would treat such images as late manifestations of a panhistorical Japanese taste or worldview. First, there is no single Japanese aesthetic tradition to be transmitted. Western accounts of Japanese aesthetic concepts are notably onesided, favoring Zen-inflected art, but this is part of the “Japan” packaged for foreign consumption over the last century. Some Japanese traditions indeed emphasize delicacy, spirituality, and understatement, but there are also traditions of roughness, bawdiness, and extravagant emotion.39 (Contrary to clichés about Japanese aesthetic restraint, film and theater audiences love netsuen, scenery-chewing overacting.) Furthermore, we have good evidence that the arrival of Western art and media in the nineteenth century made Japanese artists acutely aware of their own methods. Now they had a new set of choices. Artists might embrace Western forms (as in shingeki, the Europeanized modern theater) or blend inherited forms with imported ones (as in shinpa theater or the “I-novel”). Even sticking with a received tradition could trigger a search for novelty; by 1900 painters worked in either the “Western-style” (yoga) or the “Japanese-style” (nihonga), but the latter could never be confused with the works of centuries past.40 There are other reasons to be cautious about explaining Japanese artistic practices as

MIZOGUCHI, OR MODULATION

97

direct continuations of time-honored customs. The artistic concepts purportedly shaping all Japanese art (wabi, yugen, iki, mono no aware) often turn out to have complex and ambivalent histories, during which they were redefined for various purposes.41 More generally, a great many “distinctively Japanese” traditions, from emperor worship to the rules of sumo, were devised in the nineteenth and twentieth centuries by elite factions forging new national identities for a modernizing society.42 We are not in the habit of explaining contemporary Hollywood style by reference to northern European Renaissance painting, so why should ancient aesthetic traditions be relevant to twentieth-century Japanese film? Someone may respond that whereas we have lost touch with premodern customs and ways of thinking, the Japanese have retained a living relation to theirs. Yet this idea itself is no less an invented tradition, with sources in twentieth-century Japanese ethnology and cultural theory.43 It seems likely that the notion of a “purely Japanese” cinematic style is another invented tradition. During the 1910s the trade press called for filmmakers not only to match foreign technique but also to express “true Japanese” qualities. Writer Tanizaki Jun’ichiro observed that by making Hollywood-like films he hoped to introduce his nation’s “art and national sentiment” to audiences abroad.44 This nationalistic view emerged in other countries at the period, as US films oªered both models to copy and rivals to surpass.45 As a result, from the 1920s on Japanese filmmakers often borrowed from, and even cited, pictorial devices found in older arts—all the while inserting these elements into the Western continuity framework. The schemas devised and reworked by filmmakers, harking back to older ones in the graphic arts, are best seen as tactical choices, self-conscious displays of pictorial intelligence. In studying Japanese cinema, it seems most fruitful to treat the flamboyant visual devices we notice as strategic means for achieving specific ends— one of which may indeed be the evocation of “Japaneseness” as a distinct national/cultural essence. Mizoguchi, with his mastery of Western staging and editing, his eye for schemas circulating in his milieu, his acute awareness of rivals abroad, and his cultivated taste in Asian art, exemplifies this process with remarkable clarity. Although most of his output is missing, the films that we have oªer abundant evidence of Mizoguchi’s contributions to the art of cinematic staging.46 All manner of visual ideas poured from him from the 1930s to the 1950s. Almost every one of his works warrants in-depth analysis, and I regret not being able to explore several of my favorites. To keep things manageable, this chapter tracks his work across four periods. Roughly, we can locate a phase of eclecticism before 1936, an eªort toward purity and austerity in his major films of 1936 to 1945, a controlled pluralism between 1945 and 1949, and a phase of refinement running from 1950 to 1956. Someone asking other questions could carve up Mizoguchi’s career diªerently, but these periods enable me to highlight some important continuities and changes in his staging. To focus things further, I select two or three films from each period as signposts of distinctive and developing strategies. Some films are better known than others, but each illustrates one problem or point in his career profile.

98

MIZOGUCHI, OR MODULATION

This chapter will have served its purpose if its roaming-spotlight survey illuminates the art of a director who set his viewers, and himself, unique challenges.

A young geisha is courted by two o‹ce workers and an older patron; she is inclining toward one of the young men when she learns that he is her half-brother (Tokyo March, 1929). An opera singer is sustained by the love of a devoted young woman, but when he becomes a success, he is drawn to a conniving flapper (Hometown, 1930). A female carnival performer supports a young man through his legal studies, while she slips into criminality; when she must stand trial for murder, her case will be prosecuted by her protégé (Taki no Shiraito, 1933). As rebels terrorize the countryside, a general holds a coachful of travelers hostage, and several women among them are oªered to placate him (Oyuki the Virgin, 1935). A woman working with a gang of art swindlers escapes with an orphaned servant boy, and she prostitutes herself to send him to medical school. Years later, waiting for a night train, he discovers her on the platform, a haggard madwoman (The Downfall of Osen, 1935). With plots centered on lowly heroines who sacrifice and suªer for a man, Mizoguchi’s pre-1936 films place themselves on the melodramatic terrain of popular literature and theater. The situations evoke shinpa, and it’s no surprise that the two most famous films of the group, Taki no Shiraito and The Downfall of Osen, derive from shinpa plays based on novels by Izumi Kyoka. Taki is particularly important; frequently adapted for the screen before and after Mizoguchi’s version, it was the prototypical shinpa play.47 Not every Mizoguchi film that we have from this period relies so heavily on these conventions, but all are melodramatic to a high degree.48 Even The Poppy (1935), adapted from a prestigious novel, includes scenes that permit shinpa-like emotional reversals. Stylistically, these films are quite varied. Between the mid-1920s and mid-1930s Japanese cinema was diversifying its expressive means, and fresh influences and technologies crowded in from abroad. Mizoguchi was exploring a range of techniques and story materials, so we shouldn’t be surprised that his films include flamboyant passages of rapid cutting, which look European or even Soviet in inspiration but which were by the late 1920s common in Japanese film as well. A tennis game in Tokyo March is rendered in swift alternating cuts, and a scene showing Taki no Shiraito flinging herself oª a train presents thirty-three shots in forty seconds. As late as The Poppy, when a man is accompanying a woman shopping and meets the other woman he’s romancing, Mizoguchi renders the confrontation in static facial close-ups, trimmed to as little as eleven frames each. Mizoguchi would soon leave such assertive montage to others, but he continued to rely on the analytical cuts and shot/reverse-shot matching he had mastered while young. His earliest surviving film, The Song of Home (1925), consists of well-executed passages of continuity editing on the American model (Figs. 3.16–3.18). Even when he works with much longer takes, the catchphrase “one scene–one shot” will prove misleading. Mizo-

MIZOGUCHI, OR MODULATION

99

3.16 The Song of Home (1925): Naotaro approaches his father in an establishing shot.

3.17 The Song of Home: reverse-angle cut on the father hearing his plea.

3.18 The Song of Home: a reverse angle on Naotaro, with his mother out of focus in the background.

3.19 Taki no Shiraito (1933): a variety of angles captures the heroine facing her judges.

3.20 Taki no Shiraito.

100

MIZOGUCHI, OR MODULATION

3.21 A depth composition showing the young hero’s torment in the gangster’s household (The Downfall of Osen, 1935).

3.22 A similar “big-foreground” composition from Tanaka Shigeo’s Three Beauties (1933).

guchi almost never handles a scene in a single shot, so editing must orchestrate the flow of his extended takes.49 Like Ikeda in our scene from Cuckoo, early Mizoguchi inclines toward constantly changing camera setups. The process is visible in the scene from Song of Home: few of the shots of Naotaro or his family repeat a camera position. This prodigality is typical of the whole film; the climax, in which Naotaro is rewarded with a chance to go to the city and study, consists essentially of alternating shots of him and other characters, but the twelve images present seven diªerent setups. Similarly, the rapidly cut courtroom scene of Taki no Shiraito (1933) yields eighty-five images and forty-five setups. Mizoguchi’s exaggerated depth compositions underscore the diªerences between camera positions (Figs. 3.19– 3.20). By stringing together such striking pictures, he oªers not only a readout of the drama but also a virtuosic suite of variations. In the early 1930s he seems so eager to vary shot scale and angle that he is willing to violate precepts of eyeline matching (most obviously in The Poppy). From 1930 on, shot length increases. Often, as in the late Feuillade films, a long take is broken by several intertitles. Of the 450 or so shots in The Downfall of Osen (not counting intertitles as shots), about 370 stem from only one hundred setups, so if we clipped out the dialogue titles, several shots would run between one and two minutes.50 The talkie Hometown averages 9.9 seconds per shot, partly because of its headlong and bumpy tracking shots and a bravura crane shot zigzagging down a banquet table. Most of Mizoguchi’s silent films contain at least one unbroken take, as when in Osen several complex pansand-tracks let the action flow around the camera. On the whole, though, most Mizoguchi films before 1936 contain little camera movement, for now staging in depth is becoming more prominent. Already, in the shots from Song of Home above, we see an eªort to keep background planes active in the ongoing drama (for example, Fig. 3.18). Mizoguchi seems to have used wide-angle lenses fairly early, occasionally yielding the sort of “big-foreground” schemas that we find elsewhere

MIZOGUCHI, OR MODULATION

101

3.23 Oyuki the Virgin (1935).

3.24 Oyuki the Virgin.

during the period (Figs. 3.21–3.22). Although he seldom combines staging in depth with long takes at this phase, when he does, he can create dramatic pictorial eªects of suspense and surprise, as a scene from Oyuki the Virgin (1935) nicely illustrates. Coach passengers are taken hostage by an intransigent general, and the barmaid Okin oªers herself to him in hopes he will free the group. Mizoguchi stages her eªorts to seduce him in a series of wide shots displaying constantly changing depth relations. Once Okin is shown into General Asakura’s quarters, a ninety-second shot, taken from the high angle that Mizoguchi will come to favor, allows Okin’s sensuality full play. She brazenly circles his desk, drinks from his glass, and eventually settles in a chair, pouring another drink as he tries to ignore her (Fig. 3.23). This is the beginning of an erotic exchange that Mizoguchi’s staging will amplify. A reverse-angle cut presents a new area of space, a room beyond the study (Fig. 3.24), and Okin drifts into it, passing oªscreen (Fig. 3.25). Cut to a brief shot of her perched temptingly on the bed (Fig. 3.26), which serves as a pivot to a new depth image, a marked foreground composition putting Asakura close to the camera (Fig. 3.27). In the distance Okin stretches out on the bed and lights a pipe. A 180-degree cut reverses the depth arrangement, putting her in the foreground as he comes to the threshold (Fig. 3.28). Will he succumb to her? No; he closes the shoji (Fig. 3.29). She plops a cavalryman’s cap over her face and lies still for several seconds. The suspense would seem to be over. But then murmurs are heard, and she rises. When Okin slides open the door, over her shoulder we can see another passenger, the daughter of a rice merchant, standing sobbing before the general (Fig. 3.30). In the course of this fifty-three second shot the depth relations have squeezed from an ample foreground and full background to a distant dramatic zone wedged between the shoji and Okin’s shoulder. Overall, the narration has restricted our knowledge to Okin’s tour of the quarters, and the revelation of the new victim is made all the more pointed for obliging us to discern her in a thin slice of space. This sequence from Oyuki the Virgin also exemplifies Mizoguchi’s creative recasting

102

MIZOGUCHI, OR MODULATION

3.25 Oyuki the Virgin.

3.26 Oyuki the Virgin.

3.27 Oyuki the Virgin.

3.28 Oyuki the Virgin.

3.29 Oyuki the Virgin.

3.30 Oyuki the Virgin.

MIZOGUCHI, OR MODULATION

103

3.31 Utamaro Kitagawa, Monkey Performing in a Wealthy Household, from the series Young Ebisu, 1789. Color woodcut. Elvehjem Museum of Art, University of Wisconsin–Madison. Bequest of John H. Van Vleck. 1980.3211.

of some schemas circulating in his milieu. The slot that allows us to glimpse the weeping girl aptly illustrates Japanese directors’ proclivity for aperture framing. Tantalizing partial views were a convention of graphic art, encouraged by features of indigenous architecture, such as gridded rice-paper walls and various sizes of screens (Fig. 3.31). Mizoguchi’s contemporaries exploited this tradition (Fig. 3.32), and it’s likely that the popularity of Josef von Sternberg among Japanese filmmakers lay partly in the way his shots sacrifice narrative disclosure to such pictorial ornamentation (Fig. 3.33). Mizoguchi admired Sternberg’s films and met with him during Sternberg’s 1936 visit to Kyoto.51 Mizoguchi would employ aperture framing throughout his career. In this early period he already goes somewhat beyond the one-oª flourishes of his contemporaries, perhaps because the aperture tactic lent itself to that perceptual absorption he sought. The device allowed the director to activate any crevice of the frame and rewarded the spectator’s scanning of a dense field, while also providing expressive echoes of the main action (Fig. 3.34). In Hometown aperture framing in taxicab windows becomes a motif (Figs. 3.35–3.36). For the Oyuki scene in the general’s quarters, he at once makes the aperture daringly slender and reserves it for a climactic point in the action. Here the game of vision both mutes the melodramatic revelation and gives it subtle pictorial salience. Another move in the game of vision proposed by both von Sternberg and Japanese filmmakers is a play with intermittent visibility—letting characters pass through shadowmottled areas; or momentarily blocking the action with foreground lamps, doorways, or other items; or just letting people turn from us briefly to arouse a desire to see them again

104

MIZOGUCHI, OR MODULATION

3.32 A straightforward aperture framing in Town of Love (1928). Compare Fig. 3.4.

3.33 Shanghai Express (1932).

3.34 In The Poppy (1935), while Fujio’s mother and Munechika’s father discuss finding a groom for Fujio, the daughter of the household eavesdrops, her silhouette visible in a square on the far right.

3.35 When the opera singer of Hometown achieves success, an aperture framing shows the predatory Ayako eyeing the hero.

3.36 After he has fallen under Ayako’s spell, his faithful love, Omura, waits in vain for him to meet her.

MIZOGUCHI, OR MODULATION

105

(Figs. 3.37–3.38). Mizoguchi turned this schema into something quite idiosyncratic, pushing depth staging beyond denotative functions and toward obliquely expressive ends. Often a scene’s crucial narrative action will have concluded, but Mizoguchi continues to play out remarkably indirect staging patterns. His shinpa-flavored scenes exude shame and sorrow, which drive characters to hide their faces and slump in a heap. An early example occurs at the climax of Tokyo March, which forces three characters to collapse or turn from us, all within a setting that further conceals one of them (Fig. 3.39). This array of poses is held fairly briefly, but Mizoguchi soon began to prolong and even flaunt various ways of impeding our vision for expressive ends. One climactic scene of Taki no Shiraito makes blocked vision a source of pictorial suspense. Shiraito confronts the young man whom she has financed through law school; now he is her prosecutor. At first she faces him (Fig. 3.40); but eventually shame overcomes her, and she kneels, turning away from him (Fig. 3.41). Then, astonishingly, Mizoguchi has Shiraito bend down behind the table until only the very top of her head is visible (Fig. 3.42). The table edge acts as a benchmark measuring her act of shrinking, making her shame palpable. A less flagrant but more sustained version of this approach dominates a tearful scene in The Poppy (1935). Back in the village the young tutor Ono had pledged his love to Sayoko, but now in Tokyo he has fallen under the spell of Fujio, and his friend Asai has come to tell the bad news to Sayoko and her father. The entire scene is played out with characters reacting in anguish by doubling over. The scene’s final shot carries the principle to a climax by setting the camera at a distance and letting the father’s movements in depth create ever more minute ways of concealing his and his daughter’s misery (Figs. 3.43–3.46). By hiding the faces—the most emotionally informative part of the body—the staging forces us to find heartsickness written in nuances of posture and gesture. In most films by other directors expressive qualities of style heighten denotation, the depiction of salient story information through speech and facial expression. But Mizoguchi found ways to play down these cues in favor of other indices of emotion and a more teasing treatment of them. Perhaps he was drawn to stories of “fallen” women partly because the ready-made sentiment brought along a behavioral repertoire. The genre demanded that betrayal, humiliation, and loss yield romantic suªering, which could in turn be displayed in postures. When devastated, people who live on the floor will crumple and curl up like snails. In a culture of shame, people betrayed or humiliated will turn from the camera. They will retreat behind sliding doors and sink behind tables and chairs. More and more Mizoguchi built his mise-en-scène up to moments in which characters come together in misery but struggle to conceal their facial expressions, only to have their bodies suggest their feelings. A staging that both stifles and prolongs signs of acute sadness is only one tendency in the earliest surviving films, but it becomes a dominant strategy in those of 1936 onward. Mizoguchi’s breakthrough films of 1936 were promoted and received as works of harsh realism. Naniwa Elegy (known in English as Osaka Elegy), advertised as a “meticulous analysis of the atmosphere of Naniwa,” highlighted local color. It showcased the distinctive

106

MIZOGUCHI, OR MODULATION

3.37 Shanghai Express: characters’ faces slide unpredictably across rectangles of space, intermittently blocked and revealed.

3.38 Intermittent vision in 100,000 Ryo Mystery (1939): bold verticals slice oª a foreground body and segregate a distant figure. Compare the screens splitting the bodies of the watching women in Fig. 3.31.

3.39 The old rake Fujimoto, having told Orie she is in love with her half-brother, must now tell his son, who confronts both (Tokyo March, 1929).

3.40 Taki no Shiraito: stages of shame.

3.41 Taki no Shiraito.

3.42 Taki no Shiraito.

MIZOGUCHI, OR MODULATION

107

3.43 The Poppy: the father sees Asai out, both men’s reactions concealed by distance and position in the frame.

3.44 He hurries back to his room, where he kneels; both he and Sayoko are slumped and turned from us.

3.45 As he begins to cough, she rises and comes to him, covering him with a blanket.

3.46 They embrace, half-hidden, vowing to return to Kyoto.

Osaka dialect (identified with the crassness of local merchant culture), and it included a scene at the city’s famous puppet theater. Sisters of Gion is set in “low Gion,” the more inelegant pleasure quarters of Kyoto. “I brought to these films the things I had observed in the streets and showed them as they were—the vulgar as the vulgar.”52 If Mizoguchi and Yoda intended to shock, they succeeded. They had to negotiate with the police over Naniwa Elegy, and when it was finished, the distributor gave it little publicity; in 1940 it was banned. As for Sisters of Gion, which included prostitutes in its cast, Kyoto’s high-level geisha protested the film in the newspapers, and government censors declared it “decadent.”53 Although received well by the critics, the two films probably hastened the collapse of DaiIchi Eiga. Their bluntness was not wholly unknown in the current milieu. The “neoromanticism” of shinpa-flavored literature had lost ground to novels of “neonaturalism,” influenced by Germany’s Neue Sachlichkeit movement. “Demi-monde fiction” (karyu

108

MIZOGUCHI, OR MODULATION

shosetsu) and “love-talk literature” ( jowa bungaku) centered on waitresses, prostitutes, and geisha struggling for survival in a grim urban milieu. Unlike the innocent and passive heroines of shinpa, these cynical women fight back. Cinema had also featured such protagonists, particularly in “tendency films” like What Made Her Do It? (1930), the tale of an oppressed woman who rebels by setting fire to a chapel. American imports like Bondage (1933) were flaunting unrepentant gold diggers. Films, popular literature, and high literature were all explicitly criticizing woman’s place in society. “Under the kind of social conditions that exist at present,” writes the protagonist of the novel The Waitress (1931), “such a thing as a woman’s helplessness can be understood only by women. Only while she is the object of a man’s pleasure does the latter try to please her with a curious grinning look.”54 The authorities began to object to this bitter view of modern Japan. In 1939 an o‹cial history of current literature classified the neonaturalist trend as a branch of yuto bungaku—pornographic writing.55 Mizoguchi and Yoda’s hard-boiled heroines sauntered into cinema just as right-wing politicians were deploring the decline of traditional values. Naniwa Elegy centers on Ayako, a telephone operator at a pharmaceutical company. While the company owner Asai tries to seduce her, his wife has her eye on Ayako’s boyfriend, Nishimura. Ayako needs money to pay her shiftless father’s debts, so she succumbs to Asai, who installs her in a modern apartment. Asai’s wife finds out, and he leaves Ayako. Now working as a geisha, Ayako beguiles money out of Asai’s executive, Fujino, to help pay her brother’s college tuition. She invites Nishimura to her apartment to confess her transgressions, hoping that he will marry her anyway, but Fujino storms in to demand his money. Ayako is arrested. Nishimura abandons her, and the police release her to her father’s custody. Although she has been supporting the family, they disown her, and she indignantly stalks out to an uncertain life on the street. Sisters of Gion presents a double plot: a gold-digging moga brazenly uses men, while her geisha sister clings loyally to her bankrupt patron. Naniwa Elegy combines both sisters’ impulses in its protagonist. Like a shinpa heroine Ayako has proven herself a loyal daughter, and she longs for a respectable married life. But like her sisters in neonaturalist literature, once she is cast out, she spares no tears for her kin, her lovers, or herself. In the last shot (Fig. 3.47) she strides toward the camera, almost smiling, determined to survive. Stylistically, Naniwa Elegy is a mixture. Like the early 1930s works, it employs a good deal of cutting, but by now Mizoguchi has given up the orthodox over-the-shoulder shot/ reverse shot, and the results are sometimes choppy and disorienting. It is as if, having overlearned the norms of continuity, Mizoguchi felt confident enough to assemble a conversation out of awkwardly matched and oddly decentered images. Often, though, these images include dense compositions with bold foregrounds. (See box.) Ayako’s final meal with her family consists of nineteen shots, each from a diªerent setup, thrusting a new face under our nose (Figs. 3.48–3.49). The film also contains neatly rhyming crane shots, paralleling Nishimura’s visit to Ayako’s apartment with Asai’s. Mizoguchi’s eclectic style

MIZOGUCHI, OR MODULATION

109

3.47 Ayako stalks toward the camera in the last shot of Naniwa Elegy (1936).

3.48 The dinner-table conundrum solved in Naniwa Elegy: the family are gathered to eat, with Ayako’s brother on the right upbraiding her.

3.49 Another angle puts the father in the foreground. The variety of angles recalls the courtroom scene in Taki no Shiraito (Figs. 3.19–3.20).

has led at least one critic to judge it artistically regressive.56 Yet Naniwa Elegy pushes further the experiments with distanced framing, opacity, and muted emotion that emerge sporadically in the earlier films, letting these techniques shape both denotation and expression more fully. Japanese directors’ game with visibility included not only aperture framing and intermittent revelation. More common than big foregrounds was an inclination toward curiously empty ones. In many Japanese shots significant planes of depth begin in the middle ground and stretch into the far distance, as already seen in our shots from The Abe Clan, The Whole Family Works, and Iwami Jutaro (Figs. 3.5, 3.15). (Such shots have the advantage of putting more planes in focus; it is easier to keep focus from, say, twelve feet to infinity than from four feet to twelve feet.)57 Mizoguchi employed this distant-depth schema in his earliest surviving films (Figs. 3.30, 3.39, 3.45), and it is exploited in the first

110

MIZOGUCHI, OR MODULATION

AGGRESSIVE FOREGROUNDS AND RETICENT DEPTHS During the 1920s, filmmakers around the world tried out deep-space compositions, and experimentation continued once they regained flexible shooting options in the sound era (Fig. 3.50). Bazin credited chiefly Orson Welles and William Wyler with pioneering “deepfocus,” whereby significant dramatic material was placed fairly close to the camera while other elements were situated in zigzagging depth, and all planes were in acceptable focus.58 Certainly Citizen Kane (1941), The Little Foxes (1941), and other films helped popularize this compositional device (Fig. 3.51). Bazin did not know, however, that other traditions had already systematically exploited, and occasionally subverted, the big-foreground schema. The bold foreground was, for instance, a hallmark of late Soviet montage cinema. The device carried over into the sound era partly, it seems, through the influence of Eisenstein’s courses at the film academy. (Recall the murder of the pawnbroker as an instance of “mise-en-shot” in chapter 1, Figs. 1.31–1.32.) With the spread of Socialist Realist doctrine in the 1930s, Russian film became heavily theatrical. As in other countries, directors seem to have felt that staging in greater depth and longer takes solved the problem of dynamizing talky scenes. Few countries, however, cultivated a sense of depth as extravagant as that found in such Soviet films as Lenin in 1918 (1937) and The Great Citizen (1938–1939) (Fig. 3.52). “Wellesian” depth became a hallmark of Stalinist cinema through the 1940s and 1950s. Aggressive foregrounds can be found in Japanese woodblock prints, particularly those of Hiroshige, and they reappeared in many Japanese films of the 1920s and 1930s (Figs. 3.22, 3.53). Mizoguchi eagerly experimented with these schemas (Figs. 3.21, 3.54),59 but he seems to have largely given them up around 1937. He explores the resources of another, more prominent Japanese schema: that in which the foreground is unemphatic and the most important action occurs in background planes—some of them remarkably far oª (Fig. 3.55). This more subdued handling of depth may be taken as a rebuke in advance to the Welles-Wyler tradition, which can seem at once flashy and heavy-handed.

MIZOGUCHI, OR MODULATION

111

3.50 The flamboyant depth shot in Hollywood: a gun wired to fire into a prizefight audience in Mr. Moto’s Gamble (1938). The woman’s face is inserted via a matte eªect.

3.51 Citizen Kane (1941).

3.52 A clenched depth composition for spies and wreckers in The Great Citizen, Part 2 (1939).

3.53 As in Fig. 3.50, an optical eªect creates a gigantic foreground (Kaiki Edogawa Ranzen, 1937).

3.54 Mizoguchi experiments with the exaggerated foreground (Naniwa Elegy).

3.55 Iwami Jutaro (1937): townsfolk call on the swordsman. In a characteristically Japanese composition the foreground area is set far back, and many layers of depth stretch into the distance.

112

MIZOGUCHI, OR MODULATION

3.56 Naniwa Elegy: when Asai and Ayako confront his wife, the characters shift around a small zone, each one occupying central position in turn.

3.57 Naniwa Elegy.

3.58 Naniwa Elegy.

3.59 Naniwa Elegy.

shot of our Oharu sequence (Figs. 3.6–3.10). Such shots force the spectator not only to scan the screen but to scrutinize rather small patches of it. Again, most directors used the empty foreground to create a picturesque establishing shot that initiates an edited sequence. The composition could also serve as a prolonged epilogue for a scene, as in The Poppy (Fig. 3.46). Mizoguchi, however, saw still other possibilities. In Naniwa Elegy an entire scene may be played out quite far from the camera, as when Asai’s wife discovers him in Ayako’s apartment (Fig. 3.56). More daring is a later scene in the same setting. Ayako has just told her boyfriend, Nishimura, that she has been Asai’s mistress, and his reaction has been concealed in ways I’ll examine shortly. As Fujino arrives to wrest his money from Ayako, she disguises Nishimura as her “tough protector” and sets him down facing away from Fujino. She quarrels with Fujino in a dining room in the distance (Fig. 3.57), and the camera holds on the composition while he leaves oªscreen. Nishimura eventually rises and goes to the rear to talk with Ayako (Fig. 3.58). Here Mizoguchi holds the end of the shot long enough to make sure we see him—just

MIZOGUCHI, OR MODULATION

113

barely see him—slump slightly in his seat. He is already defeated, and a cut to another long shot confirms that this tired, frightened silhouette will never agree to marry a tarnished woman (Fig. 3.59). As in The Poppy, the long take joins forces with what we might call “distant depths” to make characters’ postures salient; but now the take serves not as coda but as revelatory climax. Staging the action at such a distance creates a psychological opacity that is deepened by the lighting. In the early 1930s films Mizoguchi had worked occasionally with patches of darkness (Fig. 3.27), and his Dai-Ichi Eiga cinematographer, Miki Minoru, was adept at providing silhouettes and chiaroscuro. Given the atmosphere of Naniwa Elegy’s story, shrouding many scenes in darkness evokes the sordid life Ayako pursues and also leaves certain things to the imagination; but as in Figs. 3.57 and 3.59 it obliges us to notice Nishimura’s posture. Now that the entire scene can be narrated through selectively lit long shots, Mizoguchi can add expressive overtones at every moment. When Asai propositions Ayako early in the film, the shot begins at his desk and shows her dodging right and left until he pursues her to the rear of the room. Her evasive embarrassment is, as so often in Mizoguchi, conveyed through a movement away from the camera. This “dorsality” tactic forms the basis of a climactic scene later, when Ayako confesses to Nishimura that she is Asai’s mistress. The shot has an empty foreground, and Ayako leaves him at the table as she turns from us, walking slowly to the rear (Fig. 3.60). Eventually she goes into her dressing room, and Nishimura follows (Fig. 3.61). Then Mizoguchi starts the process all over again, as she shrinks from him (Fig. 3.62), goes to the middle ground (Fig. 3.63), then goes to crouch in the rear, at the farthest point from him possible (Fig. 3.64). He oªers no consolation, standing—also with his back to us—mute and impotent. The whole passage yields a powerful sense of Ayako’s shame, expressed less through facial expression than through her plaintive monologue and the arrangement of bodies in space. The tone of the scene shifts to desperation. In the corner she twists to her feet and rushes to Nishimura, her face frontal and hotly lit as she begs him to marry her (Fig. 3.65). Now it is Nishimura’s facial reaction that becomes important. Many directors would handle the scene by letting Ayako draw him toward the foreground and swivel his face to the camera so that we look over her shoulder at his face. Not Mizoguchi. Ayako pulls her reluctant suitor away from us and downward, kneeling with him in the middle ground and continuing to plead (Fig. 3.66). A cutaway to Fujino’s arrival outside keeps Nishimura’s reaction unknown—although his silence and passivity suggest that he is shocked and afraid, and further hints to this eªect will soon arise from the compositions we’ve already considered (Figs. 3.57–3.59). In their sustained rigor the scenes in Ayako’s apartment go beyond Japanese norms and Sternberg’s pictorialism. They achieve both denotative richness—a suspense and uncertainty about Ayako’s confession, about Nishimura’s reaction, about Fujino’s retaliation— and pervasive emotional tints, conveying, in a film noir atmosphere, the heroine’s alternating shame and braggadocio, oblivious to her lover’s paralysis. They also require an

114

MIZOGUCHI, OR MODULATION

3.60 Naniwa Elegy.

3.61 Naniwa Elegy.

3.62 Naniwa Elegy.

3.63 Naniwa Elegy.

actress (the jaunty Yamada Isuzu) willing to forgo the glamorous close-ups most would expect. Brilliant though these scenes in Ayako’s apartment are, the staging is notably simple. It is as if Mizoguchi, hitting on a way to sustain and develop the “distant-depth” schema, wanted to limit his variables for greater control. Having set himself a unique problem, he would soon make things harder on himself and his collaborators. With Japan’s invasion of China in 1937, films like Naniwa Elegy and Sisters of Gion became politically unwelcome. The failure of Dai-Ichi Eiga drove Mizoguchi to the Shinko studio, where he accepted assigned projects, sometimes shinpa-flavored, sometimes patriotic. He filmed a story about traveling actors based on Tolstoy’s Resurrection (The Straits of Love and Hate, 1937), a home-front film about a soldier’s family (The Song of the Camp, 1938; lost), and a village romantic drama (Ah, My Home Town, 1938; lost). Mizoguchi then moved to Shochiku, where he was given higher-profile projects. He made a series of Meiji-mono (dramas set in the Meiji era, 1868–1912), which commemorated practitioners of traditional arts—kabuki performers in The Story of the Last Chrysanthemum (1939) and The Life of an Actor (1941; lost), puppeteers in Woman of Osaka (1940; lost).

MIZOGUCHI, OR MODULATION

115

3.64 Naniwa Elegy.

3.65 Naniwa Elegy.

3.66 Naniwa Elegy.

Mizoguchi has said that he ran to the Meiji-mono as sanctuary from propaganda duties, but the government smiled on these celebrations of indigenous culture and Shochiku recruited him for the two-part Genroku Chushingura (The Chushingura Tale of the Genroku Era, a.k.a. The Loyal 47 Ronin, 1941–1942). Based on the nation’s most distinguished heroic legend, featuring the formerly Communist kabuki troupe Zenshinza, adapted from a cycle of plays by the distinguished Mayama Seika,60 Genroku Chushingura was the most prestigious “national-policy” film. Although it did not garner acclaim, Mizoguchi was sent to scout China for projects. Soon, however, studio resources and film stock began to dwindle, and on tiny budgets Mizoguchi completed another kabuki biography (Three Generations of Danjuro, 1944; lost), two films honoring famous swordfighters (Miyamoto Musashi, 1944; The Famous Sword Bijomaru, 1945), and an episode in an all-star patriotic vehicle, Victory Song (1945). In the surviving films up through 1942 we can see a steady eªort to both complicate and purify the most demanding aspects of Naniwa Elegy. Now Mizoguchi builds nearly every scene on long takes. The average shot length increases dramatically, from twenty-

116

MIZOGUCHI, OR MODULATION

three seconds in Sisters of Gion to ninety seconds in the first part of Genroku Chushingura (and eighty-three seconds in the second part). Sisters of Gion pushes farther into chiaroscuro, and although it does not dispense with analytical editing, a great many scenes are played out in fixed, distantly framed shots (Fig. 3.67). With The Straits of Love and Hate, shot/reverse-shot cutting and big foregrounds are abandoned in favor of sustained long shots that develop in crowded depth (Fig. 3.68). Mizoguchi carries many figures through apertures and patches of darkness, and the results are virtuosic (Fig. 3.69). To this pictorial richness, enhanced by the multilevel backstage areas of kabuki playhouses, The Story of the Last Chrysanthemum adds a gravity of figure and camera movement. The rhythm becomes even more majestic in Genroku Chushingura, in which the canonical tale of bloody retribution becomes a tribute to patience and self-sacrifice. The forty-seven vassals’ revenge on the corrupt Lord Kira is perpetually delayed and then (against every convention of previous versions) kept oªscreen. Instead, we are given monumental tableaus, statuesque scenes of deliberation, ceremonies conducted with a serene poise, vignettes of characters enduring and fulfilling the samurai code (Fig. 3.70). Mizoguchi’s experiments with pictorial opacity and obliqueness give way to a Homeric transparency, in which every locale, gesture, and stance is articulated crisply—save the heroes’ ritual seppuku, tactfully hidden in a field of taut screens (Fig. 3.71). At once gorgeous and severe, Genroku Chushingura bathes bushido in an aura of spirituality.61 Three moments in The Story of the Last Chrysanthemum indicate how Mizoguchi sets himself the task of purifying and nuancing his staging during the war period. The plot, set in the late nineteenth century, centers on the young actor Onoe Kikugoro, whose weak performances disgrace his adopted kabuki family. None of Kiku’s hangers-on dares to criticize him, but the family maid, Otoku, is frank and oªers to help him improve. Their friendship scandalizes the family, who cast both out. With Otoku Kiku embarks on a lengthy tour of the provinces to gain more practice. He often despairs or turns lazy, but Otoku (another version of the devoted shinpa heroine) sustains him. Thanks to an uncle’s intercession, Kiku gets to display his new skills in Tokyo, and Otoku slips away to let him reconcile with his stepfather. Kiku enjoys public acclaim at the very moment that she dies in a shabby boardinghouse.62 One turning point occurs when Kiku, left alone with Otoku, joins her in caring for a baby and eating watermelon. In one shot she carries the baby inside the house, sets him down under a protective net, and walks to the distant kitchen, barely visible in a slot (Fig. 3.72). Kiku follows (Fig. 3.73). The next shot, taken at a point 180 degrees to the first, shows Kiku and Otoku kneeling and silhouetted in the kitchen (Fig. 3.74). He slices watermelon for them while she flicks out the seeds with her hairpin. As they shift a bit further back to eat, he tells her he’s never been so happy as the night she told him the truth about his acting. She is moved and replies that she’ll do all she can to help him. Their quasi-love scene consists of his moving slightly toward Otoku and her lowering her head (Fig. 3.75). At this moment the family returns, and Kiku and Otoku draw apart (Fig. 3.76). Otoku hurries out left and returns, bending again but turning to watch Kiku’s aunt behind them.

MIZOGUCHI, OR MODULATION

117

3.67 Sisters of Gion (1936): Umekichi is consoled by Omocha after her patron has abandoned her.

3.68 The Straits of Love and Hate (1937): a backstage area, with actresses making up in the nearest plane, actors passing onstage and oª in the background, and the drama played out in a distant middle zone.

3.69 Backstage clutter reminiscent of von Sternberg is revealed by the grave camera movements of The Story of the Last Chrysanthemum (1939).

3.70 Many scenes in Genroku Chushingura (1941–1942) consist of gathered samurai, shifting their positions slightly to sustain a long-take shot.

3.71 The ritual suicides that conclude Genroku Chushingura.

118

MIZOGUCHI, OR MODULATION

3.72 Story of the Last Chrysanthemum: the first kitchen scene begins as the maid Otoku passes from the baby’s play area into the depths of the house.

3.73 Kiku follows.

3.74 A reverse-angle cut shows them in the kitchen, where he slices watermelon. At moments the baby is visible far in the distance.

3.75 Facing each other to eat, Kiku tells Otoku how grateful he is for her frankness.

She pulls aside, as if worried she’s getting too close to Kiku, while Kiku (in an echo of the cowardly Nishimura) turns away, embarrassed (Figs. 3.77–3.78). In the next scene the aunt will fire Otoku. The dim foreground, characteristic of mid-1930s Mizoguchi, creates a threshold beyond which light traces a welter of crisscrossing planes. Thanks to a 180-degree camera shift, like that in Oyuki the Virgin, the same area of the house is made to yield two rich fields of corridors and hideaways. As Kiku follows Otoku, she is barely visible in a distant aperture; in the reverse shot the distant plane is occupied by the baby. The arrangement allows Mizoguchi to present a delicate play of blocking and disclosing. When Kiku slides toward Otoku, his gesture takes place at screen center, masking the baby (Fig. 3.75). The family returns, and Otoku’s head withholds, then reveals, the arrival of Kiku’s aunt (Fig. 3.76). The same ballet is repeated when Otoku pulls away from Kiku and reveals the

MIZOGUCHI, OR MODULATION

119

3.76 They edge apart as Kiku’s aunt arrives.

3.77 Otoku busies herself, coming and going, blocking the aunt . . .

3.78 . . . until she is revealed watching the young couple.

aunt, grimly unwrapping her obi (Fig. 3.77–3.78). As a whole the shot is at once busy and purified. The mazelike background and cluttered foreground are more dense with information than in most of Mizoguchi’s films before 1936, yet the middle-ground drama is a suite of microactions conveying the warming of a man and a woman to each other. Nishimura’s slight slump in Naniwa Elegy has become an entire repertory of minute shifts of position. The full shot and long take yield not just apertures but pathways. Looking at the final image of The Poppy (Fig. 3.46), we can recall the father’s and daughter’s trajectories. The stages of Ayako’s humiliating confession in Naniwa Elegy are implacably mapped out as phases of her progress through two rooms (Figs. 3.60–3.66). Thus can the dense, static image preserve the traces of the characters’ movements; cutting up the scene would wipe out these spatial memories. In our Last Chrysanthemum scene the angle makes us notice how Kiku follows Otoku’s path to the kitchen, and in the reverse shot their traversal of

120

MIZOGUCHI, OR MODULATION

3.79 Story of the Last Chrysanthemum: the second kitchen scene, without Otoku. Kiku passes the boy, now a toddler. Compare Figs. 3.72–3.73.

3.80 He comes to the kitchen area, now filled with servants. The shot recalls the one shown in Figs. 3.74–3.75.

3.81 Once more the aunt arrives in the background, echoing Figs. 3.76–3.78; but now Otoku is gone.

this space hovers, ghostlike, behind them until Kiku’s aunt works her own variation on it. The plot actually builds a second moment on this spatial memory of brief intimacy. Years later, as the family celebrates Kiku’s triumphal return, he wanders into the same room. The baby is now a little boy. Kiku goes to the kitchen (Fig. 3.79), and in reverse shot we get another repetition. A cook is chopping bread where Otoku and Kiku had sat, and maids scurry about in silhouette (Fig. 3.80). Now the aunt comes in from a diªerent angle and summons him to the party (Fig. 3.81). No need for a close-up of Kiku’s chagrin: our memory of the figures’ movement through these spaces echoes his awareness of the loss of the woman who enabled him to succeed. A third powerful moment, also foreshadowed by the first kitchen scene, comes after Kiku learns where the ill Otoku is staying. She has witnessed the performance that assures his fame, so she can die fulfilled. While her landlord Genshun fans her and a pulselike hammering comes from a neighborhood artisan’s shop, Kiku kneels beside her. In

MIZOGUCHI, OR MODULATION

121

3.82 The Story of the Last Chrysanthemum: at Otoku’s deathbed.

3.83 The Story of the Last Chrysanthemum.

3.84 The Story of the Last Chrysanthemum.

3.85 The Story of the Last Chrysanthemum.

3.86 The Story of the Last Chrysanthemum.

122

MIZOGUCHI, OR MODULATION

a shot lasting nearly six minutes, with no significant use of depth, Mizoguchi renders the tenderness of woman’s sacrifice and the man’s recognition of it, which always comes too late. Everything here turns on minute changes. Kiku passes to the left edge of Otoku’s pallet, and the camera obligingly arcs to resettle him. When Kiku says that they can now marry, she turns her head slightly and the fan stops (Fig. 3.82). Kiku insists she’ll get well, again Genshun ceases fanning (now to wipe his eyes), and Kiku moves slightly so that his shadow falls on her face (Fig. 3.83). The fan stops and starts; she tells Kiku to go to his public; he hesitates. Otoku stretches out her arm, whispering, “I’m your wife now, with your parents’ permission!” (Fig. 3.84). He bends close over her, pauses, and then moves even closer—the nearest he has been to her in the entire film (Fig. 3.85). He agrees to go, rising and pausing with his back to us (naturally), then leaves, pausing to look back. Otoku twists her head just a bit to see him as he says: “Wait for me” (Fig. 3.86). As he and Genshun go out, Genshun’s daughter resumes fanning. After all the shots draped in darkness, all the labyrinths exposed behind the kabuki stage, all the intricate blocking and revealing of barely visible depths, Kiku’s farewell to Otoku oªers a heightened simplicity. Compared to the nuances of posture, gaze, and aspect here, the couple’s byplay in the kitchen looks broad. Yet that byplay, and many others like it, have prepared us to see much in little, and now Mizoguchi tests our training. He strips his setting, his composition, and the performances to the drama’s emotional core. He seems determined to discover how much expressiveness he can summon up from the simplest ingredients—faces and barely moving bodies observed at a discreet distance. The same blend of density and purity pervades Genroku Chushingura: within massive castles and corridors, solemnly moving figures and grand tracking shots achieve very gradual changes of angle and aspect. The wartime swordplay films Miyamoto Musashi and The Famous Sword Bijomaru occasionally oªer kindred, if paler and less demanding, moments of pictorial gravity. Forged under new political demands for cultural sobriety, Mizoguchi’s style accorded well with a “national aesthetic” compounded of Zen and other signs of sacred traditions.63 For us its place in the history of cinematic staging lies in its expressive ambitions. In seeking to mesmerize the viewer by dwelling on tiny gestures within richly packed spaces, Mizoguchi’s images become at once reserved and moving, sumptuous and austere.

At the war’s end the Allied occupation, led by the United States, began to remake Japan. MacArthur and his staª reformed the Diet, strengthened political parties, broke up conglomerates, gave land to peasants, legalized labor unions, and expanded women’s rights.64 Occupation authorities forced all filmmakers to avoid “feudal” subjects (no more swordplay movies) and authorized only films that illustrated liberal and democratic values. Under this system Mizoguchi made several films that more or less overtly promoted the new ideology. Two were guardedly optimistic a‹rmations of women’s rights: Victory of Women

MIZOGUCHI, OR MODULATION

123

(1946), which centered on a woman lawyer defending a mother charged with causing her baby’s death, and My Love Has Been Burning (1949), the story of a Meiji era feminist. In somewhat the same vein was The Love of Sumako the Actress (1947), based on one of the most colorful women of the 1910s. Utamaro and His Five Women (1946) presented the woodblock artist as a man of the people, scorning aristocratic patronage and recording the fugitive charms of the “floating world” of brothels and teahouses. Believing that Japanese culture had encouraged people to suppress aªection, Occupation censors encouraged filmmakers to break sexual taboos; soon the first films to show kissing were released.65 To this trend Mizoguchi contributed Women of the Night (1948), a neorealistic cross-section of women forced to sell sex amid the rubble of a starving Tokyo. Many critics, at the time and since, have dismissed most of these films as weakened by their propagandistic purposes,66 but all are of interest to the student of filmic storytelling and style. The stories are occasionally thin and didactic, but the passions run high, and the mise-en-scène is usually admirable. To a certain extent the 1946–1950 films also constitute a new phase of experimentation. If we call the earliest films eclectic and the 1936–1945 films ascetic (as well as, curiously, replete), we can see the immediate postwar years as a phase of intelligent, somewhat extroverted pluralism. For one thing the emotions are less contained; there is less sobbing and more shrieking. Mizoguchi admitted that he was now making more “barbarous” films, claiming that he had to release the resentment he felt toward wartime authorities.67 Moreover, now this director knows what he can do best. He will never return to the wild days of quick montage. In the wake of Welles many directors were embracing aggressive foregrounds, but Mizoguchi withdrew to other terrain. Secure in his repertoire of staging strategies, he now carefully extends them. He subtly tests boundaries, tackles fresh tasks, and flexes his craft. The films of the late forties are made by a man in his late forties, and he brings to them a confidence not quite ready to settle into routine. So, for instance, he tries out new patterns of staging in relation to camera movement. In most of his prior work he had tracked characters laterally through a space or, in Genroku Chushingura, swooped and spiraled around fixed figures. One nearly eight-minute shot in Victory of Women (1946) achieves a spike of expressive power through figures driving abruptly toward the camera, forcing it back, only to retreat abruptly into the distance before returning to the foreground (Figs. 3.87–3.89). The close framing and thrusting action here oªer a remarkable reversal of the principles informing Ayako’s confession in Naniwa Elegy, yielding not muted pathos but almost hysterical passion. Another rearward tracking shot acquires diªerent dimensions in My Love Has Been Burning. Here the heroine tries to reach her lover after a meeting of his political party. As he and his cronies descend a stair and stride expansively toward us, she plays hide-and-seek with the retreating camera (Figs. 3.90–3.91). Mizoguchi’s characteristic game of vision has turned symbolic, putting the feminist, the true progressive, on the periphery of men’s making of history. No such splendid camera movements grace The Love of Sumako the Actress, which harks back to the opacity and asceticism of the late 1930s. It consists largely of long shots—

124

MIZOGUCHI, OR MODULATION

3.87 Victory of Women (1946): a woman whose baby has just died crawls toward us, frantically explaining to a female lawyer how it happened. She half-rises, clutching her breast, filling the frame with her desperation.

3.88 After collapsing below the frame line, she rushes to the rear, to fall in a heap. The shot holds on a distantdepth composition as the lawyer goes to the rear room.

3.89 The camera moves forward as she guides the woman to the middle ground, and the advance to the camera resumes, culminating in a close two-shot.

3.90 My Love Has Been Burning (1949): Eiko behind the men who make history.

3.91 My Love Has Been Burning: blocked by the shifting positions of the men, Eiko can occasionally be glimpsed looking on.

MIZOGUCHI, OR MODULATION

125

3.92 The Love of Sumako the Actress (1947).

3.93 The Love of Sumako the Actress.

3.94 The Love of Sumako the Actress.

deep, dark, uncommunicative images that swallow up the characters. But there are also a few closer views, and these participate in a carefully judged pattern across the film. One might expect that the theater scenes would be handled in wide shots, the oªscreen drama in close-ups. Almost willfully, however, Mizoguchi reserves the closest shots of Sumako for performances or rehearsals, as if to sharpen the diªerence between theater and life. After the death of her mentor and lover, Sumako kneels at her shrine to him and asks him how her performance was tonight (Fig. 3.92). Getting no answer, she doubles over, then shrinks from the camera, as if trying to press herself through the wall (Fig. 3.93). Now dorsality and a retreat from the camera express not shame but loss and fearful solitude. Sumako’s suicide will take place oªscreen, but this scene does duty for it, showing her pain pushing her into oblivion. Yet the purely private emotion seen here is abruptly countermanded. The next shot is a blunt medium close-up of Sumako onstage as Carmen, facing the camera and brandishing her rose (Fig. 3.94). Her aggressiveness seethes through the rest of the scene, with her voice and gestures trembling on the verge of hysteria. Soon she will frantically hurl an actor to the floor backstage; as Carmen she will be stabbed; and after the curtain falls, she

126

MIZOGUCHI, OR MODULATION

3.95 Utamaro and His Five Women (1947): Utamaro tries to persuade his pupil to give up licentiousness.

3.96 Drawing Oran for the first time: the screen’s halfhiding of Oran may cite the original artist’s famous technique of “veiled views,” but the schema was also used in Japanese films (see 3.38).

will hang herself. The harsh shot-change marks the stylistic registers distinguishing the oªstage drama from the performances, while immediately suggesting Sumako’s swing between despair and defiance. She is at once a Japanese woman still looking to her sensei for approval and a Carmen-like artist finding giddy liberation in a burst of creative energy. Mizoguchi’s willingness to blend his more muted staging patterns (retreats from the camera, glimpsed bits of action, distant objects of significance) with strategic close-ups is apparent in Utamaro and His Five Women. A better title would be a straight rendition of the Japanese: Five Women Surrounding Utamaro. The famous Edo print designer sits at the center of the plot, threading five lives together. The least important is Oshin, his publisher’s fiancée. There are also Yukie, a painter’s daughter thrown over by Utamaro’s disciple Seinosuke, and Orin, the lovely woman with whom Seinosuke eventually elopes. There is Takasode, whose ravishing back inspires Utamaro to design a tattoo for her. Most important is Okita, who seduces Seinosuke but also competes with Takasode for the favors of the vacant, beautiful boy Shozaburo. The shifting love triangles overlap on Utamaro and his passion for glorifying the female form. The film’s actual title not only evokes his gravitational pull on other fates but also recalls the original artist’s bijin-ga (“beauty pictures”), which included many series based on numbers (“Six Selected Types of Love,” “Ten Studies in Female Physiognomy”). Utamaro vividly illustrates the new pluralism of Mizoguchi’s early postwar work. Plenty of scenes exploit reticent depths, as when Take tries to retrieve the drunken Seinosuke from Okita (Fig. 3.95). When Uta draws Oran for the first time, the shot employs both an aperture and a screen that tease us with glimpses of this extraordinary beauty (Fig. 3.96). Yet Mizoguchi sets himself new problems as well. How, he asks in eªect, can one coordinate several people crisscrossing a fixed frame? The initial drawing duel between Utamaro and Seinosuke conveys brisk energy by stitching a welter of diªerently paced trajectories through layers of depth, with superbly synchronized pauses. The film’s final scene, as Uta-

MIZOGUCHI, OR MODULATION

127

3.97 When Utamaro returns a free man, his followers surround him.

3.98 As they regroup to bring food and wine, he pivots and slides closer to us.

3.99 They reenter; he shrugs oª his jacket and, springing up, tells them he simply must draw!

3.100 As they start to eat in the distance, he bends eagerly over his paper in the foreground.

maro returns from prison and starts to draw again, oªers another instance of eªortless group choreography (Figs. 3.97–3.100). Mizoguchi can now handle crowds without resorting to stasis or to solemn unison movements (as sometimes in Genroku Chushingura). His long shots radiate the assurance that they can render any rhythm, any dramatic turn. When Mizoguchi cuts up a scene into closer views, he is self-consciously pursuing particular ends. Granted, Yoda records that Mizoguchi said, “I hate close-ups.”68 But we do have him noting in 1952: “Close-ups can’t be avoided. Though long shots can convey psychological conflicts, close-ups are indispensable for more complicated nuances.”69 The close-ups in his earliest films were usually brief, simply composed in the American manner, and part of a string of reverse shots or eyeline matches. In the 1930s the close-up became more complex, as an aggressive foreground or, as at the close of Naniwa Elegy and at moments in Sisters of Gion, a singularized image standing out from the surrounding long shots. One scene in Utamaro shows Mizoguchi experimenting with the sustained close view “for more complicated nuances.”

128

MIZOGUCHI, OR MODULATION

3.101 Utamaro and His Five Women: Okita enters, having murdered her rival and her lover.

3.102 Okita leaves to surrender to the police.

Okita (Tanaka Kinuyo) has stabbed both her feckless lover, Shozaburo, and her rival, the tattooed Takasode. She staggers into Utamaro’s lodgings, where she confesses her crime and says farewell. The scene consists of two shots, the first running two minutes and eighteen seconds, the second running almost four minutes. The whole action is simple and symmetrical, structured on a coming/staying/departing pattern. Okita enters via the porch, and the camera follows her leftward in medium shot (Fig. 3.101) to confront Utamaro and his housemates. The bulk of the scene consists of her confession, broken by the scene’s only cut. Then Okita rises, determined to surrender to the police, and goes out, followed in a rightward tracking shot and the film’s biggest close-up (Fig. 3.102). Others rush to follow her out, but Utamaro stays behind, writhing in his handcuªs and gasping, “I want so much to draw!” Like Mizoguchi himself, Utamaro seeks the expressive image: Okita’s suªering lends her the pathetic beauty he longs to capture. The close-ups leading Okita in and out of the room bookend a remarkable stretch of staging, itself also filmed at fairly close range. Once she is in the room, Utamaro asks if she has gone mad. She drops to her knees before him and denies it (Fig. 3.103). Then Okita’s sister and Yukie come in to squat behind her. After declaring that she won’t accept a lukewarm love, Okita slides toward Utamaro, demanding: “Doesn’t your portrait of me express the same thought?” (Fig. 3.104). On the line Mizoguchi cuts to a medium close-up (Fig. 3.105). The cut underscores the dialogue point: Okita acknowledges that Utamaro’s art has captured her heedless search for perfect love. Curiously, though, the cut shifts us only a little closer to the action and without a change of angle. Eisenstein and other Soviet directors had exploited this sort of axial cut, and so had Kurosawa, but in the sound era most directors avoided it because it makes the scene jump out a bit spasmodically at the viewer. And here the change of scale is so slight that we don’t doubt that Mizoguchi could have sustained the previous shot. So why this new framing? Evidently Mizoguchi does not want Utamaro to be visible in the next phase of the action, which will center on Okita.

MIZOGUCHI, OR MODULATION

129

3.103 Okita sits.

3.104 Okita comes close to Utamaro.

3.105 Cut in to Okita; now Utamaro and his servant, Take, are merely on the edges of the action.

3.106 She speaks of her love for Shozaburo, glancing oª at each man and asking: “Is love like this?” Her answer comes not from them but from behind.

3.107 Yukie raises her head and slides a bit forward. As the image racks focus to her, Yukie says she understands and now knows “the path a woman must follow.”

3.108 Yukie lowers her head and Okita’s sister speaks up from behind: “I too.”

130

MIZOGUCHI, OR MODULATION

3.109 As her sister lowers her head, the framing tilts down to Okita, who stiªens and declares she will turn herself in to the authorities.

3.110 At this the women behind lift their heads again.

3.111 In the next phase of the scene Okita says her farewells. She bows to Utamaro, now made prominent by a slight pan and the fact that the women all hide their faces by bowing.

3.112 Then Okita lifts her head, draws back, and thrusts herself closer to Utamaro than ever before, saying: “Uta, after I die, please be nice to the woodblock print of Okita!” After this she rises, trudging out via the close tracking shot already shown (3.102).

Yet to have Utamaro pull away from her would be out of character, for he is keenly concerned about her fate. So the cut simply chops him out, at least until the climax of the shot. And for reasons that will become clear, Mizoguchi cuts directly in because he wants to preserve the slightly downward angle. The rest of the scene (traced in Figs. 3.106–3.112) reminds us of what Feuillade already knew: close framings need not simply carry one piece of information, such as a character’s facial reaction (Figs. 2.74–2.75). Mizoguchi’s prolonged medium close-up can oªer some of the same expressive resonances we find in shots like Otoku’s death in Last Chrysanthemum. He builds Okita’s confession out of minimal movements (her turning of her head or sliding a few inches left or right), with other faces concealed by being oªscreen (Utamaro, Take), by being blocked by Okita, or by being lowered (the women

MIZOGUCHI, OR MODULATION

131

behind her). The closer view can also aid the hypnotic eªect Mizoguchi seeks. He shifts our attention from point to point through focus changes and the upturned faces of Yukie and the sister before creating a climax out of Okita’s thrusting her face closer to Utamaro’s than ever before. (The tactic echoes Kikugoro’s bending over the dying Otoku.) Now we can understand the final ambivalent close-up of the departing Okita as the pictorial crescendo, aimed at clinching the “complicated nuances” of her state of mind (Fig. 3.102). And the high angle that Mizoguchi retains after the cut-in closer view keeps the other women in the action so that their reactions oªer poignant echoes of Okita’s confession. The shot may also allude to the “corner illusion” central to Japanese art: the view looking downward into a room, with figures arranged to hide the joints among walls, floors, and screens.70 This portrait of an Edo artist may also be quoting the schemas at which he was so adept.

