The Poetics of Chinese Cinema

EAST ASIAN POPULAR CULTURE
The Poetics of
Chinese Cinema
EDITED BY GARY BETTINSON
& JAMES UDDEN
East Asian Popular Culture
Series Editors
Yasue Kuwahara
Department of Communication
Northern Kentucky University
Highland Heights, USA
John A.Lent
School of Communication and Theater
Temple University
Philadelphia, USA
This series focuses on the study of popular culture in East Asia (referring
to China, Hong Kong, Japan, Mongolia, North Korea, South Korea, and
Taiwan) in order to meet a growing interest in the subject among students
as well as scholars of various disciplines. The series examines cultural pro-
duction in East Asian countries, both individually and collectively, as its
popularity extends beyond the region. It continues the scholarly discourse
on the recent prominence of East Asian popular culture as well as the give
and take between Eastern and Western cultures.
More information about this series at

http://www.springer.com/series/14958
Gary Bettinson • James Udden
Editors
The Poetics of
Chinese Cinema
Editors
Gary Bettinson James Udden
Lancaster University Gettysburg College
Lancaster, UK Gettysburg, Pennsylvania, USA
East Asian Popular Culture

ISBN 978-1-137-56608-9 ISBN 978-1-137-55309-6 (eBook)
DOI 10.1057/978-1-137-55309-6
Library of Congress Control Number: 2016947977
© The Editor(s) (if applicable) and The Author(s) 2016

This work is subject to copyright. All rights are solely and exclusively licensed by the
Publisher, whether the whole or part of the material is concerned, specifically the rights of
translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on
microfilms or in any other physical way, and transmission or information storage and retrieval,
electronic adaptation, computer software, or by similar or dissimilar methodology now
known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information
in this book are believed to be true and accurate at the date of publication. Neither the pub-
lisher nor the authors or the editors give a warranty, express or implied, with respect to the
material contained herein or for any errors or omissions that may have been made.
Cover illustration: © Rawpixel Ltd / Getty
Printed on acid-free paper
This Palgrave Macmillan imprint is published by Springer Nature

The registered company is Nature America Inc. New York
ACKNOWLEDGMENTS
We are indebted to our editor, Shaun Vigil, for his continued guidance
and encouragement and to Michelle Smith (production editor), Erica
Buchman, Robyn Curtis, and Felicity Plester at Palgrave for their invalu-
able input at various stages of the book’s production.
Ysue Kuwahara and John A. Lent offered astute comments on the man-
uscript, for which we are grateful.
Special thanks to our excellent contributors, with whom it has been a
pleasure to work.
We are grateful for permission to reprint the following work:
Rey Chow, European Journal of Cultural Studies (Volume 17 Number 1),
pp. 16–30, copyright © 2013 by The Author. Reprinted by Permission of
SAGE Publications, Ltd. Published in the special dossier, “Looking after
Europe.”
v
CONTENTS
1 Introduction: The Poetics of Chinese Cinema 1

Gary Bettinson
2 Five Lessons from Stealth Poetics 15

David Bordwell
3 Red Poetics: The Films of the Chinese Cultural Revolution

Revolutionary Model Operas 29
Chris Berry
4 Renewal of Song Dynasty Landscape Painting Aesthetics

Combined with a Contemplative Modernism in the Early
Work of Chen Kaige 51
Peter Rist
5 Poetics of Two Springs: Fei Mu versus Tian Zhuangzhuang 79

James Udden
6 Remaking Ozu: Hou Hsiao-hsien’s Café Lumière 97

Emilie Yueh-yu Yeh
vii
viii CONTENTS
7 Hong Kong Puzzle Films: The Persistence of Tradition 119

Gary Bettinson
8 Can Poetics Break Bricks? 147

Song Hwee Lim
9 Poetics of Parapraxis and Reeducation: The Hong Kong

Cantonese Cinema in the 1950s 167
Victor Fan
10 China as Documentary: Some Basic Questions (Inspired by