When the American occupiers departed in 1952, Japan’s film industry was in high-speed growth. Despite US eªorts to break up conglomerates, six major studios ruled, with five controlling theater chains. The firms quickly reinstated the “feudal” swordplay films (chanbara) and expanded their lines of romances and comedies. Output swelled from about two hundred titles in 1950 to more than five hundred in 1956, reestablishing the Japanese film industry as the world’s most prolific. Several generations competed on this crowded field. Ozu, Naruse, Gosho, and other senior directors maintained the piecemeal-découpage tradition in the protective shadow of studios like Shochiku and Toho. Men who had started directing during the China war, such as Kurosawa, Kinoshita Keinosuke, and Imai Tadashi, came to the fore, as did slightly younger ones like Ichikawa Kon, Kobayashi Masaki, and Shindo Kaneto. Many directors eagerly embraced the wide-angle, deep-focus aesthetic that was becoming an international norm, complete with aggressive foregrounds. A few younger directors occasionally incorporated dense compositions on the model of prewar pictorialism, as well as displaying one-oª visual innovations. (Kurosawa excelled in the latter.) Most of the young men proved versatile, shifting among various genres and techniques; none, it appears, probed a core of stylistic premises from film to film. In these circumstances Mizoguchi could only appear a throwback. Still widely respected, he was nonetheless obliged to shift among studios, adapting novels and often adding the sexual ingredients of which the Occupation approved. In A Picture of Madame Yuki (1950) a wife raped by her husband guiltily derives satisfaction from his brutality. Miss Oyu (1951) concentrates on a discreet ménage à trois. In Lady of Musashino (1951) a man drawn to a former lover punishes her by flirting with younger women. Mizoguchi’s work had often featured eroticism, but these films—and Ugetsu Monogatari (1953), with Lady Wakasa’s dreamworld of sensuality—flaunted it. Even with such concessions he declared himself unable to understand the tastes of the young audience.71 He was rescued by his old friend Nagata Masaichi. Nagata’s Daiei company held few

132

MIZOGUCHI, OR MODULATION

screens, so after Rashomon (1950) won the top prize at Venice and the Academy Award for best foreign film, Daiei aimed at the overseas market. Although Japan was still sunk in poverty, with one-tenth of the per capita income of the United States, it became the first Asian country to gain worldwide attention for its cinema. Festivals paid tribute to Mother (1952), Where Chimneys Are Seen (1953), Seven Samurai (1953), and The Burmese Harp (1956). Kinugasa’s Gate of Hell (1954) earned two Academy Awards, one for best foreign film, the other for best costume design—an event unthinkable even a few years before. Thanks to the festival market, Mizoguchi’s fortunes changed. Oharu ranked only ninth in the Kinema Jumpo poll, but it won the top prize at Venice in 1952. He left the Shintoho studio for Daiei, which backed all his subsequent projects. Ugetsu, placing third in the poll and earning another grand prize at Venice, clinched his status. He could now command important subjects and big budgets. After Sansho the Bailiª ’s Venice win in 1954, Mizoguchi joined the Daiei board of directors. He was not completely free of studio interference or the occasional forced project (for example, Princess Yang Kwei Fei, 1955, a coproduction with Hong Kong), and his films still attracted little attention at the local box o‹ce. But for a year or two after Oharu, critics looked more favorably on his work, which screened at festivals and began to enter overseas markets. In 1954 Mizoguchi received an Education Ministry award for lifetime achievement. The very end, however, was not so sunny. He suªered from leukemia in his final years, and he was quite feeble during the filming of Street of Shame (1956). His last three films— Princess Yang Kwei Fei, Shin Heike Monogatari (1956), and Street of Shame—did not place in the Kinema Jumpo poll because, one observer suggests, most critics considered them “purely commercial eªorts, overestimated by westerners.”72 His Daiei works struck many as kitschy exoticism catering to the growing festival circuit. Indeed, Mizoguchi’s postwar output brought his entire career into question. Writers suggested that he was a romantic pessimist, since his films on social issues dramatized ills without oªering cures.73 Many younger critics promoted left-wing social-problem directors like Imai Tadashi, whose films consistently beat Mizoguchi’s in the Kinema Jumpo rankings. Older critics continued to prefer Ozu, Naruse, and Gosho, who found poignancy shining out of commonplace lives. Mizoguchi’s aristocratic elegance, along with his taste for shinpa dramaturgy, had become old-fashioned. Those who thought him behind the times had their opinion confirmed by his admiration for classical art. He often compared film to the e-makimono, the picture scroll that unrolled a scene.74 Like other directors of monumental jidai-geki, he had already quoted the bird’s-eye view characteristic of e-makimono and picture screens (Fig. 3.71). The lakeside romp of the princess and the potter in Ugetsu was modeled on a landscape drawn from Ogata Korin, one of the most famous seventeenth-century painters.75 Unlike most of the directors praised by 1950s critics he clung to the pictorialism of the 1930s, when he and his contemporaries were exploring distinctively Japanese shot designs. In a 1952 interview he spoke of admiring the “density” of Japanese art; he found in ink paintings

MIZOGUCHI, OR MODULATION

133

of the fifteenth and sixteenth century “an enormous suggestive intensity that I often wish I could get into a film.”76 He seems not to have regarded his Japanese contemporaries as serious rivals. He saw William Wyler as his only competition on the festival scene, and he fretted that Roman Holiday would beat Ugetsu at Venice. When he watched a Wyler film, he timed the scenes and shots.77 Yet he seems not to have worried about competition in the long-take sweepstakes (his shots were already far lengthier than Wyler’s or virtually anyone else’s). He told Yoda that he admired Wyler’s “vertical framing”—presumably his low-angle depth shots, which pile faces from bottom to top.78 By studying Wyler, Mizoguchi seems to have been considering ways he could squeeze still more into every image. Is this competition or co-optation? Noël Burch has proposed that after The Love of the Actress Sumako Mizoguchi brought his work into line with “one of the dominant tendencies in Hollywood at that time: the long take à la Wyler.”79 He identifies Mizoguchi’s “representational system” with “radical camera distance, long takes, dolly movements” and then claims that these evolved into “an academically decorative, opportunistic approach to découpage, contrived by laying certain traits of his earlier system over the framework of the Hollywood codes.”80 For example, in Utamaro “the long take is largely subservient to the Western codes and there are frequent close-ups. . . . One observes both the temptation of the codes from the conquering West and the nostalgia for an exemplary past of personal and national rigour.”81 Yet the techniques Burch itemizes aren’t the whole of Mizoguchi’s style. He was for most of his career a variegated stylist (180-degree cutting patterns versus single-shot scenes, big foreground close-ups versus empty-foreground depth shots).82 For Burch, to rupture a distant shot with a close-up is to risk surrendering to Western norms; but Mizoguchi was surrendering since the 1920s, albeit in distinctive ways. The “Hollywood codes” were his point of departure throughout his career, along with more idiosyncratic schemas he shared with his colleagues. Ironically, one could make a case that Mizoguchi most fully embraced Western norms in low-budget wartime eªorts like Miyamoto Musashi and The Famous Sword Bijomaru. But then Burch’s belief that Occupation forces (“the conquering West”) steered Japanese directors toward Hollywood would come into question. We shouldn’t forget that even in the 1950s no other filmmaker, apparently not in Japan and certainly not in the United States, was making films as complex in their staging as Mizoguchi’s. The disparaging comparison with Wyler or Cukor becomes implausible when we look at the staging of The Best Years of Our Lives (1945) or The Marrying Kind (1952)— fine films but worlds away from the precision and delicacy we find even in a lesser Mizoguchi film like Gion Festival Music (1953). Let us also recall that Mizoguchi’s purportedly code-complicitous films were received with boredom or annoyance by Western critics who had not embraced the mise-en-scène aesthetic. Of Chikamatsu Monogatari Ado Kyrou complained: “The direction is flat. The camera keeps its distance; it is too polite. Good films cannot be made without close-ups.”83 Of Ugetsu Paul Rotha wrote that by developing his action in long shots Mizoguchi becomes “a pedestrian filmmaker.”84

134

MIZOGUCHI, OR MODULATION

He was nearly always more pluralistic within the long-take tradition than critics allowed: he tested his staging along many dimensions (varieties of shot scale, degrees of depth, fixity versus mobile camera, static figures versus figure movement). The idiom of his last phase, I think it fair to say, is closer to that of Taki no Shiraito and Oyuki the Virgin than the severity of Last Chrysanthemum and Genroku Chushingura. If one takes those exercises in austerity as a benchmark, then anything that is more diversified will look compromised. Although I, too, am thrilled by the pushing of limits, to elevate this to a criterion of excellence seems too narrow a view of style. Simply refusing the close-up is no guarantee of artistic purity (as witness the lumbering procession of “minimalist” films since the 1970s). Stylistic rigor can become rigor mortis. The critic’s task is to weigh functions and eªects, and in Okita’s confession to Utamaro (Figs. 3.101–3.112) we have already seen how a closer view can powerfully coordinate an ensemble of details. It is true that by 1950 Mizoguchi seems to have solved the principal problems he posed himself. He had created one “distinctively Japanese” style that could answer to various cultural demands, from pride in artistry to political nationalism. He had shown that riches lay hidden in pictorial schemas his colleagues often left unmined. He had created a hypnotic cinema centered on the dense image, governed by unpredictable spatial developments and slightly changing details. He had made the sustained shot discreetly expressive through shinpa conventions and nuanced performances. Seen from this standpoint, the last films present distillations of his characteristic staging strategies. They also continually recall his formative era. None of his peers preserved so integrally, as if in an archive, prewar achievements in shot design. Whereas Kurosawa and others created a muscular technique out of the calligraphic impulses of earlier years (freely borrowing from Soviet and Hollywood films), Mizoguchi’s long takes reactivate, with subtle variations, a panoply of pictorialist imagery. In this sense he is a conservative, but what he conserves becomes in his hands an inexhaustible resource. In the 1930s and 1940s he had refined current schemas; now he refines his refinements. Ugetsu Monogatari uses his veiling eªects to evoke the supernatural (Fig. 3.113). In 1947’s Utamaro and His Five Women he had shown Okita grabbing a knife to wreak her vengeance, treating the action in an aperture framing and an almost expressionistic shadow (Figs. 3.114–3.115). The Life of Oharu, as we’ve seen, almost completely conceals the knife, then flashes it along the frame edge, barely visible (Fig. 3.10). In other films he stretches his love of lattices in fresh directions (Figs. 3.116–3.118). He also sets himself new problems. He multiplies the number of figures whose faces he can hide while pushing the action at ever-increasing distance from the camera (Figs. 3.119–3.120). Otoku’s death scene in Last Chrysanthemum had put her feeble facial turns at the center of a sparse frame (Figs. 3.82–3.83); but what, our director seems to be asking, if Lady Musashino expires in a more crowded frame, with her face even tinier than Otoku’s and changing its position ever so slightly (Figs. 3.121–3.122)? If this be academicism, make the most of it. Let every young filmmaker take any late Mizoguchi film and watch. Let the student see how what is most significant can arrive

MIZOGUCHI, OR MODULATION

135

3.113 Ugetsu Monogatari (1953): screens and veils evoke the otherworldly life of the princess.

3.114 Utamaro: Okita rushes behind a screen, yielding Mizoguchi’s characteristic glimpse of her eye in the gap between curtain and wall, before . . .

3.115 . . . her silhouette reveals her murderous intent.

3.116 In Sansho the Bailiª (1954) the grid becomes a motif: the imprisonment of a mother . . .

3.117 . . . who is tortured in the brothel . . .

3.118 . . . is linked to the capture of her son.

136

MIZOGUCHI, OR MODULATION

3.119 The tormented ménage à trois of Miss Oyu (1951).

3.120 Miss Oyu.

3.121 Lady of Musashino (1951).

3.122 Lady of Musashino.

eªortlessly before us, often by the smallest resettling of figures or the simplest panning movement. Let the student observe how what is most dramatically important can shrink or hide itself—sometimes unobtrusively, sometimes in full view. Let the student plot how earlier phases of the action leave their traces in pathways and pigeonholes, to be reactivated when needed (often when least expected). If the student wants the camera to move (as all do), let him or her observe the timing by which a simple tracking shot permits figures, setting, and framing to align, swerve apart, and reconverge: a compact choreography of delicate changes, with no pictorial dead spots and none of that hurly-burly of today’s walk-and-talks. And above all let the student consider how almost every image grasps our attention as a splendid composition (which you want to explore in fine detail), a vessel of dramatic conflict (which you scan for glances, facial expressions, compositional confrontations, imminent arrivals), and a tensely contained emotional field (which may erupt into violence or crumple into a zone of private feeling). The “enormous suggestive intensity” that Mizoguchi admired in Japanese art lies coiled and waiting for us in scene after scene of his late films. In an era when “visual literacy” means taking for granted

MIZOGUCHI, OR MODULATION

137

3.123 Street of Shame (1956): Dreamland’s staª and habitués are introduced in distant-depth apertures.

3.124 Street of Shame: a slight shift of the brothel master’s head will block the kitchen boy far oª and left of center . . .

3.125 . . . and activate new zones of space on the right, stretching far into the distance.

3.126 Shizuko is prepared for her first night.

3.127 Street of Shame: the last shot.

3.128 Street of Shame: a last look. Compare Figs. 3.35 and 3.114.

138

MIZOGUCHI, OR MODULATION

those bursts of pointless cutting that purportedly energize a scene, Mizoguchi (like Ozu, but in a diªerent way) gives us time to see everything. We could do worse than to treat this oeuvre as an Academy for the Study of Staging. His last film makes a good postgraduate exercise. Street of Shame has converging plotlines in the vein of Utamaro, but now the point of intersection is a brothel. As the parliament debates the abolition of prostitution, six women are linked through their work at Dreamland, in the Yoshiwara district of Tokyo. One supports a sick husband and a baby; another has a grown son who is ashamed of her; another is out to bilk men. Into this sordid tangle struts Mickey, the fifties equivalent of the moga of Naniwa Elegy. Sashaying in her tight skirt, flipping her ponytail, scrambling after every customer, smoking and chewing gum and nibbling snacks (sometimes all at the same time), Mickey brings postwar vulgarity to Dreamland’s echt-neoclassical house of pleasure. While oªering bitter humor and genuine pathos, Street of Shame provides a tutorial in how to stage groups. We have adjacent rooms and apertures laying out scenes (Fig. 3.123). We have apparently casual shots that exfoliate into richly articulated planes, and we have bold compositions in flamboyant depth stretching out from a middle ground, activated by the slightest movement (Figs. 3.124–3.125). The idea of the simple close-up here loses nearly all meaning; even near shots are packed with information (Fig. 3.126). At the film’s close the young serving girl Shizuko—a minor character until now—is dressed and made up to take her first customer. She is sent outside, and Mickey shows her how to catch a man’s attention. Oªscreen a radio announcer reports that once more the antiprostitution bill has been defeated. After glancing this way and that, Shizuko timidly peeks out at us from behind a pillar (Fig. 3.127), then ducks partway back (Fig. 3.128). In finally bringing us to a close view of a character, the shot recalls the last image of Naniwa Elegy (Fig. 3.47). But this image is more sustained, and it passes through a suite of small mutations. Shizuko’s expression shifts quickly from shyness to childish excitement to fearful uncertainty to . . . what? It is a searing shot with which to end a film, let alone a career. No long take, no distant framing, no camera movement, no chiaroscuro. How Mizoguchian can this image be? Utterly. For one last time the game of vision, now borne by a single darting eye, traces exquisite modulations of emotion.

MIZOGUCHI, OR MODULATION

139

4 ANGELOPOULOS, OR MELANCHOLY

Theodoros Angelopoulos may be the last believer in a cinema of heroic statement. His films put on display characteristic agonies of modern civilization—totalitarian regimes, the misery of the stateless immigrant, the disenchantment of the Left. They search for the roots of these conditions in history or, as he sometimes puts it in interviews, History.1 Each film presents a journey, he asserts, “and the journey embraces all the human emotions from love and sexuality to hate and war.”2 He freights his scenes with mythological references (citations of The Odyssey or Attic tragedy) and portentous symbols. He draws his dialogue from the poetry of Homer, Eliot, and George Seferis. He employs allegory and iconography with shameless literalness, dramatizing political conflict through a flurried dance of street banners, or representing the end of a political movement by a drifting barge draped with flags. He doesn’t think twice about having the same performer play several parts or suggesting that the action is all taking place in the mind of a character (a filmmaker, no less)—as if he hasn’t heard that these tricks went out with platform shoes and mood rings. He works on the edge of embarrassment. In Ulysses’ Gaze (1995) a colossal head of Lenin is loaded onto a barge and then floats down the Danube to a somber musical score (Plate 4.1); the sequence, Angelopoulos tells us, marks the death of Communism for him and his generation.3 His admirers believe that he can capture modern anguish in monumental imagery; for others it is all bombast. Indeed, part of the fascination of Angelopoulos’s cinema is its almost naive anachronism. This filmmaker, so attuned to history (or History), so sensitive to post-Communist emigration and the war in the Balkans, appears unaware of how dated his artistic ambi-

140

A N G E L O P O U L O S , O R M E L A N C H O LY

tions seem. He can with a straight face propose a trilogy on nothing less than the history of the twentieth century, keyed to the myths of Oedipus and Antigone.4 Oblivious to fashion, he never seems to worry that his films will seem ponderous and overwrought. No postmodern pastiche, tone switching, or self-ironizing here. He will not make a playful reference to anything; every moment strives to resonate deeply. His lifelong passion for the poetry of Greek masters like Seferis and Cafavy and the novels of Faulkner, Joyce, and Stendhal have given his films literary aspirations. Mass culture appears as a vulgar scrawl on the landscape (the pop song in the dingy café in The Beekeeper), and cinema is cited only in its gravest mode (Murnau, Dreyer, Bergman). Growing up in an era when solemn filmmaking was the best proof that cinema could be as exalted an art as poetry, theater, and the novel, Angelopoulos has embraced all those subjects that are di‹cult to lighten up: war, repression, the failure of love. Pretentiousness is the risk he runs for believing that only high seriousness is worthy of his medium and of the questions an artist must ask. I may as well confess my divided aªection for these sprawling, majestic, irritating works. Certain of them (for example, Ulysses’ Gaze; Eternity and a Day, 1998) seem to me deeply flawed, inflating their thematic statement at the expense of narrative density. Others, such as The Travelling Players (1975) and The Hunters (1977), are remarkable accomplishments, but often they read better in critical commentary than they play on the screen. Four, however—Alexander the Great (1980), Voyage to Cythera (1983), Landscape in the Mist (1988), and The Suspended Step of the Stork (1991)—seem to me authentic masterpieces, attaining grandeur without becoming grandiose. Still, none of the films lacks awe-inspiring passages, and all sustain, sometimes brilliantly, that tradition of staging and shooting that is the concern of this book. If I treat Angelopoulos as more a synthesizer than an innovator, that takes nothing from his achievement. In today’s thintextured European cinema, where formulaic découpage and big close-ups rule nearly as much as in Hollywood, at least one stubbornly highfalutin artist tries to keep alive the rich, evocative image.

I think that other Greek directors don’t have the same problems as I do. Being Greek, I am a part of Greek cinema, but not in the localized, provincial sense; and as far as style is concerned, there’s no meeting point. THEO ANGELOPOULOS5

Born in Athens in 1936, Angelopoulos spent his youth in the years of World War II and the Greek civil war (1946–1949). After taking a law degree and serving in the military, he went to Paris in the early 1960s to study film. Back in Athens he began writing film journalism for the left-wing newspaper Democratic Change. In 1967 a cadre of generals took over the government and installed Colonel George Papadopoulos as national leader. Angelopoulos began his movie career under the junta, working as actor and associate pro-

A N G E L O P O U L O S , O R M E L A N C H O LY

141

ducer on several films and shooting his first film, the short The Broadcast, in 1968. Despite the oppressive political climate, a group of young directors emerged in the late 1960s and were able to circumvent censorship somewhat. Military factions controlled the government until 1974, the year before Angelopoulos’s international breakthrough with his third feature, The Travelling Players. Since then he has been the most famous Greek director and one of the most honored filmmakers in the world. Angelopoulos and his generation worked in conditions very diªerent from that of Greek cinema’s golden age. Out of the rubble of World War II and the civil war, entrepreneurs built a small but thriving studio system. Production boomed, with over 250 films released in the 1950s and over 1,400 during the 1960s. Comedies, musicals, melodramas, and pastoral dramas ( foustanellas), all filled with popular stars, brought audiences into the theaters. New sound stages were equipped to European standards, not only improving local quality but also luring in productions from abroad. A film school was founded in 1951, and an archive in 1957. The first national film festival was held at Thessaloníki (Salonika) in 1960. Greek films achieved export too, most notably with Never on Sunday (1961) and the dramas of Michael Cacoyannis (Electra, 1962; Zorba the Greek, 1964), and Cacoyannis and other young directors became festival regulars. Even the arrival of the generals did not immediately stifle output, since the government financed historical films and patriotic spectacles. What local critics call the “New Greek Cinema” grew out of changes in film culture in the late 1960s. Cacoyannis and some other established directors fled the Papadopoulos regime, creating space for a new generation. In response to the dictatorship, and in harmony with New Left movements throughout the world, young directors began to commit themselves to a socially and politically critical cinema—at least as critical as could be sneaked by the censors. In addition, national television arrived in 1971 and immediately triggered a steep decline in filmgoing. In 1968 annual admissions had hit an all-time high of 137 million; by 1973, admissions were only 70 million. As the commercial cinema flagged, the New Greek Cinema oªered artistic films of personal vision. The spearhead was Angelopoulos’s initial feature, Reconstruction (1970), which won four awards at Thessaloníki and became identified as the first mature work of the rising generation. Directors bypassed the studio system and garnered funding from French, Italian, and US companies, as well as the Greek Film Center, founded in 1980.6 Despite accolades and subsidies, the New Cinema did not win back the audience. Admissions continued to plummet; by 1993 the annual figure stood at seven million. One or two dozen films were made each year, but American movies, along with a few French and Italian imports, ruled the box o‹ce. Virtually no Greeks went to Greek films, let alone to the recondite works of Angelopoulos and his peers. Their films circulated primarily on the festival circuit, which, however informal and decentralized, was becoming the only international distribution network comparable to Hollywood’s. Angelopoulos no longer entered his films in competition at Thessaloníki, partly out of concern for dominating the awards and partly because his proper venue was Cannes.

142

A N G E L O P O U L O S , O R M E L A N C H O LY

Angelopoulos’s career has a self-fashioned unity characteristic of a 1970s filmmaker aspiring to high artistry. Reconstruction oªers a critical look at dying Greek villages. His next four films set out to trace his nation’s modern history. Days of ’36 (1972) examines the dictatorial rule of General Metaxas (1937–1941), and The Travelling Players loops back and forth between the period of Nazi occupation and postwar civil strife. The Hunters revives the years from the civil war (1944–1949) to the mid-1960s through an arresting narrative device: present-day hunters find the corpse of a civil war partisan, as fresh as if killed yesterday. Alexander the Great completes the cycle by surveying the period from the turn of the century until the 1930s, tracing a bandit’s tyranny over a remote village organized as a commune. Throughout this tetralogy individuals tend to stand for larger forces, both political and mythological. The traveling players are named Electra, Orestes, and the like; Alexander the bandit is a brutish version of his legendary namesake; the hunters exist only as types identified with political positions. Voyage to Cythera marks a new stage in Angelopoulos’s oeuvre, since now political forces are embodied in more traditionally drawn characters. The elderly Communist Spyros, let out of prison, returns to his Macedonian village. This simple situation is refracted through the imagination of a film director who is meditating on a film project and who will, in the film-within-the-film, find in Spyros a mythical father figure. Further meditations on solitude and age are presented in The Beekeeper (1986), about a schoolteacher and beekeeping enthusiast who abruptly abandons his family and embarks on an odyssey across a bleak, unyielding modern Greece. Landscape in the Mist oªers another journey, following two children convinced that their father is waiting for them just across the border. Boundaries and migration come to the fore again in The Suspended Step of the Stork. Visiting a town on the Albanian border jammed with war refugees, a television journalist believes he recognizes a prominent Greek politician who has disappeared. Ulysses’ Gaze projects the same concerns with borders in following an archivist (called simply “A”) in his search for a lost early film. A’s quest is at once a parallel to the Odyssey, a hazardous homecoming, a trip into his childhood via flashbacks to the 1940s, an interrogation of central European history since World War I, and a quest for the historical significance of the Balkans, all presented through A’s search for the “first image,” vanished footage from the turn of the century. The turbulence of Eastern Europe becomes even more personalized in Eternity and a Day, in which a poet in the last days of his life reflects on his past while accompanying a young Albanian boy trying to return home. Like Feuillade and Mizoguchi, Angelopoulos brooks no contradiction on the set. “He is the image of his Alexander,” notes his longtime cinematographer Yorgos Arvanitis, “who ended as a tyrant.”7 To achieve the rhythm the director demands, the actor must count silently between lines. In order to deflate the tourist ideal of Greece, Angelopoulos favors filming in winter, when rain and mists dominate; if the sun breaks out, he halts shooting. While making The Travelling Players, “I was persecuted by beautiful weather.”8 His search for majestic locales takes him to rough and remote places. The filming of

A N G E L O P O U L O S , O R M E L A N C H O LY

143

Alexander the Great in a poverty-stricken mountain village over a bitterly cold winter led to underfed extras and crews threatening to defect.9 On location in the Yugoslavian town of Mostar during bombardment in the Balkan War, Angelopoulos refused to leave as ordered by authorities, until he got the footage he wanted. He has become an Odysseus, taking cast and crew all over his region. His most daunting expedition was, appropriately, Ulysses’ Gaze, which passed from Greece through Albania, Macedonia, Romania, and the Black Sea city of Constanza before ending up in Croatia, with Mostar standing in for Sarajevo. “I believe something special happens on location, in the real place. . . . When I am in the place I have set the film, all five of my senses are working.”10 Historically, as well as geographically, the films sink their roots deep into Balkan culture, as Andrew Horton has shown in a careful study.11 Yet by rendering the history of Greece and its neighbors through imagery culled from mythology, literature, and Marxist thought, Angelopoulos joined a broader tradition. He worked in a vein that was cosmopolitan and pan-European, aimed at an elite audience. From the start his films took prizes at major continental festivals, and soon they were financed by private companies and television chains in Italy, France, and Germany, as well as by EU agencies. In the 1980s and 1990s he relied on stars like Marcello Mastroianni, Jeanne Moreau, Harvey Keitel, and Bruno Ganz. Angelopoulos has thus kept alive a strain of European modernist filmmaking that runs back to the 1950s. But his historical situation obliged him to sustain the tradition with a new degree of self-consciousness. Inevitably, a filmmaker starting out in 1970 confronted problems that directors had not faced fifteen years earlier. He or she came belatedly to the modernist tradition, one that had already acquired a history, a pantheon of masters, a theoretical program, a rhetoric, and a set of institutions. In a tradition that prized originality, the novice who wanted to be a film artist was expected to create a strongly distinctive style; but that in turn required knowing what others had done. Because reputations were made principally on the festival circuit, the filmmaker had to find international financing and distribution and settle for minor festivals before arriving at one of the Big Three (Berlin, Cannes, Venice). Because critics’ comments weighed heavily, one had to oªer them food for thought and participate in intellectually inclined interviews. Angelopoulos was well equipped to succeed in this situation. He passed through Paris in the crucial years 1961–1964, studying film and, he claims, getting thrown out of the Institut des Hautes Études Cinématographiques (IDHEC). Like most “young cinema” directors, he came under the influence of what we might call the Langlois Cinémathèque canon: auteurs such as Welles and Mizoguchi, Hollywood genres such as the crime film and the musical. The seriousness with which he takes the director’s vocation doubtless derives from the saliency of Bergman, Fellini, Antonioni, and others in the cinéphile culture of the 1960s. Gods walked the earth then. Angelopoulos also grasped the need to shape critical appropriation of his films. A for-

144

A N G E L O P O U L O S , O R M E L A N C H O LY

mer critic himself, he gave interviews that help us to interpret his works, as when he explained that his own voice as the oªscreen father at the beginning of Voyage to Cythera is related to the symbolic “search for a father” that rules the film.12 He can come up with a lapidary formula that many critics would envy: “Voyage to Cythera evoked the silence of history, The Beekeeper the silence of love, and Landscape in the Mist the silence of God.”13 He has assisted us in understanding the development of his work, tracing recurring preoccupations and new shifts of interest. He points out “some very obvious obsessions”: umbrellas, marriage ceremonies, certain names, the sea, the oscillation between dense images and empty ones, a preference for silences, and “almost always, the presence of a little boy.”14 So it is the filmmaker, not a gang of critics, who proposes that his work splits into two periods. First was a forthrightly political phase, in which under the influence of Marx and Brecht he sought to explore “epic” presentation without psychological identification. During his stay in Paris, he claims, Brecht became “the essential reference point.”15 Then, beginning with Voyage to Cythera, came films that were more emotional and existentialist, rejecting the certainties of o‹cial politics. “Art has become anthropocentric again and asks more questions for which it has no answers.”16 His declaration of an ethically grounded humanism chimed well with post-Marxist attitudes of European intellectuals in the 1980s. I’m not suggesting that in packaging his oeuvre Angelopoulos is opportunistic; an artist’s help in deciphering his work is virtually demanded by festival cinema, which makes the director’s career development one of the film’s selling points.17 A cinéphile in the Parisian mold, a thoughtful exegete of his own work, Angelopoulos also created a trademark technique. Like Bresson, Dreyer, and other classic stylists, he seems to “process” every script, passing the story action through a series of finely meshed filters to arrive at a distinctive string of images. Critics commonly point out that he relies on long shots, long takes, temps morts, and oªscreen space, and he has discussed these techniques at length in his interviews.18 At the time of his early work he claimed that the long takes and camera movements created a “dialectic” among diªerent elements in the shot.19 He thereby encouraged a Brechtian characterization of his style— one amenable to critical reception in the wake of late 1960s left aesthetics. Ten years later, however, asked if the sudden rain in Voyage to Cythera was aiming at a Brechtian distancing, he replied: “Not for such reasons, but because artificial rain is very beautiful! It is as beautiful as natural rain.”20 As Mizoguchi discovered in the 1950s, the festival circuit rewards individuality of subject and form. Angelopoulos quickly understood that his playing field was pan-European. He delved into concrete Greek issues, but he kept his films from looking provincial by using techniques already canonized in the international milieu. His eªort, then and ever since, has been to achieve a localized cosmopolitanism, casting distinctive national material in sophisticated modernist shapes. He has experimented with flashbacks and ambiguous time schemes, devices canonized since the 1950s as signs of ambitious filmmaking. He has

A N G E L O P O U L O S , O R M E L A N C H O LY

145

also cultivated a style that drains away narrative denotation to a considerable degree, pushing instead toward a muted expressivity and a monumental, near-abstract pictorial design. Angelopoulos’s response to his circumstances reflects a new stage in filmmakers’ stylistic self-consciousness. When Feuillade adopted ensemble staging in the sustained shot, he was following the norms of his milieu. When Mizoguchi embraced the technique, he was picking a secondary option, pursuing it beyond the norms of sound filming in order to refine it. But by the early 1970s the long take had a distinguished pedigree, having been canonized not only in filmmaking practice but in the critical literature. Angelopoulos, along with contemporaries like Wim Wenders, Rainer Werner Fassbinder, Chantal Akerman, Maurice Pialat, and Werner Herzog, knew the debate about mise-en-scène that Bazin had launched and that animated 1950s Cahiers criticism. In Paris, during the early 1960s, Angelopoulos studied with Georges Sadoul and Jean Mitry, and he claims that while working as an usher at the Cinémathèque he acquired his tastes. The films he admired reflect the Bazin-Cahiers consensus: Murnau, Dreyer’s Ordet (1955), Welles (for his “cutting in the camera”), Renoir (for “deep focus and parallel stories”), Mizoguchi (like Sarris he was “seeing many of his films without subtitles, just watching the image”), and Antonioni.21 His first feature, Reconstruction, turns “new cinema” devices toward political critique. Kosta, a “guest worker,” returns from Germany to his tiny village, populated by fewer than a hundred women, children, and old men; virtually all the able-bodied young have found work abroad. Kosta is murdered by his wife, Eleni, and her lover Christos, and although they try to keep their crime a secret, gossip spreads about Kosta’s disappearance. Soon the police arrive. The lovers confess, each putting the blame on the other. Both are arrested, and the village’s women scramble to attack Eleni as she is hauled oª in a police van. The story is not, however, told in the order I have just presented. The opening scenes show Kosta laboriously returning to his village and greeting his children (who don’t recognize him). The credits appear over a freeze-frame of the family having their first shared meal in years. Then we see police questioning Eleni in her home, and she declares that Christos strangled Kosta as she watched. The plot goes on to alternate fairly brief scenes of police investigation—the “reconstruction” of the title—with long episodes from the couple’s eªorts to hide their crime. They bury Kosta’s body in the cottage’s garden, they flee town and spend a bleak night in the city, and eventually the police arrive, tipped oª by a suspicious old woman. At the end, after Eleni and Christos are taken oª, Angelopoulos provides an epilogue. He flashes back to the murder but keeps it oªscreen by holding the camera on the family cottage as the children play outside. This scrambled chronology reworks a template that has characterized European art cinema since the 1950s: the police-investigation-plus-flashbacks. Like Bertolucci’s La Commare secca (1962), Reconstruction uses detection to juggle time and shift point of view, letting us know the outcome straightaway in order to explore the causes leading up to it. Instead of providing psychological motivation—Eleni and Christos are barely characterized, and the final shot carefully prevents us from ascribing passion to the crime—the structure evokes a broader social critique. The flashbacks suggest that Kosta’s death grows out

146

A N G E L O P O U L O S , O R M E L A N C H O LY

4.1 Reconstruction (1970): doorways providing multiple frames for the action.

4.2 Reconstruction: figures in a landscape.

of the village’s frustrating poverty. The film “reconstructs” the social pressures leading to domestic infidelity and murder—a point underscored in a sequence showing journalists circulating through town while we hear, as voice-over commentary, guest workers in Germany explaining why they fled Greece.22 Reconstruction exploits several visual techniques typical of the 1960s “young cinemas.” Angelopoulos insists on shooting on location, confining his camera to cramped interiors and providing glimpses of a tavern or bus terminal. The cinéma vérité quality, strengthened by the many handheld shots, links the film to the new waves of the 1960s. Still, there are hints of the signature style to come. The average shot length (nineteen seconds), although much shorter than that of any later Angelopoulos film, announces his reliance on staging within the frame. The interiors are thickly packed, with doorways providing pockets of space (Fig. 4.1). Stillness and silence play central roles; the murderous lovers stand panting over Kosta’s body for what seems a very long time. The camera dwells on landscapes, as a bus churns slowly through the mud or as the couple totter along a rainswept highway and trucks roar past them. The plot has time to linger on a quasi-abstract shot of men dotted around a field, urinating during a bus stopover (Fig. 4.2). The elliptical treatment of the murder in Reconstruction becomes a pervasive narrational strategy in Days of ’36. A political prisoner takes a hostage, and eventually the police raid his cell. But these central events occur oªscreen, hidden by stone walls and cunningly angled doorways, all the better to concentrate on the maneuverings of a corrupt regime. Once again individual motives are minimized, with most shots presenting groups of prisoners and o‹cials. The film also pushes the long-take strategy much further (the ASL is seventy-two seconds), and long shots predominate so that the prison provides its own geometrical landscapes and spare choreography (Fig. 4.3). It is, however, with The Travelling Players that the Angelopoulos look emerges most fully. In a typical scene the actors’ troupe walks from the beach and up a town street; after they have passed through the frame, the camera holds on the street, and the period

A N G E L O P O U L O S , O R M E L A N C H O LY

147

4.3 Days of ’36 (1972): the abstraction of the prison setting.

4.4 The Travelling Players (1975): the shot starts in November of 1952. The traveling players walk from the dock . . .

4.5 . . . and we follow one as they pass. The color shifts slightly toward yellow.

4.6 The troupe moves on into the distance as . . .

shifts from November 1952 to winter of 1942 (Figs. 4.4–4.10). The distant framing, the lingering on an empty vista, and the flashback occurring within a single shot all became hallmarks of Angelopoulos’s style. With The Travelling Players, which has an average shot length of 105 seconds, Angelopoulos also confirmed his commitment to the very long take as his primary stylistic vehicle. Such shots rework familiar schemas in fresh ways. Nearly empty long shots had punctuated Antonioni’s 1960s films and Wim Wenders’s Goalie’s Anxiety at the Penalty Kick (1972), but Angelopoulos builds entire, slowly developing scenes around them. Flashbacks within a single shot were known in Hollywood (Caravan, 1937; Enchantment, 1948) and European cinema (Miss Julie, 1951). Yet as Angelopoulos has pointed out, his originality lay in freeing such flashbacks from the recollections of a single character and expressing instead “collective historical memory” or even the echoes of past events lingering in the shot’s locale.23 Likewise, in his reliance on the long take Angelopoulos proved

148

A N G E L O P O U L O S , O R M E L A N C H O LY

4.7 . . . a motorcycle cart bearing a political poster passes in the foreground. The camera follows it a bit before it goes into the distance. The color tonalities shift toward gray as . . .

4.8 . . . out of the distance comes a German automobile. We are now in the winter of 1942.

4.9 The car pulls up past a checkpoint, installed at the corner we saw earlier (4.6).

4.10 A German soldier patrols the checkpoint.

a judicious adapter. Few directors had applied the technique to every scene, and none had applied it so single-mindedly to landscapes. The flagrantly prolonged shots in The Travelling Players and his later films constitute his eªorts to enrich a recognized, indeed canonized, tradition. The ostentatious style results in part from an artist’s coming belatedly to a game in which most of the moves have been made, and by very strong players.

A European filmmaker starting out in the late 1960s confronted a rich array of stylistic options. One of the major choices was this: to utilize the long take or not? “The only great problem with cinema,” Godard remarked, “seems to me more and more with each film when and why to start a shot and when and why to end it.”24 Feuillade’s generation was not obliged to reflect on shot duration as an end in itself, and in the 1930s only a few di-

A N G E L O P O U L O S , O R M E L A N C H O LY

149

rectors, like Mizoguchi, seem to have grasped its possibilities. Simultaneous with the emergence of Bazin and mise-en-scène criticism, however, a great many directors were putting the long take on the agenda. The coming of sound encouraged filmmakers to lengthen their shots, and although this was done fairly unobtrusively in the 1930s, the problem of when to cut could not be ignored after Welles’s Citizen Kane (1941), which highlighted several fixed takes running a minute or more. Throughout the 1940s and 1950s the long take, almost always in tandem with camera movement, became a salient stylistic option. In American studios, at a period when the average shot length per film was nine to eleven seconds, some directors favored very long takes: about sixteen seconds for Otto Preminger’s Angel Face and George Cukor’s The Marrying Kind (both 1952), twenty-two seconds for Preminger’s The Fan (1949) and Vincente Minnelli’s The Band Wagon (1952). The device was flaunted in the very lengthy sequence-shots in Lady in the Lake (1946) and Rope (1948). Most filmmakers thought that CinemaScope demanded long takes, so many widescreen films, especially musicals, were cut rather slowly; Oklahoma! (1955), for example, has an ASL of nearly seventeen seconds. In postwar Eastern Europe as well the long take was alive and well, particularly because the Soviet cinema had developed a rather academic wide-angle style at about the same time Hollywood had (Fig. 4.11). In the late 1950s Panavision lenses eliminated the peculiar distortions of close views yielded by CinemaScope’s anamorphic process. American filmmakers could now return to the over-the-shoulder framings and strong singles of classical découpage. A few directors—Preminger and Billy Wilder, then Roman Polanski and Mike Nichols—persisted in employing long takes, but by the 1970s it was a retrograde gesture. The accelerated cutting of the intensified continuity style had triumphed, and only “oª-Hollywood” directors like Woody Allen and the younger independents would consistently rely on lengthy shots. In the hands of Hollywood directors the long take proved capable of great economy and unobtrusiveness, but it also suited the elliptical, suppressive narration of European modernist filmmaking. With causal connections often only hinted at, with character motivations kept mysterious, the gradually unfolding image gained extraordinary weight. The spectator had to search for relevant story information, register shifts in emotional tone, or spot moments that commented on characters’ relationships—individual versus group, interest versus detachment. Two influential stylists showed that the long take could sustain this oblique approach to storytelling. Carl Dreyer had been an editing-based director in his silent films; The Master of the House (1925) is possibly the fastest-cut Danish film of the period,25 and La Passion de Jeanne d’Arc (1928), built almost completely out of suªocating close-ups, became a hallmark of montage-based silent cinema. Yet in his sound films Dreyer swung to the opposite extreme. From the 1940s he embraced an explicitly theatrical aesthetic: minimal figure movement, scanning or arcing tracking shots, and steadily lengthening takes: an ASL of fifteen seconds in Day of Wrath (1943), seventeen seconds in Two People (1944), sixty-five seconds in Ordet (1955), ninety seconds in Gertrud (1964). The precision staging in the latter two

150

A N G E L O P O U L O S , O R M E L A N C H O LY

4.12 Dreyer recasts the tableau cinema of the 1910s for Gertrud (1964); compare Figs. 2.9 and 2.35. 4.11 In The Cranes Are Flying (1958) Mikhail Kalatozov inflated the deep-focus look of the o‹cial Soviet style, creating a kind of gargantuan romanticism. Compare Fig. 1.28.

4.13 Le Amiche (1955): Momina strides into the gallery, passing the portrait of Rosetta, the woman who has just attempted suicide; the camera catches a customer in the distance.

4.14 Quickly Momina turns; the camera reframes her, and now she blocks the customer, revealing Nene, the gallery owner. Throughout the long take Rosetta’s image and Nene’s timid presence will appear on the edges of the frame.

works constitutes a retrospective revision of the schemas of the European cinema Dreyer entered in the 1910s (Fig. 4.12). His films were often denounced as boring, but his influence on younger filmmakers such as Angelopoulos was considerable. Other European directors put the long take on the creative agenda—Luchino Visconti and Max Ophuls come to mind—but particularly important for our purposes is Michelangelo Antonioni.26 Antonioni’s 1950s films exemplify the richness of postwar mise-enscène. Fluid tracking shots pick up one character just as the one we’re following drops away, and other elements in the frame regroup themselves. Emphasis shifts easily between foreground and background, with only a nudge of the camera or a slightly turned player enough to create fresh compositions.27 (See Figs. 4.13–4.14.) The ASLs range from

A N G E L O P O U L O S , O R M E L A N C H O LY

151

twenty-two seconds (the film noir Cronaca di un amore [a.k.a. Story of a Love Aªair], 1950; Il Grido, 1957) to a very unusual forty-three seconds (I Vinti, 1953). In the string of more famous films beginning with L’Avventura (1960), Antonioni’s cutting pace picks up and his camera movements become less dazzling, but these works explore other options to which filmmakers also proved sensitive. The long take soon became central to the various Young Cinemas and new waves that emerged during the 1960s. Truªaut, Godard, Alexander Kluge, Jerzy Skolimowski, Ruy Guerra, and many other directors of the new generation flaunted long takes, often with generous camera movement. Significantly, however, virtually no directors adopted the long take as her or his stylistic signature in the manner of Dreyer or Antonioni. Most young filmmakers favored an “assemblage” aesthetic, whereby some sequences were handled in very long takes and other sequences were built out of dynamic cutting. Godard’s À bout de sou›e (1960) typifies this tendency, mixing prolonged camera movements with disjunctive montage. Oshima Nagisa took eclecticism to another level, building one film out of extraordinarily long takes (Night and Fog in Japan, 1960; ASL exceeds two minutes), and another out of very short shots (Violence at Noon, 1966; ASL 4.5 seconds). In an era of dawning film consciousness such pluralism implied that the filmmaker had mastered the entire history of cinematic technique, from Eisenstein to Ophuls. At the same time, certain technological factors that had encouraged long takes (sync-sound filming, heavy studio cameras) were largely gone, so long takes could be sensed as a deliberate eªect. After the 1960s many “new cinemas” would flaunt stylistic pluralism, the 1980s cinéma du look of Léos Carax and Jean-Jacques Beineix no less than the New Hollywood of Spielberg and Scorsese. Still, the long take is a device capable of serving many diªerent purposes. In Hollywood it was used chiefly for virtuosic following shots, as in today’s extended walk-andtalk sequences. By contrast, some postwar European filmmakers exploited it for the sake of what has come to be known as dedramatization. Between neorealism, with its gapfilled narratives and unexpected longueurs, and the blank surfaces and diminuendo pacing of L’Avventura, filmic storytelling changed significantly. Dramas began excising “melodrama,” moving closer to the spacious rhythms of the modern novel. Emotions were underplayed, even suppressed. Real life, critics argued, did not parse itself into tight plots and smoothly ascending tension. Cinema could break free of Hollywood artifice by rendering in detail the anticlimaxes and wayward incidents of everyday life. Aided by an approach to storytelling that fostered uncertainty about character psychology and causal connections, dedramatization became one hallmark of ambitious filmmaking. Dedramatization took two principal forms. Instead of playing an emotionally charged situation for maximum expressiveness, the filmmaker could treat it in suppressive or oblique fashion. The paradigmatic example was Rossellini’s Voyage to Italy (1953). A marriage is fading, and during their stay in Italy the husband and wife agree to keep away from each other. She desperately tours the region while the husband seeks out romantic temptations. As he comes home after several nights away, Rossellini shows the wife lying on the sofa, listening intently to him undressing and gargling; either she is hoping

152

A N G E L O P O U L O S , O R M E L A N C H O LY

4.15 Voyage to Italy (1953): the waiting wife.

to avoid a confrontation or hoping that he will come in to find her (Fig. 4.15). The resulting scene is no domestic quarrel; instead, it is merely a tissue of polite lies and halfformed accusations. Mizoguchi’s staging often conceals the emotions that burst out of his characters, but Rossellini keeps the emotions almost unexpressed. Voyage to Italy showed that a film could mute its action, even redefine what could count as action, by keying its tone to the couple’s boredom, enervation, and uneasiness—emotional states dissected in the European novel at least since Alberto Moravia’s Time of Indiªerence (1929) and Sartre’s Nausea (1938) but seldom explored on film. Directors could dedramatize their plots in a second way: by including moments their predecessors would have trimmed as waste. Voyage to Italy, almost in the manner of a travelogue, devotes long stretches to the wife’s ramblings around Naples and Pompeii. Filmmakers began wedging in temps morts, the “dead time” between dramatic arcs. In place of the compact dialogue of Hollywood, conversations were broken by prolonged pauses, often underscored by actors frozen in place. When characters did stir themselves, they often moved with a slowness reminiscent of the Sjöström and Bauer classics of the 1910s. The simple act of walking became prime cinematic material once more; filmmakers seemed to enjoy just trailing their characters. Bazin praised de Sica’s Bicycle Thieves (1947) for playing down traditional plotting and patiently observing the two main characters’ gaits. “It would be no exaggeration to say that Ladri di biciclette is the story of a walk through Rome by a father and his son.”28 The distant long take could support both suppressed emotion and temps morts. By not breaking the scene into close-ups, filmmakers could maintain a sense of muted drama. Subtle changes in framing could alert the spectator to the nuances that were replacing histrionics. And the sustained shot could force the viewer to concentrate on the empty intervals that filled so many scenes. Dreyer’s Ordet may have initiated this trend, but for most critics it was Antonioni who brought the novelistic, dedramatized, even antidramatic, cinema to maturity. It is hard for us today to realize how revolutionary a figure he seemed

A N G E L O P O U L O S , O R M E L A N C H O LY

153

4.17 L’Avventura (1960): lovers in a landscape in the final shot. 4.16 The reunion in Il Grido (1957).

in the 1960s. For one French critic his work marked the third great epoch of film history; after thirty years of silent cinema and thirty years of sound film, we now had “a cinema of behavior.”29 As another writer put it: “Antonioni’s people are not tidy, artificial, comfortably definite bundles of motives; they are always somewhat opaque, capable of surprises. As we are made to stare at them, and at the intervals of space and time which surround them, we must assess them as we assess people in our own lives.”30 All of Antonioni’s early films narrow their emotional range, but from Il Grido to Red Desert (1964)—the films that made his international reputation—he goes farther than his contemporaries in draining the drama out of charged situations.31 He virtually eliminates nondiegetic music, shot/reverse-shot cutting, and optical point of view. More and more he stages major actions in long shots, presenting figures in landscapes or spacious interiors.32 In the beginning of Il Grido, for instance, Aldo returns to Irma after seven years, and Antonioni presents their reunion in a very uncommunicative image (Fig. 4.16). Perhaps the most famous of his distant shots is the forlorn ending of L’Avventura, which shrinks Sandro and Claudia to mere streaks in a blocked-out composition (Fig. 4.17). Antonioni likewise reduces scenes to silences and stretches of dead time in which nothing of traditional dramatic moment is occurring. “I need to follow my characters beyond the moments conventionally considered important, to show them even when everything appears to have been said.”33 “Antonioni is so elegant,” remarks Jim Jarmusch, “in the way he can let the scene go past its normal length, or the shot even, and the whole weight of the scene changes, the essence.”34 He renders acting more impassive by means of a fairly muted performance style, long-held poses, and a tendency to turn figures from the camera at moments of dramatic intensity, a “dorsality” reminiscent of Mizoguchi (Figs. 3.60, 3.63, and 4.16). The result is the “inexpressive” shot that actually expresses lassitude, anomie, suppressed pain, or emotional distances between the characters. The stasis of such scenes alternates with the dilatory walks that consume so much time in L’Avventura, La Notte (1961), and L’Eclisse (1962), all films full of drift. With Antonioni, critics remarked, the talkies became the walkies.

154

A N G E L O P O U L O S , O R M E L A N C H O LY

4.18 Play Time (1967): the waiter in the foreground describes how the sole is prepared, while the waiter in the distance uses the same gestures to glue a tile to the dance floor.

Antonioni wasn’t alone in exploring subdued emotion and empty intervals within the long shot. Jacques Tati’s Les Vacances de M. Hulot (1953), Mon Oncle (1958), and Play Time (1967) use temps morts to encourage the spectator to find humor by comparing diªerent areas of the shot (Fig. 4.18). (Tati also pioneered the “full-field” schemas that would be so important for Otar Iosseliani and the 1970s minimalists like Akerman.) Nonetheless, it was Antonioni’s middle-period work that constituted the paradigm of “dedramatized” filmmaking for Angelopoulos’s generation. “When I came to France,” he recalled, “it was the epoch of Antonioni, of L’Avventura and La Notte. When I was at IDHEC, you’d go every day to see a film by Antonioni.”35 What was so impressive? “The considerable length of his shots, which went on just a little bit longer than expected to allow for a deep breath before going on.”36 The di‹culty for newcomers was that Antonioni had apparently pushed the dedramatized long take about as far as one could. He was, to borrow Harold Bloom’s phrase, a “strong precursor,” an innovator who intimidates his successors.37 What could a young filmmaker add to a distinguished tradition that seemed to have already reached its peak? One possibility was the poetic cinema that emerged in the USSR, most palpably in the work of Andrey Tarkovsky. Whereas Antonioni concentrated on urban anomie, Tarkovsky strove to capture the spiritual imprint of natural landscape on humans who could only partially apprehend it. He insisted that the film image should be neither obviously denotative, simply illustrating a story, nor portentously symbolic. He heaped scorn on all those schemas showing lovers split by fences and erring workers returning to the factory in triumphal long shots.38 The film artist should seek out images that cannot be reduced to words or ideas— images that strike with the force of religious revelation. The long take proved crucial to this neosymbolist line of thought. Rejecting Eisenstein’s conception of montage, Tarkovsky joined Mizoguchi in holding that sound cinema had no need for découpage that spelled everything out for the audience. Like Bazin, he believed that time is “imprinted in the shots,” so a film’s rhythm was determined “by the pressure of the time that runs through them.”39 Accordingly, Tarkovsky cultivated an atmospheric long-take style. Face-to-face encounters are filmed in protracted medium shot, but they are punctuated by hallucinatory

A N G E L O P O U L O S , O R M E L A N C H O LY

155

4.19 As women follow in the foreground, a procession restages Christ’s path to Calvary in the Russian snows (Andrei Rublev, 1966).

vistas and still-life studies. Humans move through a space of primal starkness (Fig. 4.19). At the other end of the visual scale the camera caresses quivering petals, murky streams, greasy puddles glistening on floorboards, or apples scattered across a tabletop, while breezes or rippling water or drifting snowflakes measure the time running through the shot. With the triumph of Andrei Rublev (1966) in European festivals, and the praise bestowed on Solaris (1972), The Mirror (1974), and Stalker (1979), Tarkovsky became the emblem of a mystical, quietly ravishing cinema whose long takes invited the spectator to become absorbed in “the concrete, living, emotional content of the object.”40 Another route out of the Antonioni impasse emerged from the political debates of the 1960s. Antonioni, like Balzac and Flaubert, had dissected bourgeois life from within, but many believed that this was not enough. Directors all over the world sought ways to put postwar modernist experimentation at the service of more radical political critique. Some filmmakers looked to 1920s Soviet filmmaking, but they filtered the montage aesthetic through the revival of disjunctive editing in the work of the Nouvelle Vague and Brazil’s Cinema Nôvo. The result was a mordant cinema of collage seen in Fernando Solanas and Octavio Getino’s The Hour of the Furnaces (1968) and Dusan Makavejev’s WR: Mysteries of the Organism (1971). Other politicized modernists turned to the “dedramatization” tradition. Most prominent among them was the Hungarian Miklós Jancsó, already strongly influenced by Antonioni in Cantata (1963). Turning from Antonioni’s enervated couples, however, Jancsó revamped the 1920s Soviet conception of a group protagonist. Thus The Round-Up (1965), Silence and Cry (1966), and The Red and the White (1967) concentrate on specific moments of historical change but play down heroes and concentrate on large-scale forces and momentby-moment fluctuations in power. The situations are intrinsically charged with dramatic voltage—guards confronting prisoners, commanders randomly pulling captives for execution—but Jancsó short-circuits the emotional eªect. Exposition is sketchy; secondary characters aren’t individuated (a device Angelopoulos would draw on), and central players sometimes don’t get characterized until moments before they die. The walkies are here with a vengeance. Instead of dialogue we get men and women pacing, strutting, marching, trotting, and dancing, often supplying few hints about their motives or destinations. Jancsó

156

A N G E L O P O U L O S , O R M E L A N C H O LY

4.20 Politics as mass spectacle in Jancsó’s Hungarian Rhapsody (1978).

forgoes nondiegetic music and employs a camera more concerned with group dynamics than individual destiny, casually picking up and dropping figures in the course of a shot. Scenes are played out in very long takes; some of Jancsó’s shots run nine minutes or more, consuming an entire camera reel, and some of the films consist of a mere ten shots. In the late 1960s Jancsó gave up the pretext of presenting a concrete dramatic situation, pushing toward ensemble numbers that put abstract conflicts of ideology on display. Now he aims at an all-engulfing choreography of figures, landscape, and camera. The scene teems with people, usually young and often naked. They move swiftly, in circular and serpentine patterns, while horsemen circle, musicians wind through the crowd, and banners flutter past us. Color, used symbolically to sharpen political oppositions, stands out against the neutral landscape. All the while, the camera tracks, cranes, zooms, and racks focus to create a floating, plastic, constantly shrinking and swelling space. Oxymoronic as it sounds, Jancsó’s version of dedramatization is florid, even “maximalist,” as became evident with Red Psalm (1972) and the films that followed. The contending groups have become pure emblems of social forces, playing out rituals on the bare arena of the Hungarian plains— cinematic pageantry enacting a neo-Marxist reading of history (Fig. 4.20).41 Although Jancsó avoids traditional dramatic appeals like psychologized protagonists, he oªers a robust kinetic spectacle in their place. Other political modernists pursued a more austere approach. Inspired by one conception of Brecht’s Verfremdungseªekt (estrangement eªect), some took dedramatization toward simplification and severity. JeanMarie Straub and Danièle Huillet explored the long-shot technique not only in The Chronicle of Anna Magdalena Bach (1967), whose pseudo-Baroque tableaus keep the camera at a considerable distance from the drama (Fig. 4.21), but also in a series of landscape films that included Fortini/Cani (1977). Perhaps the limit of this minimalist tendency is Godard’s “blackboard films” such as Vent d’est (1969), but it emerges in less forbidding form as well. One example is Chantal Akerman’s Jeanne Dielmann 23 Quai du Commerce 1080 Bruxelles (1975). In its tracing of a few days in the life of a housewife who works prostitution into

A N G E L O P O U L O S , O R M E L A N C H O LY

157

4.22 Geometrical filming of housework in Jeanne Dielmann (1975). 4.21 A baroque composition in long take for The Chronicle of Anna Magdalena Bach (1967).

4.23 Klaus Wyborny’s distant mise-en-scène in Dallas, Texas: After the Gold Rush (1970–1971).

4.24 India Song (1975): a mirror plays perceptual tricks on the colonial elite.

her daily schedule, the film dwells on mundane activities and temps morts. Through rigorous uses of perpendicular framing and 90-degree and 180-degree cuts, Akerman gives Jeanne’s routines a sober pictorial abstraction (Fig. 4.22). Even directors not inclined toward politicized modernism opted for long takes that presented a dry, detached pictorial design. For example, during the 1970s many directors began to present action in fairly fixed camera positions from a great distance, without cutting in to nearer and clearer views. The tendency was most evident in the avant-garde, such as James Benning’s 11 × 14 (1976) and Klaus Wyborny’s “neoprimitivist” works (Fig. 4.23). Marguerite Duras’s India Song (1975) played out highly melodramatic scenes in distant, oblique interiors accompanied by languorous oªscreen tango tunes (Fig. 4.24). All these avenues for exploiting the long take—lyrical spirituality, colorful spectacle, minimalism—were open to Angelopoulos when he began his career. But there was also a problem. So many people were embracing the long-take technique that originality was hard to come by, and sheer repetition could sooner or later register as mannerism. An-

158

A N G E L O P O U L O S , O R M E L A N C H O LY

gelopoulos became, like most of his contemporaries, a synthesizer who developed a distinctive style by mixing and revising available schemas. He was attracted to the austere variant of political modernism, and his earliest works explored its resources. In the second phase of his career (often with the aid of Antonioni’s scriptwriter, Tonino Guerra), he created ambivalent psychological dramas situated within concrete political and geographical milieus. Within each stage he preserved Antonioni’s dedramatizing tactics but adapted them to spectacular panoramas. What Antonioni had occasionally done with landscapes, Angelopoulos did consistently and at length. Further, he countered the impersonality and coldness of minimalists like Straub/Huillet and Akerman by evoking feelings suitable to grim weather, forbidding topography, and the demeanor of the human body in a monumental space. His images’ characteristic emotional tone—bleakness, futility, melancholy—took some of the stringency out of political modernism. Unlike Tarkovsky, he would not let the camera hover over textures and teasing details. Unlike Jancsó, he would not maintain a restless, headlong rhythm; he would slow the action down, as he puts it, to the tempo of steady breathing.42 True, by playing out his dramas in vast landscapes he recalled Jancsó, but instead of the exuberant rolling expanse of the Hungarian plains, Angelopoulos oªered rocky mountainsides and overcast seascapes: landscapes that exuded mournfulness. In sum, he sustained the European modernist tradition not only through time-shifting narration and allegorical tales of History but also through an aesthetic of austere spectacle. By finding a way to innovate within a competitive tradition, he revived and revised staging techniques developed in the 1950s and 1960s.

Angelopoulos’s gamble succeeded; he became the foremost long-take director of his time. The technique became the basis of his style in his tetralogy, extending the shot from an average of 72 seconds in Days of ’36 and 105 seconds in The Travelling Players to 206 seconds in The Hunters, then stepping back down to 107 seconds in Alexander the Great. In the first trilogy of his “humanist” phase he cut back, as if shorter shots were better suited to rendering more personalized psychology: 66 seconds average shot length in Voyage to Cythera, 78 seconds in The Beekeeper, and 82 seconds in Landscape in the Mist. Then his shot lengths started to rise again: The Suspended Step of the Stork averages 116 seconds, Ulysses’ Gaze 103 seconds, and Eternity and a Day 121 seconds. Despite these fluctuations, every film remains extraordinary in its avoidance of editing. In a period when a two-hour commercial film would contain between 900 and 1,800 shots, all of Angelopoulos’ films contained fewer than 150, and most fewer than 100. Some of those shots run a full camera reel. As with Antonioni and Jancsó, awareness of the apparently endless length of the take becomes an ingredient of the viewer’s experience. Just as pronounced was Angelopoulos’s preference for very distant shots. Some longtake films keep the camera close to the figures (e.g., Fellini’s 8 1/2, Alexei Guerman’s Khroustaliev, My Car!), but Angelopoulos favors the long-shot and extreme-long-shot range. He doesn’t avoid medium shots entirely, but usually they are incorporated into much wider

A N G E L O P O U L O S , O R M E L A N C H O LY

159

views. Even in interiors his camera assumes a considerable distance from the action, and outdoors it can reduce the humans to mere dots. He relies, with a somewhat anachronistic idealism, on the powers of cinema to present fine detail within a large-scale image. Angelopoulos’s films depend so crucially on fine-grained photographic rendition of immense spaces that much of their impact is lost on a television monitor, where a character may be only a clump of pixels. He is one of the last directors to design his images specifically for the theater screen. Long takes, long shots: the consistency is remarkable, but recognizing these factors constitutes only a first step toward understanding Angelopoulos’s style. How does he use the long take to achieve his characteristic eªects? What functions are performed by the distant framings? And how do these stylistic patterns revise norms of staging circulating at his moment? Most generally, Angelopoulos gains what other filmmakers in this tradition gain: the simultaneous presence of many elements in the visual field, soliciting the viewer to search out revealing aspects; the gradation of emphasis that activates secondary elements with special vividness; the sustaining of rhythm through the movement and stillness of figures and their coordination with the moving camera. Yet no deep-space directors before the 1960s relied so completely on the long shot and the extreme long shot. Like Braque carrying Cubism beyond Picasso, like Hindemith taking neoclassicism past Stravinsky, Angelopoulos pursues the dramatic and expressive possibilities opened up by Antonioni’s dedramatized compositions but within a consistently distant view. Still, he must also tell his story, and the continuous long shot poses problems on this front. Angelopoulos must now make the shot’s narrative development legible—if not as evident as in other stylistic traditions, at least minimally apparent. As with Feuillade and Mizoguchi, we can thus ask: How is attention guided and maintained within these images? Angelopoulos employs several staging strategies, some of which we have seen operating since the 1910s, others of more recent vintage. Most obvious is centering. A style based in tight framings, such as Hollywood’s intensified continuity, will tend to keep the center of the widescreen frame vacant, using cutting to balance slightly oª-center close shots against one another (Figs. 1.47–1.50). But in a long-shot, long-take aesthetic the central zone becomes an important guide to what is salient. The more distant the shot, even if it is comparatively empty, the more likely that key dramatic information will rest in or pass through the geometrical center of the picture format. This principle holds good in many Angelopoulos scenes already mentioned (for example, Figs. 4.1, 4.3, 4.5). So does the corollary strategy of two rival centers of interest operating in tension across the middle of the frame (Fig. 4.2). Like other directors in the mise-en-scène tradition, Angelopoulos uses aperture framing to guide our attention. Rectangular shapes in the set highlight certain elements. When Alexander and Voula greet their father in Voyage to Cythera, a “good Gestalt” of the three figures is created by means of centering, a sparsely furnished frame, the father’s frontality, and the sheltering box of the station doorway (Fig. 4.25). Later in the film a more dense

160

A N G E L O P O U L O S , O R M E L A N C H O LY

4.25 Voyage to Cythera (1983): Alexander and Voula meet their father in the station.

4.26 In the tavern, rectangles segregate figures (Voyage to Cythera). Compare Fig. 4.1.

4.27 Alexander the Great (1980): at the temple of Poseidon, tourists pause in a hush . . .

4.28 . . . so that the guide’s small gesture, just left of center, stands out.

shot of policemen and townsmen gathered in a café is made comprehensible through several lines and grids. As two policemen, near frame center, explain the deportation of Spyros, the key action is framed by the windowpane, and a nearly right-angled stovepipe segregates a secondary batch of listeners (Fig. 4.26). Alexander the Great boldly exploits aperture framing, for the camera is often very far away from the action. As the British tourists climb to the Temple of Poseidon, from which they will be kidnapped by Alexander’s gang, the camera tilts up with them, framing the temple from afar (Fig. 4.27). Their guide explains that the Greeks have lost faith in their gods. One tourist walks to the top, apparently overcome by the aura of antiquity, and says, “Shh!” while putting his finger to his lips (Fig. 4.28). Remarkably, this gesture is discernible in our postcard view; Angelopoulos has turned the other tourists from us, raised the speaker above the others, silhouetted him against the sky, and framed his gesture in a slot formed by the columns. Just as audacious is the scene of Alexander’s bringing his captives to the Communist village he dominates. Gathered in the town square the vil-

A N G E L O P O U L O S , O R M E L A N C H O LY

161

lagers quarrel about how to deal with him, but then Alexander’s men become visible between highly placed branches in the frame (Plate 4.2). We are cued by the sudden silence of the communards, the fact that they all turn from us to look at the horizon, their stillness when they see the band, the presence of movement at the top of the frame, and the patches of bright colors visible on the kidnapped lords. Slowly the bandits come forward (Plate 4.3), clear a space among the crowd, and put the captives at frame center—a triumphal moment whose majesty is undercut by the opaque means of rendering it (Plate 4.4). For Feuillade and his peers aperture framing was swift and unforced, its artificiality often unnoticed by the viewer. Here, as with Mizoguchi, it is a self-conscious aesthetic eªect: we are aware of straining to see bits of action within crannies of the setting. In Mizoguchi that straining is often rewarded with narratively important information, however diminished and fine-grained it may be. But in Angelopoulos the momentum of storytelling is invariably dissipated by dedramatization, and the positions of the English tourists and the villagers in these examples illustrate one distinctive staging tactic in his repertoire. His people turn away from us, and in an unusual way. Since the late 1910s, mainstream cinema has characteristically positioned characters at a three-quarter frontal angle. (Interestingly, some psychologists have suggested that this is the most informative posture for recognizing faces and registering eye movements.)43 In his 1950s and early 1960s work Antonioni replaced the angled frontal view with an exactly opposite one, framing characters turned diagonally away from us. In our earlier shot from Il Grido the dialogue is played with characters’ backs to us in two three-quarter views (Fig. 4.16). The eªect is that of an incomplete shot/reverse-shot sequence, withholding the shot that would show the characters frontally. This “three-quarter dorsality” minimizes our access to character reaction—sometimes oªering only a bit of cheek or jaw to watch, often obliging us to read reactions solely from stance and posture. Tarkovsky occasionally employs such uncommunicative views, but they are building blocks of staging for Angelopoulos, who insists on holding them quite a while. In the famous shot from Reconstruction several men dotted around a drab gray field relieve themselves in variants of the three-quarter pose (Fig. 4.2), and in The Beekeeper the hitchhiking girl dances to the jukebox in the roadside cafe (Fig. 4.29). In exteriors camera movement often works to preserve the three-quarter rear view. Alexander trails the bridal procession in Eternity and a Day from a distance, the camera angled to keep everyone in extreme long shot and turned from us (Fig. 4.30). Angelopoulos’s scenic curve often displays a gradual development toward three-quarter dorsality. Accordingly, frontal or profiled positions of the figures often serve as mere transitions to the three-quarter resting point (Figs. 4.31–4.32). In Voyage to Cythera Spyros has returned to his village and meets his old neighbor on the street, first in a marked three-quarter rear view (Fig. 4.33). After they share cigarettes and reflect on their departed comrades (“We were all finished in the civil war. . . . Everyone lost”), the neighbor says he has no reason to stay in the village, and the scene concludes with a rhythmic alternation of dorsal views (Figs. 4.34–4.35). The three-quarter rear view, apart from forcing us to infer characters’ facial reactions,

162

A N G E L O P O U L O S , O R M E L A N C H O LY

4.30 Eternity and a Day (1998): the wedding procession. 4.29 The hitchhiker dances in The Beekeeper (1986).

4.31 The Travelling Players: the conversation starts by letting the three-quarter poses of Electra and Pylades favor the poet’s dialogue and facial reactions.

4.32 In the course of the scene his reactions become more inaccessible as he goes to the window and presents yet another three-quarter back view.

4.33 Voyage to Cythera: Spyros meets his old comrade.

A N G E L O P O U L O S , O R M E L A N C H O LY

163

4.34 As the old friend turns away, the camera arcs to follow him leaving, swiveling him from one threequarter dorsal view to another.

4.35 The shot ends with Spyros tentatively advancing to the fence, creating another switch of character position as the diagonals lengthen.

also helps solve the problem of legibility in the landscape shots. The posture allows Angelopoulos to mark out human presence but also to coax our eye back to the surroundings. Instead of faces, which attract our notice even in very distant images, we get backs and shoulders—uninformative aspects that can’t hold us long. We are obliged to study the body or, more accurately, study the body’s relation to the larger field into which it’s inserted. As with Mizoguchi’s distant depths and dorsal views, the three-quarter views in Angelopoulos encourage us to pass from figure to environment. So does the sheer paucity of what is present. The Angelopoulos long shot tends to be sparse. The frame is not literally empty—some human figures are usually visible—but only a few sectors of space are activated, the rest being neutralized or serving to frame the action. Thanks to the diminutive scale, the comparative absence of figure movement (the characters walk slowly or stand in place), and a lack of information about the characters (reduced in size, seen from the rear, often in three-quarter view), the image goes very still. It becomes a dead space that invites the viewer to linger over previous developments, to wait for something new to modify the stasis, or to simply contemplate vacancy. Antonioni had suggested the possibilities of such compositions (Fig. 4.36), but Angelopoulos carried the idea much further. Reconstruction makes eloquent use of the vacant shot (Fig. 4.2), and The Travelling Players is filled with such images, most notably the bookend shots of the troupe’s arrival at the film’s beginning and end (Figs. 4.37–4.38). In The Beekeeper the empty shot is often used to present Alexander tending his hives (Fig. 4.39). A more stripped-down version occurs in Landscape in the Mist, when the two runaway children part from Orestes on the motorway (Fig. 4.40). The sparseness of the frame aªects the pacing of the long shot. If the viewer is fully to take in such distant images, they must be held on the screen for some time. Yet by refusing to pack them with movement, Angelopoulos brakes the dramatic rhythm. Searching

164

A N G E L O P O U L O S , O R M E L A N C H O LY

4.36 Red Desert: three-quarter dorsality and the spaces between people. Compare Figs. 4.2 and 4.16. 4.37 The Travelling Players: the first shot, set in 1952.

4.38 The Travelling Players: the final shot, with the players starting their journey in 1939.

4.39 The Beekeeper.

4.40 Landscape in the Mist (1988): Angelopoulos’s high angle sets the figures in a wedge of unfilled space.

A N G E L O P O U L O S , O R M E L A N C H O LY

165

4.41 Alexander the Great: The communards surrender.

for gestures and telltale alignments of figures, the viewer must test a variety of ways to fill such hollow shots with meaning. The result, beyond the self-conscious pictorialism, is a suspension of dramatic progression that allows both detached contemplation and a sense of dry, understated emotion. In a distant shot communards under Alexander the Great’s command surrender their weapons. We have no access to their facial expressions, but we can read their defeat from their weary loping to obey his order (Fig. 4.41). Tactics of centering and aperture framing highlight the (already attenuated) story action while three-quarter dorsality and spare settings deflect us from it. Angelopoulos does not, however, stop with this dynamic. He goes on to “preform” his figures, interiors, and landscapes by means of a consistent set of dedramatized staging options. Here again we can see his synthesizing role. None of these schemas is unique to him, but he concentrates so single-mindedly on them and explores them so imaginatively that they have come to form his signature, stamping every film as “by Angelopoulos.” They give each film a theme-and-variations structure; one image will come to be seen as a return to or deviation from another image shown earlier in the film or seen elsewhere in the oeuvre. And these types of images can be combined, within a single sequence or even within a single shot, to maximize their visual and dramatic possibilities. Angelopoulos revises and explores the resources of two predominant image schemas. (See box.) Distanced recessiveness, emerging in Europe during the 1950s, continued to be a salient option throughout the 1970s, and Angelopoulos seized on it as a central staging tactic. He was able to fine-tune it by combining it with the empty frame, three-quarter dorsality, and strongly perspectival settings. When he organizes a recessive shot around central perspective, as in interiors or on city streets or roadways, the vanishing point can pull our attention to important action. The rape of Voula in Landscape in the Mist becomes all the more shocking because the diagonal plunge of the truck and roadway keeps forcing our eye back to the truck driver’s pursuit (Fig. 4.50). But central perspective is seldom

166

A N G E L O P O U L O S , O R M E L A N C H O LY

TWO WAYS TO DEDRAMATIZE FILMIC SPACE Between the late 1950s and the early 1970s European filmmakers explored image schemas that could mute the flow of drama within a scene. Both schemas also constituted reactions against the deep-focus style that had emerged in American cinema in the 1940s. One sort of image makes use of what the art historian Heinrich Wöl›in called “recessive” space.44 Here figures and architectural spaces present diagonals that shoot from foreground to background. Directors from the 1940s to the 1960s had explored a deep-focus variant of the recessive image by placing the foreground plane quite close to the camera. Welles made this device part of his stylistic signature, and many directors around the world embraced the plunging deep-focus look (Fig. 4.42). The neorealists and Antonioni, however, often presented less-aggressive depth, setting the foreground plane a fair distance from the camera, as Japanese directors of the 1930s had. Wellesian depth hyperdramatizes the action by thrusting one plane out at us, whereas the milder recessiveness aids dedramatization, especially when combined with dorsality and vacant areas of the shot (Fig. 4.43). This image could make characters’ expressions and relationships less easy to read at one glance, and by avoiding a looming face or prop, this compositional device gave landscape and architecture more prominence (Fig. 4.44). During the 1960s, recessive, wide-angle depth was contested by another strain of staging and shooting that, again following Wöl›in, I will call “planimetric.”45 Here the background is resolutely perpendicular to the lens axis, and the figures stand fully frontal, in profile, or with their backs directly toward us. We can find early examples—even in Antonioni (Fig. 4.45)—but the technique developed most evidently within the “new waves” of the late 1950s. Young directors gravitated toward a less volumetric, more self-consciously “modernist” image—flatter, obviously constructed, sometimes posing disconcerting optical puzzles (Fig. 4.46). Political modernists quickly picked up on the device: Godard’s to-camera monologues in La Chinoise (1967) became prototypes (Fig. 4.47). The new planimetric image was particularly suited to that ascetic strain of political modernism. It is central to Jeanne Dielmann (Fig. 4.22 above), as well as to Fassbinder’s most minimalist exercise, Katzelmacher (Fig. 4.48).

4.42 A deep-focus close view from the Egyptian film The Land (Youssef Chahine, 1969).

4.43 A zigzag use of distant depth in Luchino Visconti’s La terra trema (1948).

A N G E L O P O U L O S , O R M E L A N C H O LY

167

4.44 I Vinti (Antonioni, 1952): urban landscape and distant figures.

4.46 Reflections in a coªee-bar counter play on illusions of flatness and depth (Identification Marks: None, Jerzy Skolimowski, 1965).

4.48 Katzelmacher (R. W. Fassbinder, 1969): a planimetric lineup of loafers.

4.45 A planimetric framing of one protagonist in I Vinti.

4.47 The planimetric mug-shot framing in Godard’s Pierrot le fou (1965).

4.49 The long lens reduces depth and makes distant planes similar in size (I Even Met Happy Gypsies, Alexander Petrovid, 1967).

The sources of this schema are various. Most proximately, during the 1960s, when shooting on location became more common, filmmakers used long (“telephoto”) lenses much more frequently, and these narrowed the playing area and flattened space into parallel planes (Fig. 4.49). Once these shots became common, many directors began elaborating images that looked more shallow and perpendicular. Greenbergian conceptions of modernism as “asserting the picture plane” and “denying representational depth,” disseminated through high culture in the 1960s and 1970s, may also have steered some filmmakers to these more abstract compositions.

4.50 Landscape in the Mist: the truck driver’s rape of Voula, with the vanishing point giving the event a grim inevitability.

4.51 The Travelling Players: a monologue is rehearsed.

4.52 The Suspended Step of the Stork (1991): empty space and interrupted gesture.

this deterministic in Angelopoulos. He often creates a tension between a central vanishing point and a foreground, oª-center figure. This eªect is marked in the declamations of The Travelling Players (Fig. 4.51), as well as in the handling of the border station in The Suspended Step of the Stork (Fig. 4.52). Typically, the perspective is not so neatly centered. The endless corridors of the prison in Days of ’36 and of the hotel in The Hunters remind us that Angelopoulos can evoke a more oblique sort of depth. He uses recessional perspective to diminish our view of salient information, further dedramatizing the scene and obliging us to concentrate on very distant clues to the plot’s progress. The spectator must strain to follow the tiniest wrinkles in the action. A bold example occurs in The Hunters. Yorgos is sitting in a tavern, his back to us, and two burly men come to sit before him in the foreground (Fig. 4.53). One complains that the leftists are planning a coup. Angry, Yorgos rises and starts to sing a left-wing protest song as he threads his way into the distance (Fig. 4.54). The camera arcs left, keeping him more or less centered as he uncertainly finishes the song, his back still to us, caught

A N G E L O P O U L O S , O R M E L A N C H O LY

169

4.53 The Hunters (1977): Yorgos confronted by thugs.

4.54 Yorgos retreats, singing.

4.55 Yorgos in the distance, still singing.

in a slim aperture (Fig. 4.55). If the scene had been shot from the opposite angle, Yorgos would advance to the camera, and this, along with his defiant singing, could make him heroic. Here, his shrinking to a tiny silhouette and the weakening of his voice undercut his commitment, foreshadowing his eventual capitulation to the right-wing forces. Pictorially, the composition engages the viewer in a steadily intensifying act of visual concentration on a man who will become little more than a speck. Here the lighting and central placement highlight Yorgos, but Angelopoulos also uses architecture to shape the frame space. Recessive diagonals can slice a locale into sharply separated zones. Antonioni explored this option occasionally, but from the earliest scenes of Reconstruction Angelopoulos made it central to his handling of “empty” shots (Fig. 4.56). Assigning cells of space to characters is easily thematized as expressing division and incommunicability, but I think that the perceptual dynamics of such shots are more important, especially in the context of Angelopoulos’s distant framings. If foreground areas can be kept relatively bare, stretching to the middle ground, then far-oª gates, doors, or street corners can mark distant incidents as important (Fig. 4.57). Within interiors Angelopoulos frequently creates recessive perspectives around door-

170

A N G E L O P O U L O S , O R M E L A N C H O LY

4.56 Diagonals dividing up frame space in Reconstruction.

4.58 After doorways frame the police interrogation of Eleni (4.1), similar compositions are used in the lover’s confession . . .

4.57 As the wedding guests of The Beekeeper leave the ceremony in a characteristically open landscape shot, the wedding couple comes out in the far distance, framed in the gateway.

4.59 . . . and in the couple’s reenactment of the murder (Reconstruction).

ways. In Reconstruction a recurring use of the household’s doorway measures the progress of the action (Figs. 4.1, 4.58, 4.59). And Days of ’36 treats certain scenes as permutations of recessional views through apertures. The authorities constantly return to the cell in which Sofianos holds Kriezis hostage, allowing us the narrowest sliver of access (Figs. 4.60–4.62). This obliqueness is motivated as well by the consistently external standpoint the narration presents; we are always attached to the authorities outside the cell and are never allowed to go inside with the prisoner and his hostage. Spatial obliquity becomes narrational opacity. The zigzag scanning demanded by such recessional shots recalls the wide-angle depth associated with 1940s trends, but Angelopoulos, starting his career after Antonioni, Jancsó, and others had developed a more distanced recessiveness, knows the advantages of setting the nearest plane some distance from us. In modern European cinema the empty foreground often signals an aesthetic distance between the viewer and the characters’ sit-

A N G E L O P O U L O S , O R M E L A N C H O LY

171

4.60 Days of ’36.

4.61 Days of ’36.

4.62 Days of ’36.

uations. At first glance this strategy of medium-shot depth would seem to recall the schemas employed by Feuillade and his contemporaries, but it actually inverts the dramatic scale of values implicit in their shots. In the 1910s the rising action of a scene pulled the actors toward the camera, with greater drama occurring in the closest planes. Angelopoulos, in flagrant defiance of audience empathy, often makes his characters retreat from us, as when Yorgos leaves the table (Figs. 4.53–4.55), and the scene’s climax may be played out at the points farthest from the foreground. The distanced, recessive frame is central to Angelopoulos’s mise-en-scène from the start, but after Days of ’36 the planimetric composition starts to develop as an alternative. Throughout The Travelling Players characters are ranged in friezelike compositions, either within the theater (Fig. 4.63) or within landscapes (Fig. 4.64). Several critics have pointed out that this device turns many locales into Brechtian tableaux, paralleling historical events with the performance of the players,46 but we should also note that Angelopoulos is adapting a compositional schema that was already circulating in his milieu (see box). His subsequent films employ the planimetric image to combine both frontal

172

A N G E L O P O U L O S , O R M E L A N C H O LY

4.63 The Travelling Players: a lateral performance arrangement.

4.64 The Travelling Players: the long lens turns a landscape into horizontal bands.

and profile views or to create muted moments through dorsality. He striates landscapes by spreading ribbons of figures parallel to the horizon, as in the breathtaking wedding ceremony at the riverside in The Suspended Step of the Stork (Figs. 4.65–4.67). In addition, the “clothesline” eªect of the planimetric image seems to encourage his camera to pan or track in order to explore all those parallel vectors running through the shot. And apertures play no less important a role in the planimetric image than in the recessive one. In Landscape in the Mist, when Voula stands at the station doorway and speaks of darkness and light, railway track workers riding on their handcar are first seen as spectral blobs of yellow in the right window pane before they coast past the doorway (Plates 4.5–4.6). From first to last, the planimetric imagery of Eternity and a Day slips toward pictorial abstraction, perhaps evoking the dying poet’s yearning for a spiritual “beyond” in memory or in death. The opening, with the boy Alexander called outside, follows him from his room to a staircase; the camera lingers on a tall window (Fig. 4.68). Via a cut, the window frame gives way to another straight-on rectangle, the silhouette of the family’s beach pavilion as the boy rushes in from the front (Fig. 4.69). As he is joined by friends, they race to the surf, the camera tracking forward and zooming in to create a planimetric image (Fig. 4.70). Over the film, Angelopoulos turns streets, a construction site, even a massive ship into planimetric surfaces. Within a single shot, a quasi-documentary image of Alexander walking along the harbor (Fig. 4.71) becomes planimetric as he passes (Fig. 4.72) and turns into pure abstraction (Fig. 4.73). Again and again, through Alexander’s memories we return to the villa and beach of his childhood and marriage (Figs. 4.74–4.75), before the film’s final shot pins him to the horizon (Fig. 4.76). Although the recessional image and the planimetric one may seem to be rivals, Angelopoulos exploits both. In his work they form a pair of schemas that can be revised from film to film. We instantly recognize the Angelopoulos café (recessional, with oblique depth), the Angelopoulos roadside (recessional, often with central perspective), the Angelopoulos beach or riverside (planimetric, with clotheslined figures), and so on. Further, a

A N G E L O P O U L O S , O R M E L A N C H O LY

173

4.65 Suspended Step of the Stork: a couple separated by a border is about to marry. The bride’s retinue on one side of the river . . .

4.66 . . . faces the groom arriving on the other side.

4.67 Another planar composition shows the wedding ceremony.

4.68 Eternity and a Day.

4.69 Eternity and a Day.

4.70 Eternity and a Day.

4.71 Eternity and a Day.

4.72 Eternity and a Day.

174

A N G E L O P O U L O S , O R M E L A N C H O LY

4.73 Eternity and a Day.

4.74 Eternity and a Day.

4.75 Eternity and a Day.

4.76 Eternity and a Day.

particular image may combine the two options through camera movement or staging (Figs. 4.77–4.78). The stunning finale of Suspended Step shows the reporter starting oª along the river as the linemen ascend the poles (Fig. 4.79). But then the camera arcs rightward to create a monumentally planimetric tableau as the reporter pauses at the river and the men pull the lines taut (Fig. 4.80). A sharp diagonal has become a laminated landscape. With some sense of Angelopoulos’s salient principles of visual design we can reevaluate our first impressions about his technique. The long take can now be seen as creating, sustaining, or blending empty shots and three-quarter views, recessional and planimetric compositions. Even when the long take does not merge two time periods within the shot, as in The Travelling Players and The Hunters, it also serves to confine us to certain sorts of images; very soon we learn that we will not get cut-in close-ups or shot/reverse shots; we must work with the comparatively distant view. Similarly, the temps morts compel us to concentrate on the gradual changes that constitute the action. Yet the very scale of what counts as action in these frames poses a problem. The shots help tell the story but so slowly and sparsely that they outrun standard conceptions of narrative economy. Our next impulse is to interpret them, but that path is usually no more rewarding, since the flagrant lack of narrative economy in the images is matched by a lack of symbolic richness as well. (Hence the objection voiced by one nonadmirer: “O.K.,

A N G E L O P O U L O S , O R M E L A N C H O LY

175

4.77 One shot in The Hunters begins as a horizontal and frontal setup.

4.78 As Yannis moves, however, an arcing camera creates a recessive composition that puts the figures in three-quarter back views.

4.79 Suspended Step of the Stork.

4.80 Suspended Step of the Stork.

we got it, Theo, now turn it oª!”)47 The protracted image outruns its denotative and symbolic aspects, striving instead for both emotional expressivity and formal abstraction. On one hand this is “mood cinema,” like mood music: the story is amplified not only through the threnodic score but also through the skies, the landscapes, and the bearing of the humans dotted around the frame. On the other hand the fastidious compositions and stringent vectors of movement ask to be apprehended as monumental pattern. Lenin’s head, lashed to the barge drifting down the Danube (Plate 4.1), passing people kneeling and crossing themselves on shore, soon stops being a symbol of the death of Communism and forms an occasion for grave emotion, mixing sorrow and a guarded exhilaration. The throb of the barge’s engine is overwhelmed by Eleni Karaindrou’s briskly pulsating score, setting a mournful viola and other solo instruments against massed and rising strings. The alternating shots of the barge and the watchers on shore are cut in time to the variations worked on the brightening, dancelike melody, as if this were a weirdly spare music video.48 After nearly three and a half minutes the statue has become a massive chunk of the landscape, a drifting iceberg counterpointed to the enthralled people on the bank, themselves turned into a stream of planimetric bands. Mizoguchi’s films kept nar-

176

A N G E L O P O U L O S , O R M E L A N C H O LY

rative denotation, expressivity, symbolic meaning, and pictorial abstraction in balance, but Angelopoulos has reshifted the image’s weight. He sacrifices storytelling and moves beyond symbolic purport in order to conjure up a bleakly expressive cinema that also lays claim to overpowering beauty. The commitment to long takes, distant views, sparse frames, and temps morts stakes everything on the unfolding image. Since cuts are unlikely to come, we must wait for the shot to reveal its mysteries at its own pace. The slackened rhythm allows us to form rather explicit expectations about how the staging will develop. Feuillade and his contemporaries had no time for such niceties; they left such expectations tacit, exploiting them for the sake of fluid visual narration. Mizoguchi balanced compelling plot situations with shots that might tease us with fastidious changes in design. Angelopoulos is yet more overt, announcing himself as an impresario of vision. He will show us everything, but in good time, and at a tempo that drains normal dramatic momentum out of the scene. The shot will pursue its own spatial trajectory. During the final battle for the village in Alexander the Great, Alexander’s son is struck by a bullet as he runs for cover; he is rescued by a housewife (Figs. 4.81–4.87). His pain and her panic are played out far from the camera, and his escape is further dedramatized by taking place in the most distant reaches of the shot. Here the spatial arc does not match the dramatic one: each rising phase of the action takes place farther and farther away. Endowing the long take with its own spatial structure often leads Angelopoulos to build the shot toward a purely pictorial climax. That may be a sublime landscape or a grand enigmatic emblem, like the gigantic stone hand lifted mysteriously from the bay in Landscape in the Mist (Fig. 4.88). Or the climax may be a leap into pure abstraction, as we have already seen in the movement toward planimetric imagery in Eternity and a Day. Sometimes the pictorial epiphany will come at the end of the shot, sometimes at a point that permits a diminuendo into something more prosaic. In any case the visual climax may reinforce a dramatic climax, or even create one. Camera movement, present in almost every long take that Angelopoulos makes, provides help in creating pictorial epiphanies. Usually the camera follows characters in motion, sometimes picking up one while dropping another. This tactic allows Angelopoulos to keep the shot alive, to sustain visual interest while also linking or developing his characteristic compositions, as in the Voyage to Cythera and Suspended Step instances already mentioned (Figs. 4.33–4.35, 4.79–4.80). But his most common camera movement derives from a very old schema that he has revised for his own purposes. Recall the scene from The Travelling Players discussed at the outset of this chapter. The troupe walks to town from the waterfront, moving diagonally toward us, the camera tracking to keep them at a three-quarter angle (Fig. 4.4). Then the camera slows to let them pass rightward, panning to keep them in profile (Fig. 4.5). The camera slows still more, and the actors walk diagonally into the distance, turned from us in a three-quarter view complementary to that seen at the start of the shot. Some earlier filmmakers used this angular pickup/profiled passing/angular walkaway pattern, most commonly in scenes shot on

A N G E L O P O U L O S , O R M E L A N C H O LY

177

4.81 Alexander the Great: the camera picks up Alexander’s son along a diagonal . . .

4.82 . . . and pans right with him.

4.83 Another director would have staged the boy’s fall while he was close to the camera, but instead the young Alexander drops in the middle distance, along another diagonal.

4.84 The woman in the background comes to cradle him as he revives.

location (for example, Mizoguchi’s Sisters of the Gion, 1936), although closer versions are feasible in large sets (for example, Renoir’s Rules of the Game, 1939).49 This schema is often seen in postwar neorealist movies and in Hollywood films shot on location. (Preminger is especially fond of it in Exodus, 1960, and Advise and Consent, 1962). This come-and-go pattern allows Angelopoulos to structure his scenes around a visual arc that may or may not synchronize with the conventional dramatic one. The simplest possibility is to use the diagonal-approach phase of the shot to establish the main players in the scene and to save the visual revelation for the field revealed at the end of the panning movement. This is more or less what happens in the boy’s wounding in Alexander the Great (Figs. 4.81–4.87) and the time-shifting stroll in The Travelling Players (Figs. 4.4–4.10). The diagonal approach/diagonal withdrawal schema can also generate a certain suspense: we see characters who are looking at something oªscreen, and the camera pans with them as they move toward something revelatory, a visual/dramatic climax. A good example occurs in Ulysses’ Gaze, when the protagonist, A, follows a woman

178

A N G E L O P O U L O S , O R M E L A N C H O LY

4.85 After going inside and returning, she binds his wound . . .

4.86 . . . carries him to the patiently waiting horse . . .

4.87 . . . and shoos it away at the end of the lane.

4.88 Landscape in the Mist.

down a street (Fig. 4.89). The camera pans to let the two pass then picks them up on the opposite diagonal, now revealing a vast crowd bearing candles (Fig. 4.90). Here, as so often, the last phase of the shot shows space on a grand scale. The climax is more visual than dramatic, or, rather, by being a visual climax it can create a dramatic one, even though we may be uncertain about exactly what plot action is taking place. In this shot from Ulysses’ Gaze Angelopoulos piles on the complications, with police pouring into the midground as the camera cranes up (Fig. 4.91) and then a crowd of umbrellas pressing into the foreground (Fig. 4.92). After everything freezes into a tableau, the crowd rushes the police. The shot ends at the height of visual and dramatic intensity. So the final phase of the diagonal approach/profiled pan/diagonal retreat schema can provide a crescendo when something massive is revealed. Just as often, though, Angelopoulos reserves the depths of the shot for developments on a somewhat smaller scale. Take the scene in Voyage to Cythera in which Katerina oªers to coax the recalcitrant Spyros out of their home. The sequence starts with her, their son Alexander, and the town

A N G E L O P O U L O S , O R M E L A N C H O LY

179

4.89 Ulysses’ Gaze (1995).

4.90 Ulysses’ Gaze.

4.91 Ulysses’ Gaze.

4.92 Ulysses’ Gaze.

o‹cials approaching the cottage (Fig. 4.93). She passes fairly near the camera, but the closest moment of facial expression in the scene is presented not as the climax but as preparation for it (Fig. 4.94), leading from a three-quarter dorsal view (Fig. 4.95) to her arrival at the door of the cottage, to meet Spyros (Fig. 4.96). “Always the same,” says Katerina. “When you’re afraid, you hide.” After another moment they embrace. At this great distance from the camera, seen through the spindly gate work, Spyros accepts leaving his home for permanent exile. Whereas the 1910s director would stage this climax in the closest plane, Angelopoulos pushes the actors very far back, throwing dramatic structure out of gear with spatial patterning. The shot, however, is not over. Turning, the old couple return to the gate to stare at the men oªscreen (Fig. 4.97). After Spyros says, “Withered apple” (a phrase he uses to evoke both his lost lands and his relation to Katerina), the twominute shot ends. This is another way to round oª the action—bringing figures back to the middle ground as a diminuendo. In shots such as this, character movement and camera movement participate in a larger dynamic of opening and filling space at a tempo that allows us to anticipate how the blocking will develop. Thanks to the long take, the muted action, and dead intervals, Angelopoulos prolongs the process of staging, leaving us plenty of time to recognize that we are forming expectations about where the character or camera will go next. In this respect Angelopoulos slows down 1950s Antonioni.

180

A N G E L O P O U L O S , O R M E L A N C H O LY

4.93 Voyage to Cythera: Katerina comes to the camera in a diagonal medium shot, the camera tracking back to preserve the angle.

4.94 The camera slows to let her pass in profile.

4.95 Katerina continues past us, and the camera tracks her to the gate in a three-quarter rear view.

4.96 She goes diagonally to the cottage door (near the center of the frame), saying, “It’s me.” After a long pause the door unlatches and Spyros comes out.

4.97 The shot ends with the old couple in the middle ground.

A N G E L O P O U L O S , O R M E L A N C H O LY

181

4.98 Suspended Step of the Stork.

4.99 Suspended Step of the Stork.

4.100 Suspended Step of the Stork.

Since the early years of cinema, one prominent staging schema has relied on filling in the blanks. As we’ve seen in both Feuillade and Mizoguchi, the director may leave a vacant spot in the frame and then bring an actor into it, calling attention to the new figure while also completing the composition. Angelopoulos lays this technique bare by overtly creating gaps that he will then slowly fill, delaying the moment when the shot clicks into place. In a café scene in Suspended Step the reporter meets the young bride. The camera follows him striding toward her through the crowd (Fig. 4.98), but suddenly (and implausibly) a wide stretch of the dance floor opens to allow us an unimpeded view of her in the background (Fig. 4.99). We are given several moments to speculate on who will move next, but then the camera edges slightly leftward, clearing still more room for the couple; eventually the bride rises and comes to the dead center of the frame (Fig. 4.100). Even without camera movement, simply filling a gap can develop and climax the long take. We’ve already seen this happen in the blank foregrounds of the street in Ulysses’ Gaze, filled first by police and then, more surprisingly, by the ranks of umbrellas (Figs. 4.90–4.92). In one episode of The Hunters the police fan out across a street but charitably leave a slot that will serve as a window through which we can watch the approach of the peace demonstrators (Fig. 4.101), the arrival of the car that will snatch Yannis (Fig. 4.102), and then the aftermath of the clash, with the police dispersing and the dead victim lying in the street at the center of the shot’s perspective (Fig. 4.103). It is as if the en-

182

A N G E L O P O U L O S , O R M E L A N C H O LY

4.101 The Hunters: aperture framing with masses of people.

4.102 The Hunters.

4.103 The Hunters.

tire arc of the action were known in advance. Watch this space, the shot implies from the outset, and eventually you will be rewarded. We study the distant view, we probe those figures turned from us, we scan the horizontal stretch or wait for action at a vanishing point, we form expectations about the trajectory of the camera or the characters. Our patience is often rewarded at the end of a shot. The visual climax may be somewhat striking, as in this Hunters scene; it may be nothing short of stunning, as in linemen stretched out like notes on a staª (Suspended Step, Fig. 4.80) or a gigantic hand rising from the bay to float over the city (Landscape in the Mist, Fig. 4.88). It may also be quite muted, as when men in yellow slickers slide ghostlike through the rain, or when the boy leaving Alexander’s car for an uncertain future rushes past the taillights and his face turns brilliant red for an instant (Plates 4.7–4.8). In all cases the strategy of building a long take to a moment of heightened pictorial expressivity, in the context of a drama that mu›es the sort of emotional investment solicited by mainstream storytelling, has its source in a modernist aesthetic. Angelopoulos has blended salient techniques of two traditions of European cinema—Antonioni’s de-

A N G E L O P O U L O S , O R M E L A N C H O LY

183

dramatization and austere political modernism—in order to sustain a refined awareness of the very act of apprehending the scene, an experience crucial to the tradition of cinematic modernism. His shots, then, curiously halt the forward time of dramatic propulsion; narrative stasis becomes the norm, against which moments of splendor stand out. The plot suspends itself, like the reporter’s leg at the border, for the sake of pictorial and expressive impact— often on a stupendous scale. The eªect arises not only from a canny play with history and memory but also from fresh revisions of staging principles stretching back through the 1950s to the first years of cinema.

To what broader ends? Granting the two-phase periodization of Angelopoulos’s work, we can note that in the films before Voyage to Cythera all the techniques we have examined accord with the goals of a “political modernism.” By concentrating on groups and by staging in a manner at once minimalist and monumental, Angelopoulos blocks traditional paths to empathy. This yields a critical detachment that in turn invites us to reflect on the larger historical forces at work in a situation. The later work is more concentrated on the individual, delineating personal crises that express, as the emigrant author of Suspended Step puts it, fin de siècle melancholy. This tendency toward portraying psychological crises can be seen as well in Angelopoulos’s new reliance on major stars, on nondiegetic music, and on a more self-conscious spectacle. We might consider him to be oªering us an updated version of “Antoniennui.” Both phases exploit slowness of rhythm, the vagaries of landscape, and sparseness of composition. I think that the two-part periodization holds good, but it is worth pointing out that Angelopoulos’s later work has not relinquished political critique. He tackles concrete political issues of his moment: the death of Marxism (Voyage to Cythera, Ulysses’ Gaze), the desperate flow of populations across European borders (Landscape in the Mist, Suspended Step), and the possibility of a national politics based on honor (Suspended Step). One can hardly imagine the Antonioni of the 1950s making a film about Algeria or the Berlin Wall, but Angelopoulos unswervingly bears witness to political crisis. Conversely, even in his “Brechtian” days he gave his work a strong emotional overlay. Jean-Marie Straub and Danièle Huillet, crucially influenced by Bresson, achieve a discomfiting emotional neutrality throughout their work, but Angelopoulos’s 1970s films, operating under the aegis of Antonioni, are soaked in feelings of solitude, bleakness, and futility. In both Straub/Huillet’s Not Reconciled (1965) and The Travelling Players, cycles of national history are fractured into a complicated flashback structure, but the dedramatization is achieved in radically diªerent ways. Straub and Huillet rely on disorienting cuts, closeups of nonactors sti›y reciting monologues, and temps morts that refuse to be read as expressing an atmosphere. Angelopoulos presents long takes, distant declamations, and stillnesses pervaded by oppression and muted despair. Across his career certain emphases have shifted, but he has always combined social

184

A N G E L O P O U L O S , O R M E L A N C H O LY

critique—grounded at first in Marxism, now in humanist ethics—with a tone that, he claims, is not exactly pessimistic. “I am a melancholic. And according to Aristotle, melancholy is the source of the creative spirit.”50 He sometimes attributes this tone to the lost 1960s dream of social revolution or to the insularity of most people today, cut oª from each other and history.51 But he also suggests that his melancholy has local sources: “We live in a country of memories, of aged stone, of broken statues.”52 Angelopoulos’s achievement shows how a filmmaking tradition can renew itself in the contemporary moment. He has pushed certain insights of Antonioni to new limits, encouraged by the achievements of others (Tarkovsky, Jancsó, the minimalist radicals) and by broader aesthetic currents (Brechtianism particularly). Similarly, he has blended distinct stylistic schemas available in his milieu—the three-quarter view, the come-andgo pan, the deeply perspectival shot, and the planimetric framing—into an expressive whole. That he has won such renown suggests that for many viewers the postwar tradition has not exhausted itself, that it can endow our world of snack bars, video clips, and ethnic wars with a severe, contemplative beauty. At a moment when European cinema, both popular and elitist, seems to be aping Hollywood, Angelopoulos’s work takes on unique importance. Sometimes mannered, often majestic, almost always melancholy, Angelopoulos demonstrates that, contrary to what the prophets of postmodernity keep telling us, cinematic modernism can still open our eyes.

A N G E L O P O U L O S , O R M E L A N C H O LY

185

5 HOU, OR CONSTRAINTS

Chiu-fen, Taiwan, late 1945. Freed from fifty years of Japanese rule, a few leftist intellectuals have begun meeting in the studio of a deaf-mute photographer. Tonight they sit around a restaurant table with a mainland reporter. After some jokes he asks how they feel living under the Chinese flag again. The mainland-appointed governor, Chen Yi, is a bandit, one man remarks; but he is quickly silenced. Outside, a voice is heard singing the “Song of the Exiles.” The men join in, opening the window onto a vast night sky. As they sing of years of wandering and the longing to return home, they snack on kebabs the photographer has brought in. Thematically weighty as it is, this scene from Hou Hsiao-hsien’s City of Sadness (1989) also poses a problem of craft we’ve met before. How to stage, shoot, and cut action taking place around a dinner table? Hou’s solution is striking. He arranges his players in horizontal layers and flanking profiles, making the most important figures centered, frontal, and brightly lit. He avoids the Last Supper eªect by seating other actors in the shadowy foreground, their backs to us, spaced out to frame the central players (Fig. 5.1). He may cut in to a single speaker (as he does once in this scene, to underscore the moment when Hinoe criticizes Governor Chen), but on the whole our attention is directed to the key participants through composition, lighting, and discreet blocking and revealing. When one intellectual tells a joke about the Taiwanese flag, the foreground figures lean casually aside to allow a full view of him and the listening reporter (Fig. 5.1). When the reporter wants to know what the song is, a foreground figure blocks the first speaker so that the reporter commands the central zone (Fig. 5.2). This eªect comes oª as all the

186

HOU, OR CONSTRAINTS

5.1 City of Sadness (1989): The intellectuals gather, and the composition centers the local man on the left and the visiting mainlander on the right.

5.2 The local man resettles himself, favoring the mainlander as he gives his reply.

5.3 Later in the scene, while Wen-ching is out getting snacks, the men join in the singing outside.

5.4 As they rise to go to the window, Wen-ching returns on frame right, and Hinomi looks up to him.

more risky given the scale of the framing—a long shot, taken with a long lens that squeezes together minor aspects of setting. (During 35 mm projection you can examine the rice and vegetables on the table.) As the action develops, Hou throws the emphasis onto the song, and indeed the future of Taiwan, by having his actors rise and gather at the window. The scene becomes an emblematic tableau, with the window opening this little room onto the uncertain horizon of a nation’s future. Standing in patches of darkness, the men become mere silhouettes, as if melting into the unseen population outside (Fig. 5.3). The scene beautifully balances symbolic implication, expressive qualities (the quiet elation of camaraderie between mainlander and islander, intellectuals and populace, united in spontaneous song), and subtleties of denotation: the central characters of the story sit silent on the margin here. The deaf-mute Wen-ching and the woman Hinomi, on frame right, bear witness to the political drama, as they will throughout; yet already they exchange glances, prefiguring the love they will come to share (Fig. 5.4). The delicacy of this passage from City of Sadness can be appreciated if we compare it to a dinner scene in another Chinese film released only a year later (and produced by Hou). In Raise the Red Lantern (1990) Zhang Yimou solves the dinner-table problem in a less original way. He cuts among singles of the women dining, and in the master shot the bodies are not so tightly packed and the interplay of faces nothing like as intricate.

HOU, OR CONSTRAINTS

187

5.5 Zhang Yimou revives the classic dinner-table schema (Raise the Red Lantern, 1990).

5.6 After a distant establishing shot, Zhang builds the rest of the scene out of tight singles. Compare Figs. 1.1–1.3.

5.7 In Early Summer (1951) Ozu presents the family dining, using the wife’s head to first call attention to the grandmother . . .

5.8 . . . and then block her and favor the older brother at center as he speaks. This image functions as an establishing shot; the dialogue is largely carried in single shots of the speakers.

Indeed, the layout and découpage recall classic US silent film (Figs. 5.5–5.6). For a betterfitting precedent perhaps we have to go back to Ozu Yasujiro, whose images display some of the density found in Hou’s scene (Fig. 5.7–5.8). Yet Ozu’s jammed master shot is merely a prelude to a string of strict close-ups as the dialogue shifts around the table. Antonioni laid out schemas that Angelopoulos developed; now Ozu sketches a possibility that Hou elaborates by prolonging and mining the full shot. Another Japanese director, Kitano Takeshi, has remarked that he wishes he could film eating scenes as well as Hou does.1 Hou makes no films without them, and his interest in how people behave while wolfing down noodles or rice typifies his commitment to the mundane, the concrete, and the sensuous.2 His is a cinema of mundane detail, of almost finicky attention to small-scale matter. Audiences invariably recoil from the sons’ discovery of their dead grandmother at the end of A Time to Live, a Time to Die (1985). Her underside has begun to rot, and it is characteristic of Hou to devote the closest shot in the film to the sticky, dark stain on her matting. Even his landscapes do not have the allegorical

188

HOU, OR CONSTRAINTS

sweep we find in Angelopoulos; embedded within them are bits of life proceeding quite uncosmically—often just people walking along a winding road, plucking at their shirts or fanning themselves. The camera’s angle does not ennoble the traveler, as it does the allegorical troupe of The Travelling Players or even the tenacious children of Landscape in the Mist. No parades or demonstrations, mass meetings or defiant facedowns with authority; no spectacular executions; no onscreen confrontations with Power. Instead, we get casual-seeming minutiae. In one steep long shot of Son’s Big Doll (1983) an elated father who has finally found a job runs home so fast his sandals fly oª, becoming orange dashes against gray streets and ripe tropical greenery. When the heroine of Daughter of the Nile (1987) shrieks on awakening from a nightmare, the fish in her aquarium, sequestered in a corner of the frame, practically jump out of the water. In this world people mostly smoke, walk, watch, wait, ponder, shoot pool, drink, and eat. Hou’s life and tastes did not prepare him for the abstract ideas or political concerns pursued by Angelopoulos. He was a cinéphile, but his tastes favored mass-market movies from Japan and the West. Many shapers of the New Taiwanese Cinema had studied in the United States and assimilated the masterpieces of European and Asian cinema, but Hou, who did not travel abroad until the 1980s, was led to Pasolini, Godard, and Antonioni by friends. He did not know Mizoguchi, and his first viewing of Ozu put him to sleep. Although he is not above an occasional movie citation, he usually takes his sources from local life. The gangster ingredients that enliven his plots come not from the screen but from nostalgia for his shady youth: “I miss the old macho world, a world of competing mad dogs.”3 Eleni Karaindrou provides mournfully minimalist scores for Angelopoulos, but Hou’s sound-track albums mix sentimental love tunes, Japanese-flavored ballads, hard rock, MOR pop, and plaintive synthesizer Orientalism.4 One cannot imagine Angelopoulos cutting karaoke CDs, as Hou has, or singing soulfully while working over his scripts. His modest technique suits an examination of the rhythms of everyday life. Hou has always been interested in how people grow up, break away from family, move to the city, find a job and a mate, get money legally or illegally, and have fun. His characters, like Wong Kar-wai’s twentysomethings or the young wastrels in Fellini’s I Vitelloni (1953), are often merely hanging out, dining and drinking, quarreling and tussling, observing dramas playing out around them, trying to avoid parental harangues or the boss’s wrath, using their downtime to devise clumsy scams or to consummate a courtship. (True, the refined brothel society of Flowers of Shanghai, 1998, avoids this working-class world, but there dandies play out at a higher level the dawdling meals and boisterous drinking games, the teasing and misjudgments typical of all Hou’s characters.) Yet while they are chilling out, these youths are living through history. Taken as a whole, Hou’s films follow Taiwan through the twentieth century, from the early 1900s to the turn of the millennium (itself recalled from 2010 in Millennium Mambo, 2001). The characters may play pool, horse around on the beach, cut petty deals, serve their time in the military, or dance at a club, but oªscreen, politics and society are in turmoil, often registered through a radio broadcast or a voice-

HOU, OR CONSTRAINTS

189

over narration that has one character reflecting on events not shown to us. And although one or two of these youths may plunge into a larger social world, idealistically committing themselves to political action, most are oblivious to it. The most sensitive characters tend to be watchful observers who struggle to understand how distant, sometimes ominous, forces are shaping their lives. There were many such forces at work in Taiwan, and Hou’s treatment is all the more remarkable for emerging from a climate that stifled any record of ordinary life. Like the Greek junta of Papadopoulos, Chiang Kai-chek’s Kuomintang Party (KMT) ruled with an iron hand. Mainlanders who had fled Mao’s revolution composed a mere 15 percent of the population, but they subjugated the indigenous Taiwanese. In 1949 the KMT declared a state of martial law in preparation for an imminent reinvasion of the mainland. Throughout the 1950s and 1960s the economy grew, and there emerged a generation that saw itself as Taiwanese rather than mainland-in-exile. Yet public life played out the o‹cial charade that the little island was the one true China, poised to retake the homeland. The popular cinema oªered only distraction, barring directors from raising social issues or criticizing government policy. But Taiwan forfeited its UN seat to the People’s Republic in 1971 and lost US diplomatic recognition soon afterward. Chiang died in 1975, and his son governed somewhat less strictly. The KMT’s authority was weakening just as Hou and other artists of his generation came to maturity. It probably also helped that Hou was something of an outsider. Born in 1947 in southern China and taken to Taiwan as a baby, he arrived with the mainlanders (and his father became a civil servant), but the family was Hakka, a minority on both the mainland and the island. Hou grew up on the margins of petty crime, spending his teen years brawling and gambling. After a stint of army service he went to film school in Taipei. Unlike Angelopoulos, who, returning to Greece from Paris, faced a moribund local cinema, Hou, after graduating, found a place in an industry turning out a film per week. He worked dutifully for a decade, climbing from script assistant to assistant director and directing three buoyant musical romances between 1980 and 1982. He went on to become the most prominent member of the government-sponsored New Cinema, which sought to capture fresh aspects of local experience—and, not incidentally, win places on festival screens. Hou’s breakthrough films, an episode in the portmanteau project The Sandwich Man (1983) and the full-length Boys from Fengkuei (1983), helped define Taiwan’s new movement for critics abroad, somewhat as Angelopoulos’s Reconstruction (1970) had spearheaded the New Greek Cinema. At the urging of Cahiers du cinéma editor Olivier Assayas, The Boys from Fengkuei was put on the program at the Festival of Three Continents in Nantes and received the Grand Prize. By the time martial law was finally rescinded in 1987, the New Cinema had played a part in opening up an authoritarian culture. Hou was now a festival darling, and with overseas funding he began to craft a series of ever more demanding, occasionally obscure works. Like Angelopoulos, he chose to probe his country’s political history and contemporary ambience, although both directors found that critics abroad were far more interested than were local audiences.

190

HOU, OR CONSTRAINTS

Hou’s commitment to concreteness emerges not only from his subjects and themes. A sense of observing life as it passes is embedded in the style as well, making scenes like the intellectuals’ song seem to be straightforward recording. Hou has said that he simply wants to capture what is happening to a character in a particular place.5 He cites Confucius: “Watch but don’t intervene; observe but don’t judge.”6 Unintentionally echoing Rossellini (“Things are there; why manipulate them?”), he invites us to consider his style as issuing from a sympathetic but detached eye, an invisible observer taking in a rich reality. For Flowers of Shanghai he wanted his camera to be like “a character who strolls around these places where the characters meet, who spies on them while remaining transparent.”7 At the theoretical level, there are many di‹culties with the ideal of an invisible observer, not least the fact that it occludes the shaping force of staging.8 True, in virtually all fiction films we sense that the camera’s eye is ours, surveying an action that just seems to be there. Yet this invisible-observer eªect is produced by a coordination of shooting and staging. The action is always sculpted from within, angled to or from the camera in order to mold our moment-by-moment experience. Hou’s films, no less than those of the directors examined in earlier chapters here, guide our vision by a deliberate, often minute, movement of bodies through light and space. In our City of Sadness scene (Figs. 5.1–5.4) the blocking and revealing of the men’s faces, the well-timed retreat from the middle ground, the opening out of the room’s space onto the arena of collective history, the shifting of visual weight from the men at the table to the sky and the couple on the right—all this is no casual reality simply caught by a shrewdly placed camera. Just the opposite: no less than in Feuillade, Mizoguchi, or Angelopoulos, the action is designed to flow felicitously around our point of vantage. Hou thereby inserts himself in the tradition we’re tracing, but with diªerences. He revives tactics of minute blocking and revealing, but he puts the camera farther back than would Feuillade. He exploits long lenses (unlike Mizoguchi) and gradations of color. And unlike Angelopoulos he takes us, we might say, beyond dedramatization. He discloses a surprisingly wide range of feelings in what might seem a detached perspective on the action. His films are melodramas but refined ones; as in Mizoguchi, emotion is not erased but purified. In his staging Hou has been resolutely exploratory. He has developed his style in dialogue with the commercial cinema he entered in the 1970s, then the New Cinema of the 1980s, and most recently the directors throughout Asia on whom he had a strong influence. Across twenty years he has revised and refined current norms (some of which he had helped erect). And although he takes risks with every film, he lays down constraints: “For all true creation, you need rules and limits, and a predetermined space paradoxically permits you a much wider field for creation. The closed settings that I employ leave me few choices, and curiously create a lot of obstacles, but so much the more can I exercise my freedom to deepen what counts the most, the heart of the subject.”9 Once he establishes narrow limits, he can be intuitive and follow his instincts: “I don’t try new things

HOU, OR CONSTRAINTS

191

each time, as Western critics think, but rather I set diªerent limits.”10 However much Hou may invoke Chinese culture as the source of his inspiration, in this respect he converges with a reductivist strain in Western art, epitomized in Stravinsky’s remark that an artist’s freedom increases when he or she sets up formidable obstacles: “In art as in everything else, one can build only upon a resisting foundation; whatever constantly gives way to pressure constantly renders movement impossible. . . . Whatever diminishes constraint diminishes strength.”11

The changes within Hou’s work illustrate how stylistic innovations can spring from a director’s perception of an artistic problem. From the start he found Taiwanese film acting amateurish, yet his first jobs demanded that he pull decent performances out of pop singers and children. Working within the norms of local production, he discovered a pictorial idiom for handling these di‹culties. Later, he was able to revise this idiom to suit the sort of storytelling promoted by the New Cinema. Hou entered a film industry lurching between boom and bust. Output had soared in the late 1960s as swordplay films and costume spectacles won regional audiences, but by the mid-1970s, when Hou became an assistant director, the number of releases had sagged to fifty or so per year. Yet when he began directing in 1980, things had improved. Companies were pumping out over 130 films, although the business was still squeezed by low budgets, shortage of film stock, and cramped shooting schedules. Apart from martial-arts films the dominant genres were star-driven comedies and romances (wenyi pian). So great was the popularity of the novelist Chiung Yao that her name identified an entire cycle of melodramas centered on the Westernized rich. The government’s Central Motion Picture Corporation (CMPC) was the chief backer of films, but it also encouraged independent companies to enter the market—not least to oªset the growing popularity of Hong Kong films. Hou’s first three films came as a burst of fresh air. He made them for an independent company, carrying many players and crew members from film to film. Each was scripted, shot, and cut in two months. Set in Taipei and the villages around it, these unpretentious romantic comedies, brimming with ingratiating music, won large box-o‹ce returns. The opening of Cute Girl (a.k.a. Lovable You, 1980) gives the flavor. Out of steamy Taipei tra‹c a lemon-yellow Triumph rolls toward us while a bouncy song, sung by a boy/girl duet, floods the sound track. At the wheel is a resolutely modern-looking young woman, Wen-wen (Taiwanese pop singer Feng Fei-fei). Immediately we cut to Ku (Hong Kong pop singer Kenny Bee) on his motorbike. He honks flirtatiously at her when she stops to check her engine. Wen-wen snubs him and drives oª to a business meeting. There, out of nervousness she kicks oª her shoes and tries to rub her feet together—accidentally stroking the shoes of her boss. Cut to Ku, working in an architect’s o‹ce, absentmindedly rubbing his feet together. Obviously meant for each other, the couple keep meeting by chance before Ku’s surveying job in a nearby village brings them together.

192

HOU, OR CONSTRAINTS

Cute Girl and its successor, Cheerful Wind (1982), announce themselves at ease in the modern city, setting scenes in Shakey’s pizza parlor and developing gags involving tape recorders, TV commercials, and a French tutor who resembles Jesus. The heroines are cosmopolitan professionals. At the close of Cheerful Wind Ling (Feng Fei-fei again) temporarily sidelines her suitor (Kenny Bee again) in order to fulfill her dream of seeing Europe, and he approvingly escorts her to the plane. Yet the films also create an idyllic Taiwanese countryside, and there the characters learn what finally matters to them. Cute Girl identifies love and fecundity with a sacred tree in the village, first shown with children playing in its branches. Later the couple carve their names into the tree, and the film ends with Wen-wen, now pregnant, dragging Ku under its branches. In Cheerful Wind Ling escapes from career complications by hiding in a village and taking over her brother’s primary-school class. The Green, Green Grass of Home (1982) is set wholly in a village, where the up-and-down courtship between two primary-school teachers (Kenny Bee paired with another diva, Jiang Ling) is overshadowed by the crises in the lives of the townspeople. Blending an episodic romance with children’s family problems and the pollution of the local river, Green, Green Grass compares favorably with Ozu’s lyrical comedies of the 1930s. In their relaxed simplicity these films stand in contrast to other twentysomething romances of the period. In typical wenyi films the young people shuttle from coªee shop to mansion to college to discothèque, seldom exchanging words with the locals. They drive or bike through crowded streets, but they live in Western popular culture. Love in a Cabin (1974) is stuªed with rock albums, fluªy pillows, and posters of James Dean. In Diary of Didi (1978), when the family visits an airport, the crowds obligingly clear a path for the major players. My Lovely Neighbor (1979) centers on the romance between an athletic young woman and a disco-dancing biologist; except for mass shots of a marathon, the film confines itself to a hypermodern house and showcases frantic musical numbers (including a dance called “Kung-Fu Disco”). Hou’s willingness to take his players and camera outside and to let ordinary life crowd into his frame must have seemed refreshing. Taiwanese films of the 1970s tend to be visually florid, even aggressive. Invariably shot in 2.35:1 ratio (a cheap way to raise production values), they jazz up formulaic situations with flashy technique. One of the most famous Chiung Yao adaptations, Outside the Window (1972), flaunts fast cutting and zooming, extreme close-ups, and a willingness to spiral the camera around any character. (In many respects it looks forward to the intensified continuity of 1980s Hollywood.) And the trickery of Outside the Window is restrained compared to the deliberately unbalanced compositions, wild high and low angles, frenzied zooms, swish pans, split-screen eªects, jump cuts, slow-motion, rack focus, and upsidedown shots on display in many wenyi dramas and teen romances of the period. As a handsome violinist and a beautiful pianist play a duet (Love Love Love, 1974), their erotic attraction is conveyed through very rapid close-ups of their burning glances and fluttering finger work. The villain’s menacing walk in Love in a Cabin is chopped into Godardian

HOU, OR CONSTRAINTS

193

bits, stuttering through repeated shots and boldly mismatched positions—apparently just because he is the villain. Not that Cute Girl, Cheerful Wind, and Green, Green Grass ignore the conventions of the 1970s genre pictures. Hou works in the anamorphic ratio, uses an occasional zoom, and provides montage sequences accompanied by bouncy songs warbled by the stars. When measured against what he was to accomplish later, this trio of movies may seem mere kitsch. But compared with the overblown eªorts of Hou’s predecessors, they look daringly restrained. What makes the diªerence? Most noticeably, he avoids the stressed angles and oªcenter framings of the romances and comedies. He favors an eye-level camera, although an occasional high angle might bring out figures crossing a landscape or winding through narrow village streets. He also forswears decorative foregrounds, all those outof-focus table lamps, plant fronds, and flickering candles that clutter up the Chiung Yao films. His plainer style in his first films served as a basis for much of what he would achieve shortly in a rather diªerent mode. Taiwanese filmmaking in the 1970s was pledged to standard analytical editing, with big close-ups coming at moments of dramatic tension. Many directors never made allencompassing master shots, since film stock was too precious to devote to recording the same action twice. Accordingly, directors built up their scenes from medium shots and close-ups captured in as few takes as possible.12 For his first films Hou picked an uncommon option from the analytical-editing menu, that attenuated approach to continuity editing to which European directors adhered in the 1910s. He favored prolonged establishing shots, with comparatively few cut-ins, thereby producing somewhat longer takes. His average shot length in these early 1980s films (between ten and twelve seconds) is about twice that of the fast-cut films of the 1970s.13 He is willing to film a conversation in a shot lasting more than two minutes, and he is prone to include more onetake scenes than his contemporaries, especially in interiors. He attributed the success of his comedies to their new manner of creating more naturalistic performances: “Ordinarily, scripts were written so that one could film the actors in a series of shot/reverseshots, each one of which corresponded to a character’s line of dialogue. This device was lifeless and made the performances seem very artificial. In reaction to this we decided to film the actors together, in a manner closer to both theatre and life itself.”14 In Cute Girl, for example, the heroine’s family gathers around a tape recorder to hear the message she’s left behind. A hyped-up 1970s film would have thrust the tape recorder to the foreground, reels spinning, and family members’ heads would be seen in the background; then each one’s reaction might be underscored by quick zooms. Hou lets the scene play out simply in the full anamorphic frame (Fig. 5.9). Similarly, when the TV producer Lo answers the phone in Cheerful Wind, he stands by a partition, and the station desks stretch out behind on either side, balancing the composition (Fig. 5.10). When Hou does cut in, he tends to employ comparatively few medium shots and virtually no true close-ups.

194

HOU, OR CONSTRAINTS

5.9 Cute Girl (1981): a single-shot sequence as Fei-fei’s family listens to her taped message.

5.10 A depth-oriented anamorphic composition for Cheerful Wind (1982).

The director who did not then know the term “master shot” proves to be quite dependent on it.15 Hou turned other industrial constraints to his advantage. Small budgets obliged Taiwanese filmmakers to shoot on location, both outside (public parks, waterfronts, streets) and indoors. Apparently sets were rarely built for films taking place in modern times. These location interiors, typically Western-style homes, were usually lit to blinding brightness. The multiple-source, high-key lighting brings out saturated colors (and thus adds production values) but also creates harsh overlapping shadows. Hou’s first films scale down the 1970s extremes. Decor isn’t quite so desperately hip; lighting is subdued and even dim; some locales lie oª the beaten path (a community facility for the blind in Cheerful Wind, a rundown village movie house in Green, Green Grass). Most remarkably, Hou makes fluid use of exteriors, filming on the streets or in the countryside. The three first films contain the obligatory music montages of couples romping in landscapes, but most of the dramatic scenes take place outdoors as well. His widescreen images reveal a world beyond doily-draped living rooms—tra‹c, children scampering through the frame, perhaps only trees rustling in the background (Fig. 5.11). For Hou’s contemporaries, location filming called for a particular camera technique. Shooting in exteriors poses many problems, chiefly delays: a crowd may gather, the weather may change. As we’ve seen, by the mid-1960s filmmakers around the world found that the lens of long focal length (75 mm to 100 mm or more) allowed the filmmaker to place the camera at a great distance from the action—across the street or in an unobtrusive vehicle—and still obtain legible shots of major characters. This “telephoto” shot, often supplemented by zooming in or out (that is, changing the focal length to narrow or widen the field of view), was a mainstay of location shoots (Fig. 4.49). The opening of Cute Girl, showing the heroine’s sports car coming toward us, is typical of the period (Fig. 5.12), and Hou frequently stages his two-shot outdoor conversations in telephoto (Fig. 5.11). The long lens committed the Taiwanese filmmaker to a stationary camera, since tracking telephoto shots were very di‹cult to execute in busy locales. To create a sense of frame mobility, the pan or zoom had to su‹ce. Within the constraints of location shooting and long-lens filming, guided by his search

HOU, OR CONSTRAINTS

195

5.11 Cheerful Wind (1982): the long lens diminishes the apparent distance between foreground and background planes, making apparent size and familiar size primary cues to depth.

5.12 The opening shot of Cute Girl: a very long-lens framing that centers the heroine’s yellow sports car.

for more muted performances, or at least for ways to minimize his actors’ shortcomings, Hou’s trio of comedies explored the resources of the telephoto shot. Letting passersby drift through the shot was one possibility. The long lens elongates cinema’s visual pyramid (Fig. 5.13), so in order to cover the same breadth as a shorter lens, the camera must be set further back; hence, on location, the greater likelihood that people will cross in front of the lens. (A classic medium shot, taken with a 35 mm lens, will put the camera only six feet away from the subject, making it unlikely that anyone will pass in between.) Although the inclusion of passersby has been common in Western telephoto shots since the 1960s (Fig. 5.14), most Taiwanese directors, who worried about wasting film stock on retakes, seem to have restrained extras from entering the field of view. Even Kurosawa expressed surprise that inconsequential people slipped into Hou’s shots, occasionally blocking our view of the principals. “In [Kurosawa’s] films,” Hou remarks, “he more or less followed the Hollywood system, shooting in studios with stars. The main character is in the center of the screen, everything else is secondary. For my part I almost never shoot in the studio, almost always in natural surroundings, and most of my actors aren’t stars.”16 Another advantage of the long lens was that it minimizes the size diªerences of objects on diªerent planes. Minor action passing in the foreground or background gains more prominence by being nearly the same size as the shot’s prime subject (Figs. 5.11, 5.15). Without the long lens Hou’s casual passersby would not be so salient. The telephoto lens also allowed Hou to film economically from very few angles; sometimes only one setup su‹ced. A single distant shot worked best if you were trying to wring fluent performances from inexperienced stars. Before Chen Kun-hou and I started directing films, we used to block out our shots in advance: numbers 1 and 3 for you, 2 and 4 for me, and so on. When we came to direct, we found that it was sometimes too hard for the actors to shadow-act the reverse-angle shots they weren’t seen in, and so we started filming master shots of dialogue scenes, to serve as guides in the editing. Often we found that the reverse-angle close-ups weren’t needed at all, and so I developed a preference for continuous takes from one angle.17

196

HOU, OR CONSTRAINTS

5.13 The various angles subtended by lenses of diªerent focal lengths. The 50 mm lens, common in Feuillade’s time, takes in 45 degrees, whereas Mizoguchi’s wide-angle shots were made with lenses of 28 mm or 35 mm focal length. Note that the longer lenses used by Angelopoulos and Hou, 100 mm and up, subtend a very narrow playing space. In order to include several characters, the camera must be quite far back. On location this entails that people will be likely to pass between the subject and the camera.

HOU, OR CONSTRAINTS

197

5.15 Layers of onlookers gain prominence, thanks to the long lens (Cheerful Wind). 5.14 As two policemen walk to the camera, the long lens picks out other figures hurrying through the frame (Le deuxième sou›e, Jean-Pierre Melville, 1966).

Nonprofessionals were even less skilled at replaying action in closer views than pop stars: “Amateur actors are often very nervous in close-ups.”18 Hou’s preference for nonprofessionals led him to find one fixed position for capturing the entire scene. I let them [the nonactors] use what they were familiar with—their usual habits and gestures—so that they need not “act”: for example, having a meal or smoking in sunlight, and so forth. In order not to make them nervous, I would deliberately put the camera quite a long distance away and not move it. Hence, I had to find the right angle for every scene so that the camera could take in as much of the incidents and people as I wanted.19

Yet this way of putting the matter—finding the right spot for the camera to take in the action—plays down the active way in which Hou lays out the action itself. The first three films show that the long lens and the distant, stationary camera enable Hou to stage scenes in fresh ways. Most evidently, Hou discovers some schemas familiar to us from the ensemble-staging tradition. James Udden has pointed out a felicitous piece of depth staging in Green, Green Grass of Home.20 A boy has discovered that his father has killed his pet owl, and the shot starts with the father stretched out on the porch in the middle ground while his son runs up shrieking in dismay (Fig. 5.16). Like the reunion in Angelopoulos’s Voyage to Cythera (Figs. 4.95–4.96) this scene unfolds in the background, with the father chasing the boy to a trembling standstill (Fig. 5.17). Earlier in the film Hou has presented a more flagrant use of distant depth. The scene of the new teacher Ah-lian (Kenny Bee) arriving is filmed from inside the train. He alights but leaves his ticket behind and must run back to fetch it (Figs. 5.18–5.22). A more aggressive director would have cut from close-ups of the ticket to shots of Ah-lian groping for it at the gate, or perhaps set the ticket closer to the camera. Hou instead exploits Barr’s gradation-of-emphasis principle, letting various spectators discover the ticket on the win-

198

HOU, OR CONSTRAINTS

5.16 A long take from The Green, Green Grass of Home (1983): the boy discovers that his pet owl is dead.

5.17 The boy’s tantrum and the father’s punishment of him are played out at a considerable distance from us.

5.18 The Green, Green Grass of Home: Ah-lian stirs himself from his nap and gets oª the train.

5.19 He shambles to the ticket agent, and through the window we watch him fumble for his ticket.

5.20 Realizing he has left it on the windowsill, he comes running back for it.

5.21 Somewhat awkwardly he retrieves it . . .

5.22 . . . and runs back to the agent.

dowsill at diªerent points in the shot’s unfolding. The unemphatic foreground suits the film’s relaxed, digressive dramaturgy. Ah-lian is a little forgetful and a little resourceful, but nothing of consequence hangs on this moment or these traits. Is this director simply watching? No; he is staging. This is particularly evident in the train-ticket scene, which presents a fairly traditional wide-angle composition. What,

HOU, OR CONSTRAINTS

199

5.23 Cute Girl: in a high-angle telephoto image, space is cleared for Wen-wen to question Ku’s partners.

5.24 A surveyor in a red shirt moves aside to reveal Ku as Wen-wen tries to convince him to be treated in the village.

5.25 When Ku interrupts Wen-wen, all turn to watch him, and . . .

5.26 . . . when he doubles over in pain, the shot favors Wen-wen, who reminds the men that the road has collapsed.