Michelangelo Antonioni and Jia Zhangke) 185
Rey Chow
Bibliography 203
Index 219
CONTRIBUTORS
Chris Berry is Professor of Film Studies at King’s College London. His publica-

tions include (with Mary Farquhar) China on Screen: Cinema and Nation
(Columbia University Press and Hong Kong University Press, 2006) and Island on
the Edge: Taiwan New Cinema and After (Hong Kong University Press, 2005).
He is co-editor of Chinese Films in Focus: 25 New Takes (BFI, 2003) and Chinese
Films in Focus II (BFI, 2008). His latest book is the co-edited anthology Public
Space, Media Space (2013) from Palgrave Macmillan.
Gary Bettinson is Senior Lecturer in Film Studies and Director of Film at
Lancaster University. He is the author of The Sensuous Cinema of Wong Kar-wai:
Film Poetics and the Aesthetic of Disturbance (Hong Kong University Press, 2014)
and editor of the Directory of World Cinema: China, volume 1 (2012) and 2
(2015). His publications on Chinese-language cinema have appeared in journals
such as Jump Cut, Post Script, and the Journal of Chinese Cinemas. He is chief edi-
tor of Asian Cinema.
David Bordwell professor emeritus of Film Studies at the University of Wisconsin-
Madison, is the author of several books about the art and history of cinema.
Among those titles is Planet Hong Kong: Popular Cinema and the Art of
Entertainment (2000, 2010). With Kristin Thompson, he has written two text-
books: Film Art: An Introduction and Film History: An Introduction. He has won
a Chancellor's Distinguished Teaching Award. He and Thompson write regularly
about cinema on their blog, Observations on Film Art (http://www.davidbord-
well.net/blog).
Rey Chow is Anne Firor Scott Professor of Literature and the current Director
of the Program in Literature, Duke University, USA. The books she authored in
the past decade include The Age of the World Target (2006), Sentimental
ix
x CONTRIBUTORS
Fabulations, Contemporary Chinese Films (2007), Entanglements: Transmedial

Thinking about Capture (2012), and Not Like a Native Speaker: On Languaging
as a Postcolonial Experience (2014). Widely anthologized, her writings have
appeared in more than ten languages. The Rey Chow Reader, ed. Paul Bowman,
was published in 2010.
Victor Fan is lecturer in the Department of Film Studies, King’s College London
and Film Consultant of the Chinese Visual Festival. His articles have appeared in
journals including Camera Obscura, Journal of Chinese Cinemas, Screen, and Film
History: An International Journal. His book Cinema Approaching Reality:
Locating Chinese Film Theory was published in 2015 by the University of Minnesota
Press. His film The Well was an official selection of the São Paolo International
Film Festival; it was also screened at the Anthology Film Archives, the Japan
Society, and the George Eastman House.
Song Hwee LIM is associate professor in the Department of Cultural and Religious
Studies, The Chinese University of Hong Kong. He is the author of Celluloid
Comrades: Representations of Male Homosexuality in Contemporary Chinese
Cinemas and Tsai Ming-liang and a Cinema of Slowness. The founding editor of
the Journal of Chinese Cinemas, he is also co-editor of Remapping World Cinema:
Identity, Culture and Politics in Film and The Chinese Cinema Book.
Peter H. Rist is a Professor of Film Studies at Concordia University in Montréal,
Canada. His major research initiative is the history (or histories) of film style, with
his PhD thesis, completed in 1988 at New York University being written on style
in the early films of John Ford. He has since specialized in non-US and European
cinema, with book chapters published on Sub-Saharan African, Chinese, Taiwanese,
Hong Kong, Japanese, and Korean films/filmmakers. His second edited book was
on Canadian Cinema(s), and his third book, the 800-page Historical Dictionary of
South American Cinema, was published in 2014.
James Udden is Associate Professor of Cinema and Media Studies at Gettysburg
College. He has published the first book-length monograph on Hou Hsiao-hsien
entitled, No Man an Island: The Cinema of Hou Hsiao-hsien (Hong Kong UP,
2009), which has recently been translated into Chinese by Fu Dan University
Press. He has published extensively on Asian cinema in several journals and anthol-
ogies and is currently finishing up a book project on the parallel rise of Iran and
Taiwan in the global network of international film festivals.
Emilie Yueh-yu Yeh is Professor and Director of the Academy of Film at Hong
Kong Baptist University. Emilie Yeh’s English works have been translated to
Japanese, Hungarian, Spanish, and Chinese. Her publications include: Rethinking
Chinese Film Industry: New Methods, New Histories (Beijing University Press, 2011),
CONTRIBUTORS xi
East Asian Screen Industries (with Darrell Davis, British Film Institute, 2008),
Taiwan Film Directors: A Treasure Island (with Darrell Davis, Columbia University
Press, 2005), Chinese-Language Film: Historiography, Poetics, Politics (with
Sheldon Lu, University of Hawaii Press, Choice's 2005 outstanding academic
title), and Phantom Of The Music: Song Narration And Chinese-Language Cinema
(Taipei: Yuan-liou, 2000).
LIST OF FIGURES
Fig. 4.1 Ma Yüan’s A Mountain Path in Spring 54