5.27 When Wen-wen finally confronts Ku directly, the foreground surveyor, who has served as a pivot for the composition, swivels to let her move to frame center.

5.28 The shot ends with Wen-wen at frame center vowing to convince the village healer to help.

though, of those casual shots packed with adventitious objects and foot tra‹c? Here, too, we can watch Hou shaping his scenes through his repurposing of the long lens. Whereas Angelopoulos uses the telephoto to create fairly stark abstraction, Hou spreads a welter of details across the anamorphic frame and then engages in a delicate play of balance and counterbalance. One instance occurs during a semidramatic moment in Cute Girl. In the village the surveyor Ku has been bitten by a snake. The shot begins with a panning zoom following Wen-wen and a handful of children running toward Ku and his crew. They arrive in a densely crowded frame, with onlookers piled up by the long lens and the slightly high angle (Fig. 5.23). In a friezelike string of heads, helmets, and shoulders, the scene plays out as an argument, with Wen-wen in the frame’s right half, surrounded by surveyors, and Ku on the left, framed in the cab (Fig. 5.24). As the scene develops, the image’s vi-

200

HOU, OR CONSTRAINTS

5.29 Kurosawa employs the long lens and the high angle to render figures as stacked up the vertical dimension of the frame (Red Beard, 1965).

5.30 A common Hou compositional strategy: faces spread like beads along the frame, with depth suggested more by placement and overlapping planes than by shrinking size (Cute Girl).

sual weight shifts between halves, with one surveyor, at the bottom frame edge, becoming a fulcrum (Figs. 5.25–5.28). This sort of image, made possible by a fixed camera stationed in an open exterior far from the action, may be indebted to Kurosawa, who experimented with impacted widescreen telephotos throughout the 1960s (Fig. 5.29). Still, Hou, inclined by current craft practice to the long lens and location filming and intent on not distracting his players, might easily have discovered it on his own. Shots like the one from Cute Girl are not simply events captured by an alert, observing eye; they are precision crafted from the inside out. In his first three films Hou discovered that the long lens can make a setting teem with layers of detail; it can stack or string faces along axes that highlight slight diªerences (Fig. 5.30); it can turn transient apertures into fruitful slots in the picture space; and its exaggeration of slight shifts of figure position can yield subtle dramatic rewards. He would apply these lessons to nontelephoto images as well, in the process creating a rich, nuanced pictorial style. Angelopoulos, committed to a conception of how space and time might represent history and well aware of European film theory, set out to explore the long take as an expressive device in itself. How far could he push it? How could he enrich the tradition of Renoir, Mizoguchi, and Antonioni? How could he design shots that revealed untapped resources of the technique? By contrast, Hou knew nothing of Bazin and the Cahiers debates on mise-en-scène. He became a long-take director because he sought to refine performance by capitalizing on craft practices—location filming, long-lens shooting— normalized in his commercial milieu. The fixed long take, his stylistic signature, was the product of a cascade of solutions to stubborn practical problems.

Presumably Hou could have kept making good-natured, crowd-pleasing movies for many years, but changes in his professional circumstances gave him new opportunities. In the early 1980s Taiwan film attendance declined sharply, and Hong Kong films began to command more attention than the local product. The rash of independent companies

HOU, OR CONSTRAINTS

201

had concentrated on speculation, not long-term investment, so only the government’s Central Motion Picture Corporation could initiate recovery. Ambitious government o‹cials launched a “newcomer” program that oªered support for cheap films by fresh talents. Even if the new films could not win back the local audience, they might gain renown at foreign film festivals. At the same period, a local film culture began to emerge, relying on critics who were sympathetic to the creation of a New Taiwanese Cinema. The first CMPC project was the portmanteau film In Our Time (1982), gathering four stories, each dealing with a diªerent phase of life in a diªerent decade. Of the directors represented, Edward Yang (Yang Dechan) was to become the most famous. In Our Time was followed by another anthology, The Sandwich Man (1983), which included Hou’s Son’s Big Doll episode. Several features followed in the same year: Hou’s The Boys from Fengkuei; Growing Up, directed by Chen Kun-hou (cinematographer and producer for Hou’s first three features); and Wang Tung’s Flower in the Rainy Night. These films did well at the box o‹ce, with Chen’s earning several local awards. At the very end of 1983 appeared Edward Yang’s That Day, on the Beach. Its story of a woman’s empty marriage and her search for selfdefinition was refracted through interior monologues and complex flashbacks. That Day, on the Beach quickly became known as the most artistically audacious film ever made in Taiwan. The New Cinema provided a framework within which Hou was able to reconceive his practice. The key question for all directors, he recalls, was: “What is it to be Taiwanese?”21 Nearly all the films of the 1970s had avoided that question. The wenyi and Chiung Yao romances had treated the characters as pseudo-Westerners, whereas government propaganda pictures presented idealistic mainland emigrants striving to recover the homeland. To explore local identity, the young directors emphasized aspects of Taiwanese life long ignored by the o‹cial culture. They presented middle-class and working-class milieus; everyday life in villages and small towns, especially in the south; the island’s unique amalgam of Japanese and Chinese culture; the indigenous Taiwanese population, speaking its own dialect and not the prescribed Mandarin; and minorities like the aboriginal peoples and the Hakka Chinese.22 Because censorship was still powerful, the New Cinema often adapted acclaimed novels and short stories. Filmmakers gained some prestige by drawing on the two most significant strains of current literature: the modernist trend of the 1950s and 1960s and the nativist (“homeland”) literature of the 1960s and 1970s.23 (In this respect the “ordinary life” to which the films testified was already in circulation as an artistic construct.) Wu Nian-jen, a key screenwriter of the period, admitted that basing The Sandwich Man on three tales by the respected Hwang Chun-ming was an eªort to attract intellectuals.24 Autobiography was another realm that could be explored fairly safely. By treating the mundane microcosm they had known as part of communal memory, directors and screenwriters could delineate fresh aspects of ordinary Taiwanese experience without running afoul of the KMT. Coming-of-age stories, staples of the festival circuit, might appeal to young Taiwanese audiences as well. For his New Cinema projects Hou would develop a unique relation to literary works while exploiting the opportunities aªorded by autobiography.

202

HOU, OR CONSTRAINTS

New Cinema films broke with formal traditions as well. Many of the directors had studied abroad, and they tried out some narrative innovations associated with Bresson, Antonioni, Godard, and other European modernists. As Robert Chen has pointed out, New Cinema films often present slices of life rather than tightly structured plots; they may mix two or more story lines together; and they rely on elliptical storytelling to a degree not seen before in Taiwanese cinema. Many films, particularly those of rural realism, use nonprofessional actors to counter the 1970s star cults. Stylistically, directors gravitated toward more natural lighting and experiments with sound, using both voice-over and sound bridges in fresh ways.25 Although Edward Yang’s early films utilize rather disjunctive editing, and some New Cinema films are conventionally edited, the most prominent stylistic feature of the movement was the static long take. Most long-take directors have been exponents of camera movement as well, as we have seen in Mizoguchi and Angelopoulos, but New Cinema directors favored the fixed frame—perhaps, one director has suggested, because their inexperienced crews did not know how to execute steady traveling shots.26 For many, including Hou, the stationary camera yielded not coldness but objectivity, a refusal of the inflated emotionalism of the arcing tracking shots, musical crescendos, and intense close-ups prevailing in the 1970s. The fixed shot was also considered more subtle. Chen Kun-hou regretted using several zoom-ins to his actors in Growing Up: “I should be moved by the situation itself; you don’t have to tell me to be moved.”27 Some local critics complained that the long take fostered a ponderous mannerism,28 but the technique became a festival trademark. Far from indicating naivete, pictorial stasis seemed to commit Taiwanese directors to purity and simplicity. Hou was a central contributor to the New Cinema aesthetic, and he found many of the movement’s tendencies congenial. He had already come to rely on the fixed, distant camera and the accidents of location shooting. Shooting The Green, Green Grass of Home he had enjoyed improvising scenes with children and other nonprofessionals. His plotting, rather loose to begin with, had become more open-ended across the trilogy: the young lovers don’t unite in conventional ways at the end of either Cheerful Wind or Green, Green Grass. The latter film also has an episodic structure and openness to local color that has led some to see it as a harbinger of the New Cinema. Nonetheless, Son’s Big Doll marks a significant change. The core plot is simple. In order to aªord a child, Kun-shu gets a job advertising movies, stalking through town in a clown outfit and sandwich board. His clown scheme proves futile, but his boss gives him a new job, pedaling a cart and announcing films through a megaphone. Yet Kun-shu’s infant son enjoys him in the clown makeup and no longer recognizes him when he’s not wearing it. In the final scene Kun-shu is putting the makeup back on, happily becoming his son’s big doll. Interwoven with this ironic inversion are glimpses of the wife’s daily routines, prolonged passages of the sandwich man’s futile wanderings, and several brief flashbacks, which invoke incidents around the son’s birth and the shame that Kun-shu’s job brings on his uncle. In the margins of the action families line up for flour doled out by a Christian church, passengers alight from trains, kids climb trees and play with fishnets. Across

HOU, OR CONSTRAINTS

203

5.31 The opening of the Son’s Big Doll episode of The Sandwich Man (1983): a very long lens captures the hero, decked out in clown regalia, walking the streets. Compare Fig. 5.12.

5.32 A distant and dense composition from Son’s Big Doll: as temple workers give out rice in the foreground, the neighbors queue up in the center; and the hero, with his dunce cap, can be spotted still further back.

a mere thirty-five minutes, an entire community is sketched. Details blossom into anecdotes without quite becoming dramatic turning points. When Kun-shu pauses to defecate, he laboriously takes oª his costume, only to have it stolen by children. Yet this is no Bicycle Thieves; Kun-shu shortly recovers the costume and resumes his trudge through town. Son’s Big Doll incorporates techniques from Hou’s first three films, tries out some fresh ones, and prolongs his takes (the average shot runs seventeen seconds) in service to a drama more muted and episodic than any he had mounted before. The film starts with a to-camera telephoto shot like the openings of Cute Girl and Cheerful Wind (Fig. 5.31). In the street scenes passersby slide through the frame as unpredictably as in Hou’s musical idylls, but now there are poignant moments of stillness, as when Kun-shu slumps with weary resignation or when he and his wife sit silently contemplating the possibility of having a child. There are also relaxed compositions in depth, along with audacious aperture framings exploiting the wide format (Figs. 5.32–5.33). The result, which stylistically resembles no other segment of either The Sandwich Man or its predecessor, In Our Time, amounts to a first draft of what Hou was to accomplish in the New Cinema. Having brought touches of lyrical realism to commercial genres, he was ready to plunge into a full-blown realist mode, which would oªer him the chance to pit his pictorial intelligence against new constraints. Between 1983 and 1987 Hou directed five features, most for the CMPC: The Boys from Fengkuei (a.k.a. All the Youthful Days), Summer at Grandpa’s (1984), A Time to Live, a Time to Die, Dust in the Wind (1987), and Daughter of the Nile. Over these years he built a team of trusted collaborators: cinematographers Chen Kun-hou and Li Ping-pin (Mark Lee), along with two screenwriters, Wu Nian-jen and Chu Tien-wen. He also settled on his preferred working methods. The script that Wu or Chu prepared would be principally for the crew, not for Hou, who elaborated freely on it. The lines were only sketchily indicated in the script, and he encouraged performers to devise their own dialogue. He derived the staging after seeing the location and setting. His carefully composed scenes were built out of improvisation to a startling degree; the father’s death scene in A Time to Live was developed on the spot. “I am there to remind them not to be a stickler about the dialogue,

204

HOU, OR CONSTRAINTS

5.33 Daring aperture framing: the uncle in a noodle shop turns away to glimpse, through a distant window, his nephew with a new job, pedaling through town to announce movies (Son’s Big Doll).

5.34 A planimetric framing in telephoto depicts the attack of a rival gang (The Boys from Fengkuei, 1983).

to pay more attention to others, the space and the interactions. . . . Spontaneous responses are powerful and moving.”29 The films probed recent Taiwanese history and culture, focusing on unglamorous regions and ways of life and increasingly including the native Taiwanese dialect as a complement to the KMT-imposed Mandarin. (Dust in the Wind was the first film in many years to feature extended conversations in Taiwanese.) Hou provided quietly powerful tales of maturation and disillusionment. Summer at Grandpa’s introduces a city boy to the country, where he and his sister learn that adults are often weak, cruel, and confused; but most of the films trace the opposite movement, from country to city. Living in the city means separation from one’s family, meaningless work, transitory friendships and love aªairs, and sometimes, as in the gangster subplot of Daughter of the Nile, death. Life’s hope, it seems, rests chiefly in tradition—not the old way of life, which is passing away, but the integrity and stoicism that infused it. These values come to be incarnated in Li Tien-lu, the venerable puppet master whom Hou uses as a stern, salty grandfather in Dust in the Wind and Daughter of the Nile. Hou’s storytelling strategies display a steadily growing ambition, making bold use of oªscreen action and suppressed climaxes, as well as flashbacks and subjective sequences. Violent action is handled in a calm, almost detached, fashion, frequently in the distant long takes that had by now become Hou’s trademark (Fig. 5.34). The 1983–1987 films’ average shot lengths increase dramatically, reaching thirty-two seconds in Dust in the Wind.30 As Hou had given some stylistic gravity to the commercial romantic-comedy genre, so now he pushed the lyrical and naturalistic tendencies of New Cinema toward quietude and contemplation. In turn, these qualities fostered fairly disjunctive narratives. “My feeling for the mood, the pace, the film time and so on seemed to be working strongly in my subconscious,” he explained of Dust in the Wind. “It began to aªect not only my approach to the story, but the story itself.”31 Across these films the calm rendering of dailiness yields to a narration that withholds information. What began as tact fades into an almost stubborn reticence. Hou began to conceive his approach as one of emotion-laden objectivity. In preparation

HOU, OR CONSTRAINTS

205

for The Boys of Fengkuei Chu Tien-wen suggested that he read the autobiography of mainland author Shen Tsung-wen (a.k.a. Shen Congwen), whose work was undergoing a revival. During the 1930s and 1940s Shen composed stories of rural life, flavored with prosaic details and observed with a detachment that Hou believed had a filmic equivalent in distant shots. “I discovered a simple device, and that was to constantly tell the cinematographer to ‘keep a distance and be cooler.’” 32 He explained that this position, along with the long take, rendered the film more faithful to life; if you witness an accident or a fight, you have only one vantage point, and your experience unfolds in continuous time.33 Many of Hou’s European contemporaries justified their distant long shots on pseudo-Brechtian grounds, claiming that distancing blocks emotion and provokes the viewer to thought. But Hou wanted to capture a particular emotion made possible only by distance: “Say you’re in a street and you see something a little further along: a group of kids standing on a corner or a little kid who runs out into the road to pick something up and then runs back. Even from a distance, you can feel the tension. That’s the kind of feeling I wanted in my shots.”34 After Edward Yang suggested that he watch Pasolini’s Oedipus Rex (1967), Hou refined his rationale a bit further. He claimed that a film could represent not only the objective standpoint of the director but also that of the character. It could also portray what was passing through either the character’s or the director’s imagination. “To make a film is to play with these four points of view.”35 This conception seems to explain the deviations from objectivity to be found in the quiet flashbacks of The Sandwich Man and Boys from Fengkuei (presumably motivated by character memory), as well as the more disorienting passages of Dust in the Wind, which may present the hero’s thoughts and feelings recast by an external narrational agency (the “director’s imagination”). Significantly, whatever type of subjectivity is present, Hou still tends to render it in fairly distant long takes, as if any degree of subjectivity would still be refracted through the director’s objective “point of view.” Despite the New Cinema’s pressure to innovate and Hou’s growing awareness of what he wished to convey, his films of this period do not wholly reject what he had already accomplished. He remains an apostle of attenuated continuity. Most scenes are built out of a master shot and a very few inserts, with the cut likely to be an “axial” one (that is, straight in or out in the 1910s manner; Figs. 2.5–2.6). As in his first films, he favors the long lens for both overall views and medium shots. Now, however, the lesson of the dense widescreen image is transferred to stories with a slower rhythm, less definable plot twists, and a subtler, more oblique approach to exposition. Hou, starting out a bit like Renoir in his almost casual willingness to let life trickle in around the frame edges, was becoming more like Mizoguchi, a proponent of rigorous framing and modulated action. So we get a scene like that in A Time to Live, a Time to Die, in which standard continuity editing is revised in order to register the eªect on the family of Ah-Hsiao’s passing to middle school. He comes home as his mother, sister, and two brothers are scrubbing the floor and his father sits at his desk. A telephoto shot of the doorway shows, in depth, the boy arriving to stand silhouetted in the center, where his mother embraces him; the grandmother is sitting in the foreground, and brothers, sisters, and father are dimly visible in

206

HOU, OR CONSTRAINTS

5.35 A Time to Live, a Time to Die (1985): Hou now uses the long lens to clear the foreground and create a distant-depth composition akin to those Mizoguchi created with normal and wide-angle lenses. Compare Figs. 3.39, 3.76, and 3.123.

5.36 A Time to Live, a Time to Die.

5.37 A Time to Live, a Time to Die.

5.38 A Time to Live, a Time to Die.

5.39 A Time to Live, a Time to Die.

the darkness behind (Fig. 5.35). This tightly packed image, with information layered within the central doorway, will serve as our establishing shot. When the boy turns to come to his father, a medium shot shows him announcing that he passed. The telephoto lens and wide format allow us to see his mother striding through the frame behind him (Fig. 5.36). The reverse angle, also a telephoto, shows the father slowly reacting, then telling Ah-Hsiao: “Good. Then you must study hard” (Fig. 5.37). We return to the master shot when the grandmother asks what’s happened; after one elder brother has gone to scrub another room, AhHsiao approaches her to explain (Fig. 5.38). A long-lens medium close-up shows her rewarding him with a coin, panning from his hands (Fig. 5.39) to his face.

HOU, OR CONSTRAINTS

207

With the family aware that Ah-Hsiao has passed to middle school, Hou goes on to present, with great gentleness, the eªects of it. A long-lens cutaway shows the elder brother scrubbing in an adjacent room, while the mother’s voice explains that he has yet to hear the results of his college examination (Fig. 5.40). In the same shot we start to hear the daughter, Hui-lan, comment that Ah-Hsiao always had the best luck. What follows is a dense, prolonged shot showing the father at his desk, while in silhouette Hui-lan and her little brother scrub their way across the tatami mats in the middle ground. Over the nearly two minutes this shot consumes, Hui-lan tells of taking the train to Taipei for her college examination (Fig. 5.41). The family’s cramped finances do not allow her to attend, though, so she must be married oª, whereas the elder brother will go to college if he passes his exam. As Hui-lan recalls her dashed hopes, the mother comes into the rear of the shot, wiping down the bamboo furniture (Fig. 5.42). “What a pity!” Hui-lan exclaims, rushing out of the frame (Fig. 5.43). After holding on the father and mother, Hou cuts to a telephoto long shot of Hui-lan in the next room, framed in a doorway, squeezing out her scrub cloth and wiping away sweat, or tears, or both (Fig. 5.44). The scene ends with Ah-Hsiao, looking oª (perhaps at what we have just seen) and being told by grandma that she will take him to the mainland some day. The cuts and compositions shift the scene’s weight from Ah-Hsiao to his brother and then to his sister, all the while embedding her recollection within household routines. No close-up favors Hui-lan’s complaint or underscores the contrast between her future and her brothers’ prospects. The camera position, the silhouetted lighting, and the presence of so much beyond the immediate drama serve not to distance us from the action but rather to convey Hui-lan’s frustrated hopes with carefully judged discretion. Other scenes of A Time to Live will be presented in fewer shots, but most follow the logic of this one: sustained master shots thick with furniture, utensils, and hand props, often thrown into semidarkness or including windows opening onto still more dense zones; and closer views open to the injection of stray details. Unlike Angelopoulos, who favors the remote, spare image and works toward pictorial climaxes that may not coincide with the drama’s high points, Hou starts with a busy world, out of which dramatic climaxes will stream, sometimes unexpectedly. Spatial detail runs alongside drama in one dazzling moment in Summer at Grandpa’s. Ting-ting, a little girl visiting from the city, is spurned by her brother and his pals. She stumbles across a railroad track, dropping the toy battery-powered fan she carries everywhere (Plate 5.1). Seconds later, Hau-zi, the madwoman who stalks the fringes of the village, scoops Ting-ting out of the train’s path (Plate 5.2) and carries her to the verge. Hou holds the shot for the entire duration of the train’s passing, so that the two are intermittently visible through the flashing wheels of the cars (Plate 5.3). There is, we know, a traditional schema for filming people through the flicker of a passing train. In The Sandwich Man Hou had already experimented with this vision of a figure glimpsed through train cars, and a later telephoto shot in Summer at Grandpa’s invokes the standard usage (Plate 5.4). During Ting-ting’s rescue, though, Hou introduces a bold stroke: her little pink-

208

HOU, OR CONSTRAINTS

5.40 A Time to Live, a Time to Die.

5.41 A Time to Live, a Time to Die.

5.42 A Time to Live, a Time to Die.

5.43 A Time to Live, a Time to Die.

5.44 A Time to Live, a Time to Die.

and-green fan falls athwart the tracks, its orange blades driven to spin in the opposite direction by the train’s roaring momentum. Hou recasts the familiar schema to include a competing microevent, tucked into its own slot on screen right. Although Hou’s New Cinema films abandoned the anamorphic ratio, the 1.85 format aªorded him many opportunities to present a teeming lateral space. The Boys from Fengkuei uses the long lens to present planimetric shots of the boys hanging out (akin to that seen in Fig. 5.34). Hou also builds up extreme long shots that are quite mazelike, particularly during the scenes in which the boys are chased over village rooftops. In Summer at Grandpa’s, the first film in which Hou renders a Japanese-style house, the full shots present zigzag spaces and gridded walls (Fig. 5.45). Both films set such architectural spaces in opposition to other sorts of locales. Boys juxtaposes the crooked byways of Fengkuei with the Pescadores seafront, then the jammed streets and vertically stacked open-faced apart-

HOU, OR CONSTRAINTS

209

5.45 In Summer at Grandpa’s (1984) the house’s upper story harbors many rooms, and Hou often keeps discreetly back from them. Here, the children arrive from the city and begin playing in the room far on the left.

5.46 Summer at Grandpa’s: the boy hero sees his grandfather tend the wounded trucker, the man’s face barely visible at frame center.

5.47 In a parallel framing the stricken madwoman lies unconscious in the grandfather’s infirmary.

5.48 Dust in the Wind (1986): Ah-wan leaves his job at the printing plant, meeting his friend on a motorcycle outside.

ments of Kaohsiung. The crisscrossing interiors and crowded village houses of Summer at Grandpa’s stand in rich contrast with the broad rivers, trees, and rail corridor. And both films make use of aperture framing; Summer at Grandpa’s presents exhilarating variants of the technique, often motivated as point-of-view shots (Figs. 5.46–5.47). Aperture framing gets pushed to extreme limits of visibility in some scenes of Dust in the Wind, capturing Ah-Wan as he leaves his job at the printing company and climbs aboard a barely visible motorbike in the street (Fig. 5.48). The composition and lighting draw our eye to the bright patch outside and Wan’s framed body while also registering the full sense of the job and people he is leaving. This one-take scene boasts a greater intricacy than one finds in The Sandwich Man or Boys from Fengkuei only a few years earlier. Other New Cinema directors rely on telephoto lenses and long takes, but none so singlemindedly pursues the dense image. My Favorite Season (1985), by Hou’s cinematographer Chen Kun-hou, contains several planimetric shots, yet they are used fairly casually. It is Hou’s Daughter of the Nile that becomes a virtual textbook on planimetric composition, with the central locale, the apartment shared by the thieving brother and his two sisters, seen again and again from the same general orientation, creating variants of cubical spaces and aperture framings. The opening scenes introduce the device boldly (Figs. 5.49–5.50). Through-

210

HOU, OR CONSTRAINTS

5.49 Daughter of the Nile (1987): after Ah-yang provides a to-camera monologue, we see a flashback to her brother Ah-fang giving her a crimson Walkman.

5.50 The next scene blacks out the opposite side of the frame when Ah-fang administers an injection to their dying mother.

5.51 In a moment of harmony the family gathers for a meal, framed in the doorway and oªset by the fish tank in the right middle ground.

5.52 The murder of Ah-san: he flees the phone booth, and the killers rush past out of focus in the distance.

out Daughter of the Nile the telephoto lens and the wide format provide secondary and tertiary detail, either beside or behind the principal players (Fig. 5.51). Cheap furnishings and the clutter of popular culture are lit and framed as still-life elements around the characters, and the logo etched into a phone booth’s windowpane is more sharply visible than murderers who, bursting unexpectedly into the frame, become streaming blurs (Fig. 5.52). In these New Cinema features Hou’s long-lens compositions recall his explorations in the early films (for example, Figs. 5.12, 5.15) but add a subtler use of depth and shifts of focus. In The Sandwich Man he continued to line up figures in diagonal depth, like cutouts, each given a distinctive shape, texture, and degree of focus (Fig. 5.53). This eªect of the telephoto creates misleading narration in one scene of A Time to Live, a Time to Die. A dense medium shot presents some pictorial tensions: the mother is talking, but turned away from us; the daughter, Hui-lan, is central but out of focus; the eldest son, Ah-Chung, is barely visible in the lower left corner. In sharpest focus are the sink, the utensils, and the pantry (Fig. 5.54). The mother talks of the family’s money problems, and when she turns, she seems to be talking to Hui-lan: “You are to get married soon, even if you teach” (Fig. 5.55). As the grandmother’s voice is heard asking, “What do girls need all that education for?” Hou racks focus to Hui-lan. After a cutaway to the grandmother completing

HOU, OR CONSTRAINTS

211

5.53 Son’s Big Doll: a street crowd, layered in telephoto depth.

5.54 A Time to Live, a Time to Die.

5.55 A Time to Live, a Time to Die.

5.56 A Time to Live, a Time to Die.

5.57 A Time to Live, a Time to Die.

5.58 A Time to Live, a Time to Die.

her question, cut back to a slightly diªerent setup showing Hui-lan answering grandmother by looking up directly at the camera: “She’s not talking about me” (Fig. 5.56). She explains that it’s Ah-Chung who needs to go to the university. Then the mother turns and with a clearer eyeline cue tells Ah-Chung to study more (Fig. 5.57). Pan diagonally left to Ah-Chung, sitting bowed down by responsibility, as his mother and sister continue their kitchen chores, in slightly graduated planes of focus behind him (Fig. 5.58). The long lens oªers us a packed medium shot through subtly layered planes of focus, yet Hou’s visual narration has led us to make grandmother’s mistake as well, keeping Ah-Chung oªscreen while his mother hectored him. By elaborating schemas developed in his early films, using them to dwell on moments of muted melodrama, Hou rediscovers classic staging practices. In A Time to Live, a Time

212

HOU, OR CONSTRAINTS

5.59 The father’s death in A Time to Live, a Time to Die: as the doctor calls . . .

5.60 . . . family members enter and rearrange themselves.

to Die some key scenes are presented laterally, with the family gradually filling the empty frame (Figs. 5.59–5.60). Dust in the Wind contains some virtuoso staging in depth, going far beyond the moments in Green, Green Grass. In particular, the scene in which the young friends gather in a loft behind the movie screen to share a meal avoids the telephoto image to create a passage of unforced ensemble activity (Figs. 5.61–5.63). The shot exceeds two and a half minutes and is really only the preparation for the more dramatic moment during which Ah-Wan reacts badly to his friends’ compliment on the watch that his father has sent him (Figs. 5.64–5.65). Nonetheless, the table-setting scene nicely encapsulates the ambitions of Dust in the Wind, a film that carries the naturalistic aesthetic of the New Cinema toward vigorously self-conscious artistry. Hou protests that he does not tell his actors how to move or what marks to hit; they discover the best arrangement in the course of planning out the shot.36 This should not, it seems to me, be taken as a claim that he doesn’t arrange his actors within cinema’s characteristic pyramidal playing space. (Otherwise, the felicities of figure movement would amount to miracles.) He gives his actors everyday tasks—smoking, eating, scrubbing, setting a table, running down a street—all of which provide the concreteness so characteristic of his action. He obliges his players to concentrate on ensemble staging in accord with the setting (“to pay more attention to others, the space and the interactions”). It is Hou who, after various trials, decides when the shot reaches its proper density and what lucky accidents are to be retained. He can allow his players some flexibility, but the tightly timed choreography in such shots as the loft scene presents us once more with not Hou the observer but Hou the artificer, orchestrating screen space to display a spectrum of pictorial and dramatic nuances. “Sometimes,” he confesses, “I am too painstaking in my mise-en-scène.”37 By 1987 the director who had helped the New Cinema achieve quiet realism had become a virtuoso of ensemble staging. His five features of the period constitute extraordinary accomplishments; many critics regard all but Daughter of the Nile as his finest work. Stylistically, they display an eager exploration of the denotative and expressive resources of the fixed camera, the long shot, the long lens, and the detail-packed set. He finds almost endless variety in dense medium and long shots, far-oª silhouettes, planimetric compo-

HOU, OR CONSTRAINTS

213

5.61 Teenagers assemble for a meal in Dust in the Wind: Ah-Wan and his girlfriend, Ah-Yuen, are setting out food at the central table.

5.62 As the studio fills up, figures alternately block and clear our vision of the rear doorway, allowing us to concentrate by turns on the table and on the kitchen in the distance.

5.63 Once the guests have gathered around the table, they pivot slightly to reveal others’ reactions.

5.64 In the next shot Ah-Wan’s friends compliment him on the watch, but he knows that his father sacrificed a good deal to buy it . . .

5.65 . . . and in embarrassment he flees the table.

sitions, and carefully timed changes of the figures’ position or aspect. Hou would push still further in the trio of films to come—indeed to the edge of pictorial legibility. Again and again these films would ask, How distant, how dense, how dark, how full of nuance can a shot be, and still coax the spectator into probing the space and seizing on the story?

Along with Edward Yang’s tales of urban anxiety Hou’s cycle of autobiographies changed the world’s perception of their domestic cinema. During the 1970s, festivals had avoided

214

HOU, OR CONSTRAINTS

programming Taiwanese films out of fear of mainland protests, but the CMPC and the Government Information O‹ce launched a coordinated strategy to crack the festival circuit.38 The quality of the films, promoted by European gatekeepers and tastemakers, tipped the balance. Beginning at fringe festivals, New Cinema films won praise and prizes. By the mid-1980s Taiwanese films were routinely playing at Berlin, Locarno, Cannes, and other top events. In 1988 Italy’s Pesaro festival devoted a carefully documented book to the New Cinema.39 Ironically, over the same years the movement’s local reputation fell. With Hong Kong films dominating the market, New Cinema films were denied time to build an audience. Yang’s Taipei Story (1985) ran for only three days. Compared with the kinetic, polished Hong Kong product, the New Cinema seemed old-fashioned, especially to young viewers raised on Western popular culture. Critics also began to attack the films’ resolutely personal and experimental approach. These young auteurs, many claimed, alienated the audience with their slow-moving autobiographical tales and endless long takes.40 Hou became the figurehead of all that was intransigent in the New Cinema, and much of the criticism focused on him.41 By 1987 the movement was believed to be over. Whereas lesser-known filmmakers would have to rely wholly on government largesse, Hou could aim for bigger budgets by supplementing subsidies with funding from abroad. Thanks to an infusion of Japanese money and an ambitious distributor with an eye on the international market, he was able to make City of Sadness (1989). At Venice it won the Golden Lion for best film and became a record-breaking hit in Taipei. Almost certainly the most culturally important Taiwanese film ever made, City of Sadness appeared just after martial law had been lifted. It traces the island’s passage from Japanese colony to KMT stronghold at the end of the 1940s. The plot hinges on an incident that had been banned from public discussion—the “2-28” massacre of February 1947, in which the KMT killed thousands of Taiwanese. Hou’s film helped kindle new discussions of the turbulent postwar years and the KMT’s path to power. In addition, City of Sadness revels in Taiwan’s distinctive sounds. Characters speak Taiwanese, Japanese, Fukien Chinese, Shanghainese, and Mandarin. Even New Cinema filmmakers had routinely replaced actors’ voices with professional dubbers, but Hou shot with sync sound. He claimed that he was inspired to make the movie by listening again to Taiwanese pop music of the 1950s, and the film showcases indigenous love songs and ethnic tunes along with Occidental music, Japanese songs, and Peking Opera. This was, he said, his way of representing the “concrete character” of Taiwan’s culture.42 The film presents this culture in a rarefied way. The story centers on the Lin family’s three sons. One is a “businessman” involved in smuggling with Shanghai-based racketeers; another is a deaf-mute who is a portrait photographer; the third has been driven to a nervous breakdown by working as a translator for the Japanese on the mainland. The photographer marries a young nurse whose brother is leading an anti-KMT movement among intellectuals. Historical turning points of the 1945–1949 years, such as the 2-28 massacre and the executions that followed, take place oªscreen, reported in conversations

HOU, OR CONSTRAINTS

215

or radio broadcasts. Any scene is likely to run for some time before we understand its relation to the previous one—how much time has passed, what has happened in the interval, why these characters assume importance now. Narrational vantage points shift unpredictably. Some events are presented through the young nurse’s correspondence, some through authorial intertitles, and some through the notes the deaf-mute writes to those around him. The central locale is the Lin family restaurant, converted from a Japanese bar, which will bear the signs of the toll taken on the family. At the start of the film the “New Shanghai” restaurant bustles with activity; the closing shot shows the business shuttered and the family decimated. The eldest brother has been murdered by his Shanghai partners, the deaf-mute has been swallowed up in the KMT terror, and the impotent patriarch sits eating with his half-mad son, who is now oblivious to what has happened. The survivors are the women, softly filing in and out of the kitchen. Accompanied by a mournful synthetic score, filmed in majestically sustained shots wrapped in chiaroscuro, City of Sadness showed that Hou could create an intimate national epic.43 It arrived just as films from mainland China were also touring the festival circuit, so Hou became a celebrity like Chen Kaige and Zhang Yimou. And since the art-film circuit encourages filmmakers to create multifilm frescos (the 1960s trilogies of Satyajit Ray, Bergman, and Antonioni; the more recent cycles by Angelopoulos and Kietlowski), few were surprised to learn that City of Sadness initiated a trilogy centered on Taiwan’s twentieth-century history. The Puppetmaster (1993) traces the years of Japanese colonization through the career of the puppeteer Li Tien-lu, from his birth in 1909 to the withdrawal of the Japanese in 1945. Good Men, Good Women (1995) weaves together three time frames: the late 1930s through the 1940s, in which a Taiwanese husband and wife go to the mainland to join the clandestine resistance to the Japanese occupation; the present, in which an actress prepares for her part in a film about the couple; and the recent past, which shows the actress as a gangster’s mistress. Both movies benefited from new laws permitting filmmakers to shoot on the mainland, and both premiered at Cannes, sealing Hou’s status as an international auteur. As if to confirm Hou’s reliance on an international audience, the films also became grander in conception. Although Li’s chatty voice-over gives The Puppetmaster an earthy texture, Good Men, with its avoidance of concrete details in the 1940s strand and its puzzling film-within-a-film structure, leans toward the portrayal of History we find in Angelopoulos. Like Fellini, Hou was an intuitive, unpretentious filmmaker who stretched his ambitions in response to a market that demanded reflexivity, roundabout storytelling, and other modernist conventions.44 For example, whereas Hou’s New Cinema films had included brief flashbacks and had played with voice-over narration, none had gone so far with delayed and stingy exposition as these two. Peggy Chiao has aptly called the construction of The Puppetmaster “recapitulative” because scenes run a cryptically long time before Li Tien-lu’s voice-over explains what we are seeing and its relation to previous action.45 Good Men, Good Women carries this strategy still further: the film-within-a-film strategy is not signaled definitely until the very last shot.

216

HOU, OR CONSTRAINTS

Similarly, the last two films in the trilogy juxtapose past and present in ways familiar from European cinema since the 1960s. In The Puppetmaster the elderly Li Tien-lu of today sits in the reconstructed locale in which the past action will unfold. Good Men, Good Women plays on more disruptive ambiguities. The scenes taking place in the 1940s are presented in black and white, as if they were the events that will be dramatized in the 1990s film; yet in those scenes the historical personages are played by the actors who have been cast for the film. So is this a “virtual” version of the film yet to be made? If so, Hou complicates the distinction near the film’s end by presenting a crucial scene of Chiang Bi-yu kneeling at her husband’s corpse not in a realistic-looking morgue but on a bare theater stage; as she weeps, the film slowly bleeds into color. Hou has suggested that we are to interpret this blending subjectively: the actress Liang Ching identifies with the suªerings of Bi-yu, and the murder of her own lover sensitizes her to the execution of Bi-yu’s husband.46 At another level the sequence (along with other rehearsal scenes) probably alludes to Hou’s original source, a play about Bi-yu and her husband. Yet the ambiguities thicken because the film that is being made is called “Good Men, Good Women.” Are we then watching the fictionalized gestation of the very movie we are seeing, in the manner of Fellini’s 8 1/2? The festival circuit expects its auteurs to exhibit a distinctive style (a marketable brand, cynics would say), and Hou did not disappoint. He extended his signature long takes; City of Sadness averages 43 seconds per shot, the shots in The Puppetmaster average about 82 seconds, and those in Good Men, Good Women slightly over 106 seconds. From this base he was able to create variations: the docudrama side of The Puppetmaster, and several free-ranging tracking shots—virtually never seen in the 1980s films—for Good Men, Good Women. Now he began to claim that his long take and other stylistic features were indebted to Chinese art. In a 1993 interview he said that the landscapes in The Puppetmaster stemmed unconsciously from techniques of empty framing (liu-pai) in painting.47 These remarks lent the films a more exotic air on the festival circuit, and Hou’s declarations of his attachment to classical Chinese culture were approvingly echoed by critics.48 Nevertheless, there are pertinent reasons to consider his visual style in the light of more proximate conditions, not least his own previous accomplishments. In the features of his New Cinema period he had experimented with a range of visual resources, and as the mid-1980s films became more narrationally complex, his style moved toward ellipsis and indirection. The historical trilogy oªered him the opportunity to push certain premises to new limits, coordinating visual technique with very oblique storytelling strategies. Crucially, the postponement of scene-by-scene exposition, Chiao’s “recapitulative” strategy, meshes nicely with Hou’s propensity for the dense, static long take. By denying us a link to the previous scene through either character-based causality (goals, appointments, deadlines) or voice-over explanation, he lets the new locale register initially as a space, not a container or background for well-defined narrative action. We simply watch what’s happening (or not happening) in the frame, taking in the vastness of a

HOU, OR CONSTRAINTS

217

landscape or the details of an interior without yet knowing how it links to a larger story rhythm. Only at length—sometimes quite late in the scene—when the characters broach their projects or the voice-over explains what has occurred since the last scene do we understand what is transpiring here. In the meantime we have been obliged to study the shot in itself. While he is deferring plot updating, or even while the dramatic issues are beginning to coalesce, Hou is likely to make the ostensible subject of the shot—character action, facial expression, and the like—somewhat hard to detect. If Angelopoulos delays showing us the key pictorial element, it is only to make that element shine forth more powerfully at the climactic moment, as when the yellow slickers of the railway workers glide into Landscape in the Mist (Plates 4.5–4.6). Hou’s trilogy does not usually develop shots around such pictorial epiphanies. He will very often plant the camera so far back, bathe the shot in so much darkness, and pack so much architectural geometry and naturalistic detail into the settings that the images yield a much less definite arc of visibility. True, he does not foreswear the plainer long take—two or three people in a fairly cleared space conversing without much figure movement (Fig. 5.66)—but often we must strain just to make out what is happening. This strategy forms a stylistic equivalent of his delayed exposition and uncommunicative scene linkage. Cast into darkness, blocked by walls or props or other bodies, his characters occupy only slivers of space. Sustained concealment and brief, unemphatic revelation are central to the trilogy, a strategy as apparent in landscape shots (Fig. 5.67) as in more intimate framings. As in Mizoguchi’s films, we must be alert to just-noticeable diªerences, those barely discernible changes in the image that ordinary filmmaking overrides in favor of sweeping eªects. Presenting the life of a puppeteer was a felicitous choice for this strategy, allowing Hou to replicate his extreme long shots of rooms and landscapes in miniature spectacles of slight movements in splendid, if tiny, surroundings (Fig. 5.68). At the same time, as a director who wants to tell a story, Hou must also coax us to notice or infer some major incidents. So a tension emerges between the demands of plot, however minimal, and the barely legible image. When denotation is this elusive, we cast about for other functions the image might be fulfilling. Sometimes, as in the intellectuals’ meeting mentioned at the start of the chapter, the symbolic dimension of the shot stands out vividly. At other times, Hou like Angelopoulos delays denotation in order to summon up expressive qualities, which are likely to be reinforced by music or dramatic context. At still other moments his urge to fill the shot with detail and then render it obscure seems to ask us to appreciate sheer pictorial patterning. Most strikingly, the di‹cult image and its just-noticeable diªerences sensitizes us to small gradients of change in the image, and their potential narrative significance. To such ends, lighting openly governs the play of visibility. The Puppetmaster counterpoints three lighting textures: the soft, usually overcast daylight of the landscapes; the brilliant color of the puppet shows; and dark, rich, deep interiors. Hou commented that

218

HOU, OR CONSTRAINTS

5.66 City of Sadness: Hinomi and Wen-ching receive the news of her brother’s death.

5.67 City of Sadness: Ah-ga has stood placidly facing us in the foreground, but as the rickshaw pulls up in the distance, he turns, and as he walks from us, we realize he carries a sword.

5.68 A puppet show from The Puppetmaster (1993).

the lamps and bright windows provide “a focal point” for the action, as well as evoking “old China, its melancholia, its intimacy and density.”49 Often by using only one light source per locale (or at least appearing to use only one), Hou can plunge rooms into cavernous blackness. In both this film and Good Men, Good Women Hou and his cinematographers create alternating layers of brightness and shadow; one plane is etched in silhouette, the figures behind are volumetrically lit, the next plane falls into darkness, and so on (Fig. 5.69). The long lens helps in this slicing and stacking of planes in depth, stretched along the camera axis. Hou understands that modern color film stock allows one to set dark colors against still darker ones, even shades of black on black, with very little edge lighting. In Good Men, Good Women the arrival of the Taiwanese supporters in the KMT camp is played out as an almost completely silhouetted scene (Fig. 5.70). As we struggle to make out facial expressions and gestures, we must become sensitive to the ways bodily movements express characters’ attitudes—such as the peremptory movements of the KMT o‹cer as he turns from the emigrants to the telephone (Fig. 5.71). The greater reliance on chiaroscuro cooperates with a stringent use of setting. Shadows can hint at a welter of objects and furnishings, out of which a few may catch a stray gleam. Shadow can also define architectural shape (Figs. 5.69–5.70). In particular, shadows can create apertures that tease us toward certain zones of activity. His rural childhood,

HOU, OR CONSTRAINTS

219

5.69 Good Men, Good Women (1995): figures layered in light and darkness as Hao-tung sells his child.

5.70 Refugees are interrogated in semidarkness before . . .

5.71 . . . the o‹cer goes to his phone, with space cleared to make his silhouette stand out (Good Men, Good Women).

5.72 The Puppetmaster: Li calls on his fiancée’s family. At the start the most prominent zone is occupied by busy servants clearing a table in the courtyard, framed by the pillars of the porch.

5.73 From the street far in the distance, caught between the two women, the delegation of men accompanying Li strides into frame center, as one servant moves aside to permit a fuller view of Li himself.

5.74 The visitors come to the central courtyard, and now the doorway frames them greeting the maids.

Hou says, made him fascinated with doorways and windows opening onto tempting vistas. A fine example occurs in The Puppetmaster, when the young Li comes calling on the family of his bride-to-be (Figs. 5.72–5.78). As so often in these late films, the camera looks out from inside, letting darkened rooms frame a teeming world beyond the threshold. Initially the shot seems to have no subject, so its many centers of activity dawn gradually. Thanks to blocking and clearing, centering and decentering, and subtly layered light

220

HOU, OR CONSTRAINTS

5.75 As Li stands smiling in the entryway, members of the bride’s family come to frame center to greet him; a sidelight picks them out against the dark central post.

5.76 A slight panning movement brings Li into the cool interior, where he and his companions seat themselves on frame left.

5.77 With family members in silhouette they drink tea and begin discussion of the marriage.

5.78 The shot ends with a sound cue. Everyone turns to the street, where—again through embedded openings—we can glimpse children setting oª firecrackers.

5.79 A later scene in The Puppetmaster: from within a co‹n shop we see Li and his child set oª firecrackers to mark the departure of the Japanese.

and color, the shot traces a zigzag progress from one aperture to another, beginning and ending with the world in the street beyond. Besides providing a contemplative way of showing Li’s absorption into the family—this shot must do duty for the wedding ceremony we won’t see—the shot will have a parallel later, when Li and his family take refuge in a co‹n shop at the very end of the war, and then it will be his own children who set oª firecrackers outside (Fig. 5.79).

HOU, OR CONSTRAINTS

221

EDWARD YANG AND THE STAGING TRADITION At the start of the New Cinema movement Hou and Edward Yang seemed to have developed a neat division of labor. One was the chronicler of working-class life, chiefly in rural areas; the other concentrated on the anomie felt by intellectuals and the urban professional class. Hou turned his attention to the past, whereas Yang presented the strains of a newly modern city and the converging destinies of people within it. The Terrorizers (1986), perhaps Yang’s masterwork of the New Cinema period, traces how a random phone call triggers a couple’s breakup and a suicide. Yang’s ricocheting narrative structure seemed drastically at variance with Hou’s lyrical, diaristic accounts of growing up in postwar Taiwan. Hou cultivated long takes in depth; Yang (a proficient comic-strip artist) told his stories in fragmented montage. Both relied on long lenses, but Hou turned them toward tableaux, whereas Yang used them to pick out planimetric medium shots of isolated individuals (Fig. 5.80). Like most neat contrasts, it proved too good to be true. From time to time Hou would venture into the modern urban world, concentrating particularly on young people falling into crime (for example, Daughter of the Nile). And after the New Cinema had ended, Yang seemed to absorb new energy from Dust in the Wind and City of Sadness. With A Brighter Summer Day (1991) he shifted his sights, basing the plot on a famous murder of a teenager in the 1960s and turning to more distantly framed long takes. Yang’s first three features have ASLs of between eleven and fifteen seconds, but A Brighter Summer Day averages twenty-six seconds. Its opening shot shows a lightbulb switching on—a major motif in the film, which will play on pulsating light throughout, but perhaps also a tribute to the start of City of Sadness. Throughout, A Brighter Summer Day blends Yang’s taste for disjunctive montage (a key scene in a nighttime schoolroom is edited to create uncertainty) with a mastery of intricate staging. One scene—the arrest of the hero’s father—is typical of the film’s subtle spatial and dramatic tensions (Figs. 5.81–5.82). Yang’s A Confucian Confusion (1994), Mahjong (1996), and Yi Yi (2000) returned to modern Taipei as their milieu, and each one explored the resources of the long take and dense staging (the films average between twenty-seven and forty-nine seconds per shot).

222

HOU, OR CONSTRAINTS

5.80 The Terrorizers (1986): the husband tries to win back his wife, who is shielded by her colleague.

5.81 Dense staging in A Brighter Summer Day (1991): as the political o‹cers question the father outside in the rain (in an aperture on the far left), mother and daughter look fearfully at one another.

5.82 Coming back in to get his coat, the father turns to his inquisitors: “Can we do this tomorrow during o‹ce hours?” Soon he will be taken into custody.

HOU, OR CONSTRAINTS

223

5.83 The Puppetmaster: as the o‹cer asks about the matter, a foreground figure blocks him, stressing the boy and his father.

5.84 But when the o‹cer must decide whether to prosecute, he shifts slightly into visibility, commanding our notice.

5.85 In the shot’s finale he claps the boy on the shoulder and sighs, his expression no longer visible but his arm near frame center.

Discussing his love of deep space, Hou told a French critic in 1990: “I’ve always liked an action which hides itself—or better, an action which reveals another action.”50 Compared to his mid-1980s New Cinema films, the trilogy revels in oblique and uncommunicative images. A café conversation in Good Men, Good Women seats the central character away from us, the booth shows only the top of her head, and that in turn blocks one of the two men with whom she talks. Only when she turns her head slightly does the second man’s face briefly become visible. The image recasts the sort of schema we examined at the beginning of the chapter, where Hou solved the dinner-table problem by having diners shift their position in the near plane; yet this taciturn shot makes that one seem transparent. In The Puppetmaster Li’s son is brought into the Japanese o‹cers’ headquarters, accused of selling fish harvested from the river. The interrogation is handled through minute shifts masking the o‹cer’s reaction and highlighting his gesture of patting the boy’s shoulder (Figs. 5.83–5.85). An action revealing another action: the phrase encapsulates Hou’s use of the thinnest slices of space and just-noticeable diªerences within broad, crowded telephoto compositions. City of Sadness harnesses all these tactics to the most complex plot in all of Hou’s work. The story of the Lin family unfolds with the major historical events taking place oªscreen, most significantly the 2-28 incident of 1947—a series of riots sparked by the killing of

224

HOU, OR CONSTRAINTS

a woman who sold smuggled cigarettes. The Nationalist army put the insurgence down brutally, following a massacre with a string of executions; a 1952 report admitted that between eighteen thousand and twenty-eight thousand Taiwanese were killed. The fact that Hou did not explicitly show the KMT’s violence was counted against him in a polemical 1991 collection of critical articles, Death of the New Cinema. He was taken to task for avoiding political criticism in favor of more “universal” concerns like romance and family continuity.51 The charge has been rebutted by Lin Wen-chi, who suggests that the film often treats the KMT dream of recapturing China as a myth.52 At the same time, fastened on the concrete, Hou traces how political change reverberates through everyday life. City of Sadness registers the fluctuating routines and tensions within a family with ties to Taiwan (the Lins’ restaurant business and imbrication with local gangs), to the mainland (where the Shanghai gangsters are based), and to Japan. Lin Wen-ching’s friends Hinoe and Hinomi are close friends with the Japanese girl Shizuko, who must return to Japan at the end of the war and in a moving scene leaves them several keepsakes. Whereas Angelopoulos would create an austere, abstract spectacle out of a KMT firing squad, Hou keeps us within the cell that the deaf-mute Wen-ching shares with other prisoners. Several men are hustled out to be shot; in one of the film’s most celebrated moments, Wenching doesn’t react to the oªscreen fusillade because, of course, he can’t hear it. This, one might say, is how people experience history. The film’s first scene opposes the thinness of o‹cial pronouncement to the textures of ordinary existence. As a radio transmits Hirohito’s announcement of surrender (in a formal language no average Japanese or Taiwanese could comprehend), Wen-heung’s mistress gives birth to a child. He lights joss sticks at the family altar and drinks tea. The scene is packed with details and bits of sound rising out of the darkness. The emperor’s stiª declaration is interwoven with the cries of the mistress, the midwife urging her to rest, the advice of female relatives, various household sounds, and a plaintive, percussive synthesizer melody. “It’s a boy,” a voice declares. It is, of course, easy to read this symbolically—the nation of Taiwan is reborn after decades of colonialism—but symbolic meanings seem secondary to the textural density of the moment. For the Lins the birth is at least as big a turning point as the surrender. History is writ both large and small. One way to register the concrete changes that history brings is by returning to images of a locale we have seen before. The iteration of compositions across a film is a common strategy of long-take directors (see Figs. 3.72–3.81), and Hou has recourse to it in A Time to Live (in the kitchen, at the father’s desk, the house viewed from the garden) and Dust in the Wind (recurring framings of the village street winding down from the family’s house). No Hou film uses such scene clusters more scrupulously than City of Sadness. The hospital and an area in the depths of the restaurant become recurring locales, filmed from quite similar positions each time. Perhaps the most prominent iterated framing involves the dining vestibule in front of the altar room. In the second scene our introduction to the New Shanghai restaurant comes through this magisterial setup, with the Lin family moving through the crisscrossed planes of the shot (Fig. 5.86). Later, variants

HOU, OR CONSTRAINTS

225

5.86 City of Sadness: the “christening” camera position for the Lin household.

5.87 Hinomi has now moved in to help care for Wenleung, the half-mad brother.

5.88 A teacher arrives to help get Wen-heung released from jail.

5.89 Hinomi stops snapping beans while Wen-ching tells her, by scribbling notes on his pad, that she is unlikely to see her brother again.

5.90 The wedding of Hinomi and Wen-ching.

5.91 Our last view of the family: Grandfather Lin and Wen-leung are the only men surviving.

of this camera position will register stages of the family’s changing fortunes (Figs. 5.87–5.91). Hou, like Mizoguchi in The Story of the Last Chrysanthemum, recycles a camera setup to provide just-noticeable diªerences in setting or character presence, calling forth from us a spatial memory of the Lins’ home, the hospital, and other locales, evoking lived history through the ways in which a constant space shelters lived change. Hou soaks the particular and personal into these fixed spaces through another strategy, one more allied to editing. He will define a locale as dominated by political debate or action and then carve out of that a pocket of more intimate action. Thus the rebel

226

HOU, OR CONSTRAINTS

camp to which Hinoe has retreated is first shown in an ensemble shot, with anonymous comrades flanking Wen-ching and Hinoe for their reunion. A subsequent cut-in to the two friends eliminates one extra man, finding the personal within the political while also not losing sight of Hinoe’s mission: one of his comrades remains quietly, almost indiscernibly, in the shot. This would seem to echo an earlier scene during which Hinoe and his fellow intellectuals met at Wen-ching’s home, and a discreet cut underscores the friendship emerging between Hinoe’s sister, Hinomi, and Wen-ching, while the political discussion murmurs oªscreen. Often Hou saves his cuts in order to highlight such moments of intimacy within more impersonal—he might say, more “objective”—scenes. By contrast, the sustained long shot, or a cut backward to a broader view, can evoke a sense of the largest possible context. This principle governs the vast landscape shots that bracket phases of narrative action. They often serve to swallow up people’s petty concerns in a larger rhythm of stability. Hou has remarked that he made the landscape shots in the final phase of production, since he felt the need to counterbalance so many interior long takes.53 As he would again in The Puppetmaster and Good Men, Good Women, Hou bathes many of the most important scenes in darkness. The first sequence, the birth of Wen-heung’s son, makes the film’s commitment to chiaroscuro evident. In a pitch-dark set yielding soft silhouettes and few highlights, Wen-heung lights joss sticks and then a cigarette; at the end of the emperor’s surrender notice the bare hanging bulb bursts on. A title tells us that the son will be named “Light” (another parallel to Taiwan as a whole, which is sometimes called the “island of light”). Thereafter, as opposed to the flatly lit hospital and the overcast skies, many interiors are plunged in gloom. Perhaps most strikingly, when the police tramp into the New Shanghai looking for Wen-heung, a variant on the principal camera angle in the Lin household shows us old Lin challenging the o‹cers in a silhouetted foreground. Meanwhile the police search for Wen-heung in a deep corridor on the very left edge of the frame, and as distant figures plunge through the passageway, the “main” characters standing in the foreground realize Milton’s oxymoron of “darkness visible” (Plates 5.5–5.6). Color film finds one justification in this palpably textured blackness. The play of small diªerences is evident in Hou’s minute adjustments of actors’ positions across a scene’s development. Often it is a matter of setting and lighting that masks oª everything but a key gesture, as when Wen-leung returns to the family after his stay in prison; the dining room wall allows us to see only Wen-heung calling out to the family on the far right (Fig. 5.92). The death of Wen-heung, shot during a quarrel with the Shanghai gang, is rendered at a distance, glimpsed in semidarkness and partly blocked by the angle of the corridor (Fig. 5.93). In the rebel camp the Taiwanese woman whom Hinoe has taken as a wife hesitates to approach him and Wen-ching; almost too neatly, she pauses in the crook of a tree (Fig. 5.94). Most characteristically, it is other actors who block and reveal key information. Important as Wen-ching’s nonreaction to oªscreen gunfire is, Hou subtly prepares for it by having him bid farewell to his fellow prisoners

HOU, OR CONSTRAINTS

227

5.92 Wen-heung, bringing back his ill brother, is seen far on the right, framed in a windowpane.

5.93 Wen-heung shot by the Shanghai gang.

5.94 In the partisans’ camp Hinoe’s wife pauses in a tree branch as he talks with Wen-ching.

in a quiet clearing of the foreground. In the same way, the all-important little notepad, on which Wen-ching must write to others, is not made visible until it is needed, and the intellectuals visiting must unobtrusively step aside to reveal it. If Hou’s New Cinema works sometimes evoke Ozu, his trilogy edges closer to Mizoguchi in the articulation of minute shifts of character position. Yet Hou’s a‹nity for telephoto lenses and the widescreen format yields a more striated space than Mizoguchi aªords. The Hou prototype is the long-lens shot of several people sitting around a table, with layers of foreground heads shifting slightly to highlight this or that player. He uses this device intermittently in earlier films, for example when the gamblers are stirred up by the arrival of Ah-Hsiao’s gang in A Time to Live, a Time to Die. By City of Sadness this schema will activate all areas of the screen, since important characters need not face us but also may be placed in profile on the far left or far right, as in Fig. 5.95. City of Sadness oªers a remarkable visual variety (and not just because of its nearly three-hour running time), so any generalization about its style is likely to be a half-truth. For instance, Wen-leung, angered at the sight of Red Monkey’s mistress consorting with her lover’s killers, ignites a fistfight, and the slow lead-up to this is handled in diagonal shifts and slender apertures reminiscent of Mizoguchi (Figs. 5.95–5.96).54 As elsewhere in Hou, a simple sustained medium long shot of a few figures reacting to a situation needs no fancy detailed blocking (Fig. 5.66). This scrupulously realized film (down to the attachable sleeves Wen-ching uses when touching up negatives) shows Hou at the height

228

HOU, OR CONSTRAINTS

5.95 Wen-leung stares at Red Monkey’s old girlfriend, now hanging around with the Shanghai gang. Hou teases us in the Mizoguchi manner by having her face slide out from behind the pillar.

5.96 As Wen-leung continues to harass her, the thug on the right of Fig. 5.95 rises to push him away.

of his powers. If only because meal scenes always tax his energies, it’s worth noting one session that parallels the intellectuals’ meeting we considered at the very start of this chapter (Figs. 5.1–5.4). Wen-leung, in a period of lucidity, is fraternizing with the Shanghai gang, in the same restaurant where we had seen the intellectuals. One goal of the mise-en-scène is to delineate a two-step strategy: the young gangster will gain Wen-leung’s confidence, and then his father will intervene to suggest that Wen-leung use his brother Wen-heung’s shipping business to import drugs. The scene also introduces Red Monkey, a hanger-on who will be suspected of leaking the plan, and it characterizes Ah-ga, who works for Wen-heung but who is prepared to join the drug-smuggling conspiracy. Wen-leung and Ah-ga sit on center right, the Shanghai gang center left. Hou starts the scene by picking out the core interaction in crowded telephoto medium shots (Figs. 5.97–5.98). In the establishing shot the men toast one another, with the configuration of diners masking the gangster patriarch and highlighting his son (Fig. 5.99). The conviviality will halt when Red Monkey brings in more women and the patriarch explodes in anger. This begins the intimidation of Wen-leung, which will draw him into an alliance with the gang (Figs. 5.100–5.103). Here no window opens out onto the nation; the gang’s plotting remains tightly enclosed. Within a cramped space Hou works yet another variant on his characteristic solution to the dinner-table problem, letting the patriarch’s face pop abruptly into a tiny slot (Fig. 5.103). Yet it is also characteristic of Hou that dramatic turning points develop around meals, not only a spatially dense arena for drama but also a deeply meaningful ritual of community in Chinese culture. In The Puppetmaster, reflecting on the death of his son, Li remarks: “That’s fate. He wouldn’t eat with us. So he’s gone. What can we do about it?” The Puppetmaster’s Special Jury Award at Cannes in 1994 was to be Hou’s last major prize in the 1990s. He was now on his own. He had developed powerful enemies in the local film establishment, who made sure that City of Sadness won no major awards at the Golden Horse Festival. Audiences now avoided his work; Good Men, Good Women earned only US$35,000 at the box o‹ce. Critics, however, kept his reputation alive, de-

HOU, OR CONSTRAINTS

229

5.97 Wen-leung, now momentarily sane, sits at the table with the Shanghai gangsters in a dense long-lens medium shot.

5.98 In an answering framing they oªer friendship.

5.99 An axial cutback yields the master shot, which will define the next phase of the scene.

5.100 After the toast Red Monkey brings in a new batch of bargirls on the far right.

bating whether his films constituted valid ways of representing national politics and history. To sustain himself, Hou directed Japanese television commercials and sold his films abroad. By now the Taiwanese industry had withered. Hong Kong films had conquered the market, and local production dropped to fewer than twenty titles per year. More embarrassingly, Hong Kong films routinely won the major Golden Horse prizes. Once import quotas were lifted, Hollywood films poured in, eventually swamping the Hong Kong product. The government tried to stimulate production, mounting subsidies and negotiating coproductions. The CMPC funded films by Ang Lee, Tsai Ming-liang, and other young talents. Although local audiences liked Lee’s films, government planners recognized that the real market lay abroad. Soon festival recognition arrived, with Lee’s Eat Drink Man Woman (1993) and Tsai’s Vive l’amour (1994) and The River (1997) winning major prizes. Hou was becoming an elder statesman, producing films by his protégé Hsu Hsiao-ming (Dust of Angels, 1992; Treasure Island, 1995) and establishing a postproduction facility that allowed filmmakers to record direct sound. Tsai and less prominent filmmakers were strongly influenced by Hou. If the long take had been an identifying tag for the New Taiwanese Cinema of the 1980s, it became vir-

230

HOU, OR CONSTRAINTS

5.101 As the women take vacant seats around the table . . .

5.102 . . . the gangster patriarch’s face jerks into a slot left of center, framed by two other faces. His voice rings out: “What is this? We’re talking business here!”