Fig. 6.1 Yoko and Hajime meet up inside the train, unexpectedly 106
Fig. 6.2 Michiko greets auntie Akiko (Tokyo Twilight , 1957) 110
Fig. 6.3 Akiko buries her sobs in her hands (Tokyo Twilight , 1957) 110
Fig. 6.4 Madame Koh shows Yoko photos of Jiang Wenye (aka Koh
Bunya 1910–1983) 116
Fig. 7.1 Wu Xia: Detective Xu (Takeshi Kaneshiro, left) imaginatively
‘witnesses’ the paper-mill skirmish 125
Fig. 7.2 Spindly branches—or perhaps internal chi energy—rescue
Jinxi (Donnie Yen) from certain death in Wu Xia 127
Fig. 7.3 Blind Detective: a harsh lighting scheme and distinctive
color palette denote subjective action as Johnston
(Andy Lau) investigates a crime 137
Fig. 8.1 Zhang Ziyi holds her posture in stillness while virtual beans
circle her like orbiting comets in House of Flying Daggers.
Copyright Beijing New Picture Film Co., China Film
Co-Production Corporation, Edko Films, Elite Group
Enterprises, Zhang Yimou Studio 155
Fig. 8.2 Close-up of Jet Li’s face in profile as he breaks through six
strings of raindrops while charging toward his opponent in
Hero. Copyright Beijing New Picture Film Co., China Film
Co-Production Corporation, Elite Group Enterprises,
Sil-Metropole Organisation, Zhang Yimou Studio 157
Fig. 8.3 A nut that bolts a metal strip on the pillar shakes but remains
intact while tiny snowflakes splash from above the strip in The
Grandmaster. Copyright Block 2 Pictures, Jet Tone Films,
Sil-Metropole Organisation, Bona International Film Group 160
xiii
xiv LIST OF FIGURES
Fig. 8.4 A screw loosens alongside its bracket, with the screw then
jumping halfway out of the hole before descending back into
it in The Grandmaster. Copyright Block 2 Pictures, Jet Tone
Films, Sil-Metropole Organisation, Bona International
Film Group 161
Fig. 9.1 Verticalisation of the social dialectics in In the Face of
Demolition 179
Fig. 9.2 Development of in-group solidarity through gradually
packing dialectically conflicting characters into a single
long shot 180
LIST OF TABLES
Table 3.1 Hong Changqing discovers Wu Qinghua in Red

Detachment of Women 42
Table 3.2 Mother Du frees Lei Gang in Azalea Mountain 46
xv
CHAPTER 1
Introduction: The Poetics of Chinese

Cinema
Gary Bettinson
From Russian Formalism and Prague Structuralism to neoformalism and

cognitivism, the poetics approach to cinema has crucially advanced the
study of popular film—and yet poetics has occupied a relatively marginal
place in the study of Chinese-language cinema. Since the 1980s, Cultural
Studies perspectives have dominated the field, and the art of Chinese cin-
ema has fallen afoul of critical neglect. It is precisely the artistic dimension
of movies that poetics—the major research program to which this book
subscribes—seeks to illuminate. As schematized by David Bordwell,1 a
poetics of cinema encompasses analytical inquiry (examining a film’s visual
and aural style, narrative construction, and thematic expression), historical
explication (tracing patterns of artistic continuity and change over time),
and spectatorial theorizing (offering an account of the interface between a
film’s compositional features and the viewer’s activity). This program has
been fruitfully adopted in recent years by scholars of Hollywood cinema,2
in ways that enrich our understanding of Hollywood’s stylistic traditions.
If we are to achieve a comparable grasp of Chinese cinema—a category
encompassing the cinemas of Taiwan, Hong Kong, and the People’s
Republic of China (PRC)—then we need to usher in a poetics of Chinese-
language cinema. Such is the purview of this book.
G. Bettinson ()
Lancaster University, Lancaster, UK
© The Editor(s) (if applicable) and The Author(s) 2016 1

G. Bettinson, J. Udden (eds.), The Poetics of Chinese Cinema,
DOI 10.1057/978-1-137-55309-6_1
2 G. BETTINSON
Why is a poetics of Chinese cinema important? For critics of a cul-

turalist stripe, Chinese cinema gains much of its interest by reference to
social upheavals (e.g. the Cultural Revolution [PRC], the 1997 hando-
ver [Hong Kong], or the lifting of martial law [Taiwan]). Chinese films
acquire salience for the culturalist by embodying such cultural landmarks.
But by analyzing the films in “top-down” fashion, the culturalist subor-
dinates a film’s aesthetic qualities to an a priori conceptual scheme; thus
the film’s stylistic construction is of interest only insofar as it reflects or
embodies sociological meaning. Poetics inverts this critical emphasis, such
that the poetician examines the artwork from the “bottom up”—hence,
the critic’s point of departure is not a broad sociological premise but “the
principled regularities of form and style we can find in the films” (Bordwell
2001: 9). If existing scholarship on Chinese cinema has overwhelmingly
prioritized aspects of culture and society, the poetics approach enables us
to put the films themselves at center stage.
This is not to disdain culturalist research, or to deny that culture and
society shape filmic construction in important ways. Indeed, culturalism
and poetics are not mutually exclusive paradigms; it is feasible, for instance,
that a film’s formal design can be usefully elucidated by reference to the
social milieu from which it springs. But a film’s compositional features are
not wholly determined by cultural factors. Other kinds of factors—tech-
nological, industrial, economic, artistic—may be at least as important as
social cataclysms in shaping the finished work, and a poetics of Chinese
cinema can bring these factors to light. Proceeding from the bottom up,
the poetician examines the film’s formal and stylistic patterns, and then
asks “what real-world activities could plausibly play causal roles in creating
them” (Bordwell 2001 : 9). Without dismissing cultural concerns, poet-
ics puts formal analysis at the heart of inquiry. As such, it shifts the field
of Chinese cinema studies toward fundamental yet hitherto neglected or
marginalized areas of research. Moreover, it provides new insights that are
compatible with already existing studies of Chinese cinema.
At the same time, however, a poetics approach can redress many of the
fallacies and misconceptions in the literature. One enduring fallacy is the
essentialist notion of a distinctively “Chinese” film style, typically charac-
terized by extended takes, distanced framings, and an alternatively sump-
tuous or austere emphasis on natural landscape. This characterization, I
surmise, is chiefly informed by the Fifth Generation films of Mainland
China, whose international profile in the 1980s and 1990s greatly shaped
Western perceptions of Chinese-language cinema. Yet the notion of a
INTRODUCTION: THE POETICS OF CHINESE CINEMA 3
quintessentially Chinese film language falters on several fronts. For one

thing, it fails to distinguish among the cinemas of the three Chinas, each
of which fostered quite distinct aesthetic programs (e.g. meditative edit-
ing is hardly a hallmark of Hong Kong’s popular cinema). Nor does it
acknowledge the stylistic pluralism within each of the cinemas: a Mainland
industry that produces both the chintzy Tiny Times 3.0 and the formally
elliptical Black Coal, Thin Ice (both 2014) thwarts assumptions of a mono-
lithic Chinese film style. The poetics perspective compels us to treat skepti-
cally such univocal notions of national style and enables us to lay bare, by
means of formal analysis, the aesthetic eclecticism of Chinese cinema.
One might counter that the increasing integration of China’s movie
industries makes prospects for a dominant national style tenable. Whereas
the three cinemas had once largely developed on separate tracks, the
Mainland’s economic rise has borne witness to a surge in pan-Chinese
collaborations and Asian talent migrating across Chinese borders. As
pan-Chinese coproductions multiply, we might ask: are the cinemas of
Taiwan, Hong Kong, and Mainland China still distinguishable as sepa-
rate entities? Perhaps a homogenous Chinese aesthetic emerges in this
burgeoning joint venture trend, a trend whereby PRC, Hong Kong, and
Taiwanese filmmaking coalesces. Yet to argue that these Chinese copro-
ductions evince a national style, one would need to show that they are sty-
listically of a piece, and this is no straightforward task. The Grandmaster
(HK-China, 2013), The Rooftop (Taiwan-China, 2013), Love in the Buff
(HK-China, 2012), Ip Man (HK-China, 2008), Red Cliff (China-HK-
Taiwan, 2008), Kung Fu Hustle (HK-China, 2004)—which of these
coproductions exemplifies Chinese film style? One might reply that they
all embody a Chinese film aesthetic, but, given the aesthetic diversity
on display even in this small sample of films, the concept of a “Chinese
national style” becomes baggy, imprecise, and uninformative. Still, a
poetics of cinema can bring to light recurring norms shared by some or
most of these films; it can determine the extent to which those norms are
culturally unique; and it can seek causal explanations for these norms, for
instance, by tracing their repetition to practices standardized within the
Chinese coproduction system.
The perspective of poetics lets us amend another essentialist fallacy,
often tacit in the literature and much discussed in the filmmaking com-
munity. This fallacy holds that Chinese storytelling—its norms of nar-
rative plotting, its schemas of visual narration, and, fundamentally, the
kinds of stories it elects to tell—does not communicate cross-culturally
4 G. BETTINSON
to mass audiences in Western territories. (Hence, the efforts by Western