5.103 Ah-ga, on the far right, asks one woman to come sit next to him, but now the rest of the scene will concentrate on the old man’s pressure on Wen-leung, just right of center.

tually a national brand in the 1990s. Filmmakers seemed to compete in creating unmoving single-shot scenes lasting many minutes.55 In Tsai’s first feature, Rebels of the Neon God (1992), a scene in a video parlor makes use of simple but eªective blocking and revealing (Figs. 5.104–5.105), and his later films explore a more spare use of depth and darkness, particularly favoring the medium-shot scale. Hou’s collaborator Wu Nian-jen made sparse, evocative use of such devices in A Borrowed Life (1994; Fig. 5.106). This style spread to other parts of the Pacific, making “Asian minimalism” something of a festival cliché by the end of the 1990s. Korean films by Hong Sang-soo (the director of the drinking scene analyzed in chapter 1) and Lee Kwang-mo (for example, Spring in My Home Town, 1998) were built on the static long take, although not always in long shot.56 For decades Japanese films had imaginatively exploited the deep-space long take, but now directors were favoring planimetric, almost fashion-shoot imagery.57 The ravishing but barely changing long shots of Hirokazu Kore-eda’s Maborosi (1996) were attacked in Japan for being too much like Hou (Fig. 5.107). The breakthrough works of China’s Fifth Generation, such as Yellow Earth (1984) and Red Sorghum (1988), were vigorously edited, and neither Chen Kaige nor Zhang Yimou gravitated toward long takes or complex staging. Eventually, however, there emerged mainland directors who did. The most famous ex-

HOU, OR CONSTRAINTS

231

5.104 The hero of Rebels of the Neon God (1992) in a videogame arcade.

5.106 A Borrowed Life (1994): aperture framing for the bride stepping out of the rickshaw.

5.105 As he moves closer to study the poster of James Dean, he clears a path for the gang to come through the aisle.

5.107 Maborosi (1996): the widow and her child wait to meet her new husband in a distant planimetric shot.

ample is Jia Zhang-ke, whose Platform (2001) tells a story of traveling players in a distant, contemplative style that irresistibly recalls Hou. By 2001 a PRC product like Wang Chao’s The Orphan of Anyang could be shot in a string of dedramatized, planimetric long takes averaging half a minute apiece.58 Few of these films engage in the intricacies of staging that Hou favored; most rely on simple figure movements and attenuated scenic development. Still, now that variants of the Hou brand were available to all, what could distinguish his own work? With each film after the trilogy he seemed to take a new tack. Goodbye South, Goodbye (1996), a tale of two petty gangsters and their ill-fated schemes, marked a return to the world of contemporary young people. It was a troubled project, with an exceptionally unsettled script, three bouts of shooting, and a first cut of six hours chopped down for a Cannes deadline. Hou then retreated to the past, turning a sprawling 1893 Chinese novel into a chamber drama. Flowers of Shanghai (1998) is set wholly in the city’s brothels quarter, centering on three women and their tangled relations with their clients. The film cost US$3 million, and across sixty shooting days Hou filmed most scenes ten or twenty times.59 Flowers won no major prizes but did attract support among critics, some of whom considered it one of the great accomplishments of the 1990s. Few tastemakers, though, got behind Millennium Mambo (2001), another foray into youth culture, starring the Taiwan-born

232

HOU, OR CONSTRAINTS

ASIAN MINIMALISM: THE COMPASS-POINT SCHEMA Influential as Hou’s use of the long take has been, another cluster of stylistic options emerged in East Asian film during the 1990s. In this approach the director sets the actors facing the camera or at ninety-degree angles to it, with any background planes perpendicular to the lens axis. The image thus becomes a planimetric one (Fig. 5.107; compare Figs. 4.45 and 4.48). A scene will be built out of such images, linked by what we might call “compass-point” editing: a shot is taken at 90 or 180 degrees to the previous shot. The string of head-on or profile views, so diªerent from the three-quarter frontal angle of most mainstream world cinema (Figs. 1.20–1.26, 1.47–1.56), is reminiscent of a simple comic strip or television’s South Park. In Japan Ozu had developed many subtle variants of this schema in the 1950s and 1960s (Figs. 1.4–1.5), so it’s not surprising that we find it in the Ozu pastiches of Suo Masayuki (Figs. 5.108–5.109). Kitano Takeshi seems to have adopted the schema more spontaneously, partly, he has explained, because he had no training in film direction and found this the simplest and most natural way to arrange a scene.60 Kitano strips down the graphic design of each shot while also varying the number of figures and changing the setups slightly (Figs. 5.110–5.111). The schema is very similar to one that appeared in European film in the 1980s, and it would be worth investigating whether there was some influence—perhaps through the festival milieu—or if the two regional styles developed independently.

5.108 In Sumo Do, Sumo Don’t (1992) Suo imitates Ozu’s college comedies . . .

5.109 . . . and finds widescreen equivalents for his 180-degree shot/reverse shots (Figs. 1.4–1.5).

5.110 Using compass-point editing and linear staging, Kitano cuts from onlookers . . .

5.111 . . . to a beachfront sumo match in Sonatine (1993).

HOU, OR CONSTRAINTS

233

Hong Kong sex symbol Shu Qi. Assessing this gorgeous, thinly plotted descent into the Taipei technobar scene, one writer called Hou “a stylist in search of content.”61 Even trimmed by twenty minutes for international release, Millennium Mambo found little distribution. The very title of Millennium Mambo suggested a man striving to keep up with the times. He made a flurry of plans. He would produce a TV miniseries with a boy band. Mambo was to be the first of a trilogy on young people. No, it would be one of ten films on the subject, shot on Super-16 mm stock.62 He would launch an interactive Web site, buy foreign films for local distribution, and convert the former US ambassador’s residence into a small cinema and cafe. He announced plans for an Ozu homage, financed by Shochiku and set in Tokyo.63 To meet his company’s payroll, he would have to make lots of commercials. At the start of a new century Hou may have lacked what he had enjoyed for two decades: an energetic dialogue with other filmmakers. In the early 1980s he had accepted commercial norms but then revised them for his own ends. The most important professional filmmaker to join the New Cinema, he was willing to write scripts for newcomers while still forging an approach so distinctive that he came to typify the movement. As the autobiographical trend faded and he realized that his audience lay overseas, his trilogy undertook to speak for his nation’s history: “I try to show the life-force of the people of Taiwan in my films. Those who have fallen by the wayside or been crushed by the system are unaware that they have a right to claim our attention.”64 As his reputation grew, he may have tried to outrun his imitators through more flagrant stylistic experimentation, as in the single-shot scenes of Flowers of Shanghai; at intervals he would retreat to his old refuge, movies about urban crime (Daughter of the Nile; Goodbye South, Goodbye; and Millennium Mambo). If, early in the new century, having influenced a generation of East Asian filmmakers, Hou seemed unsure of his next step, perhaps it was because he had become anointed as a Great Filmmaker. Splendidly solitary, he was expected to conjure up each film out of a personal “worldview” rather than from tangible memories and reactions to the world he knew best, filtered through his creative response to current cinematic norms. Still, the dialogue was not completely over. At the level of style Hou’s most recent films reveal a search for new constraints that would stimulate his explorations. The process was already at work, I think, in the sweeping tracking shots of the modern tale in Good Men, Good Women, but he carried his urge to vary his style further in Goodbye South, Goodbye, where the camera occasionally tracks and even cranes over the action. Yet the foundation of his work remained dense staging, sometimes with the telephoto image (Fig. 5.112), occasionally with the wide-angle lens. One scene in Goodbye South, Goodbye, when Kao discovers his partner’s girlfriend supine from a drug overdose, shows that the wider lens allowed Hou to pack a set with subtly shifting cues, including unexpected moves into the most distant areas of the set (Figs. 5.113–5.116). By contrast, in Millennium Mambo the long lens jams together foreground objects into an out-of-focus clutter that impedes our access to the middle-ground story action in an almost Sternbergian way (Fig. 5.117).65

234

HOU, OR CONSTRAINTS

5.112 Goodbye South, Goodbye (1996). Compare Fig. 5.51.

5.113 Goodbye South, Goodbye: the gangster Kao enters Flat Head’s room and demands to know what has happened. Flat Head is unconcernedly shooting baskets, and Kao’s body blocks the crucial elements of the scene.

5.114 Kao moves aside to reveal a second woman in the middle ground, stretched out drugged in the bed.

5.115 Maddened by Flat Head’s casual bouncing of the ball, Kao circles him, comes back to the foreground, and flings a Coke can at him. He finally moves to berate him at the distant window. The woman hops up from the bed to stop Kao, setting the punching bag swinging— another distraction from what is about to come.

5.116 When Kao returns to the foreground, Flat Head quietly slips out the window in silhouette—centered and unblocked by Kao. The scene will continue with Kao and the two women.

5.117 Millennium Mambo (2001): the heroine and her boyfriend in their cluttered apartment.

HOU, OR CONSTRAINTS

235

5.118 Flowers of Shanghai (1998): small changes of position in depth carry the drama when Master Hong negotiates with Pearl to free Shuren from a marriage promise.

5.119 Gossip and drinking games around the table in Flowers of Shanghai. Compare Figs. 5.30 and 5.53.

Oddly echoing the distracting foregrounds of the 1970s wenyi romances, this contemporary recasting of some customary Hou schemas conjures up a hazy, drug-fueled hedonism.66 Also reminiscent of von Sternberg are the radiant decor, costume, and lighting of Flowers of Shanghai, a hothouse atmosphere rendered by means of the most rigorous constraints of Hou’s career. The scenes take place wholly within brothels and courtesans’ quarters, so Hou’s characteristic recapitulative narration now restricts space as well as time. Not only must skipped-over plot developments be filled in, but any action taking place elsewhere must be reported in the softly lit dining rooms and bedchambers. Moreover, every scene but one consists of a single shot, linked by slow fade-outs and fade-ins. Hou claims that the style derived naturally from the novel, which consists of journal entries kept by a prostitute: “The plan-séquence seemed to me the appropriate form for treating each moment in a separate and autonomous way.”67 As if all this weren’t confining enough, each of the sequence shots is presented by a camera in steady, almost mechanical motion. It may pan slightly or track slowly in or back; often it arcs slowly right, then left, then right again. Far from suggesting a curious onlooker strolling around the characters, this pendular movement creates its own formal pattern, a warping or twisting of space that meshes rhythmically with the glances, gestures, and dialogue of players gathered around a table. Most scenes lack music, but when music does emerge, it is a synthesizer drone, with violins, bells, drum taps and wheezing harmonium-like timbres drifting in and out. Instead of spiking at a line or gesture, the repetitive whine and wail edges, like the camera, along its own trajectory. Every scene of Flowers deserves detailed analysis, for each is a stylistic and dramatic tour de force. For my purposes I will note only how Hou refines the just-noticeablediªerence strategy. Here not just one action may hide or reveal another; many actions hide and reveal many others. Each shot balances diªerent centers of interest, pitting spatial cues against one another. Frontal and centered figures become counterweighted by a closer figure who shears them oª from time to time, or a brilliantly lit but immobile

236

HOU, OR CONSTRAINTS

face will ponder the gestures of someone turned three-quarters from us (Fig. 5.118). With a calm inevitability the side-winding camera generates an ever-changing background and foreground (Plates 5.7–5.8). The camera’s shifts are often barely discernible, like the faces of the courtesans in the gloom around a table. The habitual play with perspective, the momentary masking and revealing, and the tiny details blooming and vanishing in seconds—all are here provided not only by figure movement but also by a camera which quietly and constantly realigns an ensemble of seated characters. The eating and drinking scenes (there are many) no longer consist of closely stacked lateral planes but rather circular arrangements captured by a ceaseless rotation. Although the shots don’t employ a very long lens, the space-squeezing lessons of the telephoto are applied once more, as faces pile up in depth or along a diagonal (Fig. 5.119). In one setting a pair of oil lanterns occupies the center point of the camera’s arc, and even their sparkling distortions become a source of minute pictorial play. You can become hypnotized by those lanterns, or by the maestoso rhythm of the whole construction. This is style as denotation, expression, and decoration all at once. Across each scene the mise-en-scène creates a languid hallucination of glowing softness and slight movement out of which the characters’ behavior and speech arise and to which, at each scene’s end, they return, voices dying, figures fading, faces dimming, with only the gleam of lanterns holding on until the screen goes completely dark. Lanterns will be the first points of light to dawn in the next scene. As a virtuoso turn Flowers of Shanghai makes the stiª, stripped-down long takes of East Asia’s younger generation look like the easy option. By displaying a splendid milieu in a stringent way, Hou preserves his commitment to the sensuous particular, to behavioral concreteness, to visible darkness. From end to end Flowers of Shanghai is composed of microelements, just-noticeable diªerences. A drama precipitates out of them, but they are always available to be savored as particles floating through an evocative world. Call it precious and mannered, or sublimely filigreed, according to taste. In any event if Angelopoulos can be accused of stamping a patented array of techniques onto each subject he tackles, Hou escapes that charge. His circumstances and his creative vigor have constantly set him fresh problems. In the course of his career he has maintained the tradition of complex staging by adapting it to new technologies (telephoto, widescreen); he has explored it for powerful denotative and expressive eªects; and he has become a “strong precursor,” reintroducing pictorial intricacy to Asian filmmaking in forms that cannot be ignored. His ongoing dialogue with his contemporaries has left us many original and deeply aªecting films, and he has shown that craft practices running back to the 1910s, when subjected to the pressure of fruitful constraints, can launch new adventures in cinematic vision.

HOU, OR CONSTRAINTS

237

6 STAGING AND STYLISTICS

For André Bazin cinema fulfilled the mission given in the name of its medium—photography, light-writing—by tracing the texture of the world onto the sensitized filmstrip. I grant that recording reality is one important function of cinema, but not all films have their origins in photography. Whether the moving image is etched by the phenomenal world, painted on an animation cel, or percolated in a computer program, it oªers traces that activate our eyes and mind. It yields spoors and trajectories, and we viewers pursue them, guided by their patterning, their context, and our sense of what is likely to be significant. The word trace, derived from tractus, a track or course, should remind us that cinema is in part a pictorial art, oªering a visual array that marks out a space for our apprehension. Tractus implies both carving a path (as the filmmaker does) and following another’s path, as we do in watching the film—or in studying the image traces considered in the preceding chapters. Throughout those chapters I’ve presumed that moviemaking is an art. This is not the only worthwhile way to understand film, but it ought to be pertinent to understanding some ways in which the bodies traced onscreen can engage us. I have not considered cinema as an auditory art (although it is that) nor as a theatrical one (although it’s that too). In studying cinematic staging I have concentrated on its pictorial aspects (for it is a graphic art) and its narrative ones (for it’s that as well). Representational cinema, like figurative painting, relies on perspective projection, that infamous optical pyramid inscribed into the lens and the camera. The filmmaker who wishes to tell a story is obliged to pursue pictorial strategies. He or she narrates visually, and in this project the deployment of figures in space is critical.

238

STAGING AND STYLISTICS

I’ve examined some traditions guiding filmmakers’ arrangements of figures in space— Eisenstein’s “mise-en-shot”—considering both the images and the ways they shape our understanding of the stories they tell. Since most of what we see (and hear) in a film resists abstract interpretation, we need other coordinates to track the pervasive flux of those pictures and stories. Instead of reading through style to theme, I have dropped down to the level of small-scale changes, the better to gauge how the filmmaker, who must manage time as well as space, sculpts an unfolding action. I’ve sought to show that beyond those (fairly uncommon) thematic-symbolic aspects, we may study staging for its “experiential logic,” the patterning of denotation and expressivity moment by moment. My expository strategy has been to focus on single scenes in order to see how a dramatic development may unfold pictorially. Unfortunately, this has slighted the overarching structure of the entire film. At times I’ve been able to show how a sequence echoes, modifies, or contrasts with earlier or later ones; but we should try to go farther still, tracing holistic principles of patterning more thoroughly than I have been able to do here. Our inquiry has spanned nearly the history of cinema, from Feuillade’s debut in 1908 to the still-ongoing oeuvres of Angelopoulos and Hou, with side trips into early sound filming and the work of Antonioni. We have seen remarkable changes in our filmmakers’ creative milieus. Feuillade flourished in a studio-driven production system that rewarded speed and mastery of craft. Japan’s studio system permitted Mizoguchi to create a daringly distinctive style that was in his last years recognized on the emerging festival circuit. That circuit provided crucial support for Angelopoulos and Hou, but their immediate circumstances oªer fascinating contrasts. Both emerged from a “New Cinema,” but whereas Angelopoulos was from the start geared to the tastes of the international art-film audience, Hou began in popular cinema and had to adapt himself to the festival scene. All these institutional diªerences presented the filmmakers with unique options and challenges. Stylistically, we have passed from a time when long-take ensemble shots were the norm to a time when they were seen for the most part in festival cinema. (As one historian has put it, “The higher the pretensions, the longer the take.”)1 We have moved from fixed camera to moving camera, from moderately wide-angle imagery (Feuillade) to bold wideangled shots (Mizoguchi) and then to long lenses with narrow angles of view (Angelopoulos, Hou). Each filmmaker has taught us something fresh about the resources of staging: the clean e‹ciency of Feuillade, the infinitesimal modulations of Mizoguchi, the brooding monumentality of Angelopoulos, and the density of the mundane provided by Hou. Yet all four artists have tapped into some of the same compositional resources. Each has explored the possibilities of depth in the frame, both as veering diagonals and planes set perpendicular to us. Each has shown the power of creating pockets or slivers of space that, sooner or later, become filled. Each has shown how aperture framing can beguile our eye and asterisk a minor action. Each has demonstrated the force of frontality and dorsality, of centering and decentering, of packed images and stark ones, of distant framings and really distant framings. Each has shown that a large part of the director’s work

STAGING AND STYLISTICS

239

is judiciously concealing elements and then revealing them at the right moment, then concealing them again, only to reveal something else . . . indefinitely. Remarkably, these tactics have gone largely unnoticed by conventional criticism. We’ve been able to trace them by watching closely, by assuming that directors typically steer our attention to salient story material, and by adopting a comparative method, bringing each filmmaker into alignment with the norms of the cinema of his time. Yet despite our chronological span, we haven’t arrived at a large-scale account of the evolution or development of film style taken as something over and beyond the work of all filmmakers. In On the History of Film Style I traced three principal research programs in this domain. How might my account here square with those? What I called the Standard Version of stylistic history appeared in the silent era. It presupposed a linear unfolding of cinema’s potential: not the recording of reality but the reshaping of reality to present a “purely filmic” space and time. This conception served to valorize the expressive imagery and experimental montage that emerged during the 1920s. Proponents of this view often went on to claim that once cinema’s array of specific resources had been revealed, they were repressed by the more “theatrical” style ushered in by sound films. A second research program, emblematized by Bazin’s dialectical history of film style, saw not one but two developmental lines running through film history: a “faith in the image,” which he believed hit its apogee in the late silent filmmaking, and a “faith in reality,” a trend that was subordinate in the silent period but gained strength with the coming of sound. That trend, basing itself on photographic recording and techniques of depth staging, deep-focus cinematography, and long takes, was said to develop in the 1930s and 1940s, with the work of Renoir, Welles, Wyler, and the Italian neorealists. A third research program, developed by Noël Burch in the 1960s and 1970s, I called the “oppositional” one. Burch proposed that we could conceive the history of film style as a constant clash between the “Institutional Mode of Representation” (basically, the storytelling techniques of mass-audience commercial cinema) and various alternatives: the “Primitive Mode of Representation” that preceded it, the distinctive cinematic approach of Japanese filmmakers, and a range of avant-garde practices in the West. Influenced by the theories of modernism in other arts, Burch suggested that we could consider much stylistic change as a reply to, critique of, or attack on the dominance of the IMR. All three research programs have left us a wealth of ideas, observations, and insights, but On the History of Film Style also found fault with them. I argued that there’s no reason to expect that the whole of stylistic history can be captured in one sweeping narrative. When we look closely, we find both continuity and change at many levels, not an inexorable river but rather streams, eddies, and pools. Moreover, both the Standard Version and Bazin’s dialectical account presuppose that the film medium has an essence that unfolds over time, with human action serving as the vehicle of this gradual emergence: the important film artists move ever closer to discovering the truly cinematic properties (“pure cinema” or faithful recording of reality). Above all, I argued that all three research programs are characterized by what E. H. Gombrich called a “Hegelianism without meta-

240

STAGING AND STYLISTICS

physics,” an eªort to find a vast pattern of change (linear improvement, dialectical surpassing, permanent opposition) across a multitude of distinct works, artists, and traditions.2 Rising far beyond individuals and concrete institutions, this account makes the individual work merely the projection of the a priori categories of an aesthetic system. This frame of reference removes agency, diversity, nuance, and sheer unpredictability from history. For such reasons our filmmakers don’t fit snugly into these grand schemes. The Standard Version critics considered Feuillade retrograde and theatrical, despite the fact that some of his contemporaries believed the truncated pyramid of filmic staging to be as “specifically cinematic” as the Standard Version theorists considered editing to be. Mizoguchi’s aggressive foregrounds predate those of Welles and Wyler, but he seems to have quickly felt the limits of what he could do with this device. So he pursued other strategies, such as the empty foreground and the information-packed depths, which find no place in Bazin’s “dialectical” account. Angelopoulos might accord with Burch’s oppositional theory somewhat, but to conceive of his work as simply a negation of Hollywood ignores the Antonioni-based tradition that he extends and enriches. Hou’s crowded, spreadout frames might seem a reversion to Burch’s Primitive Mode (one critic calls Hou a “neoprimitivist”),3 but ironically, the director seems to have arrived at his style by revising schemas of Taiwanese commercial filmmaking. Instead of treating any large-scale history of style as a template into which directors must be squeezed, I’ve sought to show that the stylistic continuities and changes we find in film history are most adequately understood as the result of human agents working within institutions and exploring, often through trial and error, the medium’s capacity to fulfill certain functions. Those functions vary, and what resources are revealed depend on filmmakers’ purposes (themselves often in a process of change). We find convergence among stylistic practices, but this isn’t because a common essence of the medium has been disclosed. The convergences spring from filmmakers facing similar tasks, or having similar proclivities, or sharing repertoires of schemas, or making congruent choices, or just using the same technology (notably the camera’s optical projection system). In particular, filmmakers who employ the long take and distant framing confront certain problems regardless of where and when they work. Hence schemas known in one period and place can be rediscovered independently, as Hou rediscovered visual tactics employed by Feuillade (while turning them to diªerent ends). Rather than embracing large-scale histories of Style, we can work comparatively and inductively. We can try to chart the variety of stylistic manifestations at particular periods and in specific production contexts, always remaining alert for both normalized practices and transformations of those norms. Accordingly, the chapters here have hewed pretty closely to concrete cases, moving to broader trends when they helped explain a shared range of choice or a likely influence. Some readers, however, might find it helpful to reflect on the theoretical presuppositions and implications of these studies, so I end the book with a more abstract explanation of my approach. How, we might ask, do

STAGING AND STYLISTICS

241

we go from an inductive style-centered criticism to a more principled stylistics, or theory of style? Readers with no appetite for theory can skip to the closing section of this chapter, but some will want to know how this approach stacks up against other theories in academic film studies. I’m also obliged to counter some objections to this orientation, and these I take up first.

While writing this book I heard over my shoulder a reader asking in exasperation: “But what role does culture play in all this? You consider the community of creators, but what about film’s relation to the audience and to wider social collectivities—ethnicity, subcultures, and the like?” The complaint isn’t completely well-founded: I do consider cultural factors (for example, the tough heroine emerging as a character type in 1930s Japan or the nativist impulse that aªected Taiwan’s New Cinema). Still, my analyses’ emphasis on internal patterns and the filmmaker’s craft context do run counter to the dominant trend in film studies. The prevalent approach conceives of history on a broad scale and seeks to interpret filmic phenomena in large-scale cultural terms. It’s worth recalling the question informing this book. I’m asking what aesthetic possibilities emerge when cinematic staging is employed in particular ways, in relation to a‹liated trends. Scholars working in the culture-based vein pursue more macroscopic and less film-specific questions, such as “How do national cultural practices aªect media?” or “How have technological and social change altered human sense perception?” By pitching questions at this level of generality, proponents of the culturalist perspective don’t present it as an alternative to the account I’m setting out. A social history of media is not necessarily a rival to a poetics of cinema; each simply studies diªerent aspects of a rich phenomenon. As confirmation of this point, the great bulk of cultural inquiry into cinema ventures no claims at all about film style. And this seems unobjectionable. Not every question we ask about cinema need consider aesthetic issues. Yet many scholars do seem to expect that a broad cultural perspective ought to yield insights into how films work. Perhaps the reader prodding my shoulder is a “culturalist,” one who believes that the explanation of any cinematic phenomenon should look first to broad cultural factors. Now taken strictly, this verges on dogmatism. In no domain of inquiry does a single theory answer every intriguing question. Still, the objection is so persistent that I’m obliged to assess to what extent a cultural perspective can analyze and explain the historical dynamic among film styles. There’s no doubt that broad social and cultural factors shape artworks in many ways. Subjects and themes come directly or indirectly from a cultural milieu, and these aspects of an artwork can be studied systematically, as through iconography in art history and rhetoric in the verbal arts. A poetics of cinema can usefully treat subject and theme as materials that are given distinct forms by filmmaking traditions and individual films, as I’ve indicated here and elsewhere.4 Social and cultural factors can also be considered preconditions for artworks. Without capitalism there would have been no French film industry

242

STAGING AND STYLISTICS

and thus no Feuillade films. Still, it would be odd to say that capitalism created the films in the same sense that Feuillade and Gaumont and a group of craftspeople created them. Capitalism is thus a precondition for more local causes. Sometimes social and cultural institutions mandate the forms or materials or functions of artworks. A given milieu can encourage, even dictate, particular styles, as with Socialist Realism under the Soviet regime. Such a “social command” situation can often be understood within the ends/means, problem/solution, choice-based model I have been advocating. For example, I suggested that after 1937, when the Japanese government asked filmmakers to conform to a policy of celebrating “national spirit,” Mizoguchi’s reverent treatments of artistic traditions at the level of the shot proved appropriate. Yet not all filmmakers’ work sought the poised intricacies of The Story of the Last Chrysanthemum and Genroku Chushingura. Ozu, Naruse, Kurosawa, and others responded to government dictates quite diªerently. The institution dictates the ends or the problems, and sometimes the means; but the artist may achieve the ends in distinctive ways. Things get more complicated when we try to show that patterns of continuity and change in film style are not mandated by explicit policies but are caused by less clearly defined forces within a culture or period. Many scholars would hold that artistic developments result from a complex interaction of social, economic, cultural, and even socialpsychological processes, all lying outside filmmaking practice or its craft traditions. In this sense we can think of culture as a distal or ultimate cause of style, whereas I have been advocating explanations that operate proximately, bearing more directly on the historical agents and their artistic circumstances. The artworks bear the traces of decisions. The artist makes choices about forms, materials, tools, and techniques, and many of these factors are defined within the traditions of the craft milieu. The culture-based historian will reply that each of these proximate factors is shaped, even determined, by processes extending far beyond the individual or the art-making institution. So the question is, How can this happen? Theorists provide diªerent sorts of explanations on this score, but within film studies I think two broad strains are relevant. Some scholars would emphasize national cultures as pervasively shaping style. We might think of Mizoguchi’s style as distinctively Japanese or Hou’s as distinctively Chinese. In what respects? Not in terms of “national character,” a shaky notion at best. Instead, the usual candidates are national artistic traditions. For example, we might hypothesize that Hou’s use of landscape has been caused, through some process yet to be described, by classical Chinese landscape painting. Alternatively, instead of invoking national cultures, we might account for stylistic features by invoking sociocultural processes pervading a period and crossing national boundaries (although the processes may be inflected diªerently in various places). I call this second perspective epochal. A typical epochal explanation would seek to show that some features of Feuillade’s style are caused by aspects of urban modernization. Another possibility would be to see Angelopoulos’s or Hou’s style as caused by conditions that have been identified as “postmodernity.” Let me take the two broad approaches in turn.

STAGING AND STYLISTICS

243

National cultural traditions, particularly as manifested in adjacent media such as theater or painting, are often relevant to causal and functional explanations of style. They are most tenuous, however, when they are oªered in collective and impersonal terms. For instance, it is unlikely that Mizoguchi’s inclination toward high-angle views or blockage of information stems straightforwardly from his belonging to a culture with such traditions in its graphic art. Many of his Japanese contemporaries display no particular inclination toward such images; and these techniques may be found in non-Japanese films as well. We also have to explain why his films don’t display other features of classic Japanese art, such as unshaded and unshadowed surfaces. It is more plausible, I think, to postulate that Mizoguchi—trained as a Western-style painter, sensitive to East Asian graphic arts—occasionally opted for certain pictorial devices because they served the dramatic purposes of particular scenes or achieved certain ends, including that of recalling traditional Japanese art. In other words, instead of unreflective transmission of a tradition we have more or less deliberate choice, perhaps even knowing citation. (See my discussion of Genroku Chushingura and Utamaro and His Five Women in chapter 3.) Similarly, when Hou claims that his use of empty space recalls classical Chinese compositional techniques, we need not commit ourselves to the idea that the pictorial tradition, flowing throughout Chinese culture, suddenly surges up in one man’s films. For then we would have the problem of explaining why these pictorial traditions don’t dominate Chinese cinema as a whole. Rather, it seems more likely that Hou pragmatically discovered some compositional devices in response to his own purposes and procedures, and he realized that these devices have analogues in other media.5 It’s also worth recalling that cinema seldom straightforwardly reproduces techniques taken from adjacent arts. Someone might argue that Hou’s magnificent landscapes are indebted to the great traditions of Chinese classical painting. But then we should have to allow that cinema cannot achieve the perspectival deformations of space found in Chinese pictorial design.6 Hou as a filmmaker is bound to monocular perspective, whereas the painters often present multiple perspectives in the same frame. Moreover, Hou’s films contain many images that are not landscapes; will there be a painterly source for each type of shot? It’s more likely that filmmakers snatch up schemas opportunistically, bending this or that element to particular purposes. The historian’s aim is to move from cultural factors to stylistic features not in one broad leap but in graceful short steps. Of course filmmakers’ cultural tastes, their susceptibility to fashions, and their awareness of belonging to a tradition do impinge on their work. But such factors carry us from the realm of distal causes to proximate ones. When considering any national-cultural tradition as it makes its way into film, I would argue that we should take the filmmakers, their peer groups, their medium, and their institutions as the “final filters,” not just gatekeepers but transformers, always inclined to recast and repurpose the materials coming their way. Once more, the filmmaker, not culture, switches on the camera. Some similar points can be made about epochal explanations of style. The most elab-

244

STAGING AND STYLISTICS

orated line of argument has developed around the idea that cinema as a medium is part of the arrival of modernity between 1850 and 1920. That vast process shaped (some would say determined) not only films’ subjects and themes but also early film form and style. How? Many theorists believe that modernity radically changed experience. They adhere to the belief, first fully articulated by Walter Benjamin, that over long periods “the mode of human sense perception changes with humanity’s entire mode of existence.”7 In the modern era, Benjamin argued, people experience the world in a fragmented, distracted way.8 Accordingly, some film scholars have proposed that early film technique corresponds in some regards to changes in ways of seeing and sensing the world within the advanced industrial societies. How much will this explain about style? The theorists seem undecided. For example, in replying to criticisms of the modernity thesis’s explanatory power, Tom Gunning writes: Bordwell’s and [Charlie] Keil’s claim that the modernity thesis cannot explain stylistic change is probably correct, but seems to defeat a claim that no scholar of early cinema has ever made. . . . Reference to the broader contexts of modernity cannot, and does not desire to, explain everything. Changes in film style derive from many immanent causes: changes in technology, industry realignment, cycles of innovation and canonization, as well as transformations in film’s relation to society.9

Yet when Gunning seeks to explain stylistic change, he summons up exactly those “broader contexts of modernity,” not the features he claims are most “immanent.” Consider his discussion of crosscut last-minute rescues in D. W. Gri‹th’s early films: An urban and industrial working-class audience might find this race against the clock and rhythmic division of time strangely familiar. . . . The syncopated rhythm of ragtime and the mechanically produced sensations of speed and force in amusement park rides reproduced the new and often repressive experiences urban workers encountered and transformed them—temporarily—through play. . . . Likewise, Gri‹th’s parallel editing invokes the split-second timing of industrial production and workers’ enslavement to an oppressive temporality.10

Presumably there are several causal forces at work in Gunning’s account—including Gri‹th as a historical agent—but urbanization, mechanization, and the industrial division of labor clearly play the starring roles. Routinized, time-pressured work processes were “reproduced” in ragtime, amusement park rides, and Gri‹th’s crosscut rescues. Stylistic changes—Gri‹th’s innovations—are explained by appeal to broad social and economic forces characteristic of modernity. No intervening variables are invoked, nor is there any suggestion that they have the power to shape the phenomenon to be explained. Likewise, what Gunning calls the “cinema of attractions,” that type of filmmaking that aims to appeal to the spectator through surprise, shock, and acknowledgment that s/he

STAGING AND STYLISTICS

245

is being addressed, owes its distinctive manifestations, in part at least, to modernity’s creation of consumers. “The cinema of attractions develops out of a visual culture obsessed with creating and circulating a series of visual experiences to stimulate consumption.”11 Ben Singer, who has applied the modernity thesis to early film most systematically, echoes Gunning in declaring that modernity is not the ultimate source of everything that might interest us in early cinema: No proponents of the modernity thesis purport to isolate the only historical force governing cinema. It goes without saying that a great many forces shape cultural artifacts like cinema (although, to be fair, the criticism is not undeserved, given that articulations of the thesis generally have made little eªort to include such qualifications, or acknowledge potential counterarguments or alternative explanations). Movies were responsive to developments in cinematographic technology, stylistic and generic cycles, industry practices, shifts in cinema’s social functions, and so on. All the modernity thesis aims to do is add “environmental context” or “sensory experience” to that list.12

Yet in the next sentence Singer indicates that the last-named items have explanatory priority: “If some films do not fit well with the modernity thesis (for example, a slow film with minimal editing, simple scenery, and subdued acting), we can assume it is because other determinants took precedence.”13 Now modernity is not one item on an open-ended list; it is the default, yielding the preferred explanation unless you have a clear-cut instance that doesn’t fit. When we find a film with fast cutting, elaborate sets, or extroverted acting, no explanations but modernity need be considered. And those other “determinants,” which would yield slow films and minimal editing, are not identified. How productive is the epochal perspective formulated by the modernity thesis as an explanation of stylistic matters? At present the results don’t seem to me promising. When modernity theorists focus on style, the analyses are usually not very precise or detailed. Any film passage that can be described in terms of powerful fleeting impressions, kinetic rapidity, and the like can be fitted into the modernity thesis. (So rapid editing, or a frantic chase, or sudden plot twists, or images of tra‹c, trains, and skyscrapers all confirm that modernity’s hand is at work.) Similarly, if we start with the concept of a cinema of attractions, any new and striking stylistic tactic can be described as an attraction. It can then be explained as another instance of the drive toward consumption—a process that can include an indefinitely large set of phenomena. In short, the very broadly conceived features associated with modernity skew the inquiry from the outset. They shape the examples supplied and the descriptions written, thereby giving the whole exercise an arbitrary air.14 Let’s go back to Feuillade. Someone could describe, in general terms, his staging as packed with visceral jolts and thus present him as an avatar of modernity. Someone else could, in equally general terms, describe his work as somnolent and backward-looking, out of touch with an audience whose nervous systems beat to the rhythms of boulevard tra‹c. Both claims would be equally ad hoc hermeneutic gestures, producing no new

246

STAGING AND STYLISTICS

knowledge but simply using the case at hand to rea‹rm commitment to a large-scale doctrine. The same sort of mismatch between theoretical scale and explanatory task would, I think, face someone trying to argue that a given batch of films—say, those by Angelopoulos—is emblematic of postmodernity.15 Of course we could study what Feuillade often presents—the city as an endless maze of false appearances and sudden revelations—and interpret it as exemplary of modernity. But in order to explain how Feuillade represents this material, the scholar would need to show how forces of modernity, taking Feuillade as their vehicle, produced his staging strategies, and at his moment. This wouldn’t be easy because modernity, as a large-scale cultural process, is necessarily a distal cause of any particular human action. When a historical explanation of a unique event, like shooting a scene a certain way, posits only distal causes, it courts vacuity. (“Culture made it happen” isn’t much more precise than “God made it happen.”) The historian asking a fine-grained question should try to sketch a plausible chain of midrange causes relevant to the event in question. The modernity account, being top-down and macroscopic, is not designed to explain phenomena at this level of delicacy. It oªers a chisel where we need a scalpel. The appeal of the modernity thesis seems to spring from two sources. First, epochal accounts in general exercise a commonsense power; they promise ultimate explanations. As a causal force, modernity looks more significant than the camera lens’s optical pyramid. Second, the modernity thesis relies on two basic strategies of informal cognition: picking out prototypes (unusually vivid instances) and reasoning by association (the dominant approach to interpreting literary texts).16 Writers in this tradition have typically started by asserting the power of modernity and then illustrated their claim with striking examples. No proponents have sampled a wide range of early films to find out how many are shocking, fragmented, or distracting. Charlie Keil has argued that “attractions” are far less common in the earliest years (1895–1908) and the transitional period (1908–1915) than modernity theorists indicate: The untold number of pedestrian comic episodes, staged vaudeville turns, or confusing brief dramas from the earliest years is virtually never mentioned, and their omission is not simply a function of being unmemorable. These films serve no illustrative purpose. While many films from the period embody the prized aspects of attractions— emphatic visuality, self-conscious employment of cinematic technique, solicitation of a direct, even visceral response—probably even more do not. And the reason these ordinary films tend to be ignored dictates that an even larger number of films from the transitional period will be overlooked, as their formal features don’t help substantiate the modernity thesis.17

After studying more than six hundred films from the transitional years, Keil concludes that both on the level of style and on the level of subject matter, the average film oªers slim support for the modernity thesis.18 At this point the thesis seems less a default con-

STAGING AND STYLISTICS

247

dition than a dogma—exactly a thesis bound to a doctrine, rather than a hypothesis to be investigated. This suggests that proponents of this approach have not been as methodologically selfconscious as they might be. They have not made clear what question they seek to answer, and they have not weighed their preferred answer against rival alternatives. They have not, in other words, been su‹ciently dialectical.19 It’s unlikely that scholars in the modernity camp launched their inquiry with open-ended research questions of the following sorts. What sorts of explanation might we find for feature X in these films? How might one explain the variety of concrete stylistic features of the period? More specifically, how prevalent are attractions or sensations in early film? How might one explain the range of the sample, from attraction-packed films to unsensational ones? And what would constitute feasible alternative explanations of the phenomena? Instead, it seems likely that many theorists have started with Benjamin’s strictures (as he started with Baudelaire’s) and then applied them selectively to early cinema, elaborating them to take in this or that intriguing case. For all the fascinating historical information brought to light about trains and museum layout and international expositions, the results often seem oddly unempirical because the process of inquiry isn’t inductive and data-driven. It is doctrine-driven, working from the top down, seeking arresting illustrations rather than a systematic body of evidence. At best, it seems, modernity might be a precondition for a great many cinematic phenomena, but even granting that, finer-grained causal stories will be needed. Well, someone might say, perhaps the modernity thesis is not particularly powerful as an explanation, but might it not have other values? Perhaps it can activate aspects of films that other perspectives have overlooked. Again, this seems a fair defense when addressing matters of subject matter and theme. Vivid prototypes can reconfigure our interests, and locomotives in early films look a lot more intriguing than they did a decade ago. Yet I think it fair to say that the modernity thesis has proved of no help in making fresh stylistic discoveries. It has revealed no previously undisclosed film techniques, no hitherto unnoticed innovations. All the stylistic concepts used by modernity theorists— close-ups, continuity editing, Gri‹th as the exemplar of crosscutting, and the like—are inherited from straightforward aesthetic analyses and craft practice itself. The theorists have assimilated the discoveries of more inductive research through redescription, often at a high level of generality. For example, Gunning recast his initial functional-formal account of the “cinema of attractions” in top-down Frankfurt School terms: this cinema presents a “miming of the drama of the fascination and mystification of commodities.”20 It seems likely that the modernity thesis is poorly equipped to identify the sort of finegrained qualities that stylisticians can bring to light and explain by proximate functions and causes. The modernity thesis was created to answer broader questions. When we examine certain stylistic features, a commitment to modernity theory, with its assumptions about change and perceptual upheaval, may turn out to be downright blinkered. In assuming that image composition can draw attention to this or that scenic element, Feuillade and his peers were relying on very old artistic traditions. Centering,

248

STAGING AND STYLISTICS

balance, comparative size, foreground/background relations, aperture framing, and the like are centuries-old strategies for composing images.21 At the same time, as we’ve seen, Feuillade employed a variety of ways to adjust those traditions to motion pictures, creating a harmonious visual flow that shaped the viewer’s uptake in subtle ways. He gives us some sensational moments—a poisoning, a snake slithering out of a ventilator—but they emerge from a background of brisk but not perceptually overpowering entrances and exits, looks and gestures. Attractions, if they are there, take their place within stable principles of visual design. A visual-experience-in-modernity account does nothing to explain why filmmakers adopted these sturdy pictorial strategies and how they recast them to the needs of telling a story in time. Indeed, an advocate of the modernity approach has no reason to notice these devices at all, exactly because their sources predate the modern era. I conclude there is good reason to think that “culture” in either national or epochal guise is an answer that does not yet fit the questions we are asking about continuity and change in the history of film style. This is not, one more time, to attack all cultural explanations. (Chisels are handy tools for certain jobs.) We should not rule out broad national or epochal factors as causes of style, but we should try to define them with a precision commensurate with the level of delicacy revealed by inductive functionalist analyses. Ideally, our account would disclose stylistic strategies or tactics hitherto unnoticed. The historian should also argue for such cultural causes explicitly and in stepwise fashion, not simply assume that they operate invariably and pervasively, and should show their influence to be more likely than that of other candidates. Not surprisingly, I expect that the most convincing case will trace how broad sociocultural factors are likely to have impinged proximately on the decision-making agents who shape the film. Again, filmmakers are the final filters, and cultural influences are channeled through filmmaking institutions, craft norms, and the means/ends reasoning so central to cinematic creation.

What is presupposed by putting the craft context at the center of studying film style? (At least style in fictional theatrical cinema, although the idea of craft might illuminate industrial, documentary, and avant-garde cinema as well.) As I’ve treated it, the idea of craft raises questions about conceiving tasks as problems and solutions; conceiving the filmmaker as an active, deciding historical agent; and conceiving filmmaking as a transcultural activity. Each of these deserves a bit more explanation.

THE PROBLEM/SOLUTION PARADIGM The craft context can be understood as a situation posing problems that filmmakers must solve. These are not only momentary production di‹culties, such as coaxing a drunken leading man out of his trailer, but also broader problems of artistic representation. How,

STAGING AND STYLISTICS

249

to revert to my earliest example, should the actors be arrayed around a table? Or how should we bring the hero into the scene so that we notice his clenched fists? How to show the moment when he confronts his adversary? The eªects will vary if we see both characters at once, or one first, or perhaps only one of them throughout the scene. Since artists rarely provide an explicit account of their practice, positing the problem/ solution dynamic is an exercise in inference and idealization. It depends on theoretical assumptions, such as the belief that directors working in a mainstream narrative tradition want us to notice, among other things, pertinent story information. Guided by our awareness of the image’s function, we engage in a sort of reverse engineering. When Feuillade moves a figure to the foreground and then has that character assume an expression and posture (Fig. 2.47), we postulate that conveying the expression and posture was the task he confronted. The likeliest reason to put Philippe’s desk in the foreground and bring him up to it is that we can thereby see his reaction clearly when he discovers the missing Vampires dossier. The figure movement was less an end in itself than a solution to the visibility problem. The whole process, as we’ve seen, can get fairly complicated. One solution can trigger a new problem.22 Advancing to the foreground solves the visibility problem, but it may pose another one—wasting time by presenting a baldly functional walk to the camera. The filmmaker might solve the new problem by endowing the walk with arresting qualities, such as briskness, or by breaking the advance into phases, each with a bit of business that is itself appealing. Such a solution would have another payoª, that of delaying the full arrival at the foreground and perhaps building suspense. As it turns out, Feuillade opts for this alternative, for he makes something energetic happen at each moment: Philippe strides in, hangs up his hat, greets his colleagues—all actions that give us time to notice Mazamette’s anxiety. This solution sets terms for the problem/solution pair that follows. If robust movement characterizes the first part of the scene, how to accentuate Philippe’s discovery of the rifled dossier? Feuillade solves it by shifting rhythm, providing a slow movement of discovery, having Mazamette cautiously retreat, and providing a moment of stasis as Philippe calls him to account (Fig. 2.48). Along these lines, the last four chapters have assumed that the craft context of problem-solving is the most pertinent and proximate causal force acting on the creation of style. Yet isn’t this model too tied to the concrete activity of filmmaking? By thinking in these terms, don’t we limit ourselves to the narrow, even anti-intellectual horizon of commercial moviemakers? “Bordwell’s cognitivism,” John Belton writes, “makes him the soulmate of every Hollywood filmmaker I’ve ever met. ‘All this theory stuª is a crock. We do what we do in an attempt to solve specific problems. You academics just don’t understand how films get made.’”23 At one level, of course, we should try our utmost to understand how films get made, and it would be an incurious scholar who didn’t want to know more about the creative process. More abstractly, in studying any historical situation, we should try to see things from the social agent’s perspective. But that doesn’t mean that we are limited to that perspective. Our inquiry doesn’t stop with a reconstruction of the film-

250

STAGING AND STYLISTICS

maker’s context of choice. We also seek to disclose a broader tradition of linked problems and solutions within which a given stylistic feature becomes intelligible; we try to compare alternative traditions and to find historical precedents. The larger story we tell is not likely to lie within the ken of most filmmakers. Further, I would propose the problem/ solution model as a candidate for understanding many filmmaking traditions, including recondite experimental schools. Those traditions aren’t part of the intuitive worldview held by LA types. And, contrary to what Belton’s imaginary filmmaker thinks, the problem/solution model is a theory, with presuppositions and conclusions that can be challenged.24 More broadly, does the problem/solution model concede too much to “instrumental reason”? James Hurley objects to the approach because he believes that this model would please his students, “especially those from non-humanities disciplines.” That’s because it “accords, all too neatly, with the instrumental approach to reason that is the operational logic of these students’ business, economics, and pre-Law courses—the logic they will then take to the life-world that awaits them outside the academy.”25 If Hurley is suggesting that this “logic” is always unwelcome, I would reply that it’s of great value in certain situations. If you hire a tax accountant, you will be best oª with one vigorously committed to problem-solving. (You don’t want one who will produce a Lacanian reading of your IRS audit notification.) If Hurley is suggesting that instrumental reason may be appropriate to business and law but not to the arts, he oªers no support for making the distinction. Nor is the problem/solution matrix specific to modern capitalism or today’s university, as Hurley implies. Marx thought in exactly these terms.26 In the history of the visual arts the problem/solution paradigm dates back to Renaissance thinkers like Vasari, and I think it can also be plausibly attributed to Aristotle in the Poetics. It’s telling, I think, that Belton and Hurley propose no alternative to the problem-solving model. For at bottom it isn’t just movie moguls and MBAs who treat it as a guide to human action. We all do, a great deal of the time. Try, for instance, to explain the division of labor in hunter-gatherer societies, or the increasing memory power of computers, or even why you shovel your sidewalk after a snowstorm, without appealing to problems that people seek to solve. This paradigm oªers a robust and reliable way to explain purposive human action. And anticipating problems and solutions can be a remarkably eªective real-world strategy. Relief agencies that adopt this framework in planning ways to ship food to a drought-stricken population will have a better chance of success than those that don’t. It might not be too much to say that in any art a tradition consists largely of successful solutions to recurring problems, a range of preferred choices guided by task and purpose. Of course we cannot capture every twitch in an artist’s decision making. Guided by our research questions, however, we can produce a plausible account, at a level of appropriate generality, of how films’ particular stylistic regularities function and how they might have arisen in a problem/solution context. At the same time, we should be prepared to recognize when solutions produce felicitous by-products. Take the moment in

STAGING AND STYLISTICS

251

Summer at Grandpa’s when the little girl is almost run over by the train (Plates 5.1–5.3). I’ve argued that Hou takes a schema, the figure glimpsed through the rhythmic bursts of a train flashing past, and revises it by including a vivid detail where we don’t normally find one. As Ting-ting is snatched away and her fan remains behind on the rail bed, some unpredictable visual eªects occur. Instead of being crushed, the fan tumbles over. Then (and this eªect may not be visible on video copies) the blades of the fan start turning in reverse, sucked backward by the train’s momentum. When the train has passed, the amazingly unscathed fan starts whirring its blades in their original direction. Who could have predicted this arresting touch? Hou’s solution to the problem of dramatizing the scene brings in its wake not only a piercing detail but a lesson about the behavior of things. Like the painter who tries to capture an eªect and then makes us newly aware of how something looks, the problem-solving filmmaker may achieve visual discoveries. Fortunately for filmmakers, many problems crop up again and again. Norms supply tested and preferred solutions. Across this book we’ve had occasion to note several norms, as manifested in well-tried pictorial schemas: frontality vs. dorsality, aperture framing, the advance to the camera, aggressive foregrounds, planimetric composition. Conceived this way, norms are not codes in any strict sense. In most semiological theories codes are systems of rules that match signifiers (a pattern realized in a physical material) to signifieds (concepts). Artistic norms are far more fluid; they are not fixed rules and often bear no conceptual meaning. What is the “signified” of a musical piece in sonata form, or a centered compositional design? What is the “signified” of aperture framing or a planimetric shot? These devices fulfill functions; they achieve eªects in dynamic relation to other technical choices; but they cannot be isolated as “signifiers.” Norms are pressures to make movies in one way or another, but they are merely regulative, not constitutive. They oªer options, but the presence or absence of any norm will not make something “unfilmic” in the way that flouting syntax will create word salad. True, many schemas, such as shot/reverse shot, are sometimes referred to as part of “film grammar,” but cinema has nothing like the rules determining verbal expression. Filmic norms are like social norms in general—not interrupting, passing the salt, holding the door open for people who follow you. If we want a linguistic analogy for film style, the closest would be not grammar but pragmatics, the loose and tacit conventions that enable us to use language for manifold purposes. To speak of norms suggests constraint, even conformity, and who wants to be considered normal? Before getting to the pleasures of the abnormal, however, let’s recall that the style of most fiction films since around 1920 is norm-abiding through and through. For good historical reasons, films outside the most extreme experimental traditions present recognizable images in legible compositions that are assembled in obedience to continuity-editing principles. That is, a film’s stylistic innovations stand out against a thick background of routine practice. This is what we should expect if filmmakers have a cumulative, schema-based conception of stylistic continuity and change. Most filmmakers

252

STAGING AND STYLISTICS

quietly accept the norms they inherit; those who don’t usually tweak a few rather than overhaul them all. Norms guide the less talented and challenge the ambitious. Handed a task, an imaginative artist probes the privileged schemas, revealing unexpected resources. This is what Feuillade and some contemporaries did with the fixed long shot in depth, what Hou did with the exterior telephoto framing, what Angelopoulos did with the simple come-andgo pan. The norm is not a thou-shalt-not but rather a cluster of preferred practices that often harbor fresh possibilities. Similarly, going against some norms may lead to obedience to others. By giving up analytical editing for the most part, Angelopoulos and Hou, consciously or not, joined a tradition that oªered both the possibilities and the constraints of staging in the long-held image. Angelopoulos explored how these constraints could be enriched by forbidding landscapes and solemn camera movements; Hou explored how he might create just-noticeable diªerences with dense staging before a static camera (thereby, unwittingly, joining the tradition of the 1910s). Innovation is norm bound, and although norms constrain action, they also foster originality. Sometimes, the artist abandons all norms, forgets problem-solving, and just fools around. She throws paint at a canvas or bangs out some random notes on the piano, if only to see what a casual gesture will yield. Yet artists often fool around in order to explore forms and materials, and this exploration will be haunted by norms. Studying the splash of paint, the artist may begin to see an intriguing or powerful shape, which will have to be articulated by principles of color and design entrenched in her expertise. The random piano chords may suggest an expressive pattern, which in turn will bring into play expectations about structural schemas. The artist comes up with a goal—making something of aesthetic value out of a shot in the dark—and what is this but a problem to be solved by craft knowledge? This open-ended trial and error is more common in literature, music composition, painting, and other arts in which solitary creators can at little cost try anything that occurs to them. Experimental filmmakers have this luxury, whereas commercial filmmaking doesn’t usually permit it. Yet those directors who improvise on the set or who shoot from several angles or who fiddle with diªerent editing arrangements or sound mixes are playing around to a degree. Testing diªerent solutions allows them to understand their problem more intimately and to discover finer-grained aspects of the schemas they inherit. Radically experimental artists face a bigger challenge. In their institutional context they get no credit for merely obeying norms, so they must set their own tasks, erect their own problems, launch individualized “research programs.” Burdened with a heavy sense of what’s been done, they must find or create their own niches, perhaps seeing a problem where nobody saw one before. Before the 1950s even experimental filmmakers assumed that out-of-focus images were useful only for suggesting extreme sensory states. Stan Brakhage realized that blurred patches of color could form the basis for extended sequences, even an entire film. Unlike abstract shapes painted on film, his photographed

STAGING AND STYLISTICS

253

images of people and household objects would always pull us in two directions: we see them as pure shapes and movements, but we also try to imagine what real scene created them (for example, Scenes from Under Childhood, 1970). This discovery created new problems of organization: how to build such images into an entire film and how to link one shot to another? Brakhage solved these di‹culties through a “lyrical,” motivic form and through unpredictable camera movements and imperceptible cuts (as in, for example, Anticipation of the Night, 1958). Whether norms are obeyed or cast aside, replicated or enriched, their centrality reminds us that the craft context is a social one. All of our four directors have had accomplices, trusted assistants, constant collaborators, and influential models. Filmmakers pick up norms and practices from others; they compete; they bend others’ accomplishments to their own purposes. Feuillade, it seems likely, assimilated continuity devices from American and European films. Mizoguchi felt rivalry with Wyler and perhaps von Sternberg. Angelopoulos watched films at the Cinémathèque and borrowed schemas from Antonioni. Partnering with Chen Kun-hou, Hou Hsiao-hsien learned what he could and couldn’t achieve in Taiwan’s 1980s film industry. Hou’s juniors competed with him on the terrain of norms he helped articulate, and this seems to have spurred him to modify his style. Problems and solutions are shared by the community of creators.

FILMMAKERS AS ACTIVE AGENTS The problem/solution paradigm presumes social agents: one or more individuals who face a problem and seek to surmount it. In our domain of interest the agents are filmmakers. Since social agents exist in a state of freedom—one can always choose to do this or that or the other, or choose to do something one way or another—the problem and solution lie within a set of possibilities from which the agents select. Yet artistic freedom isn’t infinite. There are always constraints of money, time, resources, and technology. Moreover, circumstances tend to favor one set of options over others; artistic traditions, social conventions, personal proclivities, and current tastes and fashions will steer artists toward certain choices. Because of this matrix of forces, tackling one problem or finding one solution may commit the filmmakers to a further set of choices. For example, the director wants a close-up of the performer standing on a busy street corner, but she can’t put the camera in tra‹c. So the cinematographer uses a long lens, our friend the telephoto, which provides the close framing. But if another actor is to approach the performer from the rear, the filmmaker can’t focus on both, because such a long lens doesn’t yield su‹cient depth of field. Every choice takes place within a cluster of constraints. Faced with a familiar problem, the filmmaker has at least one stylistic schema available, and she can choose to use it. Borrowing Noël Carroll’s term, we can call this stylistic replication; my telephoto close-up would be a good example.27 Or the filmmaker may revise the schema, adjusting it to a new purpose or situation, as Hou revised the

254

STAGING AND STYLISTICS

telephoto framings at work in contemporary Taiwanese cinema. The filmmaker may also reject the schema, as many of our directors reject the cluster of schemas associated with analytical editing. But rejection entails inventing a new stylistic device or, more often, picking an alternative schema (which may then be replicated or revised). Once the filmmaker picks a schema, she or he must implement it and face further choices about means and ends. As with problem and solution, the interplay of means and ends can become complicated. Sometimes, an end is conceived and a means found to achieve it, as when Hou settled on the long take as a way to preserve continuity of performance. But often the presence of a means will lead to a new purpose. When filmmakers encounter a new gadget or technical process, they may embrace it because it serves an existing function or because it suggests a new one. The first uses of the Steadicam, in Bound for Glory (1976) and Rocky (1976), were as substitutes for normal tracking shots, allowing the camera to move more smoothly through crowds or up flights of steps. Stanley Kubrick saw that the Steadicam could also be set at any height, so for The Shining (1980) he devised shots that follow a boy on a Big Wheel tricycle in fast, very low-lying movements not feasible before. A couple of quick clarifications are in order. In speaking of filmmakers, I have in mind not only the director but others involved in the work of film production. Naturally, many decisions are made by writers, producers, art directors, cinematographers, crew, editors, sound mixers, and the like, although their contributions often go unrecorded. In this book I’ve attributed decisions to directors because (a) we have good reason to believe that these four directors did govern the features highlighted in my analyses; (b) in general, final decisions about staging are likely to be made by the director, whatever the creative input from other members of the team; and (c) as a default option we assume that the director is the “final filter” for many, if not most, of the choices faced in the course of production.28 By emphasizing choice I may be seeming to make filmic creation more intellectual than intuitive, more a matter of engineering than of spontaneous creation. This is not my aim. Some films or scenes or moments are wholly planned in advance and executed with virtually no change. Others come into being in a freewheeling fashion, thanks to happy accidents and on-the-set revelations. In both cases choices are involved. The filmmakers choose to map things out or to leave things to chance. You can choose to wing it and trust your instincts, but you can’t choose not to choose. Even when trusting their instincts, filmmakers must decide between this or that improvisatory moment, this or that camera movement, this or that take. Renoir was famous for encouraging unexpected moments to bloom, but the unused footage from Partie de campagne (A Day in the Country, 1936) consists of a great many retakes, repetitions, and developments of simple gestures like lighting cigarettes and eating. In each shot the actor “acts naturally,” but some bits suited Renoir’s purposes and some did not.29 Mizoguchi made his choice earlier in the process, rehearsing interminably but shooting very few takes. Hou builds up his scene in situ and encourages the actors to rework

STAGING AND STYLISTICS

255

the scripted dialogue; but he still chooses what reworkings are to be included, and he coordinates how the players interact with each other, the space, and the camera. Filmmaking, like any human practice, involves trial and error; problems are seldom solved on the first go, failures call for escape hatches, and plans can change. The point is that the entire process calls for the filmmakers to find schemas that can be replicated or revised for the task at hand. We needn’t think that filmmakers always ponder or brood over such choices. Indeed, skillful filmmakers, at least in mass-market cinema, have so smoothly absorbed the norms and schemas, have become so adept at sizing up the problem, that the whole process of decision making is quite intuitive. A psychologist might say that they have “overlearned” their craft, as adults have overlearned the skill of driving a car. One does it without noticing. Here is Howard Hawks: “You know which way the men are going to come in, and then you experiment and see where you’re going to have Wayne sitting at a table, and then you see where the girl sits, and then in a few minutes you’ve got it all worked out, and it’s perfectly simple, as far as I am concerned.”30 Even directors trying to stand out as innovative need not articulate a large-scale plan. “It’s what you call style in retrospect only,” says Ethan Coen, who produces and shares screenwriting credit with his brother. “At the point of actually making the movie, it’s just about making individual choices . . . ” “. . . about the best way to tell the story, scene by scene,” Joel interjects, finishing his brother’s sentence, as he often does. “You make specific choices that you think are appropriate or compelling or interesting for that particular scene. Then, at the end of the day, you put it all together and somebody looks at it and, if there’s some consistency to it, they say, ‘Well, that’s their style.’”31

Stylistic choices are deliberate but not necessarily deliberated, and filmmakers need undertake no deep meditation on the ways they achieve their results. What counts, as the brothers Coen suggest, is the functional unity achieved. For such reasons we need not treat this as a full-blown rational-agent model of the kind utilized by economists. Whatever the strengths and shortcomings of that conception in the economic sphere, it may not transfer comfortably to art, where preferences and payoªs have less clear-cut meanings.32 The sort of rational agency that will serve us well enough is folk psychology. We ascribe beliefs, intentions, desires, and plans to others with a high degree of success. Visiting New York, I phone a friend, and we agree to meet at Angelo’s Pizzeria. She turns up and so do I. This is not a miracle. It simply depends on a (fairly deep) background of assumptions about means and ends. If she doesn’t show up, I run various scenarios. Is she delayed? Did she forget? Is she snubbing me? Did she go to a diªerent Angelo’s? Each scenario involves hypothesizing that she had certain goals, made certain plans, and confronted potential choices. As with the problem/ solution framework, rationality in this thin, means/ends sense is fundamental to un-

256

STAGING AND STYLISTICS

derstanding our activities as social creatures. It is not all there is to intelligent action, but it seems an indispensable part of it. Someone might object that this perspective reinstates the “intentional fallacy,” the objection to using artists’ intentions as a criterion for interpreting the correct meaning of artworks. But that view conceives of intention as a discrete mental episode, an eªort to mean this rather than that. This framework does not claim access to intentions as mental episodes, only to intentions as posited sources of patterns of action. Again, we reverse-engineer. Given our general conception of intentional action and the stylistic regularities we find in a film, we do not think that it all happened by chance. Instead we ask what function(s) the stylistic phenomenon fulfills. As a default we presuppose that the filmmakers sought to achieve such functions. Perhaps they did not, but as a methodological point of departure, functionality suggests purposiveness—that is, “intention” in the sense that the film is the product of intentional activity. Sometimes, of course, functions arise that were not initially intended. The little spinning fan in Summer at Grandpa’s behaves in a way that might not have been foreseen by the filmmakers, but its behavior is a by-product of a very calculated shot design. And although we do not know the precise chain of thought that led Hou to his design decision, we can say that he intentionally left the detail in the shot. Art making does add one kink to folk psychology, though. Often the artist seeks out constraints, limiting her range of choice quite drastically. In his early and middle years Feuillade accepted the long take as a constraint imposed from without, since analytical editing had not yet become a likely option for scene construction. But Mizoguchi, Angelopoulos, and Hou all embraced the long take as a self-imposed limit. This choice probably sprang from several sources—the demands of sound cinema, a temperamental a‹nity, a desire to create novelty or a signature style—but surely they also sought to explore the resources aªorded by the technique. They created a problem in order to challenge themselves to find provocative and promising solutions. By cutting out whole realms of choice (editing,varied camera positions), they could probe more refined choices within a narrow range. Jon Elster calls this the “focus-enhancing” eªect of freely chosen limits.33 We are back with Mizoguchi’s modulation and Hou’s constraints. Within a problem/solution framework it is reasonable to explore a narrow set of options. Perhaps they can fulfill your larger purposes, such as telling the story a certain way. Or perhaps you want to enrich an artistic technique—that is, to bring out fresh ways of denoting or expressing or symbolizing. All the filmmakers we’ve considered pursue both possibilities. Finally, does concentrating on the filmmaker in this way reinstate the “auteur” approach to film studies? If that is taken to require ex nihilo creation, whereby the auteur simply transcribes his or her worldview onto film, the answer is no. The filmmaker creates out of the norms and forms available in the craft milieu or out of possibilities in adjacent media that can be brought into that milieu. If, however, auteurism means that we may legitimately ask how particular filmmakers can create intriguing artworks, the answer is yes: the theory I’m proposing does provide a way to ground the very widespread intuition that individual creators matter.