distributors, such as The Weinstein Company, to render Chinese imports
“accessible” by means of extensive reediting, expository intertitling, and
other simplifying strategies.) Yet this view neglects salient counterex-
amples, including House of Flying Daggers (2004), Hero (2002), Infernal
Affairs (2002), and Crouching Tiger, Hidden Dragon (2000), all
of which found commercial success in the West. Worse, it recycles an
Orientalist stereotype of Chinese opacity and inscrutability. If we are
to demonstrate the cross-cultural intelligibility of Chinese cinema—and
here we alight on the terrain of spectatorial poetics—then we would
do well to undertake what Bordwell (2001) calls a “transcultural” com-
parison of Chinese and Western storytelling strategies. We might, for
instance, identify transculturally shared stylistic patterns in a Chinese and
a Hollywood movie of the same genre, the better to isolate those textual
schemas familiar to and comprehended by culturally diverse audiences.
In such ways, the poetician can qualify (or disqualify) the cultural essen-
tialism that underlies widely held assumptions about Chinese storytelling
and spectatorship.
Poetics can highlight aspects of Chinese cinema neglected in the litera-
ture. Slighted by the prevailing sociological hermeneutics is the precise
nature of Chinese film practice, the variety of craft practices within and
among the three Chinas, and the ways in which standardized work rou-
tines shape the Chinese film’s style and form. Under what production
circumstances are Chinese films typically made? What institutional and
economic constraints shape the finished work? What are the characteristic
modes of production, and how have they changed over time? These are
phenomena about which culturalism has had little to say, but a historical
poetics of Chinese cinema can posit, at a broad level, both the institu-
tional factors governing Chinese film production and the systematic craft
techniques and traditions that underpin Chinese film style. These broad
principles, in turn, constitute a ground of conventions against which the
exceptional or maverick case—for instance, the aleatory work habits of
Wong Kar-wai—stand out as legitimately distinctive.
Researching habitual practices and institutional norms may also enhance
our understanding of different modes of production. The Sino-US
coproduction model, for instance, has intensified in recent years, but pre-
cisely how this model is constituted remains opaque. (Indeed, there is no
uniform coproduction strategy, but several available partnership options.)
As the North American film market shrinks and the Mainland market
blossoms, American studios court PRC producers for collaborations and

a greater share of the foreign film quota. Consequently, a flurry of official
Sino-US movies has emerged in recent years, and altered the landscape of
Chinese film production—titles include The Karate Kid (2010), Looper
(2012), Iron Man 3 (2013), Transformers: Age of Extinction (2014),
Furious 7 (2015), Crouching Tiger, Hidden Dragon: The Green Legend
(2016), Kung Fu Panda 3 (2016), and The Great Wall (2016). By con-
trasting Mainland Chinese and Hollywood institutional norms and work-
ing situations, and by examining the formal features of Sino-US films, a
poetics approach is best placed to account for the synergies, as well as the
points of tension or incoherence, characterizing both the various copro-
duction systems and the films they beget. Most generally, the formal prop-
erties of a Chinese-language film—no matter its mode of production—can
be causally explicated, at least in part, by the institutional, economic, and
practical specificities of its production.
Culturalists prioritize social reflection exegesis, but there may be
other influences bearing on the work besides social ones. Moreover,
these influences may inform the film’s aesthetic more directly than, say,
the Tiananmen massacre, the lifting of martial law, or the severe acute
respiratory syndrome (SARS) crisis. A poetics of Chinese cinema—by
operating historically, comparatively, and from the bottom up—can
expose the pertinent sources from which the work draws, explicating
intertextual precursors (e.g. the preexisting filmic styles to which a par-
ticular Chinese film is indebted) or cross-media ones (e.g. the influence
of landscape painting on Chinese filmmakers’ pictorial design). The
transcultural dimension of poetics, meanwhile, can shed light on inter-
national as well as pan-Asian influences (e.g. Bordwell [2001] proposes
that Chinese-language cinemas adopted Hollywood’s continuity system
as a stylistic point of departure). The poetician’s standard set of heuristic
devices—the tool of average shot length (ASL), the Formalist concepts
of norms, deviations, and backgrounds—is apt to reveal the intercul-
tural flow of cinematic influence and innovation. Further, transcultural
analysis acquires additional importance in light of Hollywood’s appro-
priation of Chinese movies. American remakes such as The Departed
(2006), The Eye (2008), and Tortilla Soup (2001) invite stylistic com-
parison with their Chinese-language sources (respectively, Infernal
Affairs; The Eye, 2002; and Eat Drink Man Woman, 1994). Likewise,
Chinese remakes of American movies—A Woman, a Gun, and a Noodle
Shop, 2009 (Blood Simple, 1984), Connected, 2008 (Cellular, 2004),
6 G. BETTINSON
What Women Want (2011/2000), Bride Wars (2015/2009)—demand