STAGING AND STYLISTICS

257

FILMMAKING AS A TRANSCULTURAL ACTIVITY In caves on several continents we find images going back ten thousand to thirty thousand years. Apart from being astonishingly beautiful, they ought to give pause to anyone who believes that visual representation is a matter of transitory or arbitrary cultural conventions. The makers of these Ice Age images lived in circumstances very diªerent from any literate cultures since then; they diªer more from contemporary humans than today’s non-Western cultures diªer from the West. These Paleolithic people are a kind of ultimate Other. Yet the cave images they have left us are for the most part easily recognizable today. The paintings and carvings depict wild horses, bulls, bison, bears, panthers, rhinos, and mammoths, as well as men and women. No culture-specific code or training is required to recognize them.34 In the 1960s many scholars in film studies embraced conceptions of representation derived from language (specifically, Saussurean linguistics). They proposed that film images should be studied as arbitrary conventions. Representation was conceived as signification, a closed system of rule-governed relations of signifier to signified. To claim that many images are faithful to perception was to risk being called a “naive realist.” The semiotic move was in some ways a liberating gesture. It induced us to think about cinematic conventions as artifice, which to some degree they are. But when recruited by a sweeping cultural constructivism, this line of thought has blocked full understanding. Insofar as cinema represents the world pictorially, it relies heavily on transcultural regularities in people’s perceptual systems. A normally sighted person in any culture sees the world in three dimensions and color, with shapes, edges, and shading providing information about the layout of surfaces. As the observer moves through the environment, the spatial coordinates change in predictable fashion, according to the principles of projective geometry. Given these common features of visual perception, we should expect to find a lot of overlap among people’s ways of picturing the world in two dimensions. We do. Although the world’s cultures have produced several systems of pictorial representation, there are not infinitely many; and each provides some accurate information about the real-world appearance of things.35 In particular, outline drawings of familiar objects, such as our Ice Age animals, are readily recognized in diªerent times and places.36 Cinema, thanks to photography and the illusion of movement, can capture certain regularities in perceptual reality with great exactitude and vividness. It has consequently become a powerful transcultural medium of pictorial communication.37 Members of cultures without pictures reliably recognize people, places, and things portrayed in films.38 Granted, films also contain what we want to call conventions. At a cut the camera changes position instantly, something we can’t do in real life. To recall an example from chapter 1, in film scenes conversing people stare unblinkingly at each other. Aren’t these techniques culture driven? Yes, but in a way diªerent from the arbitrary conventions governing the lexicon and syntax of a language. A great many film conventions serve to stylize, highlight, or purify ordinary visual experiences or social interactions. In all cultures

258

STAGING AND STYLISTICS

people normally converse face to face. When film actors gaze intently at one another, they make unusually prominent the normal signals for attention and mutual interest passing between partners in a dialogue. The film convention of shot/reverse-shot cutting presents the conversational exchange in an amplified form, with the change of camera position highlighting the turn-taking and facial reactions that are salient parts of the social exchange.39 Even a cut to a closer view mimics, while also streamlining, our approach to objects from a distance.40 A great many conventions piggyback on our nonfilmic experiences of the visible world and social action taking place in it. That is likely to be the main reason why these conventions are easy to learn. Very quickly we recognize the situation represented (two people are talking) and thereby grasp the representational convention (in order to make speaker or listener more visible, the movie cuts from one to the other). Because many features of perceptual and social reality recur across cultures, international audiences can grasp the corresponding conventions quickly. This consideration points us toward a moderate constructivism: cinematic conventions are built out of recurring perceptual and social regularities, some—perhaps even most—of them to be found in many cultures. These regularities constitute what we might call contingent universals.41 How then do we learn the more recondite conventions, like dissolves and wipes? If most cinematic devices are intelligible as transformations of widely practiced perceptual and social activities, then occasional devices with no clear analogues in ordinary experience can be learned fairly easily. They ride on the more widespread patterns of comprehension. Christian Metz points out that some people believed that they could understand the film because of its syntax, whereas one understands the syntax because one has understood, and only because one has understood, the film. The inherent intelligibility of a dissolve or double exposure cannot clarify the plot of a film unless the spectator has already seen other films in which dissolves and double exposures were used intelligibly. On the other hand, the narrative force of a plot, which will always be understood only too well—since it communicates with us in images of the world and of ourselves—will automatically lead us to understand the double exposure and the dissolve, if not in the first film we see them in, at least by the third or fourth.42

Gombrich puts it more succinctly: “It is the meaning which leads us to the convention and not the convention which leads us to the meaning.”43 Filmmakers in various cultures face kindred choices about style, then, partly because their audiences share some perceptual and social skills. Other factors favor stylistic convergence as well. Some derive from the medium itself: the film frame, whatever its proportions, has to be filled; the optical pyramid (Fig. 2.50) provides both constraint and opportunity. There are also common tools and practices. Filmmakers the world over use the same cameras, film stock, camera supports, light sources, sound recording formats, and editing equipment. Production labor is divided in comparable ways in most film in-

STAGING AND STYLISTICS

259

dustries (although the variants are often significant). Historical circumstances also oªer filmmakers a shared menu of stylistic choices. After the emergence of continuity editing in the late 1910s, filmmakers around the world could benefit from their predecessors’ trial and error and borrow continuity schemas if they wished.44 Thus a craft tradition binds filmmakers across cultures. Common problems emerge and analogous solutions develop. Sometimes two filmmakers converge on similar solutions without mutual influence, as we’ve seen with Hou Hsiao-hsien and Mizoguchi Kenji. In addition, because cinema has been a global medium from its inception, it has been easier for filmmakers to learn from each other than, say, writers in diªerent languages could. In their home markets Feuillade and Mizoguchi could view films from the United States and Europe. Today massive theatrical distribution, television and cable, home video, archival screenings, and film festivals allow filmmakers unprecedented access to artistic strategies and tactics from many points in film history. Thanks to Langlois’ Cinémathèque, Angelopoulos could watch not only Antonioni but also Mizoguchi. Hou has attested that visiting festivals triggered his awareness of much contemporary cinema. Along several dimensions, filmmakers hold craft traditions in common, and this helps their films cross boundaries. Despite all the convergences within film practice, media theorists often resist the idea of transcultural norms, so it’s worth taking a moment to consider some objections raised by Slavoj Yizek to the line of reasoning I’ve traced here and elsewhere.45 Yizek’s objections are framed idiosyncratically, but their premises are probably shared by other scholars. By replying to his objections and correcting some misinterpretations, I hope to quell misgivings some readers might have. Yizek says that this argument seeks to distinguish “trans-cultural universal features (part of our evolutionary heritage and the psychic structure of human beings) from features that are specific to particular cultures and periods—i.e., to operate with a simple pyramid from natural or other trans-cultural universal features to more and more specific characteristics that depend on localized contexts.”46 But in the work he cites, and in this book as well, I’m at pains to remain agnostic about whether many transcultural regularities stem from evolution or from those cultural practices that converge when humans confront the same environmental regularities (such as the need to find food or shelter). Surely biological evolution and normal maturation account for certain propensities we have, like scanning the environment or detecting movement in a static array. Still, I’m not committed to evolutionary explanations for all the cultural commonalities we might detect. The sources of most convergences remain unknown. As for the pyramid he ascribes to me, I sketch no such geometry of cultural practices. In the relevant essay I say the opposite: “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 diªerences.”47 It seems to me that we find convergences among cultures at all levels of specificity (string is one example I invoke in the article), just as we find divergences in practices as perva-

260

STAGING AND STYLISTICS

sive as marriage rituals. The existence of loose, overlapping networks of convergences and divergences, driven not only by biological inheritance (which I don’t underrate) but also by recurrent environmental and social demands, is exactly why I speak of contingent universals. Yizek also objects to some more specific claims that I’ve restated in this book. He suggests that the idea that film style fulfills storytelling needs is somewhat ethnocentric: “Is not modern (post-Renaissance) Western culture characterized by its own specific notion of narrative (which is why, say, Chinese or Japanese novels often strike us Western readers as ‘dull’ and ‘confused’)?”48 It seems odd, in an argument devoted to criticizing homogenizing tendencies among film theorists, for Yizek to posit something called “modern (post-Renaissance) Western culture.” Racine and Cervantes, Goldoni and Virginia Woolf, Milton and James Joyce, Mark Twain and Voltaire, Richard Wright and Stephen King, Henry James and Lily Tomlin, and thousands of other storytellers apparently share a single notion of narrative—and a “specific” one at that (although we aren’t told what it is). Set alongside this monolith, my eªorts to trace one-oª overlaps look modest indeed. Still, if my claims did hold good across this stupendous category, they might be worth a good deal. To answer Yizek’s question: I take narrative to be a diªuse and complex phenomenon, with many crisscrossing features spread across the world’s storytelling traditions. Virtually all narratives seem to me to share some components, such as agents and temporal sequence. A great many narratives, across a variety of cultures, share other features, such as goal-oriented protagonists, causal consequence, and a psychology of beliefs and desires. Certainly these features are present in a large amount of non-Western literature. Yizek does not cite particular Chinese and Japanese novels, but the ones with which I’m familiar do not seem on the whole “confused.” A great many Westerners read these novels and find them intelligible. Oddly, Yizek adds that these classics are perceived as “boring,” but this is beside the point: the issue is comprehensibility, and a dull story may be intelligible. Further, such novels would not be translated (indeed, probably could not be translated) if their narrative premises were radically out of alignment with those of a Western readership. It seems evident that readers outside these cultures find such novels’ stories largely comprehensible, thanks to cross-cultural understanding of agents, causation, means/ends reasoning, and the like. When portions are not comprehensible, readers may need to know some items from the cultural context; but if that relevant information can be supplied, it typically makes the story more intelligible—exactly by fitting the puzzling element into a cross-cultural schema. Thus some obscure motivations in The Tale of Genji can be clarified by fitting them into standard belief/desire psychology.49 Normally, however, we don’t need huge amounts of such information. One reason is that many story patterns recur across cultures. After studying a very wide range of storytelling traditions, Patrick Hogan has cogently argued that the bulk of highly respected narratives in all or most traditions adhere to a few quickly recognized plot structures. Some of Hogan’s primary examples—

STAGING AND STYLISTICS

261

The Tale of the Heike, The Story of the Stone, and The Three Kingdoms—are likely to be the sort of Japanese and Chinese novels that Yizek would say “we” find boring and confused.50 Yizek goes on to ask: And is there a neutral, global notion of what is “comprehensible”? The status of narrative in cinema is much more fragile than it may appear: su‹ce it to recall the recent crisis of narrative, where we witness a kind of unexpected return to the early “cinema of attractions”—big blockbusters have to rely more and more on the wild rhythm of spectacular special eªects, and the only narrative which seems still to be able to sustain the viewer’s interest is, significantly, that of the conspiracy theory. (Although Cameron’s Titanic [1997] is praised as the return to the good old pre-deconstructionist romantic narrative, it can also be seen as the ultimate proof of narrative failure: one way to read the film is that the iceberg strikes in order to save us from unavoidable narrative deadlock—imagine what a boring film Titanic would be if it just continued as a love story between Jack and Rose.)51

What is it to be “comprehensible”? Our garden-variety sense of a comprehensible story depends on activities like grasping the state of aªairs at the outset of the tale, tracking the characters’ actions, ascribing beliefs and desires to them in such a way that their activities cohere under some means/ends explanation, recognizing obstacles and reversals, being able at any point to hypothesize the state of a character’s knowledge, being able to summarize the story, and so on. Suppose you hear the tale about Wacici told by the Kikuyu tribe of Kenya. If you understand that Wacici is very beautiful, that her friends are jealous of her beauty, that this leads them to push her into a porcupine hole and bury her alive, that her absence leads her parents to become worried and send her brother out to search for her—if, that is, you grasp this cluster of circumstances, motives, plans, actions, reactions, and consequences—to that extent you comprehend the story.52 What else could “comprehensible” mean? This is not an absolute, theory-neutral sense of comprehension, but it relies on an intuitive theory about stories and human action that cultures share. Yizek oªers no argument against this widely held theory. All he provides, irrelevantly, is the commonplace opinion that recent films overwhelm “narrative” with special eªects. Put aside the fact that several scholars have argued persuasively that this “crisis of narrative” has not happened, and that many contemporary Hollywood films are as narratively coherent as their predecessors.53 Even if Yizek’s observation were accurate, it wouldn’t show that the stories have become incomprehensible. Again, throughout the passage from which this paragraph is taken Yizek conflates boring stories with unintelligible ones. Titanic might be duller without the iceberg, but nothing in the story would necessarily become incomprehensible. Yizek asks one more rhetorical question: “Does the fact that a non-Western (or even a medieval Western) painting can appear to us extremely confusing not indicate that there are no simple trans-cultural functions of guiding attention?”54 This objection is am-

262

STAGING AND STYLISTICS

biguously stated. Yizek might be saying that in some artistic practices the function of guiding attention operates in ways very diªerent from its operation in others. That is, the “confusing” picture obeys principles of attention-guiding not found in our pictures. This reading is supported by Yizek’s rejection of “simple” transcultural functions of guiding attention (implying that there are more complex local ones). I believe he doesn’t mean to say this, however, for it would entail granting that attention-guiding is to some degree a transcultural process. I’m inclined to read Yizek’s question as suggesting that in some cultural traditions the guiding of attention is not a vehicle for pictorial storytelling. This would chime with his belief that “modern (post-Renaissance) Western culture” has its own conception of narrative. But clearly, the guiding of attention does concern artists across a great many cultures. Yizek supplies no examples, but consider just the traditions he mentions. Many medieval images (especially Byzantine ones) are centered and symmetrical in ways that make Christ or the Virgin dominate the space. Japanese picture scrolls (emakimono) are often packed with incident, but they direct us to salient actions through centering, color, compositional devices like diagonals, and blown-away roofs that reveal occupants inside buildings. To go beyond Yizek’s range of reference, an ancient Egyptian tomb relief may portray a hunt, a banquet, or a harvest with many human figures, but the wall’s composition is organized in a hierarchy that facilitates understanding. The registers, those stacked rows of figures, create similarities and variations in postures and sometimes present actions in chronological order, like a comic strip. Within registers, a more important figure will be segregated, colored diªerently, framed in an aperture, or given a varying attitude. At one end of the relief we find a large image of the tomb owner, as tall as most or all the registers combined—a powerful cue for his or her significance.55 Most to the point here, an examination of the images from non-Western films in this book will find compositional strategies best explained as highlighting the shot’s most important information. Of course filmmakers sometimes want to conceal crucial information. Yet they do so exactly by making the crucial item easy to miss (placing it oª center, in the background, in darkness, turned from us, and so forth) or by calling attention to the fact that it is hidden (as Mizoguchi does in placing Oharu and the fateful message behind the hanging kimono, Fig. 3.7). The norms for compositional emphasis function as defaults, operating as most likely, other-things-being-equal presuppositions. Yizek forgets that directing attention isn’t always a matter of a simple centering, like a bull’s-eye (although it often is; glance back at many frames in this book). Artists may provide a profusion of splendors, not least as a display of imagination and virtuosity. But these splendors are usually organized for cogent uptake; otherwise we could hardly appreciate them. Artists rely on pictorial balance, competing centers of interest, rivalry between depth and proximity, and other factors I’ve considered in earlier chapters, in order to elicit guided exploration of the pictorial space. Our saccadic sweeps are anchored in points of major interest, but from them we can launch probes into other regions. This process yields Barr’s “gradation of emphasis” and Hogarth’s wanton chase of the eye.

STAGING AND STYLISTICS

263

Further, Yizek cites no medieval or non-Western tradition in which utterly undirected attention is a working principle. Perhaps abstract painters such as Jackson Pollock had such a purpose with their “all-over” compositions. (Of course there are also abstract painters who want to direct our attention, such as El Lissitzky.) But my concern is with visual storytelling, and abstract painting doesn’t tell a story. It seems highly likely that artists whose paintings present a narrative want us to notice pertinent information in the picture. Indeed, telling a story comes down to making certain information about the story world salient. How—and this is not a rhetorical question—could one tell a story pictorially without guiding our attention to the key information about agents, actions, and the rest? Yizek does not say. Yizek expresses a skepticism about transcultural convergences that I think is widely shared among contemporary film scholars. Yet he frames his most sweeping objection to my account in an equivocal way. He asserts that “the very notion of a trans-cultural universal means diªerent things in diªerent cultures.” You might construe this to be saying that the very concept of universality is culturally variable. This is a pretty radical claim, possibly an unintelligible one; but it turns out not to be what Yizek means. Rather, Yizek means to say only that the contents of transcultural categories vary from culture to culture: “While one can claim that all cultures recognize some kind of diªerence between subjective imagination and reality—things as they exist out there—this assertion still begs the question of what ‘objective reality’ means in diªerent cultures: when a European says that ‘ghosts don’t exist in reality’ and when a Native American says that he communicates with them and that they therefore do exist in reality, does ‘reality’ mean the same thing for them?”56 Yizek drifts from the point in asking about epistemic conditions for reality (or “reality”), but that shouldn’t distract from his concession that there are transcultural convergences in certain fundamental ways of construing the world. Yizek thinks the subject/ object split is not just an artifact of the post-Renaissance West. Good thing too, for without such a distinction it’s impossible for anyone in any culture to conceive of personal identity, veridical perception, and inner states like pain and memories. It turns out that many cultural universals are “hollow,” or structural, in this sense. We have reason to believe that all cultures spontaneously distinguish between living and nonliving things, although what is put in each category will vary.57 Still, there is strong evidence for “substantive” transcultural regularities too—not only beliefs (the concept of person or of kin, in-group amity/out-group enmity) but also emotions (fear, sexual jealousy), practices (etiquette, play, facial communication), and tools (string, devices for cutting and pounding).58 Strangely, when Yizek turns to the examination of film techniques, such as depth of field or crosscutting, he presupposes that these are common across various cultures and periods. He does insist that within the context of a given film, a technique will be inflected in more specific ways; but of course this is exactly what I’ve argued in this book and the other writings he invokes.59

264

STAGING AND STYLISTICS

Yizek’s rebuttal is idiosyncratic, but some general lessons issue from its confusions. Analysis of many art forms reveals cross-cultural regularities, and these may be plausibly attributed to universals.60 If poststructuralist cultural theory is to retain any intellectual integrity, it must face the mounting challenges to its sweeping relativism. Perhaps, as Adam Kuper suggests, the concept of “culture” is useless for explaining social action.61 But if the notion retains any explanatory force, its adherents in the academic humanities should respond to the new ways of understanding it that are emerging in the social sciences. To take one that’s relevant here: rather than counterposing the individual to culture, we should consider whether the individual carries significant aspects of culture within him- or herself, as part of a panhuman nature.62 Practitioners of Cultural Studies, putting the stress on diªerence, have assumed that every entity that we call a culture is somehow sui generis. They study not culture but cultures. It appears likely, however, that there is a “metaculture” from which particular societies draw many essential resources. Like islands in an archipelago, social formations are bound by an underlying unity, part of which we may as well call culture.

I propose that we can fruitfully analyze and explain the historical dynamic of film style by inferring, on the basis of the films and what we know about their making, some pertinent craft traditions. The traditions preserve favored practices, practices that are the result of choices among alternatives. In choosing, filmmakers exercise their skill and judgment, thereby replicating, revising, or rejecting options supplied by their predecessors and peers. Lots of causes escape the artist’s awareness, and many eªects can’t be foreseen. But we discover such unintended causes and consequences only by initially positing filmmakers as goal-seeking, intentional social agents, constantly moving between tasks and purposes, problems and solutions, schemas and revisions. Our inquiry is helped by the fact that many craft traditions are shared by filmmakers working in diªerent times and places. This theoretical perspective on stylistic continuity and change is well suited to a poetics of cinema, that enterprise that seeks to understand the principles by which films are designed to achieve certain eªects. These views are not widely shared within academic film studies. In art history and musicology, as far as I can detect, the study of style remains at the center of the discipline; but film stylistics may be even less common and less well regarded than literary stylistics. True, film studies outside the academy began with a focus on film history and on film aesthetics, particularly style. The pioneering writers in film theory—Hugo Münsterberg, Rudolf Arnheim, the French impressionists, the Soviet montage filmmakers— examined film form closely. Similar impulses were visible in the 1940s and 1950s in Europe, particularly around Cahiers du cinéma. Yet academic film studies followed the example of literary criticism and developed as primarily a hermeneutic discipline. The basic operating procedure has long been to treat individual films as exemplars of theoretical posi-

STAGING AND STYLISTICS

265

tions, usually via some symptomatic reading guided by a “hermeneutics of suspicion.”63 This has more recently become a “cultural hermeneutics,” bringing in historical data (as the modernity thesis does); but this wrinkle doesn’t alter the reasoning routines involved. Style is seldom mentioned, and when it is, one-oª stylistic devices are for the most part invested with symbolic meanings. For many film scholars and students, movies exist less as parts of an artistic tradition than as cultural products whose extractable ideas about race, class, gender, ethnicity, modernity, postmodernity, and so forth can be applauded or deplored. So strong is this position that even the arrival of home video, which ought to have triggered studies of form and style, has largely fostered the proliferation of “readings.” In these circumstances, to ask questions centered on film style is at best to miss the point (which is to propose general interpretations) and at worst to engage in a dangerous aestheticism (“blind,” as the saying goes, to the cultural construction of everything that matters). This is not to complain about theorizing, in which I’ve often indulged myself. It is to suggest that the theoretical concepts that guide our inquiries should be treated as hypotheses, delicate enough to respond to fluctuations in the material but crisp enough to be corrigible. Ideas guide our observations, but observations challenge our ideas. We put questions to the material, and often we must stand corrected. Most theories on oªer in film studies, however, are not hypothesis driven. They are simply bodies of doctrine (usually not as complex as their baroque presentations make them appear) concerning how society works (usually to the misery of its members) or how the mind works (usually to the unhappiness of its owner). Most of these theories have slender empirical support, but they are immune to testing or refutation because they tend to be vague, equivocal, truistic, or all three at once. They are incorrigible. Naturally the commitment to doctrinal theories informs pedagogy too. “All too often,” Brian Boyd remarks, “university literary studies invite students to operate as literary critics before they have become sensitive readers,” and a parallel point could be made about media studies.64 Typically students are provided with theories about culture and then asked to apply them to this or that movie; and since it is easier to juggle generalizations than to weigh the factors impinging on concrete creative choices, students leave class knowing little more about cinema. This is a pity. Many students want to become filmmakers, and many more want to appreciate the fine points of movies, just as they want to grasp the fine points of music and sports. They could profitably be taught about cinema’s historical traditions, about technical options, about norms and how they can be fulfilled or creatively denied. Students might, for instance, be shown that careful description can trigger discoveries about craft: simply setting down everything you notice about a shot or sequence can lead to “Why?” or “How?” or “To what purpose?” questions. At the very least an approach highlighting style and form can help viewers appreciate the films they see, while also making aspiring filmmakers more conscious of their range of choices. A stylistics of cinema faces a further obstacle in the “death of cinema” lamentations

266

STAGING AND STYLISTICS

that arose as cinema approached and passed its centenary. Was not the golden age truly over? After the Nouvelle Vague, and certainly after Fassbinder, Duras, and perhaps The Godfather, what was there left to respect? The multiplex and the summer locomotive picture; the Europuddings and hyphenate productions that had neither local flavor nor radical ambitions; postmodernism, the new mannerism, cinéma du look, roller-coaster cinema, theme-park cinema, lightshow cinema—to film lovers everything conspired against the tradition of high artistry emblematized by Hollywood classicism and postwar European art cinema. Historically, inquiry into film style has often followed from bursts of formal innovations. Stimulated by the creative accomplishments of Soviet montage cinema, Eisenstein and the Russian Formalist theorists speculated on the poetics of film; two decades later Bazin took his cue from Welles, Wyler, Renoir, and the Neorealists. It seems likely that the interest in style in Cahiers, Movie, Monogram, Film Culture, and other magazines from the 1950s to the 1970s owed some of its momentum to the flagrant experiments on display in current cinema. In the 1960s a masterpiece seemed to arrive every week; we (for this is my generation) burned to understand the Art of Cinema, for only then could we appreciate 8 1/2 or Jules and Jim or Alphaville. Even the classicists of Movie, I suspect, spotted the long takes in Preminger because they could not help noticing the long takes in Dreyer or Antonioni, or perhaps the jump cuts in Godard. So one might argue that when filmmakers’ aesthetic ambitions cooled in the 1980s, so too did an interest in stylistics. Who wants to analyze Spielberg’s compositions or the camera movements in Jackie Brown? Yet the picture is not as bleak as I’ve painted it, for some scholars do pursue stylistic studies. Relying on production documents, Lutz Bacher has given us a remarkable reconstruction of the choices and constraints facing Max Ophuls in directing his American films.65 Scholars of silent film continue to examine its relation to theater techniques (Ben Brewster, Lea Jacobs) and its mutating strategies of editing and storytelling (Charlie Keil, Don Fairservice).66 In France stylistic analysis came to the fore in the “textual analysis” thrust of the 1970s but waned fairly quickly; it now seems to be reemerging.67 The rapid adoption of DVDs should encourage a closer attention to technique, not only because of the format’s fidelity to the original film but also because filmmakers’ commentary tracks sometimes take us into the problem-solving process quite tangibly. Opportunities to study film aesthetically have multiplied—through video and through greater access to archives, to private collections, and to restoration events such as those sponsored by Italy’s Giornate del Cinema Muto and Cinema Retrovato festivals and by Los Angeles’s American Cinémathèque. At the same time, filmmakers continue to explore form and style. The surge to prominence of East Asian cinemas in the 1980s and 1990s—Hong Kong, Taiwan, mainland China, Japan, Korea, Thailand, and Singapore— is the most evident regional trend, but striking work has emerged from Europe as well. Even the overpublicized American independent scene has yielded some intriguing experiments in narration (for example, Memento and 13 Conversations about One Thing) and vi-

STAGING AND STYLISTICS

267

6.1 Iosseliani tackles the dinner-table conundrum: a diner on frame right responds to the foreground gentleman who declares hunting essential to human nature . . .

6.2 . . . before another diner, on center left, joins the debate, his gestures echoing those of the speaker (Favorites of the Moon, 1984).

6.3 Whispering Pages (1993): the mysterious walker and the woman come to foreground, her arms outstretched.

sual technique (such as the work of Hal Hartley and Paul Thomas Anderson). In any event, if we study style as an outgrowth of craft practices, there is no reason why we cannot try to understand the traditions that inform the work of contemporary filmmakers. “Any new thing that appears in the made world,” writes a historian of technology, “is based on some object already in existence.”68 As an artifact, even the most banal film has intriguing sources, which the stylistician can bring to light. So, yes: along with understanding Angelopoulos and Hou, we should try to make historical sense of Spielberg’s compositions and Tarantino’s camera movements and, indeed, the whole range of contemporary film styles.69 To return to the compass of this book, today we find creative revisions of staging schemas in the work of Otar Iosseliani, Manoel de Oliveira, Béla Tarr, György Fehér, and Alexander Sokurov (Figs. 6.1–6.3).70 It may be significant that one critic who declared the “irreversible decline” of cinema in 1996 has since found hope in these and other directors, such as Angelopoulos and Hou.71 Perhaps staging-centered directors

268

STAGING AND STYLISTICS

will bring us back to the cinematic image in ways that CGI and intensified continuity cannot. Perhaps, too, there is an advantage in the fact that these directors’ films are, well, slow. By stepping down the tempo, they give us time to let the image soak in. “You must,” notes Jean-Marie Straub, “give some pieces of reality a certain block of time to vindicate themselves against the film, and for the audience just to see. At first, they might not see much or maybe they see a lot, and then, as the shot goes on, they see less and less and then they see more and more.”72 How we manage to see more and more depends on us.

STAGING AND STYLISTICS

269

Notes

Essays supplementing issues raised in this book may be found at www.davidbordwell.net.

CHAPTER 1. STAGING AND STYLE 1.

2. 3. 4. 5.

6. 7.

See, e.g., the sequence of Nola Darling’s Thanksgiving dinner in Spike Lee’s She’s Gotta Have It (1987), which is analyzed in David Bordwell and Kristin Thompson, Film Art: An Introduction, 7th ed. (New York: McGraw-Hill, 2003), 318–320. Quoted in Jon Stevens, Actors Turned Directors: On Eliciting the Best Performance from an Actor and Other Secrets of Successful Directing (Los Angeles: Silman-James, 1997), 375. Takeshi Kitano, Rencontres du septième art, trans. Silvain Chupin (Paris: Arléa, 2000), 36. Not Ozu’s, though. Quoted in Mario Falsetto, Personal Visions: Conversations with Independent Film-Makers (London: Constable, 1999), 55. See also “Shooting Troublesome Table Scenes,” in Richard L. Bare, The Film Director: A Practical Guide to Motion Picture and Television Techniques (New York: Macmillan, 1971), 62. Jon Boorstin, Making Movies Work: Thinking like a Filmmaker (Los Angeles: Silman-James, 1990), 39. Indeed, Tarantino seems to acknowledge this by shifting from the floating-camera technique in the first phase of the Reservoir Dogs opening to a passage where the style becomes “a bit more normal, two-shots and masters,” before concluding the scene with the dispute about tipping, handled in singles. Michel Ciment and Hubert Niogret, “Interview at Cannes,” in Quentin Tarantino Interviews, ed. Gerald Peary (Jackson: University Press

NOTES TO PAGES 2–4

271

8.

9. 10. 11. 12. 13.

14. 15. 16.

17. 18. 19. 20. 21. 22. 23. 24. 25.

26. 27.

272

of Mississippi, 1998), 22; see also Gavin Smith, “When You Know You’re in Good Hands,” in ibid., 100. I borrow this term from E. H. Gombrich, Art and Illusion: A Study in the Psychology of Pictorial Representation (Princeton, NJ: Princeton University Press, 1972), 30, 73–75, 183–191. See Michael Argyle, Bodily Communication, 2nd ed. (London: Methuen, 1988), 118, 209. Charles Barr, “CinemaScope: Before and After,” Film Quarterly 16, no. 4 (summer 1963): 18–19. A helpful review of acting traditions and techniques can be found in James Naremore, Acting in the Cinema (Berkeley: University of California Press, 1988), chaps. 2–5. Svetlana Alpers and Michael Baxandall, Tiepolo and the Pictorial Intelligence (New Haven, CT: Yale University Press, 1994). André Bazin, “The Evolution of the Language of Cinema,” in What Is Cinema? vol. 1, trans. and ed. Hugh Gray (Berkeley: University of California Press, 1967), 23–40; Bazin, Orson Welles: A Critical View, trans. Jonathan Rosenbaum (New York: Harper and Row, 1978), 64–82. See also David Bordwell, On the History of Film Style (Cambridge, MA: Harvard University Press, 1997), chap. 3. Alexandre Astruc, “The Birth of a New Avant-Garde: Le Caméra-Stylo,” in The New Wave, ed. Peter Graham (Garden City, NY: Doubleday, 1968), 20. François Truªaut, “Sacha Guitry, cinéaste,” in Sacha Guitry, Cinéma et moi, ed. André Bernard and Claude Gauteur, 2nd ed. (Paris: Ramsay, 1977), 13–14. See Alexandre Astruc, “L’évolution du cinéma américain,” in Du Stylo à la caméra . . . et de la caméra au stylo: Écrits (1942–1984) (Paris: Archipel, 1992), 291–292; and Antoine de Baecque, Les Cahiers du cinéma: Histoire d’une revue, vol. 1, À l’assaut du cinéma, 1951–1959 (Paris: Cahiers du cinéma, 1991), 175. Quoted in de Baecque, Les Cahiers du cinéma, 1:183. See ibid., 193–205; and David Bordwell, Making Meaning: Inference and Rhetoric in the Interpretation of Cinema (Cambridge, MA: Harvard University Press, 1989), 45–48. Michel Mourlet, “Sur un art ignoré,” in La mise en scène comme langage (Paris: Veyrier, 1987), 42. Ibid., 46. Andrew Sarris, “Mise-en-scène,” in The Primal Screen: Essays on Film and Related Subjects (New York: Simon and Schuster, 1973), 85. For discussion of these critical traditions see Bordwell, Making Meaning, 48–53. Ian Cameron, “Michelangelo Antonioni,” Film Quarterly 16, no. 1 (fall 1962): 2–58; reprinted in Ian Cameron and Robin Wood, Antonioni (New York: Praeger, 1968), 6–105. V. F. Perkins, Film as Film: Understanding and Judging Movies (Harmondsworth: Penguin, 1972), 74. Andrew Sarris, “Notes on the Auteur Theory in 1970,” in The Primal Screen: Essays on Film and Related Subjects (New York: Simon and Schuster, 1973), 60–61. The film is A Picture of Madame Yuki (1950). Thierry Jousse, “Renoir vivant,” Cahiers du cinéma, no. 482 (July-August 1994): 5. Dominique Rabourdin, quoted in Jean-François Houben, Feux croisés sur la critique: Dix-

NOTES TO PAGES 4–13

28. 29. 30. 31.

32. 33.

34. 35.

36.

37.

38.

39. 40. 41.

sept entretiens (Paris: Cerf, 1999), 355–356; see also 27, 63–64, 120–121, 167, 212, 231, 256, 271–272, 294–295, 298, and 318. See Bazin, Orson Welles, 77–81. Quoted in Ellen Oumano, Film Forum: Thirty-Five Top Filmmakers Discuss Their Craft (New York: St. Martin’s, 1985), 81. Bazin, “Evolution,” 35. See Vladimir Nizhny, Lessons with Eisenstein, trans. and ed. Ivor Montagu and Jay Leyda (New York: Hill and Wang, 1962), 100–139; and the discussion in my Cinema of Eisenstein (Cambridge, MA: Harvard University Press, 1993), 150–155. See Kuleshov on Film: Writings by Lev Kuleshov, ed. and trans. Ronald Levaco (Berkeley: University of California Press, 1974), 108–115. Quoted in Roger Ebert and Gene Siskel, The Future of the Movies: Interviews with Martin Scorsese, Steven Spielberg, and George Lucas (Kansas City, MO: Andrews and McNeel, 1991), 72–73. Quoted in Stevens, Actors Turned Directors, 375. See also Ken Dancyger, The Technique of Film and Video Editing (Boston: Focal Press, 1993), 257. See my “Intensified Continuity: Visual Style in Contemporary American Film,” Film Quarterly 55, no. 3 (spring 2002): 16–28; Kristin Thompson and David Bordwell, Film History: An Introduction, 2nd ed. (New York: McGraw-Hill, 2002), 688–689; and Bordwell and Thompson, Film Art, 327–331. The assistant director believed that if the audience liked this scene, they’d probably like the rest of the film. See Jerry Ziesmer, Ready When You Are, Mr. Coppola, Mr. Spielberg, Mr. Crowe (Lanham, MD: Scarecrow, 2000), 425. The concept of average shot length derives from Barry Salt, Film Style and Technology: History and Analysis, 2nd ed. (London: Starword, 1992), 142–147. My ASLs here are based on watching the entire film and dividing its running length, given in seconds, by the number of shots. Both images and intertitles are counted as shots, but production credits aren’t. The averages are based on complete viewing of more than three hundred films from the period 1961–2000. My estimates of studio-era norms are taken from David Bordwell, Janet Staiger, and Kristin Thompson, The Classical Hollywood Cinema: Film Style and Mode of Production to 1960 (New York: Columbia University Press, 1985), 60–63. Salt’s results from the 1960s onward, which largely correspond to the claims set out here, can be found on pp. 214–215, 236–240, and 249 of Film Style and Technology. Whereas Salt seeks to summarize average shot lengths of a period in a “Mean Average Shot Length,” Bordwell, Staiger, and Thompson argue for thinking of ASLs as occupying a range of probable choice. See, e.g., Patrick Tucker, Secrets of Screen Acting (New York: Routledge, 1994), where the author teaches actors how to react in close-up and “cheat” the face toward the camera in two-shots. Note also the claim that “Blocking is a way of getting the camera to see your face—sometimes with another face at the same time” (129). See David W. Samuelson, Panaflex Users’ Manual (Boston: Focal Press, 1990), xiii-xiv. Quoted in Ebert and Siskel, Future of the Movies, 73. See Boorstin, Making Movies Work, 90–97.

NOTES TO PAGES 13–28

273

42.

43.

44. 45. 46. 47.

48. 49. 50.

51. 52. 53. 54. 55.

274

Manhattan (1978) has an ASL of more than eighteen seconds, and that of Mighty Aphrodite (1995) is thirty-six seconds. “I don’t like to do all that cutting because it’s cheaper my way—it’s quicker and the actors like it. This way I can sit down with them, let them talk, put the camera on them, and live things happen. They can do it diªerently each time. Nothing has to match” (quoted in Douglas McGrath, “If You Knew Woody like I Knew Woody,” New York Magazine, October 17, 1994, 44). See also Richard Schickel, Woody Allen: A Life in Film (Chicago: Ivan R. Dee, 2003), 161–162. An intriguing case is Jim Jarmusch, who started out using very long takes, with Stranger Than Paradise (1984) averaging over forty seconds per shot. Yet in the period in which intensified continuity became the dominant style, his reliance on long takes diminished steadily, from Mystery Train (1989; ASL 23 seconds) to Ghost Dog: The Way of the Samurai (1999; ASL 6.8 seconds). Dwight MacDonald, “Tom Jones [1964],” On Movies (Englewood Cliªs, NJ: Prentice-Hall, 1969), 393. Andrew Sarris, Confessions of a Cultist: On the Cinema, 1955–1969 (New York: Simon and Schuster, 1971), 206, 261–262. Pauline Kael, “Trash, Art, and the Movies [1969],” in For Keeps (New York: Dutton, 1994), 205–206. It’s often said that film scholars pay more attention to visual style than to sound, but I think the claim is overstated. Today we can name a dozen or so prominent scholars who have film sound as a primary research interest (notably Rick Altman, Claudia Gorbman, Kathryn Kalinak, and Jeª Smith), but many fewer study visual style. The scholars of film music easily outnumber students of composition, color, or camerawork. A lively survey of these perceptual processes is Steven Pinker’s How the Mind Works (New York: Norton, 1997), chaps. 1–2. For a review of theories of expression in the arts see Noël Carroll, Philosophy of Art: A Contemporary Introduction (New York: Routledge, 1999), chap. 2. An in-depth discussion of this process can be found in E. H. Gombrich, The Sense of Order: A Study in the Psychology of Decorative Art (Ithaca, NY: Cornell University Press, 1979). See also James Trilling, Ornament: A Modern Perspective (Seattle: University of Washington Press, 2003), where Trilling distinguishes between decoration in general (Michelangelo’s Last Judgement decorates the Sistine Chapel) and ornament, “in which the visual pleasure of form significantly outweighs the communicative value of content” (23). What I am considering the most palpable and evidently “decorative” uses of film style would be ornament in Trilling’s sense. I try to make this case for Ozu Yasujiro in Ozu and the Poetics of Cinema (Princeton, NJ: Princeton University Press, 1988), chaps. 4–6. See David Bordwell, Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985), chap. 12. Quoted in Jay Holben, “Sole Survivor,” American Cinematographer 82, no. 1 (January 2001): 40–41. See Bordwell, Narration in the Fiction Film, chap. 3. See Argyle, Bodily Communication, 98–100, 178, 305–306. For a study of gesture’s relation to speech see David McNeill, Hand and Mind: What Gestures Reveal about Thought

NOTES TO PAGES 30–37

56.

57.

58.

59.

60.

61.

(Chicago: University of Chicago Press, 1992). On cultural diªerences see Jan Bremmer and Herman Roodenburg, eds., A Cultural History of Gesture (Ithaca, NY: Cornell University Press, 1991); and Roger E. Axtell, Gestures: The Do’s and Taboos of Body Language around the World, rev. ed. (New York: Wiley, 1998). On basic expressions and display rules see Paul Ekman, “Cross-Cultural Studies of Facial Expression,” in Darwin and Facial Expression: A Century of Research in Review, ed. Paul Ekman (New York: Academic Press, 1973), 169–222; Paul Ekman, “Afterword,” in The Expression of the Emotions in Man and Animals, by Charles Darwin, 3rd ed. (New York: Oxford, 1998), 375–385. See also the essays collected in What the Face Reveals, ed. Paul Ekman and Erika Rosenberg (New York: Oxford University Press, 1997). A wonderful and detailed study is Gary Faigin, The Artist’s Complete Guide to Facial Expression (New York: Watson-Guptill, 1990). See Ed Tan, “Three Views of Facial Expression and Its Understanding in the Cinema,” in Motion Picture Theory: Ecological Considerations, ed. Joseph D. Anderson and Barbara Anderson (Carbondale: University of Southern Illinois Press, 2005). See Argyle, Bodily Communication, 35–41, 109, 153–154. A survey of the relevant research is Vicki Bruce and Andy Young, In the Eye of the Beholder: The Science of Face Perception (Oxford: Oxford University Press, 1998). Daniel McNeill, The Face: A Natural History (Boston: Little, Brown, 1998), oªers an entertaining and wide-ranging discussion. For a related argument see Joseph Anderson, The Reality of Illusion: An Ecological Approach to Cognitive Film Theory (Carbondale: University of Southern Illinois Press, 1996), 90–134. I discuss this performance tactic in “Who Blinked First? How Film Style Streamlines Nonverbal Interaction,” in Film Style and Story: A Tribute to Torben Grodal, ed. Lennard Højberg and Peter Schepelern (Copenhagen: Museum Tuscalanem Press, 2003), 45–57. On the importance of eyes in the earliest films see Janet Staiger, “The Eyes Are Really the Focus: Photoplay Acting and Film Form and Style,” Wide Angle 6, no. 4 (1985): 14–23. Many humanistic scholars are uneasy with the idea of universals; in the last chapter I try to relax them. But one reservation is relevant to the argument in this chapter. Miriam Bratu Hansen has suggested that the transcultural appeal of mainstream or “classical Hollywood cinema” (which she calls “an international modernist idiom”) relies on something other than regularities of perceptual and cognitive pickup. She characterizes the position she rejects in this way: “Once ‘the system’ is in place (from about 1917 on), its ingenuity and stability are attributed 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.” This isn’t an accurate account of my argument in the essay she cites, or in this book, because 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 she cites is that of face-to-face interactions in conversation. It isn’t for me to say whether the cross-cultural constants in such interactions are strictly a biological adaptation, a byproduct of biological adaptation, a pure product of culture, or a mix of all of these. That’s why I call such regularities “contingent universals.” The sources of such universals don’t matter for my argument; my point is that such face-to-face interaction is a cross-cultural

NOTES TO PAGES 37–38

275

62. 63.

64. 65. 66.

276

universal, and cinema has found stylistic figures (e.g., shot/reverse-shot framing and cutting) that amplify and streamline this for easy uptake. However, it does seem plausible that many aspects of perception activated by style, such as the detection of edges and surfaces, or the recognition of other humans on the screen, are biologically hardwired (requiring, of course, exposure to environmental regularities in order to mature normally). Without arguing against the existence of contingent universals, Hansen goes on to assert that a better explanation for the cross-cultural legibility of mainstream cinema is that it oªers 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 [not the same as style—D. B.] but because it meant diªerent things to diªerent people and publics, both at home and abroad.” “Meaning diªerent things” is not glossed, but the questions Hansen goes on to ask (about how films were programmed, how genres were received in diªerent countries, etc.) suggest that we have moved far beyond the level of face recognition and the construction of space. The irregularities of interpretive reception she mentions are of a diªerent order than the regularities I sought to explain. Furthermore, it would seem that before distinct audiences can construe a story’s meanings in various culture-driven ways, they would have to understand at least some aspects of the story, along with the stylistic presentation of those aspects. So not only are the stylistic transformations of contingent universals on a diªerent logical level than the plurality of meanings issuing from reception; the former are preconditions of the latter. There seems to be no getting around the ubiquity of the sort of perceptual-cognitive skills that stylistic features prompt. The essay is “The Mass Production of the Senses: Classical Cinema as Vernacular Modernism,” in Reinventing Film Studies, ed. Christine Gledhill and Linda Williams (London: Arnold, 2000), 332–350; passages cited are on 339–341. Anderson articulates this line of argument well in Reality of Illusion, chaps. 2 and 3. I try to make this case in “Convention, Construction, and Cinematic Vision,” in PostTheory: Reconstructing Film Studies, ed. David Bordwell and Noël Carroll (Madison: University of Wisconsin Press, 1996), 87–107. For more information on eye movements see Stephen E. Palmer, Vision Science: Photons to Phenomenology (Cambridge, MA: MIT Press, 2002), 520–533. James E. Hoªman, “Visual Attention and Eye Movements,” in Attention, ed. Harold Pashler (London: Psychology Press, 1998), 119–120. For a useful survey of eye movements and pictorial composition see Robert L. Solso, Cognition and the Visual Arts (Cambridge, MA: MIT Press, 1994), 129–155. An early summary is Julian Hochberg, “The Representation of Things and People,” in E. H. Gombrich, Julian Hochberg, and Max Black, Art, Perception, and Reality (Baltimore: Johns Hopkins University Press, 1972), 47–94; see also Julian Hochberg, “Visual Art and the Structures of the Mind,” in The Arts, Cognition, and Basic Skills, ed. Stanley Madeja (St. Louis, MO: Cemrel, 1978), 169–171. This body of research is applied to film in Julian Hochberg and Virginia Brooks, “The Perception of Motion Pictures,” in Cognitive Ecology, ed. Morton P. Friedman and Edward C. Carterette (San Diego, CA: Academic Press, 1996), 252–254; and Julian Hochberg and Virginia Brooks, “Movies in the Mind’s Eye,”

NOTES TO PAGES 38–39

67.

68. 69. 70. 71. 72.

in Post-Theory: Reconstructing Film Studies, ed. David Bordwell and Noël Carroll (Madison: University of Wisconsin Press, 1996), 368–387. A superb application of eye-movement theory to the study of a painting is Michael Baxandall’s “Fixation and Distraction: The Nail in Braque’s Violin and Pitcher (1910),” in Sight and Insight: Essays on Art and Culture in Honour of E. H. Gombrich at 85, ed. John Onions (London: Phaidon, 1994), 398–415. A helpful discussion is in Thomas Puttfarken, “Composition, Perspective, and Presence: Observations on Early Academic Theory in France,” in Sight and Insight: Essays on Art and Culture in Honour of E. H. Gombrich at 85, ed. John Onions (London: Phaidon, 1994), 287–297. See also Puttfarken’s Roger de Piles’ Theory of Art (New Haven, CT: Yale University Press, 1985); and his magisterial The Discovery of Pictorial Composition: Theories of Visual Order in Painting, 1400–1800 (New Haven, CT: Yale University Press, 2000). William Hogarth, The Analysis of Beauty, ed. Ronald Paulson (New Haven, CT: Yale University Press, 1997), 33. The quotation is from Billy Williams, in Screencraft: Cinematography, ed. Peter Ettedgui (Crans-Pres-Celigny, Switzerland: Rotovision, 1998), 105. Quoted in David E. Williams, “Shooting to Kill,” American Cinematographer 76, no. 11 (1995): 56. I owe this suggestion to Noël Carroll. Quoted in “The Films of David Zucker, Jim Abrahams, and Jerry Zucker,” in Robert J. Emory, Directors: Take One (New York: TV Books, 1999), 341.

CHAPTER 2. FEUILLADE, OR STORYTELLING 1. 2. 3. 4.

5.

6. 7.

Quoted in Francis Lacassin, Louis Feuillade (Paris: Seghers, 1964), 109. Louis Feuillade, “Introduction,” in French Film Theory and Criticism, 1907–1939, ed. Richard Abel (Princeton, NJ: Princeton University Press, 1988), 1:225. Henri Fescourt, La Foi et les montagnes (Paris: Paul Montel, 1959), 83. For discussions of the installment film in France see Richard Abel, French Cinema: The First Wave, 1915–1929 (Princeton, NJ: Princeton University Press, 1984), 69–85; Francis Lacassin, Pour une contre-histoire du cinéma (Paris: Union Générale d’Editions, 1972), 113–126; and Monica Dall’Asta, “Le film à épisodes entre système primitif et système classique,” in Les vingts premières années du cinéma français, ed. Jean A. Gili et al. (Paris: Sorbonne, 1995), 249–259. In what follows I draw my examples from some shorts and the major installment films and serials. These were made available to me principally through the Royal Film Archive of Belgium. After this chapter was substantially completed I was able to view another thirty-six Feuillade films—shorts, short features, and Vendémiaire—at the 2000 edition of the Giornate del Cinema Muto in Sacile, Italy. These works tend to confirm the conclusions presented here, but I don’t discuss them because I’ve had no chance to study them on an editing table. Musidora reports this incident in Georges Sadoul, Le Cinéma devient un art: L’avant-guerre, 1909–1920, vol. 3 of Histoire générale du cinéma (Paris: Denoël, 1973), 431. René Navarre cited in ibid., 252.

NOTES TO PAGES 39–45

277

8. 9.

10. 11.

12.

13.

14.

15. 16.

17.

18.

278

Robert Florey, La Lanterne magique (Lausanne: La Cinémathèque Suisse, 1966), 12–13. Anecdotal information in the preceding paragraphs comes from two biographies by Francis Lacassin, Louis Feuillade and Louis Feuillade: Maître des lions et des vampires (Paris: Bordas, 1995). See also Marcel Levesque, “Cinq Ans à la Maison Gaumont avec Louis Feuillade,” Archives, no. 8 (November 1987): 1–11; and no. 9 (December 1987): 1–11. Some special numbers of journals are also useful: “Louis Feuillade,” Cahiers de la Cinémathèque, no. 48 (1988); “Feuillade and the French Serial,” Velvet Light Trap, no. 37 (spring 1996); and 1895, hors-série (October 2000). Lacassin, Louis Feuillade, 235–236; and Lacassin, Maître des lions, 68, 227. See Richard Abel, “Before Fantômas: Louis Feuillade and the Development of Early French Cinema,” Post Script 7, no. 1 (1987): 4; Jean-Jacques Meusy, “Palaces and Holes in the Wall,” Velvet Light Trap, no. 37 (spring 1996): 93–94; and Jean-Jacques Meusy, “How Cinema Became a Cultural Industry: The Big Boom in France between 1905 and 1908,” Film History 14, nos. 3/4 (2002): 418–429. See David Bordwell, Janet Staiger, and Kristin Thompson, The Classical Hollywood Cinema: Film Style and Mode of Production to 1960 (New York: Columbia University Press, 1985), 42–69, 184–230; Barry Salt, Film Style and Technology: History and Analysis, 2nd ed. (London: Starword, 1992), 40–147; Charlie Keil, Early American Cinema in Transition: Story, Style, and Filmmaking, 1907–1913 (Madison: University of Wisconsin Press, 2001), 45–174; David Bordwell, On the History of Film Style (Cambridge, MA: Harvard University Press, 1997), 129–139. I’ve taken my example from the most widely available Feuillade film, Les Vampires, and the timings cited are from the video version (Image DVD 5960WBDVD), which appears to be run at the proper speed. Here are some ASLs for 1915: The Italian (10 sec.); The Innocence of Ruth (9 sec.); The Cheat (9 sec.); Trilby (13 sec.); Regeneration (4.5 sec.). For 1916: The Social Secretary (7 sec.); The Scarlet Road (9 sec.); Hoodoo Anne (7 sec.). I use here the version of Intolerance released on laserdisc by Image Entertainment, ID8074DS. There is one jump in the last tableau of the Vampires’ party, which I assume to be a missing title; I have not counted the hypothetical title as a shot, but because of the distinct change of character position and lighting, I have counted the final tableau as two shots. For further comparison of Gri‹th and Feuillade on these grounds see Nathalie Leplongeon, “Les Vampires de Feuillade, une logique de transition,” in Les vingts premières années du cinéma français, ed. Jean A. Gili et al. (Paris: Sorbonne, 1995), 261–268. Interestingly, in 1920 Gaumont wrote to Feuillade from an American trip and described current US films as “built out of an infinite number of brief scenes [shots?]. Each represents an original treatment, a fresh idea, an eªect of lighting, etc., so good that from beginning to end, the spectator finds himself satisfied.” See Francis Lacassin, “Lettres de Léon Gaumont à Louis Feuillade,” Cinéma, no. 95 (1965): 88. A more fanciful treatment of Feuillade’s relation to Gri‹th can be found in Philippe Azoury and Jean-Marc Lalanne, Fantômas, style moderne (Paris: Centre Pompidou, 2002), 49–50. For examples of complex staging in American films see Ben Brewster and Lea Jacobs,

NOTES TO PAGES 45–48

19.

20.

21. 22.

23. 24.

25. 26. 27. 28. 29. 30. 31. 32. 33.

34.

Theatre to Cinema: Stage Pictorialism and the Early Feature Film (Oxford: Oxford University Press, 1997), 164–187; and Bordwell, On the History of Film Style, 184–187. Barry Salt’s measurements of average shot lengths support this conclusion. He finds that in American films from 1912 to 1917, the most common ASL (the mode) was seven seconds, whereas for European films of the same period, thirteen seconds was most common. In addition, most of the American films he examined from the period had ASLs between four and eleven seconds, whereas none of the European films had ASLs of fewer than six seconds, and the European ASLs were far more uniformly distributed between six seconds and twenty-two seconds. See Salt, Film Style and Technology, 146–147. As I noted in chapter 1, we need not claim that directors determine where spectators look or when they look there; the director cannot program the spectator’s vision absolutely. (We can, if we want, turn away from the screen, close our eyes, or stare at a corner of the frame.) Nonetheless, directors create the conditions under which spectators are prompted (cued, encouraged) to concentrate on certain parts of the frame. All other things being equal, our predispositions make pictorial factors operate as default values, a line of least resistance. Cited in Lacassin, Louis Feuillade, 115. The models might seem only to confirm Feuillade’s “bourgeois” tastes. (If he had mentioned Courbet and Redon instead, he might be redeemable as a protomodernist.) Still, both Millet and Puvis de Chavannes exploit a clarity of outline and a purity of composition that have a‹nities with Feuillade’s style. Kristin Thompson, “The Formulation of the Classical Style, 1909–28,” in Bordwell, Staiger, and Thompson, Classical Hollywood Cinema, 200–223. I adopt the phrase “visual weight” from Rudolf Arnheim’s fine book, The Power of the Center: A Study of Composition in the Visual Arts: The New Version (Berkeley: University of California Press, 1988). I discuss some early instances of forward-thrusting movement in On the History of Film Style, 171–174. Yhcam, “Cinematography,” in Abel, French Film Theory and Criticism, 1:73–74. Colin N. Bennett, The Guide to Kinematography for Camera Men, Operators, and All Who “Want to Know” (London: Heron, 1917), 55. Adrian Brunel, Filmcraft: The Art of Film Production (London: Newnes, 1935), 40. For more on the trapezoidal playing space of cinema see Brewster and Jacobs, Theatre to Cinema, 170–174; and Bordwell, On the History of Film Style, 163–198. Abel, “Before Fantômas,” 6–9. Robert Florey, “Louis Feuillade,” Cinémagazine, September 2, 1921, n.p. Georges Sadoul, Les pionniers du cinéma, 1897–1909, vol. 2 of Histoire générale du cinéma (Paris: Denoël, 1947), 540. See, e.g., Barry Salt, “L’espace d’à coté,” in Les vingts premières années du cinéma français, ed. Jean A. Gili et al. (Paris: Sorbonne, 1995), 201; and Richard Abel, The Ciné Goes to Town: French Cinema, 1896–1914 (Berkeley: University of California Press, 1994), 249– 251. See also Roland Cosandey, “Le plan de l’éscalier: L’Assassinat du duc de Guise (Film d’Art, 1908),” iichiko, no. 64 (autumn 1999): 36–64. Quoted in Sadoul, Le Cinéma devient un art, 252.

NOTES TO PAGES 48 –76

279

35. 36.

37.

38. 39.

40. 41.

42. 43. 44.

I discuss staging in one scene of Ingeborg Holm in On the History of Film Style, 191–195. For a more detailed discussion of Reinert’s precision staging see my article “Taking Things to Extremes: Hallucinations Courtesy of Robert Reinert,” Aura 6, no. 2 (2000): 4–19. See Yuri Tsivian, “Portraits, Mirrors, Death: On Some Decadent Clichés in Early Russian Film,” Iris, nos. 14–15 (autumn 1992): 75–77. I’m grateful to Yuri for introducing me to this sequence. On another scene in The King of Paris see Brewster and Jacobs, Theatre to Cinema, 124–126. Gombrich has discussed this process on several occasions; an early statement is in his The Story of Art (London: Phaidon, 1950), 131–208. See Thompson, “Formulation of the Classical Style,” 167–173. For a discussion of how Feuillade’s narrative construction responded to the demand for complex films see Nathalie Leplongeon, “Les Vampires de Feuillade, une stratégie de coopération spectatorielle,” Iris, no. 17 (fall 1994): 167–182. See William Uricchio and Roberta Pearson, Reframing Culture: The Case of the Vitagraph Quality Films (Princeton, NJ: Princeton University Press, 1993). Researchers into early cinema seem to disagree on the extent of theater’s influence. For Ben Brewster and Lea Jacobs, 1910s films became in most respects repositories of theatrical techniques of staging and performance (see Theatre to Cinema, 212–216). For Yuri Tsivian, filmmakers of the period strove to dissociate themselves from theater by exploiting precision staging and developing fine-art iconography, thereby creating compositions that were “paintings in motion.” See “Two ‘Stylists’ of the Teens: Franz Hofer and Yevgenii Bauer,” in A Second Life: German Cinema’s First Decades, ed. Thomas Elsaesser (Amsterdam: Amsterdam University Press, 1996), 264–276. Gombrich, Story of Art, 3. Ibid., 190. Thompson, “Formulation of the Classical Style,” 162.

CHAPTER 3. MIZOGUCHI, OR MODULATION 1. 2.

3. 4.

280

See Kristin Thompson, Exporting Entertainment: America in the World Film Market, 1907–1934 (London: British Film Institute, 1985), chap. 4. A thorough discussion of this trend can be found in Joanne Bernardi, Writing in Light: The Silent Scenario and the Japanese Pure Film Movement (Detroit, MI: Wayne State University Press, 2001). See also Aaron Andrew Gerow, “Writing a Pure Cinema: Articulations of Early Japanese Film” (PhD diss., University of Iowa, 1996). Bernardi, Writing in Light, 102. Hiroshi Komatsu surveys aesthetic developments in early Japanese cinema in “The Fundamental Change: Japanese Cinema before and after the Earthquake of 1923,” Gri‹thiana, nos. 38/39 (October 1990): 186–196. He points out, “By this period [the mid-1920s] the Japanese cinema in general had been completely assimilated to the Western mode” (193). See also Peter B. High, “Japanese Film and the Great Kanto Earthquake,” International Studies (Research Institute for International Studies, Chvbu University), no. 3 (1985): 71–84.

NOTES TO PAGES 77– 83

5.

6.

7. 8. 9.

10. 11. 12.

13.

14. 15.

Overviews of this period can be found in Marius B. Jansen, The Making of Modern Japan (Cambridge, MA: Harvard University Press, 2000), chaps. 15 and 16; and James L. McClain, Japan: A Modern History (New York: Norton, 2002), chaps. 10 and 11. See also Sharon A. Minichiello, ed., Japan’s Competing Modernities: Issues in Culture and Democracy (Honolulu: University of Hawai’i Press, 1998); and Elise K. Tipton and John Clark, eds., Being Modern in Japan: Culture and Society from the 1910s to the 1930s (Sydney: Australian Humanities Research Foundation, 2000). A vivid pictorial record of these tumultuous years can be found in James Fraser, Steven Heller, and Seymour Chwast, Japanese Modern: Graphic Design between the Wars (San Francisco: Chronicle Books, 1996). The work of this generation lies at the center of most synoptic histories of Japanese cinema. See Shinobu and Marcel Giuglaris, Le cinéma japonais (1895–1955) (Paris: Cerf, 1956); Joseph L. Anderson and Donald Richie, The Japanese Film: Art and Industry, rev. ed. (Princeton, NJ: Princeton University Press, 1982); Noël Burch, To the Distant Observer: Form and Meaning in the Japanese Cinema (Berkeley: University of California Press, 1979). For more on these trends see my Ozu and the Poetics of Cinema (Princeton, NJ: Princeton University Press, 1988), 23–24. I argue for this in ibid., 88–102. I discuss this tradition in “A Cinema of Flourishes: Japanese Decorative Classicism of the Prewar Era,” in Reframing Japanese Cinema, ed. David Desser and Arthur Noletti (Bloomington: Indiana University Press, 1992), 327–345; and “Visual Style in Japanese Cinema, 1925–1945,” Film History 7, no. 1 (spring 1995): 5–31. See also Peter Rist, “Camera Movement in Japanese Silent Films,” Asian Cinema 14, no. 2 (fall/winter 2003): 197–205. Quoted in Margaret Iverson, Alois Riegl: Art History and Theory (Cambridge, MA: MIT Press, 1993), 67. A good survey is Shvichi Kato, A History of Japanese Literature, vol. 3, The Modern Years (New York: Kodansha, 1983), 223–249. Biographical information in the previous two paragraphs comes from Akira Iwazaki, Kenji Mizoguchi, 1898–1956 (Paris: Anthologie du cinéma/Editions l’Avant-scène, 1968); Shindo Kaneto, Aru eiga kantoku no sho: Mizoguchi Kenji no kiroku [The Life of a Film Director: Record of Kenji Mizoguchi] (Tokyo: Eijin-sha, 1975); Tumura Hideo, Mizoguchi Kenji to iu onoko [A Man Called Kenji Mizoguchi] (Tokyo: Yoshiga Shuten, 1977); Yoshikata Yoda, Souvenirs de Kenji Mizoguchi, trans. Koichi Yamada, Bernard Béraud, and André Moulin (Paris: Cahiers du cinéma, 1997); Dudley Andrew and Paul Andrew, Kenji Mizoguchi: A Guide to References and Resources (Boston: G. K. Hall, 1981); and Keiko McDonald, Mizoguchi (Boston: Twayne, 1984). The last three books named have been especially valuable in the writing of this chapter. Ian Buruma provides a thought-provoking survey of these motifs in Behind the Mask: On Sexual Demons, Sacred Mothers, Transvestites, Gangsters, and Other Japanese Cultural Heroes (New York: Pantheon, 1984), 31–34, 53–85. Quoted in Hazumi Tsuneo, “Self-Assessment: An Unsatisfied Filmmaker [1950],” in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 12. Quoted in Tadao Sato, “On Kenji Mizoguchi,” Film Criticism 4, no. 3 (spring 1980): 4.

NOTES TO PAGES 84–88

281

16. 17. 18. 19.

20.

21.

22. 23.

24. 25. 26.

27.

28. 29.

282

Tsuji Hisakuzu, quoted in John Gillett, “Japanese Notebook,” Sight and Sound 42, no. 1 (winter 1972/73): 27. Yoda, Souvenirs, 82. Quoted in editorial note to Narusawa Masashige, “No Fakes Allowed,” in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 14. A helpful anthology on changing attitudes in the intelligentsia is J. Thomas Rimer, ed., Culture and Identity: Japanese Intellectuals during the Interwar Years (Princeton, NJ: Princeton University Press, 1990). For an overview of how artists and intellectuals were brought into line with wartime policy see Ben-Ami Shillony, Politics and Culture in Wartime Japan (Oxford: Oxford University Press, 1981), 91–156. Quoted in Iwazaki, Kenji Mizoguchi, 463. Peter B. High oªers a detailed account of Japanese filmmakers’ roles in promoting military ideology in The Imperial Screen: Japanese Film Culture in the Fifteen Years’ War, 1931–1945 (Madison: University of Wisconsin Press, 2003); he discusses Mizoguchi’s military projects on 34, 53, 187, and 330–333. Nihon kaiki is discussed in relation to writers in Nago Nishikawa, “Occidentalisation et ‘Retour au Japon,’” Corps écrit, no. 17 (1986): 81–90; and Donald Keene, Dawn to the West: Japanese Literature in the Modern Era: Poetry, Drama, Criticism (New York: Holt, Rinehart, and Winston, 1984), 613. A useful overview of such trends is Bob Tadashi Wakabayashi, ed., Modern Japanese Thought (Cambridge, UK: Cambridge University Press, 1998); see esp. chap. 4, “Socialism, Liberalism, and Marxism, 1901–31,” by Peter Duus and Irwin Scheiner, and chap. 5, “Japan’s Revolt against the West,” by Tetsuo Najita and H. D. Harootunian. Vê-Hô, Mizoguchi (Paris: Editions Universitaires, 1963); Michel Mesnil, Kenji Mizoguchi (Paris: Seghers, 1965). Jean Douchet, “Conoscenza di Mizoguchi,” in Il Cinema di Kenji Mizoguchi, ed. Adriano Aprà, Enrico Magrelli, and Patrizia Pistagnesi (Venice: Mostra internazionale del cinema, 1980), 78. “Mizoguchi,” in Godard sur Godard, ed. Alain Bergala (Paris: Éditions de l’Étoile, 1985), 123. “Mizoguchi vu d’ici,” Cahiers du cinéma, no. 48 (June 1955): 28. Surviving copies of Taki no Shiraito and The Downfall of Osen lack important scenes, and transitional bits of The Poppy seem to be missing. The Straits of Love and Hate (1937) survives in a deeply scarred, sometimes illegible copy. Apart from his eighty-three features, Mizoguchi contributed episodes to two portmanteau films, The Earth Smiles (1925), which is lost, and Victory Song (1945), which survives. In my total I count the two-part Genroku Chushingura (1941–1942) as a single film. I base this figure on a sample of 106 Japanese films from the period 1932–1945 in all genres, not including Mizoguchi’s own works. In this sample the mean ASL is 12.4; the median falls between 11.0 and 11.9; the mode falls between 9.0 and 9.9. The range of 9.0 to 12.9 seconds ASL accounts for 50 percent of the films in the group. Ariane Mnouchkine, “Entretien avec Yoda Yoshikata et Miyagawa Kazuo,” Cahiers du cinéma, no. 185 (August-September 1964): 24. Quoted in Kishi Matsuo, “A Talk with Mizoguchi [1952],” in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 12.

NOTES TO PAGES 88–95

30. 31.

32. 33. 34. 35. 36.

37. 38. 39. 40.

41.

42.

43.

44. 45.

Ibid.; and Hazumi, “Self-Assessment,” 11. Mizoguchi claimed that Okichi, Mistress of a Foreigner (1930; lost) was his first eªort in the one scene–one cut method. See Mizoguchi’s remarks in “Tavola rotonda con Kenji Mizoguchi [1937],” in Il Cinema di Kenji Mizoguchi, ed. Adriano Aprà, Enrico Magrelli, and Patrizia Pistagnesi (Venice: Mostra internazionale del cinema, 1980), 130. Quoted in Kishi, “Talk with Mizoguchi,” 11. Quoted in Shinoda Masahiro, “Far from Mizoguchi,” in “The Masters of Japanese Film,” ed. and trans. Leonard Schrader (unpublished ms., n.d.), 121. Quoted in Hazumi Tsuneo, “Trois interviews de Mizoguchi,” Cahiers du cinéma, no. 116 (February 1961): 17. Quoted in Mesnil, Kenji Mizoguchi, 77. On the 1910s accomplishments see Kristin Thompson, “The International Exploration of Cinematic Expressivity,” in Film and the First World War, ed. Karel Dibbets and Bert Hogenkamp (Amsterdam: Amsterdam University Press, 1995), 65–85. André Bazin, “The Evolution of the Language of Cinema,” in What Is Cinema? vol. 1, trans. and ed. Hugh Gray (Berkeley: University of California Press, 1967), 31–33. Philippe Demonsablon, “La Splendeur du vrai,” Cahiers du cinéma, no. 95 (May 1959): 3. See Buruma, Behind the Mask. Background on this development may be found in Ellen P. Conant, in collaboration with Steven D. Owyoung and J. Thomas Rymer, Nihonga: Transcending the Past: Japanese-Style Painting, 1868–1968 (New York: Weatherhill, 1995); and Michiyo Morioka and Paul Berry, Modern Masters of Kyoto: The Transformation of Japanese Painting Traditions (Seattle, WA: Seattle Art Museum, 1999). In Ozu and the Poetics of Cinema I argue that mono no aware, “the sadness of things,” exemplifies this process of ongoing revision (28–29). On iki see Leslie Pincus, Authenticating Culture in Imperial Japan: Kuki Shvzo and the Rise of a National Aesthetic (Berkeley: University of California Press, 1996). See Irokawa Daikichi, The Culture of the Meiji Period, ed. Marius B. Jansen (Princeton, NJ: Princeton University Press, 1985); Carol Gluck, Japan’s Modern Myths: Ideology in the Late Meiji Period (Princeton, NJ: Princeton University Press, 1985); and Irmela Hijiya-Kirschnereit, “The Concepts of Tradition in Modern Japanese Literature,” in Tradition and Modern Japan, ed. P. G. O’Neill (Tenterden, Kent: Norbury, 1981), 206–216. Since the 1980s this line of argument has been pursued by many scholars. An important collection is Stephen Vlastos, ed., Mirror of Modernity: Invented Traditions of Modern Japan (Berkeley: University of California Press, 1998). Vlastos’s introduction mounts a vigorous case for this research program. See also Ian Buruma’s Inventing Japan: 1853–1964 (New York: Modern Library, 2003). See Miriam Silverberg, “Constructing the Japanese Ethnography of Modernity,” Journal of Asian Studies 51, no. 1 (1992): 30–54; Tessa Morris-Suzuki, Re-Inventing Japan: Time, Space, Nation (Armonk, NY: M. E. Sharpe, 1998), chap. 4; Hashimoto Mitsuro, “Chiho: Yanagita Kunio’s ‘Japan,’” in Mirror of Modernity: Invented Traditions of Modern Japan, Vlastos, 133–143. Quoted in Bernardi, Writing in Light, 203. In France Louis Delluc popularized the slogan, “Let French cinema be French. Let French

NOTES TO PAGES 95–98

283

46.

47.

48.

284

cinema be cinema.” A writer in the German trade journal Licht-bild-bühne observed: “If our industry takes the position that it has to make German films, and not slavishly imitate foreign productions, then we must and will open the path onto the world market” (February 26, 1921, 9). The European positions are discussed in Kristin Thompson, “National or International Films? The European Debate during the 1920s,” Film History 8, no. 3 (1996): 281–296. Here is as good a place as any to consider an issue that occasionally surfaces in discussions of Mizoguchi’s style. “He never looked through the viewfinder during the whole time I worked with him,” reports his assistant director of the years 1934–1944 (Ariane Mnouchkine, “Six entretiens: Monsieur Takagi,” Cahiers du cinéma, no. 158 [August-September 1964]: 11). His camera operator Miku Shigeru confirms this, and Miyagawa Kazuo, the cinematographer on many of his late films, claims that Mizoguchi left the camera position to him (Shindo, Aru eiga kantoku no sho, 172–173; Miyagawa Kazuo, Watashi no eiga jinsei: Kyamera man ichidai [Sixty Years of My Film Life: Life of a Cameraman] [Kyoto: PHP, 1985], 80). Should we then be studying his staª rather than the great auteur? Not necessarily. Ford, Hitchcock, Hawks, and most other filmmakers of Mizoguchi’s generation seldom looked through the camera, yet all found ways to communicate their purposes to the crew. When shooting with Miyagawa, once the actors were on the set, Mizoguchi reblocked their movements so frequently that Miyagawa could never determine one position until just before the take. It seems that Mizoguchi narrowed his framing options in the course of his rehearsals. Also, it seems likely that Miyagawa learned the sort of shots Mizoguchi preferred. “Miyagawa chose the position of the camera because Mizoguchi had great confidence in him,” reports one actress (Max Tessier, “Une vraie jeune fille: Entretien avec Ayako Wakao,” Positif, no. 369 [November 1991]: 90). The result, however achieved, is clear: Mizoguchi films that were shot by various cinematographers have an undeniable singularity of style—and few a‹nities with the films that the cinematographers shot for other directors. For example, Miyagawa filmed Rashomon and Ozu’s Floating Weeds (1959), both of which diªer drastically from his work for Mizoguchi. Miyagawa claims that Ichikawa Kon gave him even more latitude than Mizoguchi did, yet the look of such films as Enjo (1958) and Kagi (Odd Obsession, 1959) is worlds away from Mizoguchi’s style. See Lyla Dusing, “Kazuo Miyagawa Honored by A.S.C.,” American Cinematographer 62, no. 5 (May 1981): 450, and Vê-Hô, Mizoguchi, 149–150. For a discussion of Taki’s importance see M. Cody Poulton, Spirits of Another Sort: The Plays of Izumi Kyoka (Ann Arbor, MI: Center for Japanese Studies, 2001), 31–37. See also Joseph Murphy, “Mizoguchi, Kinugasa, and the Melodrama of Izumi Kyoka: Visual Pleasure in the Japanese Cinema,” in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 69–71. The relation between Western melodrama and Japanese tradition is explored in Mitsuhiro Yoshimoto, “Melodrama, Postmodernism, and the Japanese Cinema,” East-West Film Journal 5, no. 1 (January 1991): 28–55; and Joseph Murphy, “Approaching Japanese Melodrama,” East-West Film Journal 7, no. 2 (July 1993): 2–36.

NOTES TO PAGES 98–99

49.

50.

51. 52. 53. 54. 55. 56. 57. 58. 59.

60. 61.

62.

63.

64.

65. 66.

See Fujiki Hideaki, “Mizoguchi ni katto saseta mono wa nanika? Naniwa ereji to Ugetsu monogatari ni miru kattingu to rongu teiku” [What Made Mizoguchi Cut? Cutting and the Long Take in Naniwa Elegy and Ugetsu Monogatari], F.B., no. 12 (winter 1998): 195–210. There is a belief that because dialogue titles broke up a long take, Mizoguchi reduced their number. (See Yoda, Souvenirs, 39.) Yet his surviving silent films contain at least as many dialogue titles as do the films of his peers, and in some cases more: 29 percent of the shots in Taki no Shiraito are dialogue titles, and a remarkable 35 percent in The Downfall of Osen are. I have found no Japanese films of the period with such a high proportion of intertitles. Mizoguchi, “Tavola rotonda,” 123–124. Quoted in McDonald, Mizoguchi, 39. Yoda, Souvenirs, 51–52; Peter Morris, ed., Mizoguchi Kenji (Ottawa: Canadian Film Archive, n.d.), 35. Quoted in Ryohei Shioda, “Jokyu (The Waitress), by Kazuo Hirotsu,” in Introduction to Contemporary Japanese Literature (Tokyo: Kokusai Bunka Shinkokai, 1939), 358. Ibid., xx. Burch, To the Distant Observer, 224. During the 1930s Mizoguchi tended to push his luck; in Naniwa and Sisters a close foreground figure is likely to be slightly out of focus. See Bazin, “Evolution,” 33–40; and Bazin, “Montage,” in Twenty Years of Cinema in Venice (Rome: Ateneo, 1952), 371–377. I oªer some examples in my essay “Mizoguchi and the Evolution of Film Language,” in Cinema and Language, ed. Stephen Heath and Patricia Mellencamp (Los Angeles: American Film Institute, 1983), 107–117; also available in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 21–23. See Brian Powell, Kabuki in Modern Japan: Mayama Seika and His Plays (London: Macmillan, 1990), 147–179. Darrell Davis considers the film as part of a cycle of “monumental jidai-geki” in chapter 6 of Picturing Japaneseness: Monumental Style, National Identity, Japanese Film (New York: Columbia University Press, 1996). For an in-depth analysis of the film, and of other Mizoguchi films from the 1930s, see Donald Kirihara, Patterns of Time: Mizoguchi in the 1930s (Madison: University of Wisconsin Press, 1992). On Zen’s appropriation by the military see Brian Victoria, Zen at War (New York: Weatherhill, 1997), pt. 2. The political use of traditional motifs is discussed by Emiko OhnukiTierney in Kamikaze, Cherry Blossoms, and Nationalisms: The Militarization of Aesthetics in Japanese History (Chicago: University of Chicago Press, 2002), 27–58. The most comprehensive and incisive survey of the Occupation years in English is John W. Dower, Embracing Defeat: Japan in the Wake of World War II (New York: New Press, 1999). See also Robert E. Ward and Sakamoto Yoshikazu, eds., Democratizing Japan: The Allied Occupation (Honolulu: University of Hawai’i Press, 1987). Kyoko Hirano, Mr. Smith Goes to Tokyo: Japanese Cinema under the American Occupation, 1945–1952 (Washington, DC: Smithsonian Institution, 1992), 154–165. See, e.g., Iwazaki, Kenji Mizoguchi, 466–468; McDonald, Mizoguchi, 71–82.

NOTES TO PAGES 101–124

285

67. 68. 69. 70. 71. 72. 73. 74.

75.

76.

77. 78. 79. 80. 81. 82.

83. 84.

Hazumi, “Self-Assessment,” 12. Quoted in McDonald, Mizoguchi, 28; see also Yoda, Souvenirs, 40. Quoted in Takizawa, “Before Filming,” 13. On the corner illusion see Donald Weismann, The Visual Arts as Human Experience (Englewood Cliªs, NJ: Prentice-Hall, 1974), 184–188. Harold Strauss, “My Aªair with Japanese Movies,” Harper’s 211, no. 8 (August 1955): 57. Iwazaki, Kenji Mizoguchi, 457–458. For a sampling of critical reactions to the late films see Andrew and Andrew, Kenji Mizoguchi, 185–216. Quoted in Miyagawa Kazuo, Film Center monograph no. 82 (Tokyo: Kokuritsu Kindai Bijutsu-kan Film Center, June 1984), 12. See also Joanne Bernardi, “Ugetsu Monogatari: The Screenplay,” in Mizoguchi the Master, ed. Gerald O’Grady (Toronto: Cinematheque Ontario, 1996), 29. Saso Tsutomu and Nishida Nobuyoshi, Eiga dokuhon Mizoguchi Kenji: Joen no hate no onna tachiyo, genmu eno riarisumu [Reading Kenji Mizoguchi: Women at the End of Passion, Realism for Phantasms] (Tokyo: Firumu Atosha, 1996), 35–40. Quoted in Takizawa, “Before Filming,” 13. Even in the later years, however, Mizoguchi remained cosmopolitan in his tastes, claiming that Women of the Night was influenced by Picasso’s postwar work (Morris, Mizoguchi Kenji, 12). Shindo, Aru eiga kantoku no sho, 63–64. Yoda, Souvenirs, 116. According to Yoda, after paying Wyler this compliment, he added, “That is all that I learned from him.” Burch, To the Distant Observer, 244. Ibid., 276. Ibid., 243. Robert Cohen points out several of Mizoguchi’s uses of point-of-view editing in “Textual Poetics in the Films of Kenji Mizoguchi: A Structural Semiotics of Japanese Narrative” (PhD diss., UCLA, 1983). “Le Festival de Cannes au jour le jour,” Positif, nos. 14/15 (November 1955): 78. Paul Rotha, “Films and Filming,” in Ugetsu, ed. Keiko McDonald (New Brunswick, NJ: Rutgers University Press, 1993), 127.

CHAPTER 4. ANGELOPOULOS, OR MELANCHOLY 1.

2. 3. 4. 5.

286

In his epic films, he says, “history, with a capital H,” takes center stage. See Andrew Horton, “‘What Do Our Souls Seek?’ An Interview with Theo Angelopoulos,” in The Last Modernist: The Films of Theo Angelopoulos, ed. Andrew Horton (Westport, CT: Praeger, 1997), 109. Quoted in ibid., 98. Ibid., 104. Andrea R. Vaucher, “Famed Greek Helmer Sets Trilogy,” Variety, September 16–22, 2002, 21. Quoted in Tony Mitchell, “Animating Dead Space and Time: Megalexandros,” Variety, September 16–22, 2002, 31.

NOTES TO PAGES 124–141

6.

7. 8.

9. 10. 11. 12. 13. 14. 15.

16.

17. 18.

19. 20. 21.

The information in the previous two paragraphs comes from Aglae Mitropoulos, Découverte du cinéma grec (Paris: Seghers, 1968), 42–122; Ninos Fenek Mikelides, “Brève histoire du cinéma grec (1906–1966),” in Le cinéma grec, ed. Michel Démopoulos (Paris: Centre Georges Pompidou, 1995), 43–63; Yannis Bakoyannopoulos, “Le Nouveau Cinéma Grec en marche,” ibid., 65–79; and Max T. Roman, “Introduction,” in Dimitris Koliodimos, The Greek Filmography, 1914 through 1996 (Jeªerson, NC: McFarland, 1999), 11–23. Jeune cinéma grec, 1951–1982 (Brussels: Cinémathèque Royale, 1982) usefully collects several reviews. Quoted in Helene Petrakis, “Une essai constant pour une inconnue: Entretien avec Yorgos Arvanitis,” Positif, no. 370 (December 1991): 14. Quoted in Michael Demopoulos and Frida Liappas, “A Journey through Greek Landscape and History: The Travelling Players,” in Theo Angelopoulos: Interviews, ed. Dan Fainaru (Jackson: University Press of Mississippi, 2001), 22. See Brizio Montinaro, “Notes sur le tournage d’Alexandre le grand,” Positif, no. 250 (January 1982): 5–9. Quoted in Horton, “‘What Do Our Souls Seek?’” 103. Andrew Horton, The Films of Theo Angelopoulos: A Cinema of Contemplation (Princeton, NJ: Princeton University Press, 1997). Jacques Gerstenkorn, “L’art des équilibres: Entretien avec Théo Angelopoulos sur Voyage à Cythère,” Études cinématographiques, nos. 142–145 (1985): 5–6. “Entretien,” in Michel Ciment and Hélène Tierchant, Théo Angelopoulos (Paris: Edilig, 1989), 145. Michel Ciment, “Entretien avec Theo Angelopoulos: ‘Raconter ce que l’on n’arrive pas à vivre,’” Positif, no. 453 (November 1998): 83. Theo Angelopoulos, “In My End Is My Beginning,” Positif, no. 463 (September 1999): 61. See also Noureddine Ghali, “T. Angelopoulos: ‘Mes films sont des appels à la discussion . . . ’” Cinéma 75, nos. 201–202 (September-October 1975): 197, 206, 208. Quoted in Michel Grodent, “‘Pomme fanée’: Autour du ‘Voyage à Cythère,’” Revue belge du cinéma, no. 11 (spring 1985): 57, 59. A cogent discussion of the two phases can be found in James Quandt, “Landscape in the Mist,” in Theo Angelopoulos, ed. Fabiano Canosa, George Kaloyeropoulos, and Gerald O’Grady (New York: Museum of Modern Art, 1990), 25. Interestingly, Bakoyannopoulos, “Le Nouveau Cinéma Grec en marche,” 75–76, finds this split to be present in the work of most New Cinema directors. See my Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985), 228–233. For helpful discussions of Angelopoulos’s style see Nikos Kolovos, “Le voyageur,” in Démopoulos, Le cinéma grec, 163–173; Michael Wilmington, “The Power and the Glory,” Film Comment 26, no. 6 (November-December 1990): 32–37; Raymond Durgnat, “The Long Take in Voyage to Cythera: Brecht and Marx vs. Bazin and God,” ibid., 43–46. “De Forminx Story à Voyage des comédiens,” in Ciment and Tierchant, Théo Angelopoulos, 55. Quoted in Gerstenkorn, “L’art des équilibres,” 12. Quoted in Dan Fainaru, “ . . . And about All the Rest,” in Fainaru, Theo Angelopoulos: Interviews, 130.

NOTES TO PAGES 142–146

287

22.

23. 24. 25. 26.

27.

28. 29. 30. 31.

32. 33. 34. 35. 36. 37. 38. 39. 40. 41.

288

On the social critique proªered by the film see Horton, Films of Theo Angelopoulos, 96–101. The mythic dimensions of one film are explored in depth in Sylvie Rollet, Voyage à Cythère: La poétique de la mémoire d’Angelopoulos (Paris: L’Harmattan, 2003). Quoted in Gerald O’Grady, “Angelopoulos’s Philosophy of Film,” in Theo Angelopoulos: Interviews, ed. Dan Fainaru (Jackson: University Press of Mississippi, 2001), 71. Jean-Luc Godard, “Pierrot My Friend,” Godard on Godard, ed. Jean Narboni, trans. Tom Milne (New York: Viking, 1972), 214. Caspar Tybjerg, “Honor Thy Craft: Questions of Style in Dreyer’s Master of the House,” Aura 6, no. 2 (2000): 67. Because of the influence of Bazin and the Cahiers critics we’re inclined to think of Roberto Rossellini as the prototypical long-take director, but until the 1960s his shots aren’t as consistently sustained as those of Visconti and Antonioni. For analyses of Antonioni’s staging strategies in these works see Kristin Thompson and David Bordwell, Film History: An Introduction (New York: McGraw-Hill, 1994), 427–429, 504–506; and David Bordwell, On the History of Film Style (Cambridge, MA: Harvard University Press, 1997), 233–237. André Bazin, “Bicycle Thief,” in What Is Cinema? vol. 2, ed. and trans. Hugh Gray (Berkeley: University of California Press, 1971), 55. Pierre Leprohon, Michelangelo Antonioni, trans. Scott Sullivan (New York: Simon and Schuster, 1963), 12. Ernest Callenbach, “About This Issue,” Film Quarterly 16, no. 1 (fall 1962): 1. Interestingly, in Il Grido and later films Antonioni uses less intricate camera movements and relies somewhat more on cutting. The average shot length of I Vinti (1953) is a remarkable forty-three seconds, and the shots in Le Amiche (1955) average twentysix seconds. Il Grido (1957) has an ASL of twenty-two seconds, about that of Cronaca di un amore (1950). During the 1960s the average starts to drop, with L’Avventura (1960) having a mean shot length of twenty seconds and La Notte (1961) about sixteen and a half seconds. For a thoughtful overview of Antonioni’s compositional strategies see José Moure, Michelangelo Antonioni: Cinéaste de l’évidement (Paris: L’Harmattan, 2001). Quoted in Penelope Houston, The Contemporary Cinema (Baltimore: Penguin, 1963), 32. Quoted in Jane Shapiro, “Stranger in Paradise,” in Jim Jarmusch: Interviews, ed. Ludvig Hertzberg (Jackson: University Press of Mississippi, 2001), 62. “De Forminx Story à Voyage des comédiens,” 51. Quoted in Fainaru, “ . . . And about All the Rest,” 130. Harold Bloom, The Anxiety of Influence: A Theory of Poetry (New York: Oxford University Press, 1973), 5. Andrey Tarkovsky, Sculpting in Time: Reflections on the Cinema, trans. Kitty Hunter-Blair (London: Faber and Faber, 1989), 25. Ibid., 117. Ibid., 69. A detailed discussion of the film can be found in Graham Petrie, Red Psalm (Wiltshire: Flicks, 1998). See also my discussion of The Confrontation (1969) in Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985), 130–146.

NOTES TO PAGES 147–157

42. 43.

44. 45. 46.

47. 48. 49.

50. 51.

52.

See Angelopoulos, “In My End,” 62. See Vicki Bruce, Tim Valentine, and Alan D. Baddeley, “The Basis of the 3/4 View Advantage in Face Recognition,” Applied Cognitive Psychology 1 (1987): 109–110; Robert H. Logie, Alan D. Baddeley, and Muriel M. Woodhead, “Face Recognition, Pose, and Ecological Validity,” Applied Cognitive Psychology 1 (1987): 53–69. Heinrich Wöl›in, Principles of Art History: The Problem of the Development of Style in Later Art, trans. M. D. Hottinger (New York: Dover, 1950), 75–82. Ibid. See, e.g., Susan Tarr and Hans Proppe, “The Travelling Players: A Modern Greek Masterpiece,” in Theo Angelopoulos, ed. Fabiano Canosa, George Kaloyeropoulos, and Gerald O’Grady (New York: Museum of Modern Art, 1990), 12–13; Isabella Jordan, “Pour un cinéma épique,” Positif, no. 174 (October 1975): 15–22. Richard Peña, quoted in Peter Brunette, “A Poetic Director Becomes a Cause to Champion,” New York Times, May 30, 1999, Arts and Leisure section, 20. For comments on Karaindrou’s scores for Angelopoulos see the interview with her in “Musique ou logique,” Positif, no. 370 (December 1991): 17–19. Kristin Thompson discusses Renoir’s tactics in “An Aesthetic of Discrepancy: The Rules of the Game,” in her Breaking the Glass Armor: Neoformalist Film Analysis (Princeton, NJ: Princeton University Press, 1988), 231–243. Quoted in Fainaru, “ . . . And about All the Rest,” 148. See Michel Ciment, “Entretien avec Theo Angelopoulos: Le regard perdu, emprisonné et qui cherche à être libéré,” Positif, no. 415 (September 1995): 27; and Ciment, “Entretien . . . : ‘Raconter ce que l’on n’arrive pas à vivre,’” 87. Angelopoulos, “In My End,” 61.

CHAPTER 5. HOU, OR CONSTRAINTS 1. 2.

3. 4.

5. 6. 7. 8.

Quoted in Yuki Sato, “Venice ’97,” in “Beat” Takeshi Kitao, ed. Brian Jacobs (Edgware, Middlesex: Tadao Press, 1999), 86. For an admirable discussion of Hou’s commitment to concreteness see Godfrey Cheshire, “Time Span: The Cinema of Hou Hsiao-Hsien,” Film Comment 29, no. 6 (NovemberDecember 1993): 56–63. Quoted in Olivier Assayas’s documentary film, HHH: A Portrait of Hou Hsiao-Hsien (Hsu Hsiao-Ming Film Corporation, 1997). Compare Eleni Karaindrou, Musique originale de la trilogie de Theo Angelopoulos (Milan CH299), with Hou Hsiao-Hsien Movie Family (Rock Records/TM Music TM0023-1 and TM0023-2). Quoted in Emmanuel Burdeau, “Rencontre avec Hou Hsiao-hsien,” in Hou Hsiao-hsien, ed. Jean-Michel Frodon (Paris: Cahiers du cinéma, 1999), 79. Quoted in ibid., 68. Quoted in Michel Ciment and Yann Tobin, “Entretien avec Hou Hsiao-Hsien: ‘Les contraintes et la liberté,’” Positif, no. 453 (November 1998): 9. For arguments against the invisible observer see my Narration in the Fiction Film (Madison: University of Wisconsin Press, 1985), 9–12.

NOTES TO PAGES 159–191

289

9. 10. 11. 12.

13.

14. 15. 16. 17. 18. 19. 20. 21. 22.

23.

24. 25. 26.

290

Quoted in Ciment and Tobin, “Entretien,” 9. Remark in panel discussion, Taipei Film Festival, November 1999. Stravinsky, The Poetics of Music in the Form of Six Lessons (New York: Vintage, 1947), 68. See James Udden, “Taiwanese Popular Cinema and the Strange Apprenticeship of Hou Hsiao-hsien,” in Modern Chinese Literature and Culture 15, no. 1 (spring 2003): 120–145. This method was reminiscent of Hong Kong directors’ segment-shooting approach, which also did not yield a master shot. For more on segment shooting see my Planet Hong Kong: Popular Cinema and the Art of Entertainment (Cambridge, MA: Harvard University Press, 2000), 129–131, 165–167. By my count Cute Girl has an average shot length of 10.4 seconds; Cheerful Wind averages 12.1 seconds; and Green, Green Grass of Home averages 10.4 seconds. Most of the 1970s films I have been able to see have average shot lengths between 5 and 7 seconds, although there are exceptions: Pai Ching-jui’s Love in a Cabin (1974) has an ASL of 9.0 seconds, and that of Song Cun-shou’s Diary of Didi (1978) is 10.8 seconds. See James Udden, “Hou Hsiao-hsien and the Aesthetics of Historical Experience” (PhD diss., University of Wisconsin–Madison, 2003), 98–99, for more information on typical shot lengths of the period. Quoted in Burdeau, “Rencontre,” 66–68. In Assayas’s documentary, HHH. Quoted in Burdeau, “Rencontre,” 81. Quoted in Tony Rayns, “The Sandwich Man,” Monthly Film Bulletin, no. 653 (June 1988): 164. Michel Egger, “Rencontre avec Hou Hsiao-hsien,” Positif, no. 334 (December 1988): 37. Quoted in Peggy Chiao Hsiung-ping, “History’s Subtle Shadows: Hou Hsiao-hsien’s The Puppetmaster,” Cinemaya, no. 21 (autumn 1993): 9. See Udden, “Hou Hsiao-hsien,” 113. Quoted in Burdeau, “Rencontre,” 78. For concise overviews of the movement see Tony Rayns, “Countdown: Steps towards Taiwan’s ‘New Wave,’” in Taiwanese Cinema, 1982–2002: From New Wave to Independent (Pusan: Pusan International Film Festival, 2002), 20–29; and Ti Wei, “Reassessing the Historical Significance of Taiwanese New Cinema in the Context of Globalization,” in Twentieth Anniversary of Taiwanese New Cinema (Taipei: Golden Horse Film Festival, 2002), 14–20. See also Yeh Yueh-yu, “Poetics and Politics of Hou Hsiao-hsien’s Films,” Post-Script 20, nos. 2–3 (winter/spring and summer 2001): 61–76. For a history see Sung-sheng Yvonne Chang, Modernism and the Nativist Resistance: Contemporary Chinese Fiction from Taiwan (Durham, NC: Duke University Press, 1993). Stephanie Hoare provides a detailed account of the New Cinema’s treatment of these trends in “Innovation through Adaptation: The Use of Literature in New Taiwan Film and Its Consequences,” Modern Chinese Literature, nos. 1–2 (spring-fall 1993): 33–55. Hoare, “Innovation through Adaptation,” 39–40. Ru-shou Robert Chen, “Taiwan Cinema,” in Encyclopedia of Chinese Film, ed. Yingjin Zhang and Zhiwei Xiao (New York: Routledge, 1998), 58. Yee Chin-hyen, interview by author, Taipei Film Festival, November 1999.

NOTES TO PAGES 191–203

27. 28. 29. 30.

31. 32.

33. 34. 35. 36. 37. 38.

39. 40. 41.

42. 43.

44. 45. 46. 47.

Quoted in Hoare, “Innovation through Adaptation,” 41. See Chen Kuo-fu’s 1987 remarks quoted in Michel Egger, “Cinema Made in Taiwan: Une nouvelle vague en quête d’identité,” Positif, no. 311 (January 1987): 31. Quoted in Assayas’s documentary HHH. The average shot lengths of the films are as follows: Summer at Grandpa’s, 12.2 seconds; A Time to Live, a Time to Die, 23.7 seconds; Dust in the Wind, 32 seconds; Daughter of the Nile, 27.6 seconds. There are at least three versions of The Boys from Fengkuei: an international version (ASL, 18.6 seconds); a French version (19.1 seconds); and a third version of uncertain source (17.7 seconds). Quoted in Evans Chan, “Homegrown in Taiwan,” undated pressbook [1988?]: 1. Quoted in Ta-yi Lee, “Hou Hsiao-hsien on [sic] Interview,” in Taiwan Film Festival 1999: A Tribute to Hou Hsiao-Hsien, ed. Ruby Lau (Hong Kong: Hong Kong Arts Centre, 1999), 19. Olivier Assayas, “Notre reporter en République de Chine,” Cahiers du cinéma, no. 366 (December 1984): 59. Quoted in Rayns, “The Sandwich Man,” 164. Quoted in Burdeau, “Rencontre,” 74. Remarks in Assayas’s documentary HHH. Quoted in Law Kar, “Director’s Note [Daughter of the Nile],” in The 12th Hong Kong International Film Festival (Hong Kong: Urban Council, 1988), 92. Chang Shih-lun provides an account of this process in “Taiwanese New Cinema and the International Film Festival Approach,” Twentieth Anniversary of Taiwanese New Cinema, 29–39. Marco Muller, ed., Taiwan: Nuove ombre elettriche (Venice: Marsilio, 1988). Lee Tain-dou and Chen Pei-chih, “A Sociological Investigation of Taiwanese New Cinema in the Eighties” (unpublished manuscript), 34–36. For summaries of the controversies see Yeh Yueh-yu, “The Poetics of Hou Hsiao-Hsien’s Films: Flowers of Shanghai,” Cinedossier: The 35th Golden Horse Award-Winning Films (Taipei: Golden Horse Film Festival, 1999), 94; see also Udden, “Hou Hsiao-hsien,” 181–197. Quoted in Michel Ciment, “Entretien avec Hou Hsiao-hsien,” Positif, no. 358 (December 1990): 8. For a more detailed account of the film’s plot and a sensitive critical essay, accompanied by a rich interview with Hou, see Tony Rayns, “Beiqing Chengshi (A City of Sadness),” Monthly Film Bulletin, no. 677 (June 1990): 152–155. See also Bérénice Reynaud’s insightful monograph, A City of Sadness (London: British Film Institute, 2002). A helpful Web site devoted to the film has been created by Yeh Yueh-yu and Abé Marcus Nornes: http:// cinemaspace.berkeley.edu/Papers/CityOfSadness/table.html (accessed November 8, 2003). For a review of some of these conventions see Bordwell, Narration in the Fiction Film, 205–233. Peggy Chiao, Hou Hsiao-hsien (Taipei: Variety Publishing, 1993), 61–62. Good Men, Good Women pressbook for Cannes Film Festival 1995, n.p. Quoted in Chiao, Hou, 63.

NOTES TO PAGES 203–217

291

48. 49. 50. 51. 52.

53. 54.

55.

56.

57.

58.

59. 60. 61. 62. 63.

292

See, e.g., the essays in Frodon, ed., Hou Hsiao-hsien; and comments by Assayas in his documentary HHH. Quoted in Chaio, Hou, 48. Quoted in Ciment, “Entretien,” 10. For a brief summary see the City of Sadness Web site cited above. “‘Returning,’ ‘Motherland,’ ‘228’: Taiwanese History and National Attributes in City of Sadness,” in Passionate Detachment: The Films of Hou Hsiao-hsien, ed. Lin Wen-chi, Shen Hsiao-ying, and Jerome Chen-ya Li (Taipei: Rye Field, 2000), 157–179; in Chinese. See Yann Tobin, “Fenêtre sur Chine (Cité des douleurs),” Positif, no. 358 (December 1990): 9. For more comparisons with Mizoguchi’s use of depth see Shen Hsiao-ying’s essay “We Should Watch These a Couple More Times to Begin With: Film Aesthetics and Hou Hsiao-hsien,” in Passionate Detachment: The Films of Hou Hsiao-hsien, ed. Lin Wen-chi, Shen Hsiao-ying, and Jerome Chen-ya Li (Taipei: Rye Field, 2000), 61–92; in Chinese. Consider these representative ASLs: Rebels of the Neon God (Tsai Ming-liang, 1992): 18.9 seconds; Dust of Angels (Hsu Hsiao-ming, 1992): 46.3 seconds; Vive l’amour (Tsai, 1994): 35.8 seconds; Lonely Hearts Club (Yee Chin-hyen, 1994): 37.9 seconds; Treasure Island (Chen Kuo-fu, 1994): 15.6 seconds; A Borrowed Life (Wu Nian-jen, 1994): 51.4 seconds; Heartbreak Island (Hsu Hsiao-ming, 1995): 52.4 seconds; The River (Tsai Ming-liang, 1997): 42.4 seconds; Sweet Degeneration (Lin Cheng-seng, 1997): 40.2 seconds; Murmur of Youth (Lin Cheng-seng, 1997): 21.6 seconds; Darkness and Light (Chang Tso-chi, 1998): 26 seconds; The Hole (Tsai Ming-liang, 1998): 52.8 seconds; What Time Is It There? (Tsai, 2001): 64.0 seconds. See, e.g., Hong’s The Day the Pig Fell in the Well (1996): ASL 25 seconds; The Power of Kangwon Province (1998): 40.1 seconds; Oh, Soo Yeong! (2000): 52.6 seconds; Lee Kwangmo’s Spring in My Home Town (1998): 45.3 seconds; Jang Song-woo’s Lies (1999): 36 seconds; and Hur Jin-ho’s One Fine Spring Day (2001): 37.1 seconds. See, e.g., Kore-eda Hirokazu’s Maborosi (1996): ASL 24 seconds; Hashiguchi Ryosuke’s Like Grains of Sand (1996): 29 seconds; Kurosawa Kiyoshi’s Barren Illusion (1999): 16.5 seconds; Haginda Yoji’s Paradise Sea (1999): 19.5 seconds; Suwa Nobuhiro’s M/Other (1999): 103.1 seconds; Kurosawa Kyoshi’s Charisma (1999): 21.6 seconds; and Aoyama Shinji’s Eureka (2000): 24.8 seconds. Interestingly, Wang Chao talks of preferring to let random incidents intrude into his shots (as when a passing bus blocks a key action in The Orphan of Anyang, 2002) but sees this as quite inimical to Hou’s style. See Matthieu Darras, “Entretien Wang Chao: Ouvrir le film à la vie,” Positif, no. 493 (March 2002): 30. Burdeau, “Rencontre,” 88. Kitano Takeshi, interview by author, Kyoto Film Festival, November 1997. Derek Elley, “Millennium Mambo,” Variety, May 28–June 3, 2001, 19. See Lee Erickson, “Preparing to Live in the Present: An Interview with Hou Hsiao-hsien,” Cineaste 27, no. 4 (fall 2002): 18. Nelson H. Wu, “Helmer Aims to Acquire Arthouse Pics,” Variety, August 12–18, 2002, 8; “Market Pressures: Untitled Hou Hsiao Hsien project (Taiwan-Japan),” Screen International, October 31–November 6, 2003, 16.

NOTES TO PAGES 217–234

64. 65.

66. 67.

Quoted in Chan, “Homegrown,” 2. Hou employed lenses of 85 mm and 135 mm for this film. See Michel Ciment and Hubert Niogret, “Entretien Hou Hsiao-hsien: Un travail à la loupe,” Positif, no. 489 (November 2001): 32. I’m indebted to Peter Rist for this idea. Quoted in Burdeau, “Rencontre,” 98.

CHAPTER 6. STAGING AND STYLISTICS 1. 2.

3. 4. 5.

6.

7. 8.

Barry Salt, Film Style and Technology: History and Analysis, 2nd ed. (London: Starword, 1992), 283. E. H. Gombrich, “In Search of Cultural History,” in his Ideals and Idols: Essays on Values in History and in Art (London: Phaidon, 1979), 24–59. See also his essay “‘The Father of Art History’: A Reading of the Lectures on Aesthetics of G. W. F. Hegel,” in E. H. Gombrich, Tributes: Interpreters of Our Cultural Tradition (London: Phaidon, 1984), 51–70. Tony Rayns, “Beiqing Chengshi (A City of Sadness),” Monthly Film Bulletin, no. 672 (June 1980): 154. I argue for treating subject and theme as materials in Ozu and the Poetics of Cinema (Princeton, NJ: Princeton University Press, 1988), 31–33. For a more detailed critique of hasty assimilations of Hou to Chinese artistic traditions see James Udden, “Hou Hsiao-hsien and the Question of a Chinese Style,” Asian Cinema 13, no. 2 (fall/winter 2002): 54–75. On these deformations see T. C. Lai, Understanding Chinese Painting (Hong Kong: Kelly and Walsh, 1983), 91–112; François Cheng, Empty and Full: The Language of Chinese Painting, trans. Michael H. Kohn (Boston: Shambhala, 1991), 93–96; Jerome Silbergeld, Chinese Painting Style: Media, Methods, and Principles of Form (Madison: University of Wisconsin Press, 1982), 37–38. Walter Benjamin, “The Work of Art in the Age of Mechanical Reproduction,” in Illuminations, ed. Hannah Arendt (New York: Schocken, 1968), 222. Benjamin’s claim has occasionally been interpreted with startling literalness. Jonathan Crary, for example, proposes that attentional shifts in our vision are of recent origin. Crary starts by describing attention as a research topic for nineteenth-century psychology, a problem “elaborated within an emergent economic system that demanded attentiveness of a subject in a wide range of new productive and spectacular tasks.” But he goes on to say: “Part of the cultural logic of capitalism demands that we accept as natural switching our attention rapidly from one thing to another.” You might think that he is suggesting that the idea of rapid attentional shifts was attractive to capitalist ideology, but the next sentence makes it plain that the act of shifting attention is created by capitalism itself: “Capital, as accelerated exchange and circulation, necessarily produced this kind of human perceptual adaptability and became a regime of reciprocal attentiveness and distraction” (Suspensions of Perception: Attention, Spectacle, and Modern Culture [Cambridge, MA: MIT Press, 1999], 29–30). Yet Crary makes this remarkable statement without backup. He never confronts the obvious objection that rapid shifts of attention, as I indicated in discussing saccadic eye movements in chapter 1, are of great evolutionary advantage and indeed form part of our mam-

NOTES TO PAGES 234–245

293

malian heritage. The child-development literature abounds in accounts of this biologically inherited phenomenon; see, e.g., Simon Baron-Cohen, Mindblindness: An Essay on Autism and Theory of Mind (Cambridge, MA: MIT Press, 1995), 97–120. Too often cultural commentators assume that in positing evolutionary sources of human capacities we surrender to “ahistorical” thinking. Not at all; it’s just that evolution takes place over natural history, which stretches over rather long periods. Typical of the confusion is Crary’s remark that [t]here is also a persistent, and most often unexamined, Kantian prejudice that perceptual and cognitive capacities are ahistorical; that is, they are unchanging and permanent, and most significantly are independent of an external social/technological milieu that is in constant flux. . . . For the last hundred years perceptual modalities have been, and continue to be, in a state of perpetual transformation or, some might claim, of crisis. If vision can be said to have any enduring characteristic within twentiethcentury modernity, it is that it has no enduring features. (“Dr. Mabuse and Mr. Edison,” in Art and Film Since 1945: Hall of Mirrors, ed. Russell Ferguson [Los Angeles: Museum of Contemporary Art, 1996], 263–264; emphasis mine)

9.

10. 11. 12. 13. 14.

15.

294

Crary never explains what perceptual modalities are, but they clearly do not include scanning the environment for information, or shifting visual focus by accommodation and convergence, or reliably detecting movement, color, and shape—for these are among the enduring characteristics of vision within modernity and for many thousands of years before. A perpetual transformation of vision, let alone a crisis of it, would leave us even more forlorn creatures than we already are. Tom Gunning, “Early American Film,” in The Oxford Guide to Film Studies (Oxford: Oxford University Press, 1998), 268. Actually, I never said that the modernity thesis couldn’t explain stylistic change, just that at present the prospects were unencouraging. See Bordwell, On the History of Film Style (Cambridge, MA: Harvard University Press, 1997), 146. Tom Gunning, D. W. Gri‹th and the Origins of American Narrative Film (Urbana: University of Illinois Press, 1991), 105. Tom Gunning, “The Whole Town’s Gawking: Early Cinema and the Visual Experience of Modernity,” Yale Journal of Criticism 7, no. 2 (1994): 194. Ben Singer, Melodrama and Modernity: Early Sensational Cinema and Its Contexts (New York: Columbia University Press, 2001), 127–128. Ibid., 128. Or even an air of invulnerability. Anything in a film that cannot be described as an aspect of modernity will have to be explained by other means, and the temptation will be to claim that the ill-fitting properties result from filmmaking that is out of tune with the modern era. This will leave us with a heads-I-win-tails-you-lose argument, quite invulnerable to counterexample. For example, Fredric Jameson, the most influential theorist of postmodernity, treats Angelopoulos’s films in a reflectionist fashion characteristic of epochal approaches. Alexander the Great “belatedly coincides with a worldwide change in cultural temperature with the emergent Reagan era in 1980.” (A “worldwide change in cultural tem-

NOTES TO PAGES 245–247

16.

17.

18. 19.

20. 21.

22.

23.

perature” is a pretty broad explanatory construct, and Jameson neither defines it nor explains how such a state of aªairs could shape or “belatedly coincide” with a particular film.) Likewise, the symptoms of postmodern cultural products delineated elsewhere by Jameson—pastiche, flatness, nostalgia, and the like—will not by themselves lead us toward Angelopoulos’s staging principles, for such symptoms are too general and unrefined. Consequently, when seeking to analyze Angelopoulos’s style, Jameson oªers only loose and sometimes contradictory intuitions: “In this work a remarkable living realism is attained in which, at least for one more long moment, the traditional opposition between the fictive and the real or documentary seems to have been suspended or neutralized. . . . Angelopoulos banishes all the conventional habits and taboos of normal filmic and narrative time.” (All of the habits? Including chronology, flashback reordering, ellipsis, and the like?) But then Jameson says a few pages later, “In these great early films of Angelopoulos (and later on, in Ulysses’ Gaze), it is chronology that is called upon to motivate the logic of transition.” In any narrative film this would not be unusual. See Fredric Jameson, “Theo Angelopoulos: The Past as History, the Future as Form,” in The Last Modernist: The Films of Theo Angelopoulos, ed. Andrew Horton (New York: Praeger, 1997), 82, 87. A summary of how these processes work can be found in Richard Nisbett and Lee Ross, Human Inference: Strategies and Shortcomings of Social Judgment (Englewood Cliªs, NJ: Prentice-Hall, 1980); and Gerd Gigerenzer, Peter M. Todd, and the ABC Research Group, Simple Heuristics That Make Us Smart (New York: Oxford, 1999). I try to make the case for film criticism as a process of informal reasoning in Making Meaning: Inference and Rhetoric in the Interpretation of Cinema (Cambridge, MA: Harvard University Press, 1989), chaps. 5–8. Charlie Keil, “‘To Here from Modernity’: Style, Historiography, and Transitional Cinema,” in American Cinema’s Transitional Era: Audiences, Institutions, Practices, ed. Charlie Keil and Shelley Stamp (Berkeley: University of California Press, 2004), 55. See also Malcolm Turvey, “How Modernity Shaped Early Film Melodrama,” Framework 44, no. 2 (fall 2003): 173–78. Keil, “‘To Here,’” 59. On film studies and the dialectical play of rational inquiry see Noël Carroll, “Prospects for Film Theory: A Personal Assessment,” in Post-Theory: Reconstructing Film Studies, ed. David Bordwell and Noël Carroll (Madison: University of Wisconsin Press, 1996), 56–61. Gunning, “The Whole Town’s Gawking,” 196. See Thomas Puttfarken, The Discovery of Pictorial Composition: Theories of Visual Order in Painting, 1400–1800 (New Haven, CT: Yale University Press, 2000). A detailed study of aperture framing in Dutch art is Martha Hollander, An Entrance for the Eyes: Space and Meaning in Seventeenth-Century Dutch Art (Berkeley: University of California Press, 2002). For an example of how this could happen in another art form see Kristin Thompson, “Frontal Shoulders in Amarna Royal Reliefs: Solutions to an Aesthetic Problem,” Journal of the Society for the Study of Egyptian Antiquities 27 (1997): 79–102, plates iv–vii. John Belton, “On the History of Film Style,” Film Quarterly 52, no. 4 (summer 1999): 57.

NOTES TO PAGES 247–250

295

24.

25. 26.

27.

28.

29. 30. 31. 32.

33. 34.

35. 36.

37. 38.

296

Incidentally, “cognitivism” is not one of those presuppositions. As Vasari’s Lives of the Painters shows, you need not be a cognitivist to find the problem/solution model fruitful. James Hurley, “David Bordwell’s Iron Cage of Style,” Film-Philosophy, http://www .filmphilosophy.com/vol2–1998/n26hurley. “Mankind always takes up only such problems as it can solve, since, looking at the matter more closely, we will always find that the problem itself arises only when the material conditions necessary for its solution already exist or are at least in the process of formation” (preface to A Contribution to the Critique of Political Economy, in Karl Marx and Frederick Engels, Basic Writings in Poetics and Philosophy, ed. Lewis S. Feuer [Garden City, NY: Anchor, 1959], 44). Noël Carroll, “Film History and Film Theory: An Outline for an Institutional Theory of Film,” in Theorizing the Moving Image (Cambridge, UK: Cambridge University Press, 1996), 275–291. See V. F. Perkins, Film as Film: Understanding and Judging Movies (Harmondsworth: Penguin, 1972), chap. 8, for an argument for the director as the principal force in shaping the look and sound of the final film. A sampling of the footage is provided in Alain Fleischer’s film Un tournage à la campagne (Cinémathèque Française/Films du Panthéon, 1994). Quoted in Eric Sherman, Directing the Film: Film Directors on Their Art (Boston: Little, Brown, 1976), 99. William McDonald, “Brothers Live in a Movie World of Their Own,” New York Times, March 3, 1996, sec. 2, p. 24. Jon Elster makes an intriguing attempt, arguing that as rational agents artists seek to maximize creativity. See Ulysses Unbound: Studies in Rationality, Precommitment, and Constraints (Cambridge, UK: Cambridge University Press, 2000), 200–212. Ibid. Someone might suggest that most people know the look of rhinos or reindeer only through pictures, not through firsthand acquaintance, so we construe the cave imagery according to some canon of representation, not as a copy of reality. But the Ice Age painters did not have access to our pictures, and they did know what these creatures looked like. If our pictures resemble theirs, the most economical explanation is that their pictures captured relevant features of the look of objects, features that our representational systems also try to capture. For a review see Margaret A. Hagen, Varieties of Realism: Geometries of Representational Art (Cambridge, UK: Cambridge University Press, 1986). See Paul Messaris, Visual “Literacy”: Image, Mind, and Reality (Boulder, CO: Westview, 1994), chap. 2. John Baines suggests that “object-centered” representational systems, rather than perspectival projection of an entire space, constitute a cultural universal of pictorial art. See “Theories and Universals of Representation: Heinrich Schäfer and Egyptian Art,” Art History 8, no. 1 (March 1985): 18. See Noël Carroll, “The Power of Movies,” in Theorizing the Moving Image (Cambridge, UK: Cambridge University Press, 1996), 78–93. See Messaris, Visual “Literacy,” 168–180.

NOTES TO PAGES 251–258

39.

40. 41.

42.

43.

44.

45. 46. 47. 48. 49. 50. 51. 52. 53.

See David Bordwell, “Convention, Construction, and Cinematic Vision,” in Post-Theory: Reconstructing Film Studies, ed. David Bordwell and Noël Carroll (Madison: University of Wisconsin Press, 1996), 87–107. See Carroll, “The Power of Movies,” 84–87. See Bordwell, “Convention, Construction,” 91–93. See also Leonard B. Meyer, “A Universe of Universals,” in The Spheres of Music: A Gathering of Essays (Chicago: University of Chicago Press, 2000), 281–303. Christian Metz, “The Cinema: Language or Language System?” in Film Language: A Semiotics of the Cinema, trans. Michael Taylor (New York: Oxford University Press, 1974), 41. In his later work Metz argued for a more code-based conception of the image. E. H. Gombrich, “Image and Code: Scope and Limits of Conventionalism in Pictorial Representation,” in his The Image and the Eye: Further Studies in the Psychology of Pictorial Representation (Ithaca, NY: Cornell University Press, 1982), 289. Someone might suggest that these schemas were imposed on the world because of the greater circulation of American films after World War I. But overseas audiences could not have accepted these schemas, and their filmmakers could not have absorbed them, if the schemas somehow lay irrevocably outside their comprehension. In other words, for this style to be “imposed,” it would have to mesh with audiences’ and filmmakers’ perceptual proclivities to a considerable degree. If it didn’t, it would register as mere noise. A media-imperialism argument presupposes some cross-cultural convergence. (In this respect it resembles the diversity-of-interpretations argument I discussed in chapter 1, note 61.) In point of fact, we have no reason to think that the continuity style was imposed. For one thing, aspects of it were developed in several European countries, as well as in the United States. Given the speed with which directors in many countries assimilated the rudiments of the style, it seems evident that they saw advantages in it. In other words, as rational deciders, many considered that its solutions to storytelling problems were worth pursuing. See Bordwell, On the History of Film Style, chaps. 5 and 6; and Bordwell, “Convention, Construction,” 87–107. Slavoj Yizek, The Fright of Real Tears: Krzystof Kieslowski between Theory and Post-Theory (London: British Film Institute, 2001), 17. Bordwell, “Convention, Construction,” 104. Yizek, Fright, 16. Royall Tyler supplies dozens of such gap-filling explanations in his superb edition of Genji (New York: Penguin, 2001). Patrick Colm Hogan, The Mind and Its Stories: Narrative Universals and Human Emotion (Cambridge, UK: Cambridge University Press, 2003). Yizek, Fright, 16–17. A version of the tale can be found in Ruth Finnegan, Oral Literature in Africa (Nairobi: Oxford University Press, 1970), 358–360. See, e.g., Warren Buckland, “A Close Encounter with Raiders of the Lost Ark: Notes on Narrative Aspects of the New Hollywood Blockbuster,” in Contemporary Hollywood Cinema, ed. Steve Neale and Murray Smith (London: Routledge, 1998), 166–177; Kristin

NOTES TO PAGES 259–262

297

54. 55. 56. 57.

58.

59.

Thompson, Storytelling in the New Hollywood: Understanding Classical Narrative Technique (Cambridge, MA: Harvard University Press, 1999), 8–10. Yizek, Fright, 17. For examples see Baudouin van de Walle, La Chapelle funéraire de Neferirtenef (Brussels: Musées Royaux d’art et d’histoire, 1978), esp. plates 9 and 12. Yizek, Fright, 17. Scott Atran, Cognitive Foundations of Natural History: Towards an Anthropology of Science (Cambridge, UK: Cambridge University Press, 1990). For a summary see Atran’s essay, “Modes of Thinking about Living Kinds: Science, Symbolism, and Common Sense,” in Modes of Thought: Explorations in Culture and Cognition, ed. David R. Olson and Nancy Torrance (Cambridge, UK: Cambridge University Press, 1996), 216–260. Atran’s In Gods We Trust: The Evolutionary Landscape of Religion (Oxford: Oxford University Press, 2002) confirms that belief in supernatural beings is indeed a cross-cultural universal (see 51–79). In his Human Universals (Philadelphia: Temple University Press, 1991), the anthropologist Donald E. Brown lists dozens of practices, conceptual categories, linguistic patterns, and other features that are confirmed to be or are likely to be universals of human culture. See particularly chapter 6. Yizek may seem to be disagreeing with me on stylistic matters, but that is due only to his rather imprecise way of putting things. Of depth of field he says: It is misleading to conceive of these concrete figurations of depth of field as subspecies of the universal genus: the totality which accounts for their specific meaning is the “mediated” totality of each historical epoch of cinematic style, the way depth of field is located in the articulated whole of stylistic procedures. (Fright, 18)

An “articulated whole of stylistic procedures” sounds something like norms, but the mention of an epoch’s “mediated totality” suggests that the network of forces is far wider than mere film forms. Something truly wide-ranging and complex is promised here. But the trumpet blast is followed directly by: Su‹ce it to recall how a simple fact of soundtrack can totally change the situation of the visual depth of field, allowing the director to focus the attention of the spectator by vocal information; or how, in the famous shot from Wyler’s The Little Foxes (1941), a minus dialectically reverts into a plus, i.e. the very blurred, out-of-focus background, instead of signalling the relative unimportance of what goes on there, is the place where, tantalizingly inaccessible to the spectator’s clear view, the crucial event—the fatal heart attack—takes place. (Fright, 18)

The “totality” turns out to be merely the immediate context of the shot and scene, in which depth of field is coordinated with other film techniques, like sound. Of course the specific meaning of the technique arises from the specific context in which it appears, but that does not throw doubt on the category “deep-focus composition.” And weirdly, Yizek’s points are ones I’ve made in the works he is criticizing. On the History of Film Style is not about a “universal genus” but about how depth staging assumed diªerent forms at various points in history; it explains how dialogue channels attention in 1930s and 1940s deep-space staging; and it discusses André Bazin’s argument about our frustration in trying to see the out-of-focus staircase in The Little Foxes (67, 221). “Su‹ce

298

NOTES TO PAGES 262–264

60.

61. 62.

it to recall” is a nice touch: who is reminding whom? Note, incidentally, that now Yizek unhesitatingly explains a device with reference to the way it directs the spectator’s attention. One would expect that a cultural critic who thinks that “the very notion” of X means diªerent things in diªerent cultures would not speak of a film technique’s reappearing in various periods or places or even from film to film. Yet Yizek again cashes in the radical-sounding idea in the most uncontroversial way, asserting of deep-focus shooting in the 1910s and the 1940s that “this (abstractly) ‘same’ procedure is not simply ‘the same,’ since it is each time ‘trans-functionalized,’ included in a diªerent, historically specific totality” (Fright, 18). Now that we know that this totality is just the film’s context, I can immediately agree. The type is not the token, and each instantiation of the technique gains its most concrete meaning from its local context. But deep-focus compositions can be considered as instantiating the same technique in a meaningful sense if your purpose is to trace stylistic regularities at a particular level of generality through history. You might also expect a skeptic about universals to deconstruct “the very notion” of cross-cutting. Yizek instead concludes that “Cross-cutting functions as a ‘floating signifier’: although generated in this [Gri‹th’s] concrete fantasmatic scenario it cut itself oª and was reappropriated for a series of other paradigms which are in no way grounded in the last-minute rescue scenario” (21). It is not news to be told that even after Gri‹th used cross-cutting for last-minute rescues, other directors could still use it for diªerent purposes. I am reminded of John Ellis’s observation that a contemporary theorist often launches a daring, strategically obscure generalization but, when pressed, restates the point as a truism (John M. Ellis, Against Deconstruction [Princeton, NJ: Princeton University Press, 1989], 106–108). And once more, On the History of Film Style, like the present book, traces how film techniques were revised, modified, and reshaped in order to fulfill various functions. Yet Yizek invokes this idea as if it were a corrective to my view, whipping out one of the most banal tropes of film history (Gri‹th = crosscutting) as his trump card. I must conclude that he agrees with me even when he says he doesn’t. See, e.g., David Epstein, “Tempo Relations in Music: A Universal?” in Beauty and the Brain: Biological Aspects of Aesthetics, ed. Ingo Rentschler, Barbara Herzberger, and David Epstein (Basel: Birkhäuser, 1988), 91–116; Brian Boyd, “Jane, Meet Charles: Literature, Evolution, and Human Nature,” Philosophy and Literature 22, no. 1 (1998): 1–30; Brett Cooke and Frederick Turner, eds., Biopoetics: Evolutionary Explorations in the Arts (Lexington, KY: ICUS, 1999); Ellen Dissayanake, Art and Intimacy: How the Arts Began (Seattle: University of Washington Press, 2000). For moving-image media the standard work is Messaris, Visual “Literacy.” Adam Kuper, Culture: The Anthropologists’ Account (Cambridge, MA: Harvard University Press, 1999), chap. 7. See John Tooby and Leda Cosmides, “The Psychological Foundations of Culture,” in The Adapted Mind: Evolutionary Psychology and the Generation of Culture, ed. Jerome H. Barkow, Leda Cosmides, and John Tooby (New York: Oxford University Press, 1992), 19–136; Robin Dunbar, Chris Knight, and Camilla Power, eds., The Evolution of Culture: An Interdisciplinary View (New Brunswick, NJ: Rutgers University Press, 1999); Henry

NOTES TO PAGE 265

299

63. 64. 65. 66.

67.

68. 69.

70. 71. 72.

300

Plotkin, The Imagined World Made Real: Towards a Natural Science of Culture (London: Allen Lane, 2002); Steven Pinker, The Blank Slate: The Modern Denial of Human Nature (New York: Viking, 2002); Matt Ridley, Nature via Nurture: Genes, Experience, and What Makes Us Human (New York: HarperCollins, 2003). See Bordwell, Making Meaning, chap. 4. Brian Boyd, Vladimir Nabokov: The American Years (Princeton, NJ: Princeton University Press, 1991), 179. Lutz Bacher, Max Ophuls in the Hollywood Studios (New Brunswick, NJ: Rutgers University Press, 1996). Ben Brewster and Lea Jacobs, Theatre to Cinema: Stage Pictorialism and the Early Feature Film (Oxford: Oxford University Press, 1997); Charlie Keil, Early American Cinema in Transition: Story, Style, and Filmmaking, 1907–1913 (Madison: University of Wisconsin Press, 2001); Don Fairservice, Film Editing: History, Theory, and Practice (Manchester: Manchester University Press, 2001). Keil explicitly utilizes the problem/solution model in chapters 2–3. See, e.g., Daniel Serceau, Kenji Mizoguchi: Un art de la condensation (Berne: Peter Lang, 1991); Jacques Aumont, “Vanités (Migrations, 2),” Cinémathèque, no. 16 (autumn 1999): 6–21; Jacques Aumont, ed., La Mise en scène (Brussels: De Boeck, 2000); Frank Curot, Du style au cinéma, Études cinématographiques, no. 65 (Paris: Minard, 2000). George Basalla, The Evolution of Technology (Cambridge, UK: Cambridge University Press, 1988), 44. I have tried to contribute to this enterprise in “Intensified Continuity: Visual Style in Contemporary American Film,” Film Quarterly 55, no. 3 (spring 2002): 16–28; and in Planet Hong Kong: Popular Cinema and the Art of Entertainment (Cambridge, MA: Harvard University Press, 2000), chaps. 6 and 8. On Tarr see the essays collected in Béla Tarr (Budapest: Filmunio Hungary, 2001). Susan Sontag, “The Decay of Cinema,” New York Times Magazine, February 25, 1996, 60–61. Quoted in Ellen Oumano, Film Forum: Thirty-Five Top Filmmakers Discuss Their Craft (New York: St. Martin’s, 1985), 112.

NOTES TO PAGES 266–269

Index

Page numbers in italics refer to figures. The Abe Clan (Abe ichizoku, 1938), 85, 110 À bout de sou›e (1960), 152 actors: Angelopoulos’s treatment of, 143; in dinner-table scenes, 3–4; Feuillade’s treatment of, 44–45; in Hou’s films, 196, 198; Mizoguchi’s treatment of, 88 Advise and Consent (1962), 178 Les agents tels qu’on nous les représente / Les agents tels qu’ils sont (1908), 50, 51, 52 Ah, My Home Town (Aa kokyo,1938), 115 Akerman, Chantal, 146, 155, 157–58, 159 Akira (1988), 17 Alexander the Great (Megalexandros, 1980), 141, 294–95n15; aperture framing of, 161, 161– 62; average shot lengths of, 159; location for shooting, 144; protracted rhythm of, 177, 178–79; themes of, 143; vacant shots of, 166 Allen, Woody, 4, 29, 30, 150, 274n42 All the Youthful Days. See The Boys from Fengkuei Alpers, Svetlana, 10 Alphaville (1965), 267 American cinema. See Hollywood cinema

American shot (plan-américain), 27 Le Amiche (1955), 288n31 Anderson, Paul Thomas, 268 Andrei Rublev (1966), 156 Angel Face (1952), 150 Angelopoulos, Theo, 9, 10, 218; aperture framing device of, 161–62; average shot lengths of, 147, 148, 159; background of, 141–42; cinematic themes of, 140–41, 143; climate/locale preferences of, 143–44; culture-based perspectives on, 294–95n15; diagonal approach/retreat schema of, 177– 80, 178–81; festival circuit’s reaction to, 144– 45; flashback device of, 146–48, 148–49; long shot/long take strategies of, 159–61; long-take location shots of, 147–49; pictorial expressiveness of, 175–77, 182–84, 218; planimetric composition by, 172–73, 174, 175, 176; recessive staging devices of, 166, 169– 72; three-quarter dorsality of, 147, 162, 163, 164; trademark style of, 145–46; two-phase periodization of, 184–85; vacant shot’s pacing by, 164, 165, 166

INDEX

301

Antonioni, Michelangelo, 13, 144, 146, 148, 189; average shot lengths of, 151–52, 288n31; dedramatization technique of, 153–55, 159; long-take technique of, 151–52, 288n26; three-quarter dorsality of, 162 aperture framing: in Angelopoulos’s Alexander the Great, 161–62; of Angelopoulos’s planimetric images, 173, 174; of Angelopoulos’s recessive images, 170–71; in Angelopoulos’s Voyage to Cythera, 160–61; in Hou’s Dust in the Wind, 210; in Hou’s Son’s Big Doll, 204, 205; Mizoguchi’s compositions, 104, 105, 119, 162 Aphrodite (1978), 274n42 Are Parents People? (1925), 2 Arnheim, Rudolf, 265 Arvanitis, Yorgos, 143 Asian minimalism, 231–32, 233 box, 292nn56–58 ASLs. See average shot lengths L’Assassinat du Duc de Guise (1908), 75, 76 Assayas, Olivier, 190 Astruc, Alexandre, 12 attention. See directing attention average shot lengths (ASLs): of Woody Allen films, 274n42; of Angelopoulos’s films, 147, 148, 159; of Antonioni’s films, 151–52, 288n31; of Dreyer’s films, 150; of European/ American silent films, 47–48, 278n14, 279n19; of Feuillade’s films, 73; of Hollywood films, 26, 150, 273n37; of Hou’s films, 194, 205, 217, 290n13, 291n30; of Japanese films, 282n27, 292n57; of Jarmusch films, 274n43; of Korean films, 292n56; of Mizoguchi’s films, 94, 101, 116–17; sound’s impact on, 86, 150; of Taiwanese films, 290n13, 292n55; of Yang’s films, 222 box L’Avventura (1960), 152, 154, 288n31 axial cut-in: European use of, 47; in Mizoguchi’s Utamaro, 129, 130, 131 Ayme, Jean, 44 Bacher, Lutz, 267 Bacon, Kevin, 3 The Band Wagon (1952), 150 Barr, Charles, 7, 263 Barrabas (1920), 44, 73, 75 Barren Illusion (Oinaru genai, 1999), 292n57 Bauer, Yevgenii, 18–19, 21, 77–79, 153 Baxandall, Michael, 10 Bazin, André, 22, 96, 150, 238, 267; on Bicycle Thieves, 153; on deep-focus composition, 111

302

INDEX

box; style concepts of, 10–11; stylistic history scheme of, 240, 241 Bébé apache (1910), 52, 53 Bee, Kenny, 192, 193, 198 The Beekeeper (O melissokomos, 1986), 141; average shot lengths of, 159; recessional framing in, 171; themes of, 143, 145; three-quarter dorsality in, 162, 163; vacant shots of, 164, 165 Beineix, Jean-Jacques, 152 Belton, John, 250 Benjamin, Walter, 245 Benning, James, 158 benshi (Japanese film commentators), 86 Bergman, Ingmar, 16, 144 Berkeley, Busby, 34 The Best Years of Our Lives (1945), 134 Bicycle Thieves (Ladri du biciclette, 1947), 153 blink rate, 38 blockage/reveal device: of Bauer’s Revolutionary, 19, 19–20, 21; cinematic playing space of, 60, 62–63 box, 64; with Feuillade’s closer cut-ins, 70–72; of Feuillade’s Judex, 67, 68; of Feuillade’s La Nouvelle Mission de Judex, 58, 59–61, 60, 64, 65, 68; of Feuillade’s Tih Minh, 67–68, 69–70, 70; of Feuillade’s Les Vampires, 58, 60, 61; of Hou’s City of Sadness, 191, 218, 219, 227–28; of Hou’s Flowers of Shanghai, 236–37; of Hou’s trilogy, 224; by Mizoguchi, for expressiveness, 106, 107–8; of Mizoguchi’s Last Chrysanthemum, 117, 119–20; of Mizoguchi’s Life of Oharu, 91– 93 Blood and Soul (Chi to rei, 1923), 93 Bloom, Harold, 155 bold foregrounds. See deep-focus composition Bondage (1933), 109 A Borrowed Life (Duosan, 1994), 231, 232, 292n55 Boyd, Brian, 266 The Boys from Fengkuei (aka All the Youthful Days; Fengguei lai de ren, 1983), 190, 202, 204, 205; average shot lengths of, 291n30; planimetric/ long lens shots of, 209; point of view in, 206 Brakhage, Stan, 253–54 Brecht, Bertolt, 145, 157 Brewster, Ben, 267, 280n41 A Brighter Summer Day (Guling jiie shaonian sharen shijian, 1991), 222 box, 223 The Broadcast (I ekpombi, 1968), 142 Brown, Donald E., 298n58 Buñuel, Luis, 42 Burch, Noël, 134, 240

The Burmese Harp (Biruma no tategoto, 1956), 133 Cacoyannis, Michael, 142 Cahiers du cinéma (magazine), 146, 267; miseen-scène concept of, 11–12; Mizoguchi’s canonization in, 89–90 calligraphic style (Japanese film), 85, 86 Cameron, Ian, 12–13 Campion, Anna, 4 Campion, Jane, 30 Cantata (Oldás és kötés, 1963), 156 Caravan (1937), 148 Carax, Léos, 152 Carroll, Noël, 254 cave images, 258, 296n33 centering: Angelopoulos’s strategies of, 160– 62; with diagonal staging, 55–58; Feuillade’s dramatic use of, 49–50; with Feuillade’s horizontal staging, 51, 52, 53–54, 55 Central Motion Picture Corporation (CMPC, Taiwan), 192, 202, 204, 215, 230 Champreux, Maurice, 45 chanbara (swordplay films), 85, 132 Charisma (1999), 292n57 Cheerful Wind (Fenger tita cai, 1982), 193, 194, 195, 203, 290n13 Chen, Robert, 203 Chen Kaige, 216, 231 Chen Kun-hou, 202, 203, 204, 210 Chiang Kai-chek, 190 Chiao, Peggy, 216, 217 chiaroscuro: in Hou’s trilogy, 219–21, 227–28; in Mizoguchi’s films, 113, 114, 115–16, 117 Un chien andalou (1928), 42 Chikamatsu Monogatari (The Crucified Lovers, 1954), 89, 134 China’s Fifth Generation, 231–32, 292n58 La Chinoise (1967), 167 box Chiung Yao, 192, 193, 194 The Chronicle of Anna Magdalena Bach (Chronik der Anna Magdalena Bach, 1967), 157, 158 church paintings, 80 Chushingura (1913), 84 The Chushingura Tale of the Genroku Era. See Genroku Chushingura Chu Tien-wen, 204, 206 cinema: academic studies of, 265–66; cultural legitimization of, 80–81, 280n41; “death of,” 266–67; economic/production issues of, 81– 82; mise-en-scène concept of, 11–13; stylistics approach to, 10–11, 238–39; transcultural

conventions of, 258–60, 297n44. See also Hollywood cinema; Japanese cinema; New Taiwanese cinema; Soviet cinema Cinema Nôvo (Brazil), 156 CinemaScope framings, 27, 29, 150 cinematic staging: as artistic/theatrical solution, 80–81, 280n41; of Bauer’s dinner-table scene, 19, 19–21, 21; blockage/reveal device of, 60, 64; current directors using, 268– 69; cutting’s cooperation with, 9, 16–17, 66–68, 69; in depth, by Eisenstein, 17–18; as economic choice, 81–82; explanatory/ interpretive approach to, 10, 238–39; by Feuillade’s contemporaries, 76–79; Feuillade’s principles of, 75–76; gradation of emphasis in, 6–7, 8–9, 21–22; of Hong’s dinner-table scene, 5–7; invisible-observer eªect on, 191; mise-en-scène vehicle of, 11–13, 16; Mizoguchi’s pictorialism of, 90, 92–93, 96–97; neglected appreciation of, 7–9; playing space of, 18, 60, 62–63 box, 64; of three-quarter dorsal poses, 162, 163, 164 ciné-romans (“film novels”), 44 Citizen Kane (1941), 28, 111 box, 112, 150 City of Sadness (Beiqing chengshi, 1989), 222 box; average shot lengths of, 217; blockage/reveal device of, 218, 219, 227–28; chiaroscuro device of, 227; dinner-table scenes of, 186– 87, 228–29, 230–31; historical event of, 224– 25; iterated framings of, 225–27; plot of, 215–16 classical continuity. See continuity style close-ups: of Bergman’s long takes, 16, 17; of conversation scenes, 26–28, 273n38; Feuillade’s cut-ins to, 67–68, 68–69, 70– 72; Hartley’s elliptical use of, 30–31; and Mizoguchi’s expressiveness, 131–32, 138, 139; of Mizoguchi’s long takes, in Utamaro, 128–29, 129–31, 134; of Mizoguchi’s theater scenes, 126–27; in widescreen format, 27 CMPC (Central Motion Picture Corporation, Taiwan), 192, 202, 204, 215, 230 Coen, Ethan, 30, 256 Coen, Joel, 30, 256 compass-point editing, 233 box composition in depth. See depth composition A Confucian Confusion (Duli shidai, 1994), 222 box continuity style: American emergence of, 45–46; conversation norms of, 22–23; cross-cultural use of, 260, 297n44; economic/production issues of, 81–82; Europeans’ sparing use of,

INDEX

303

continuity style (continued) 46–48, 67, 278n17; Feuillade’s adoption of, 73, 75; of Hou’s early films, 194, 290nn12,13; Japanese mastery/embellishments of, 83–86, 280n4; long takes blended with, 14–15 box; of Mizoguchi’s Song of Home, 99, 100, 101; of shooting dinner-table scenes, 2–4; sound’s impact on, 86; standardized schemas of, 4–5. See also intensified continuity style conversation scenes: artificiality of, 37–38; elliptical close-ups of, in Simple Men, 31; of Jerry Maguire’s airport scene, 23, 24–25, 26; of Mizoguchi’s Naniwa Elegy, 109–10; representational conventions of, 258–59, 275–76n61; singles and close-ups of, 26– 28, 273n38; stand-and-deliver staging of, 22–23; walk-and-talk staging of, 29. See also dinner-table scenes Cottafavi, Vittorio, 13 craft tradition. See problem/solution model The Cranes Are Flying (Letjat zhuravli, 1958), 151 Crary, Jonathan, 293–94n8 Cresté, René, 44 Crime and Punishment (Dostoyevsky), 18 Cronaca di un amore (Story of a Love Aªair, 1950), 152, 288n31 crosscutting styles, 47–48, 278n17 The Crucified Lovers (Chikamatsu Monogatari, 1954), 89, 134 The Cuckoo (Hototogisu, 1922), 83–84, 84 Cukor, George, 11, 86, 134, 150 culture: explanatory role of, 242–43, 249; film style in context of, 244–49, 266, 294nn9,14,15; narrative universals of, 261– 62; national/epochal processes of, 243; perceptual/social universals of, 38–39, 258–60, 275–76n61, 297n44; substantive universals of, 264–65, 298nn58,59 Cute Girl (aka Lovable You; Jiushi liuliu de ta, 1980), 201; average shot lengths of, 290n13; long lens composition of, 195, 196, 200– 201; plot of, 192–93; single-shot sequence of, 194, 195 cutting: in American/European silent films, 45–48, 278n17; competing explanations for, 40–41; economic/production issues of, 81–82; by Feuillade, to closer views, 67– 68, 68–70, 70–72; Mizoguchi on, 95–96; privileged status of, 8; sound’s impact on, 86; staging’s cooperation with, 9, 16–17, 66–67. See also continuity style; intensified continuity style

304

INDEX

Daiei studio (Japan), 89, 132–33 Dai-Ichi Eiga studio (Japan), 88, 108, 115 Dalí, Salvador, 42 Dallas, Texas: After the Gold Rush (1970–1971), 158 Darkness and Light (Heian zhi guang, 1998), 292n55 Daughter of the Nile (Niluohe de nuer, 1987), 204, 205, 222 box, 234; average shot lengths of, 291n30; mundane details of, 189; planimetric composition of, 210–11 A Day in the Country (Partie de campagne, 1935), 255 Day of Wrath (Vredens Dag, 1943), 150 Days of ’36 (Meres tou ’36, 1972), 143, 147, 148, 159, 171, 172 The Day the Pig Fell in the Well (Daijiga umule pajinnal, 1996), 292n56 dead time (temps morts): in Antonioni’s films, 153–54; in Tati’s films, 155 decoration: as function of style, 34–35, 274n50; as Japanese hallmark, 85–86. See also pictorialism dedramatization: by Angelopoulos, 159, 162, 164, 166, 169–73; by Antonioni, 153–55; by Jancsó, 156–57; minimalist approach to, 157–58; with planimetric images, 172– 73; with recessive images, 166, 169–71; by Tarkovsky, 155–56; with three-quarter dorsality, 162, 163, 164; two forms of, 152– 53; two image schemas of, 167–68 box; with vacant shots, 164, 165, 166 deep-focus composition: in Akira, 17; in Mizoguchi’s Naniwa Elegy, 109–10, 112; in Mizoguchi’s Song of Home, 101–2; origins/ popularization of, 111, 112; reactions against, 167–68 box; Yizek on, 298–99n59 Defendant and Counsel (Yeames), 80, 81 Delluc, Louis, 283–84n45 DeMille, Cecil B., 13 Demme, Jonathan, 41 Demonsablon, Philippe, 97 denotation: Angelopoulos’s sacrifice of, 175–77; directing attention principle of, 36–37, 39– 40; narrative function of, 33–34, 35–36 dense images: of Japanese art, 133–34; long take’s link to, 94; of Mizoguchi’s films, 95– 96, 120–21, 121–22, 123; of Naruse’s Whole Family Works, 94, 95 De Palma, Brian, 34, 41 depth composition: Bazin’s focus on, 11; blockage/ reveal device of, 58, 60, 64; with bold foreground, 111 box, 112; with diagonal staging, 55–56; Eisenstein’s exercise on,

17–18; with empty foreground, 110, 113; of Feuillade, with forward-thrusting action, 56–58; of Feuillade, with lateral staging, 55, 56–57; independent variables of, 16; as mise-en-scène prototype, 16; of Mizoguchi, with long takes, 102, 102–3; of Mizoguchi’s ensembles, 127–28, 138, 139; with recessive images, 166, 167 box, 169–72; as sound cinema technique, 86; Yizek on, 298– 99n59. See also deep-focus composition; distant-depth staging de Sica, Vittorio, 153 Détective (1985), 31 Les Deux gamines (1921), 45, 73, 73–74 diagonal approach/diagonal retreat schema, 177–80, 178–81 Diary of Didi (Didi de riji, 1978), 193, 290n13 dinner-table scenes: of Bauer’s Revolutionary, 19, 19–21, 21; Hong’s long take of, 5–7; of Hou’s City of Sadness, 186–87, 228–29, 230– 31; of Hou’s Dust in the Wind, 213, 214; of Hou’s Flowers of Shanghai, 236, 237; Hou’s prototype shot of, 228; problems/solutions in shooting, 1–4, 271–72n7; of Zhang’s Raise the Red Lantern, 187–88. See also conversation scenes directing attention: with Angelopoulos’s long-shot/long-takes, 160–62; within cinematic playing space, 18, 62–63 box; denotative strategy of, 36–37, 39–40; with Feuillade’s blockage/reveal tactic, 58, 60, 61, 64, 65; with Feuillade’s centering variances, 49–50, 51, 52, 53–54, 55; with Feuillade’s closer cut-ins, 67–68, 69, 70– 72; with Feuillade’s depth stagings, 55– 58; within human visual system, 38–39; perceptual universals of, 37; with pictorial cues, 48–49, 81, 279n20, 280n41; response to Yizek on, 262–64 directors: as agent/decision maker, 40, 254– 57, 265, 296n32; Bazin’s two types of, 11; convergent film styles of, 239–40, 241, 260; culture-based perspectives on, 243– 45, 246–47, 294–95n15; neglected skills of, 7–8; normative practices of, 41–42, 253– 54; problem-solving process of, 249–52, 296n24; as staging-centered, 268–69 distanced recessiveness. See recessive composition distant-depth staging: of Angelopoulos’s Voyage to Cythera, 180, 181; definition/purpose of, 110, 113; of Hou’s Green, Green Grass of Home, 198–200; of Hou’s Son’s Big Doll, 204, 204–

5; of Hou’s A Time to Live, 206–8, 209; of Mizoguchi’s Naniwa Elegy, 113–15, 115–16; of Mizoguchi’s Street of Shame, 138, 139 Do the Right Thing (1989), 29, 30 Douchet, Jean, 89 The Downfall of Osen (Orizuru Osen, 1935), 99, 101, 282n26, 285n50 Dreyer, Carl, 83, 146, 150–51, 152, 153 Duras, Marguerite, 158, 267 Dust in the Wind (Lianlian fengchen, 1987), 204, 222 box, 225; aperture framing in, 210; average shot lengths of, 205, 291n30; point of view in, 206; table-setting scene of, 213, 214 Dust of Angels (Shaolin Ye, An La!, 1992), 230, 292n55 The Earth Smiles (Daichi wa hohoemu, 1925), 282n26 Eat Drink Man Woman (Yinshi nannu, 1993), 230 L’Eclisse (1962), 154 editing. See cutting 8 1/2 (film), 159, 217, 267 813 (1923), 93 Eisenstein, Sergei, 9, 77, 129, 267; bold foreground device of, 111 box; depth staging approach of, 17–18 Ekman, Paul, 37 Electra (1962), 142 11 × 14 (1976), 158 Ellis, John, 298–99n59 Elster, Jon, 257, 296n32 e-makimono (picture scroll), 133 Enchantment (1948), 148 ensemble staging: by Feuillade, 64, 66; by Hou, 212–14; by Mizoguchi, 127–28, 138, 139 Epstein, Jean, 83, 96 Eternity and a Day (Mia eoniotita ke mia mera, 1998), 141; average shot lengths of, 159; dorsal angles of, 162, 163; planimetric imagery of, 173, 174–75; themes of, 143 Eureka (2000), 292n57 European modernist film: Angelopoulos’s adaptations of, 145–46, 158–59, 183–84; dedramatization trends in, 152–55; image schemas of, 167–68 box; long takes of, 150–52, 155–56, 288n26; of politicized filmmakers, 156–58 European silent film: average shot lengths of, 47, 48, 278n16, 279n19; French foreground of, 62 box; minimal editing of, 46–48, 66–67, 278n17

INDEX

305

Exodus (1960), 178 expressiveness: of Angelopoulos’s protracted images, 175–77, 182–84; dedramatization of, 152–57; of Feuillade’s films, 76; as function of style, 34; of melodrama genre, 97; of Mizoguchi’s blockage/reveal strategy, 106, 107–8; of Mizoguchi’s close-ups, 131–32, 138, 139; of Mizoguchi’s pictorialism, 90, 92–93, 96–97; of Mizoguchi’s rearward tracking shots, 124 extreme long shots, 159–60 eyes: and eyeline matching device, 37–38; pictorial cues for, 48–49, 279n20; visual search mechanism of, 38–39, 293–94n8 facial expressions, 37 Fairservice, Don, 267 The Famous Sword Bijomaru (Meito Bijomaru, 1945), 116, 123, 134 The Fan (1949), 150 Fantômas (1913–1914): average shot lengths of, 73; centering variances in, 49–50, 52, 54, 55; closer cut-ins of, 71, 72; depth composition of, 55, 56, 57–58; plot of, 43–44 Fassbinder, Rainer Werner, 146, 167 box, 267 Favorites of the Moon (Les Favoris de la lune, 1984), 268 Fehér, György, 268 Fellini, Federico, 144, 159, 189, 216 festival circuit: Angelopoulos’s films on, 144– 45; in Greece, 142; Hou’s films on, 190, 215– 16, 217, 229; Mizoguchi’s films on, 89, 133, 134, 239, 267; Taiwanese films on, 214–15, 229–30 Feuillade, Louis, 9, 10, 162, 177, 239; American continuity devices of, 73, 75; average shot lengths of, 73; blockage/reveal devices of, 58, 59–61, 60, 64, 65; centering strategy of, 49–50; closer cut-ins by, 67–68, 68–69, 70–72; contemporaries of, 76–77; culturebased perspectives on, 247–48; diagonal depth staging by, 55; e‹ciency/discipline of, 44–45, 82; ensemble timing tactics of, 64, 66; epochal perspectives on, 246–47; European staging principles of, 75–76; forward-thrusting action device of, 56–58; and Gri‹th’s style, 48, 278n17; horizontal staging by, 50, 51, 52, 53–54, 55, 56; installment films/serials of, 43–44, 277n5; pictorial composition of, 49, 279n22; problem-solving approach of, 248–49, 250 Fifth Generation (China), 231–32, 292n58

306

INDEX

Film d’Art cycle, 75 film festivals. See festival circuit filmmakers. See directors film style: academic neglect of, 32–33, 265– 67, 274n47; agent/director’s choice of, 40, 254–57, 265, 296n32; with constraints, 257, 296n33; culture-based perspectives on, 98, 242–49, 283–84n45, 294nn9,14,15; current interest in, 267–68; four functions of, 33– 35, 274n50; normative practices of, 41–42, 252–54; perceptual universals of, 38–39, 258–60, 275–76n61, 297n44; problem/ solution approach to, 41–42, 249–51, 296n26; task-oriented convergences of, 239–40, 241, 260; three historical accounts of, 240–41; viewer’s response to, 35; Yizek’s premises on, 260–65, 298–99n59 Le Fils du Flibustier (1922), 45 fixed long takes: of Hong’s dinner-table scene, 5–7; of Hou’s location shots, 196, 198– 201; of Korean films, 231, 292n56; of New Taiwanese cinema, 203, 230–31, 292n55; rarity of, 29–30 flashback device: of Angelopoulos, 146–48, 148–49; of Hou, 216–18 Flower in the Rainy Night (Kanhai de rizi, 1983), 202 Flowers of Shanghai (Haishang hua, 1998), 189, 191, 232, 234, 236–37 Foggy Harbor (Kiri no minato, 1923), 93 Fortini/Cani (1977), 157 forward-thrusting action: in Feuillade’s Fantômas, 56–58; by Mizoguchi, with camera movement, 124 Four Devils (1928), 93 Frankenheimer, John, 32 Furie, Sidney J., 32 Gance, Abel, 83, 96 Ganz, Bruno, 144 Gate of Hell ( Jigokumon, 1954), 133 Gaumont, Léon, 45, 55, 278n17 Gaumont company (France), 43, 44 Genroku Chushingura (1941–1942), 88, 123, 124, 135, 282n26; legendary source of, 116; long-take shots of, 117, 118 Gertrud (1964), 150–51 gestures, 37 Getino, Octavio, 156 Ghost Dog: The Way of the Samurai (1999), 274n43 Gion Festival Music (Gion bayashi, 1953), 134

Goalie’s Anxiety at the Penalty Kick (Die Angst des Tormanns beim Elfmeter, 1972), 148 Godard, Jean-Luc, 31, 90, 149, 152, 157, 167 box, 189 The Godfather (1972), 267 Golden Horse Festival, 229, 230 Gombrich, E. H., 79–80, 81, 240–41, 259 Goodbye South, Goodbye (Nanguo zaijian, nanguo, 1996), 232, 234, 235 Good Men, Good Women (Haonan, haonu, 1995): average shot lengths of, 217; blockage/reveal tactic of, 224; box o‹ce earnings of, 229; chiaroscuro of, 219, 220; flashback strategy of, 216–18 Gosho Heinosuke, 85, 132, 133 gradation of emphasis: in Angelopoulos’s long shots, 160–61; in Bauer’s dinner-table scene, 19, 21; in Hong’s drinking scene, 6–7, 8– 9; in Hou’s long lens shots, 198–201; with pictorial cues, 49, 279n20; Spielberg on, 22 The Great Citizen (Velikij grazhdanin, 1938–1939), 111 box, 112 Greek Film Center, 142 The Green, Green Grass of Home (Zaina heban quingcao quing, 1982): average shot lengths of, 290n13; gradation-of-emphasis staging of, 198–201; New Cinema elements of, 203; plot of, 193 Il Grido (1957), 152, 154, 162, 288n31 Gri‹th, D. W., 48, 245–46, 248, 298–99n59 Growing Up (Xiaobi de gushi, 1983), 202, 203 Guerman, Alexei, 159 Guerra, Ruy, 152 Guerra, Tonino, 159 Gunning, Tom, 245–46, 248 Hannah and Her Sisters (1985), 4 Hansen, Miriam Bratu, 275–76n61 Happiness (1998), 30 Hark, Tsui, 30 Hartley, Hal, 30–31, 38, 268 Hawks, Howard, 11, 12, 13, 14–15, 256 Heartbreak Island (Qunan dongtian, 1995), 292n55 Hepburn, Katharine, 39 Herzog, Werner, 146 Hirokazu Kore-eda, 231 Hiroshige, 111 box Hiroshi Komatsu, 280n4 His Girl Friday (1940), 14–15, 23, 27, 31 Hitchcock, Alfred, 11, 12 Hogan, Patrick, 261–62 Hogarth, William, 39, 263 The Hole (Dong, 1998), 292n55

Hollywood cinema: American foreground of, 62 box; average shot lengths of, 26, 47–48, 273n37, 274nn42,43, 278n14, 279n19; close cut-ins of, 27–28; continuity style of, 45–46; intensified continuity style of, 30–32; long takes of, 150; mise-en-scène vehicle of, 11–13; Mizoguchi linked to, 134; modern heroines of, 109; moving camera technique of, 28–29; as transculturally appealing, 83, 275–76n61 Hometown (Furusato, 1930), 93, 99, 101, 104, 105 Hong Sang-soo, 5–7, 6–7, 8–9, 231, 292n56 horizontal staging schema, 50, 51, 52, 53–54, 55, 56 Horton, Andrew, 144 Hou Hsiao-hsien, 9, 10; aperture framings of, 204, 205, 210; average shot lengths of, 194, 205, 217, 290n13, 291n30; chiaroscuro device of, 219–21, 227–28; on constraints of style, 191–92; culture-based perspectives on, 243, 244; dinner-table strategy of, 186–87, 228– 29, 230–31; ensemble staging practices of, 212–14; festival circuit’s reaction to, 190, 215, 216, 217, 229–30; filmic themes/plots of, 188–90, 205; flashback device of, 216–18; gradation-of-emphasis staging by, 198–201; historical trilogy of, 215–17; on improvisation, 204–5; iterated framings of, 225–27; long lens location shots of, 195–96, 198, 204; long lens/wide format shots of, 206–8, 209; narrative objectivity of, 205–6; New Cinema tendencies of, 203; passing train schema of, 208–9, 252; planimetric composition of, 209–11; problem/solution approach of, 237, 252; recent films/projects of, 232, 234, 235– 36, 236–37; romantic comedy films of, 192– 93; team of, 204; and Yang’s style, 222 box Hou Hsiao-hsien, trilogy of: average shot lengths of, 217; blockage/reveal device of, 218, 219, 224, 227–28; chiaroscuro of, 219–21, 227; flashback strategy of, 216–18; time frames of, 215–16. See also City of Sadness; Goodbye South, Goodbye; The Puppetmaster The Hour of the Furnaces (Hora de los hornos, 1968), 156 Howard, Ron, 22, 41 How to Marry a Millionaire (1953), 29 Hsu Hsiao-ming, 230 Huillet, Danièle, 157, 159, 184 Hungarian Rhapsody (Magyar rapszódia, 1978), 157 The Hunters (I kynighi, 1977), 141; average shot lengths of, 159; blank space schema of, 182– 83; recessive images of, 169–70, 176; themes of, 143

INDEX

307

Hurley, James, 251 Hwang Chun-ming, 202 Ichikawa Kon, 86, 132, 284n46 Identification Marks: None (Rysopis, 1965), 168 box I Even Met Happy Gypsies (Skupljaci perja, 1967), 168 box Imai Tadashi, 132, 133 Imamura Shohei, 86 India Song (1975), 158 Ingeborg Holm (1913), 77 In Our Time (Guangyin de gushi, 1982), 202 intensified continuity style: average shot lengths of, 26, 273n37; close-ups/singles of, 26–28, 273n38; of directing attention, 39; emergence/ pervasiveness of, 30–32; exaggerated interpersonal cues of, 38; explanatory frameworks for, 40–41, 42; of Jerry Maguire’s airport scene, 23, 24–25, 26, 273n36; long lens shots of, 28; moving camera of, 28–29 intertitles, of Mizoguchi films, 101, 285n50 Intolerance (1916), 48 Iosseliani, Otar, 155, 268 Ito Daisuke, 85 I Vitelloni (1953), 189 Iwami Jutaro (1937), 110, 112 Izumi Kyoka, 99 Jackie Brown (1997), 267 Jacobs, Lea, 267, 280n41 Jameson, Fredric, 294–95n15 Jancsó, Miklós, 156–57, 159 Japanese cinema: aperture framing device of, 104, 105; average shot lengths of, 282n27, 292n57; blockage/reveal device of, 104, 106, 107–8; bold foregrounds of, 111 box, 112; dense composition of, 94–95; distant-depth staging in, 110, 113; invented aesthetic traditions of, 97–98, 283nn41,45; during Occupation, 89, 123–24; post-Occupation expansion of, 132–33; three styles of, 85–86; Westernization of, 83–85, 280n4 Jarmusch, Jim, 154, 274n43 Jeanne Dielmann 23 Quai du Commerce 1080 Bruxelles (1975), 157–58 Jerry Maguire (1996), 31; average shot lengths of, 26, 273n37; blink rate in, 38; close-ups of, 27; intensified continuity of, 23, 24–25, 26, 273n36; long lens shots of, 28 Jiang Ling, 193 Jia Zhange-ke, 232

308

INDEX

Jousse, Thierry, 13 jowa bungaku (“love-talk literature”), 109 Judex (1917), 44, 67, 68, 72, 73 Jules and Jim (1962), 267 Kael, Pauline, 32, 34 Kaiki Edogawa Ranzen (1937), 112 Karaindrou, Eleni, 176, 189 karyu shosetsu (“demi-monde fiction”), 108–9 Katzelmacher (1969), 167–68 box Kazan, Elia, 32 Keil, Charlie, 245, 247, 267 Keitel, Harvey, 144 Khroustaliev, My Car! (Khroustaliev, machinu!, 1998), 159 Kinema Jumpo poll, 87, 88, 133 The King of Paris (Korol parizha, 1917), 77–79 Kinoshita Keinosuke, 132 Kinugasa Teinosuke, 85, 133 Kitano Takeshi, 3, 188, 233 box Klercker, Georg af, 77 Kluge, Alexander, 152 Klute (1971), 29 KMT (Kuomintang Party), 190, 215–16, 224–25 Kobayashi Masaki, 132 Korean films, 231, 292n56 Kriemhild’s Revenge (Kriemhilde’s Rache, 1924), 17 Kubrick, Stanley, 29, 32 Kuleshov, Lev, 18, 21, 72 Kuomintang Party (KMT), 190, 215–16, 224–25 Kuper, Adam, 265 Kurosawa Akira, 86, 89, 129, 132, 135, 196, 201 Kyrou, Ado, 134 Lady in the Lake (1946), 150 Lady of Musashino (Musashino fujin, 1951), 132, 137 Lam, Ringo, 30 The Land (El ard, 1969), 167 box Landscape in the Mist (Topio stin omichli, 1988), 141, 184; aperture framing in, 173; average shot lengths of, 159; pictorial climax of, 177, 179, 218; recessive images of, 166, 169; themes of, 143, 145; vacant shots of, 164, 165 Lang, Fritz, 12, 13 Late Autumn (Akibiyori, 1960), 2–3 lateral staging schema, 50, 51, 52, 53–54, 55, 56 Lee, Ang, 230 Lee, Mark (Li Ping-pin), 204 Lee, Spike, 29 Lee Kwang-mo, 231 Lenin in 1918 (Lenin v 1918 godu, 1937), 111 box Lester, Richard, 32

Letter That Was Never Sent (Neotpravlennoye pismo, 1959), 17 Levesque, Marcel, 44 Liang Ching, 217 Licht-bild-bühne (journal), 283–84n45 Lies (Gojitmal, 1999), 292n56 The Life of an Actor (Geido ichidai otoko, 1941), 93, 115 The Life of Oharu (Saikaku ichidai onna, 1952), 89, 113; critical reception of, 133; knife scene of, 90–92, 135; pictorial field of, 92–93 Like Grains of Sand (Nagisa no shindobaddo, 1996), 292n57 Lin Wen-chi, 225 Li Ping-pin (Mark Lee), 204 Lissitzky, El, 264 Li Tien-lu, 205, 216, 217 The Little Foxes (1941), 111 box, 298–99n59 Lonely Hearts Club (Gi mo fan shin gu le bu, 1994), 292n55 Long, Fritz, 83 long lens technique: advantages of, 196; angle subtended by, 197; flattened space of, 168 box; of Hou’s dense/wide format, 206–8, 209; of Hou’s dinner table scenes, 186– 87, 228, 229, 230–31; Hou’s gradation-ofemphasis with, 198–201; Hou’s misleading narration with, 211–12; of Hou’s outdoor locations, 195–96, 198, 204; of Hou’s planimetric images, 209–11; of Jerry Maguire, 28; of Klute, 29 long shots: of American continuity style, 46; of Angelopoulos’s locales, 147–49; of Angelopoulos’s long takes, 159–61; of Angelopoulos’s vacant frames, 164, 165, 166; for punctuation, 28. See also long lens technique long takes: Woody Allen on, 274n42; of Angelopoulos’s landscapes, 147–49; of Angelopoulos’s long shots, 159–61; of Angelopoulos’s pictorial epiphanies, 158, 175– 79, 182–84; of Asian minimalism, 231–32, 292nn56–58; of Bauer’s dinner-table scene, 18–19, 19–21, 21; of classical continuity style, 14–15 box; dedramatized approach to, 152– 55; of European modernist film, 146, 150– 52, 288n26; gradation of emphasis in, 6–7, 21–22; of Hollywood film, 150; of Hong’s dinner-table scene, 5–6, 8–9; of Hou’s historical trilogy, 217; of Hou’s long lens shots, 196, 198–201, 208; of Jarmusch, 274n43; of Mizoguchi’s close-ups, 128–29,

129–31; of Mizoguchi’s dense images, 95– 96, 120–21, 121–22, 123; of Mizoguchi’s depth staging, 102, 102–3, 116–17; of politicized modernists, 156–58; production di‹culties of, 82, 94; of sound cinema, 86; as Taiwanese trademark, 203, 230–31, 292n55; Tarkovsky’s style of, 155–56; of Young Cinemas, 152. See also average shot lengths; fixed long takes Losey, Joseph, 13 Lovable You. See Cute Girl Love, Love, Love (Chun Chun de ai, 1974), 193 Love in a Cabin (Baiwu zhi lian, 1974), 193–94, 290n13 The Love of Sumako the Actress ( Joyu Sumako no koi, 1947), 124, 126–27, 134 The Loyal 47 Ronin. See Genroku Chushingura Lumet, Sidney, 32 Lupino, Ida, 13 Lynch, David, 30 Maborosi (Maborosi no hikari, 1996), 231, 232, 292n57 MacDonald, Dwight, 32, 34 MacMahonists, 12, 13 Magnolia (1999), 35 Mahjong (Maijang, 1996), 222 box Makavejev, Dusan, 156 Makino Masahiro, 85 Makino Shozo, 83 Manhattan (1978), 274n42 Mann, Michael, 30 The Marrying Kind (1952), 134, 150 Marx, Karl, 251, 296n26 The Master of the House (Du skal aere din hustru, 1925), 150 Mastroianni, Marcello, 144 Mayama Seika, 116 medium shots: Feuillade’s use of, 71–72, 73; of Jerry Maguire’s airport scene, 24, 26–27; replaced by close framings, 27, 273n38 Meet Me in St. Louis (1944), 3 Meiji-mono films, 115–16 Méliès, Georges, 75 melodrama, 87–88, 97, 152–53 Memento (film), 467 mettre en scène (“arrange the action onstage”), 11 Metz, Christian, 259 Miki Minoru, 114 Millennium Mambo (Qianxi manbo, 2001), 189, 232, 234, 235, 293n65 Minnelli, Vincente, 4, 150

INDEX

309

The Mirror (Zerkalo, 1974), 156 mise-en-cadre (mise-en-shot), 9, 17 mise-en-scène: as artistic/theatrical solution, 80–81, 280n41; of classical continuity cinema, 14–15 box; depth staging of, 16–18; as economic choice, 81–82; interpretations of term, 11–13; technical elements of, 16 mise-en-shot (mise-en-cadre), 9, 17 Miss Julie (Fröken Julie, 1951), 148 Miss Oyu (Oyusama, 1951), 132, 137 Mitry, Jean, 146 Miyagawa Kazuo, 94, 284n46 Miyamoto Musashi (1944), 116, 123, 134 Mizoguchi Kenji, 9, 10, 13, 144, 146, 189, 228; aperture framing device of, 104, 105, 119, 162; average shot lengths of, 94, 116–17, 282n27; background/traits of, 87–89; blockage/reveal strategy of, 91–93, 106, 107–8, 119–20; bold foreground compositions of, 101–2, 109–10, 112; career phases of, 98; and cinematographer issue, 284n46; close-up strategies of, 126–27, 128–29, 129–31; continuity devices of, 99, 100, 101, 134; culture-based perspectives on, 243, 244; dense images of, 95–96, 120–21, 121–22, 123; depth staging of, with long takes, 102, 102–3; distant-depth staging of, 110, 113–15, 115–16; expressive blockage/reveal strategy of, 106, 107–8; expressive close-ups of, 131–32, 138, 139; expressive pictorialism of, 90, 92–93, 96–97; festival circuit’s reaction to, 89, 133, 134, 239, 267; French canonization of, 89–90; intertitles used by, 101, 285n50; last film of, 138, 139; long take pluralism of, 135, 137, 139; lost films of, 93, 282n26; neonaturalist films of, 106, 108– 9; Occupation projects of, 89, 123–24; pictorialist approach of, 85–86, 90, 92–93, 133–34; post-Occupation projects of, 132; rearward tracking shots of, 124, 125; shinpa themes of, 99; war period projects of, 115–16 modernity thesis: appeal of, 247; discounted explanatory role of, 247–49; and perceptual change, 245, 275–76n61, 293–94n8; and stylistic change, 245–47, 294nn9,14 Monday (1999), 30 Mon Oncle (1958), 155 Moravia, Alberto, 153 Mother (Okkasan, 1952), 133 M/Other (1999), 292n57 Mourlet, Michel, 12 Movie (magazine), 12–13, 267

310

INDEX

moving camera: Bazin’s focus on, 11; diagonal approach/retreat schema of, 177–80, 178–81; dinner-table shots of, 4, 271n7; free-ranging technique of, 28–29; rearward tracking shots of, 124, 125; as television device, 30 Mr. Moto’s Gamble (1938), 112 Münsterberg, Hugo, 265 Murata Minoru, 88 Murmur of Youth (Mei li zai chang ge, 1997), 292n55 Murnau, F. W., 11, 13, 83, 93, 146 Musidora, 44 My Favorite Season (Zhui xiang nian de jijie, 1985), 210 My Love Has Been Burning (Waga koi wa moenu, 1949), 124, 125 My Lovely Neighbor (Wo ai fanglin, 1979), 193 Mysterious Night of the 25th (Mysteriet natten till den 25:e, 1917), 77 Mystery Train (1989), 274n43 Nagata Masaichi, 88, 89, 132–33 Nang Nak (1999), 30 Naniwa Elegy (a.k.a. Osaka Elegy, Naniwa ereji, 1936), 88, 124; bold foregrounds of, 109– 10, 112; distant-depth schema of, 113–15, 115–16; harsh realism of, 106, 108; pathway repetitions in, 120; plot of, 109 narrative: style’s relation to, 33–34, 36; transcultural features of, 261–62; Yizek on crisis of, 262–63 Naruse Mikio, 85, 86, 89, 94, 95, 132, 133 Nausea (Sartre), 153 Navarre, René, 76 Nerven (1919), 77 Neue Sachlichkeit movement, 108 New Greek Cinema, 142, 190 New Taiwanese Cinema, 16, 189; on festival circuit, 214–15, 230; fixed long take of, 203, 230–31; Hou’s status in, 190; nativist themes of, 202; Yang’s masterwork of, 222 box New Tales of the Taira Clan (Shin Heike Monogatari, 1956), 89, 133 Nichols, Mike, 32, 150 Night and Fog in Japan (Nihon no yoru to kiri, 1960), 152 Nikkatsu studio (Japan), 87 Not Reconciled (Nicht versöhnt, 1965), 184 La Notte (1961), 154, 288n31 La Nouvelle Mission de Judex (1917), 44, 55, 73, 80; blockage/reveal device of, 64, 65, 68; depth composition strategies of, 58, 59–60 Nouvelle Vague, 156, 267

Une Nuit agitée (1908), 46, 49 Oedipus Rex (1967), 206 Ogata Korin, 133 Oh, Su-jeong! (The Virgin Stripped Bare by Her Bachelors, 2000), 5–7 Okichi, Mistress of a Foreigner (Tojin Okichi, 1930), 93 Oklahoma (1955), 150 Oliveira, Manoel de, 268 One Fine Spring Day (Bomnaleun ganda, 2001), 292n56 100,000 Ryo Mystery ( Jyuman ryo hibun, 1939), 107 On Golden Pond (1981), 39 On the History of Film Style (Bordwell), 9, 240, 298–99n59 The On-the-Square Girl (1917), 46, 70, 72 Ophuls, Max, 29, 86, 151, 267 Ordet (1955), 146, 150–51, 153 The Orphan of Anyang (Anyangde guer, 2002), 232, 292n58 L’Orphéline (1921), 45, 73 L’Orphéline de Paris (1924), 45 Osaka Elegy. See Naniwa Elegy Oshima Nagisa, 86, 152 Outside the Window (Chuanwai, 1972), 193 Oyuki the Virgin (Maria no Oyuki, 1935), 119, 135; aperture framing in, 104; changing depth relations of, 102, 102–3; plot of, 99 Ozu Yasujiro, 4, 85, 86, 89, 132, 133, 189; compass-point schemas of, 233 box; dinner-table strategy of, 1, 188 Pai Ching-jui, 290n13 Panavision, 27, 150 A Paper Doll’s Whisper of Spring (Kaminingyo haru no sasayaki, 1926), 93 Paradise Sea (Rakuen, 1999), 292n57 Parisette (1922), 45 Partie de campagne (A Day in the Country, 1935), 255 Pasolini, 189, 206 La Passion de Jeanne d’Arc (1928), 150 Pat and Mike (1952), 28 Perils of Pauline (1914), 44 Perkins, V. F., 13 Persona (1966), 17 The Phantom Carriage (Körkalen, 1921), 77 Pialat, Maurice, 146 pictorialism: Angelopoulos’s epiphanies of, 175–77, 182–84; of Japanese art, 133–34; of Japanese film style, 85–86; long takes of, 94–96; and Mizoguchi’s expressiveness, 90, 92–93, 96–97

A Picture of Madame Yuki (Yuki Fujin Ezu, 1950), 132 piecemeal découpage style, 85, 86 Pierrot le fou (1965), 168 box plan-américain (American shot), 27 planimetric composition: by Angelopoulos, 172–73, 174, 175, 176; compass-point editing of, 233 box; definition/sources of, 167–68 box; of Hou’s long lens shots, 205, 209–11 Platform (Zhantai, 2001), 232 Play Time (1967), 155 Polanski, Roman, 150 Pollock, Jackson, 264 The Poppy (Gubijinso, 1935), 99, 282n26; aperture framing in, 105; blockage/reveal device of, 106, 108; distant-depth schema of, 113, 114; pathway repetitions in, 120 Portrait of a Lady (1996), 30 The Power of Kangwon Province (Kangwon-do ui him, 1998), 292n56 Preminger, Otto, 11, 12, 13, 150, 178 Présence du cinéma (magazine), 12 Princess Yang Kwei Fei (Yokihi, 1955), 133 problem/solution model: agent/filmmaker’s role in, 254–57, 265, 296n32; craft context of, 249–51; on DVDs, 267; experimentation with, 253–54; instrumental logic of, 251, 296n26; with narrowed options, 257, 296n33; normative practices of, 41–42, 252–53 Pudovkin, Vsevolod, 72 The Puppetmaster (Xi meng renshing, 1993): average shot lengths of, 217; award won by, 229; blockage/reveal device of, 224; chiaroscuro of, 219–21; flashback strategy of, 216–18; time frame of, 216 Pure Film movement (Japan), 83 Raimi, Sam, 30 Raise the Red Lantern (Da hong denglong gaogao gua, 1991), 187–88 Rashomon (1950), 133 Ray, Nicholas, 11 Rebels of the Neon God (Qingshaonian nezha, 1992), 231, 232, 292n55 recessive composition: with centered perspective, 166; dedramatization eªects of, 167 box; with oblique perspective, 169–72; with planimetric images, 173, 175, 176 Reconstruction (Anaparastasi, 1970), 143; award won by, 142; flashback device of, 146–47; recessive diagonals of, 170–71; three-quarter dorsality in, 147, 162

INDEX

311

The Red and the White (Csillagosok, Katonák, 1967), 156 Red Beard (Akahige, 1965), 201 Red Desert (Deserto rosso, 1964), 154 Red Psalm (Még kér a nép, 1972), 157 Reinert, Robert, 77 Renoir, Jean, 11, 86, 146, 178, 240, 255 Reservoir Dogs (1992), 4, 28–29, 271n7 The Revolutionary (Revolutsioner, 1917), 19, 19–21, 21 Ride the Pink Horse (1947), 29 The River (1997), 230, 292n55 Rivette, Jacques, 11, 12, 13, 90, 97 Rohmer, Eric, 11, 13, 16 Roman Holiday (1953), 134 Rope (1948), 150 Rossellini, Roberto, 13, 152–53, 191, 288n26 Rotha, Paul, 134 The Round-Up (Szegénlegények, 1965), 156 Rules of the Game (La Règle du jeu, 1939), 178 Runaway Bride (1999), 26 saccadic eye movements, 39, 293–94n8 Sadoul, Georges, 75, 146 Salt, Barry, 273n37, 279n19 The Sandwich Man (Erzi de da wan’ou, 1983), 190, 202, 206, 208, 211, 212 Sansho the Bailiª (Sansho dayu, 1954), 89, 133, 136 Sarris, Andrew, 12, 13, 32, 34, 146 Scarface (1932), 29 Scenes from Under Childhood (1970), 254 Schlesinger, John, 32 Scorsese, Martin, 152 Sea Gull (1926), 93 Secrets of Screen Acting (Tucker), 273n38 semiotic codes, filmic norms versus, 252, 258–59 Sequence (journal), 12 sex, lies, and videotape (1989), 30 Shakespeare in Love (1998), 26 Shanghai Express (1932), 105, 107 Shen Tsung-wen (Shen Congwen), 206 Shimazu Yasujiro, 85 Shimizu Hiroshi, 85, 86 Shindo Kaneto, 132 Shin Heike Monogatari (New Tales of the Taira Clan, 1956), 89, 133 Shinko studio (Japan), 88, 115 shinpa plays, 88, 99 Shiri (1999), 30 Shochiku studio (Japan), 83, 88, 115, 132 Shu Qi, 234 Silence and Cry (Fényes szelek, 1966), 156

312

INDEX

Simple Men (1992), 31, 38 Singer, Ben, 246 single shots: of conversation scenes, 27–28; of dinner-table scenes, 1–4, 2; Hou’s use of, 194, 195; of Jerry Maguire’s airport scene, 26–27 Sisters of Gion (Gion no shimai, 1936), 88, 108, 109, 117, 118, 178 Sjöström, Victor, 77, 96, 153 Skolimowski, Jerzy, 152 Sokurov, Alexander, 268 Solanas, Fernando, 156 Solaris (1972), 156 Song Cun-shou, 290n13 Song of Home (Furusato no uta, 1925), 99, 100, 101 The Song of the Camp (Roei no uta, 1938), 115 Son’s Big Doll (Erzi de da wan’ou, 1983), 189, 202, 203–4 Soviet cinema: 1920s montage style of, 96, 111 box, 156; wide-angle style of, 150, 151 Spielberg, Steven, 22, 27, 152 Spring in My Home Town (Areumdawoon sheejul, 1998), 231, 292n56 staging. See cinematic staging Stahl, John, 86 Stalker (1979), 156 stand-and-deliver staging: defined, 22; of Jerry Maguire’s airport scene, 23, 24–25, 26; singles/close-ups of, 26–28, 273nn37,38 Sternberg, Josef von, 34, 93, 104, 236 Le Stigmate (1924), 45 Story of a Love Aªair (Cronaca di un amore, 1950), 152, 288n31 The Story of the Last Chrysanthemum (Zangiku Monogatari, 1939), 88, 97, 115, 135, 226; blockage/reveal device of, 117, 119–20; pathway repetitions in, 120–21; plot of, 117; simplicity/expressiveness of, 122, 123 The Straits of Love and Hate (Aien kyo, 1937), 115, 118, 282n26 Stranger Than Paradise (1984), 274n43 Straub, Jean-Marie, 157, 159, 184, 269 Stravinsky, Igor, 192 Street of Shame (Akasen chitai, 1956), 133, 138, 139 style. See film style Summer at Grandpa’s (Dongong de jiaqi, 1984), 204; average shot lengths of, 291n30; pointof-view shots of, 210; spinning fan image of, 208–9, 252, 257; themes of, 205 Sumo Do, Sumo Don’t (Shiko funjatta, 1992), 233 Suo Masayuki, 233 box The Suspended Step of the Stork (To meteoro vima

tou pelargo, 1991), 141, 177, 184; average shot lengths of, 159; blank space schema of, 182; planimetric imagery of, 173, 174, 175, 176; recessive images of, 169; themes of, 143 Sweet Degeneration (Fang lang, 1997), 292n55 swordplay films (chanbara), 85, 132 symbolism: academia’s focus on, 266; Angelopoulos’s reliance on, 175–77; as function of style, 34, 36 tableau framing: of L’Assassinat, 75, 76; of City of Sadness’s dinner-table scene, 187; with minimized cut-ins, 47 Taipei Story (Qingmei zhuma, 1985), 215 Taiwanese films: average shot lengths of, 290n13, 292n55; festival circuit’s reaction to, 214–15, 229–30; fixed long take of, 203, 230–31, 292n55; interior lighting in, 195; New Cinema movement of, 190, 202–3, 222 box; 1970s style/genres of, 193–94, 290n12; and 2-28 massacre, 215–16, 224–25 Taki no Shiraito (White Threads of the Waterfall, 1933), 87, 135, 282n26; blockage/reveal device of, 106, 107; camera setups in, 101; intertitles of, 285n50; plot of, 99 Tanaka Kinuyo, 89, 129 Tanizaki Jun’ichiro, 98 Tarantino, Quentin, 4, 271n7 Tarkovsky, Andrey, 155–56, 159 Tarr, Béla, 268 Tati, Jacques, 155 telephoto lens shots. See long lens technique television techniques, 30 Tell Me Something (1999), 30 temps morts (“dead time”): in Antonioni’s films, 153–54; in Tati’s films, 155 La Terra Trema (1948), 167 box The Terrorizers (Kongbu fenzi, 1986), 222 box, 223 That Day, on the Beach (Haitan shang de yitian, 1983), 202 Thessaloníki film festival, 142 13 Conversations about One Thing (2001), 267 Thompson, Kristin, 55, 81 Three Beauties (San ren ka, 1933), 101 Three Generations of Danjuro (Danjuro sandai, 1944), 116 three-quarter dorsal poses, 147, 162, 163, 164 three-quarter frontal poses, 162 Tih Minh (1919), 44, 49, 55, 67–68, 69–70 Time of Indiªerence (Moravia), 153 A Time to Live, a Time to Die (Tongnian wangshi, 1985), 188, 204, 225; average shot lengths of,

291n30; lateral staging of, 213; long lens/ wide format shots of, 206–8, 209, 228; misleading narration in, 211–12 Titanic (1997), 262 To, Johnnie, 30 Toho studio (Japan), 132 Tokyo March (Tokyo koshin kyoku, 1929), 87, 93, 99, 106, 107 Touch of Evil (1958), 29 Town of Love (Ai no machi, 1928), 105 The Travelling Players (O thiassos, 1975), 141, 142, 184; average shot lengths of, 159; diagonal approach/retreat schema of, 177–78; long shot/long takes of, 147–48, 148–49; planimetric composition of, 172, 173; recessive images of, 169; themes of, 143; three-quarter dorsality of, 163; vacant shots of, 164, 165 Treasure Island (Zhi yaowei ni huo yi tian, 1995), 230, 292n55 Trilling, James, 274n50 Truªaut, François, 11, 12, 152 Tsai Ming-liang, 230, 231 Tsivian, Yuri, 280n41 Tucker, Patrick, 273n38 Two People (Två manniskor, 1944), 150 2-28 massacre (1947), 215–16, 224–25 Uchida Tomu, 85 Udden, James, 198 Ugetsu Monogatari (1953), 89, 132, 134, 135, 136 Ulysses’ Gaze (To vlema tou odyssea, 1995), 141, 184, 294–95n15; average shot lengths of, 159; Lenin image of, 140; locations for shooting, 144; pictorial epiphany of, 178–79, 180; themes of, 143 universals: directing attention as, 262–64; narrative features as, 261–62; perceptual/ social features as, 38–39, 258–60, 275– 76n61, 297n44; substantive categories of, 264–65, 298n58; Yizek’s objections to, 260–61, 264, 298–99n59 Utamaro and His Five Women (Utamaro o meguru gonin no onna, 1946), 89, 97, 124; close-up strategies of, 128–29, 129–31, 131–32, 134; group choreography of, 127–28; knife scene of, 135, 136; plot of, 127 Les Vacances de M. Hulot (1953), 155 vacant shots: Angelopoulos’s pacing of, 164, 165, 166; Angelopoulos’s recessive diagonals of, 166, 169–72; pictorial epiphanies of, 182–84

INDEX

313

Les Vampires (1915–1916), 45, 80; blockage/reveal devices of, 58, 60, 61, 64; centering variances in, 49, 50; minimized cutting in, 47–48, 278n16; plot/characters of, 44 Vasari, Giorgio, 251, 296n24 Vendémiaire (1919), 44, 277n5 Vent d’est (1969), 157 Verfremdungseªekt (Brecht), 157 Victory of Women ( Josei no shori, 1946), 123–24, 125 Victory Song (Hisshoka, 1945), 116, 282n26 Vindicta (1923), 45 I Vinti (1953), 152, 168 box, 288n31 Violence at Noon (Hakuchu no torima, 1966), 152 The Virgin Stripped Bare by Her Bachelors (Oh, Su-jeong!, 2000), 5–7 Visconti, Luchino, 151 Vive l’amour (Aiqing wansui, 1994), 230, 292n55 Voyage to Cythera (Taxidi sta Kythira, 1983), 141, 177, 184, 198; aperture framing in, 160– 61; average shot lengths of, 159; diagonal approach/retreat schema of, 179–80, 181; themes of, 143, 145; three-quarter dorsality of, 162, 163–64 Voyage to Italy (Viaggio in Italia, 1953), 152–53 walk-and-talk staging, 29 walking scenes: of Antonioni’s films, 154; of de Sica’s Bicycle Thieves, 153; diagonal approach/ retreat schema of, 177–80, 178–81; of Jancsó’s films, 156–57 Wang Chao, 232, 292n58 Wang Tung, 202 Welles, Orson, 11, 150, 240; and Angelopoulos, 144, 146; deep-focus style of, 111 box, 167 box; Mourlet on, 12 Wenders, Wim, 146, 148 wenyi pian (romantic dramas and comedies), 192–94, 290n13 What Made Her Do It? (Nani ga kanojo o so saseta ka, 1930), 109 What Time Is It There? (Ni neibian jidian, 2001), 292n55 Where Chimneys Are Seen (Entotsu no miero basho, 1953), 133

314

INDEX

Whispering Pages (Tikhie stranitsy, 1993), 268 White Threads of the Waterfall. See Taki no Shiraito The Whole Family Works (Hataraku ikka, 1939), 94, 95, 97, 110 widescreen format: close-ups in, 27, 29, 150; Hou’s long lens shots in, 206–8, 209, 228 Wife, Be like a Rose! (Tsuma yo bara no yo ni, 1935), 86 Wilder, Billy, 150 The Wind (1928), 77 Wö›in, Heinrich, 87, 167 box Woman of Osaka (Naniwa onna, 1940), 93, 115 Women of the Night (Yoru no onnatachi, 1948), 124 Wong Kar-wai, 189 Woo, John, 30 Wood, Robin, 12 WR: Mysteries of the Organism (W.R.—Misterije organizma, 1971), 156 Wu Nian-jen, 202, 204, 231 Wyborny, Klaus, 158 Wyler, William, 11, 86, 111 box, 134, 240 yakuza (gambler) movies, 86 Yamada Isuzu, 115 Yang, Edward (Yang Dechan), 202, 206, 215, 222 box Yeames, William Frederick, Defendant and Counsel, 80, 81 Yi Yi (2000), 222 box Yoda Yoshikata, 88, 108, 128, 134 You Only Live Twice (1967), 29 yuto bungaku (pornographic writing), 109 Zemeckis, Robert, 35 Zenshinza kabuki troupe, 116 Zhang Yimou, 187–88, 216, 231 Yizek, Slavoj: on deep-focus composition, 298–99n59; on directing attention, 262– 64; on status of narrative, 261, 262; on transcultural convergences, 260, 264, 298n58 Zorba the Greek (1964), 142 Zsigmond, Vilmos, 39 Zucker, David, 41

text: display: compositor: printer and binder:

9.5/14 Scala Scala Sans Integrated Composition Systems Thomson-Shore, Inc.