comparative analysis. As the remake trend flourishes, the poetician is
provided neat opportunities to discover not only patterns of innovation
and indebtedness but also—contrary to the cultural essentialist position
limned above—stylistic and narrative schemas that are readily grasped
across cultures.
The value of poetics for the study of Chinese cinema obtains, too, in
this research program’s historical dimension. Though the existing litera-
ture contains historical discoveries of enduring import, the heritage of
Mainland, Hong Kong, and Taiwanese cinemas is far from fully chronicled.
The limited availability of certain Chinese films, some of which languish
in archives, or lack subtitle tracks, has stymied the progress of Western
historians. Many early Chinese-language films have not survived; some
that existed may never have been documented. Still today, it is a matter of
debate as to when Hong Kong filmmaking began. A historical poetics of
Chinese cinema won’t resurrect lost films, but its formalist emphasis on
historical backgrounds—one aspect of which involves viewing particular
films against other related films—encourages scholars to go beyond the
canon, examine less familiar artworks, and thereby “fill in” historical lacu-
nas in the literature. Not that the poetician’s task is simply to spotlight
neglected or forgotten movies. Rather, by charting the development of
stylistic norms over time, the poetician can reveal patterns of continuity
and change, identify innovations, and mount historical comparisons with
other national cinemas (as well as among those of the three Chinas). In
such ways, the poetics program makes an important contribution to the
historiography of Chinese film.
Perhaps above all, poetics brings us to a clearer understanding and
appreciation of the art of Chinese cinema. By placing questions of form
and style at the center of inquiry, poetics undercuts the culturalist assump-
tion that a film’s interest inheres chiefly in its manifestation of social anxi-
eties and crises. The poetics approach allows us to contextualize Chinese
cinema’s relation to international film style, laying bare those transcultural
artistic conventions on which popular storytelling and cross-cultural com-
prehension rely. It also allows us to recognize and celebrate the originality
of Chinese filmmaking. A poetics can facilitate fine-grained distinctions
among the three Chinese cinemas; it can foreground stylistic novelty; and
it can shine a light on boundary-pushing films and filmmakers. Not least,
it reminds us that Chinese cinema has fostered artistic traditions to match
any on the international stage.
POETICS AND PRECEDENTS
The Poetics of Chinese Cinema is the first book to treat Taiwanese,
Mainland Chinese, and Hong Kong filmmaking from a poetics perspec-
tive. Nevertheless, it builds on a small but seminal body of work. No
scholar has contributed more to a poetics of Chinese-language cinema than
David Bordwell—indeed, he mapped its terrain. Across a host of publica-
tions, Bordwell has mounted historical and theoretical analyses of filmic
construction within each of the three Chinese cinemas. He has explored
King Hu’s “aesthetic of the glimpse” (2002), tracing Hu’s abbreviated
combat scenes to pertinent stylistic traditions and revealing the ingenuity
with which the director recasts inherited schemas. He has scrutinized the
nuanced staging strategies of Hou Hsiao-hsien, and the laconic gunplay
sequences of Johnnie To (2005, 2003). He has examined the house style
of the Shaw Brothers film studio and compared the action genre traditions
of Hollywood and Hong Kong (2009, 2001a). And he has provided blog
commentaries on influential figures such as Ann Hui, Li Han-hsiang, Jia
Zhangke, Fei Mu, Wong Kar-wai, and Tsai Ming-liang.3 This body of lit-
erature pursues and exemplifies, in various ways, Bordwell’s transcultural
poetics of Chinese cinema.4
Most extensively, Bordwell has furnished a historical poetics of
Hong Kong film. In Planet Hong Kong: Popular Cinema and the Art of
Entertainment (2000, 2011), he posits a tradition of popular filmmaking
typified by a set of (more or less stable) institutional practices, generic
conventions, and norms of story and style. At various levels of general-
ity, he details the local, regional, and international contexts for Hong
Kong film production and consumption; the modes of film practice and
the customary craft habits adopted by local filmmakers; the indigenous
“norms of genre, stars, stories, and style” (17); and the ways that these
general forces impinge on the films themselves. He alights on striking
cases, dwelling on notable films (Chungking Express, 1994), directors
(John Woo, Tsui Hark), and stars (Bruce Lee). And he crystallizes a set of
tendencies peculiar to Hong Kong films, such as episodic plotting, tonal
ruptures, pictorial legibility, postsynchronized sound, and sentimentality.
At the same time, the Hollywood continuity style serves as a ground of
(transcultural) comparison against which the popular Hong Kong movie
stands out in relief. Bordwell’s enterprise also harbors an empirical dimen-
sion: the book’s theses are buttressed by primary interviews with Hong
Kong personnel working at all levels of the industry. By investigating how
8 G. BETTINSON
Hong Kong films are designed, Bordwell ventures into the territory of
actual film production—territory seldom approached by cultural theorists.
As such, he provides fresh knowledge and opens up new areas of investiga-
tion, unavailable (or at least untapped) in the culturalist program.
In recent years, other scholars have pursued a poetics of Chinese-
language cinema. James Udden’s No Man an Island: The Cinema of Hou
Hsiao-hsien (2009) charts Hou’s unlikely rise from a moderately popular
director in Taiwan’s fading commercial industry to one of the most ven-
erated auteurs on the international festival circuit. Udden traces Hou’s
stylistic and thematic tendencies, now crystallized as authorial traits, to a
peculiar confluence of historical factors in 1980s Taiwan. He also critiques
the widespread ascription of quintessential “Chineseness” to Hou’s aes-
thetic style, disparaging such views as uncritical and politically problematic.
Similarly, in The Sensuous Cinema of Wong Kar-wai: Film Poetics and the
Aesthetic of Disturbance (2014), I take issue with the dominant approach
to Wong Kar-wai’s oeuvre—in this case, culturalist criticism—and mount
a poetics analysis of Wong’s sumptuous yet challenging audio–visual style.
This analysis demonstrates that an aesthetic of sensuousness and “distur-
bance” permeates every dimension of Wong’s films, from plotting and
characterization to narrational strategy and genre engagement. Emilie Yeh,
meanwhile, has investigated both Hou and Wong from the perspective of
poetics, sketching the narrative and visual tendencies of the former and
the musical practice of the latter (Yeh 2005, 2008). Elsewhere, scholars
have gestured toward a poetics of performance.5 Mette Hjort (2010), for
instance, tracks the performative behavior of Mainland star Ruan Lingyu
throughout several scenes from The Goddess (1934), effectively demon-
strating how Ruan’s studied activity works in concert with filmic param-
eters of editing and cinematography to create meaning and elicit emotion.
Despite these interventions, however, the research program of poet-
ics—as brought to bear on Chinese-language cinema—is in its infant
stage. Few scholars explicitly mount a poetics of Chinese film. The pres-
ent volume, then, is intended as a step further in the development of this
research tradition, as well as a fresh perspective within the field of Chinese
cinema studies.
AVENUES OF INVESTIGATION
The chapters in this anthology demonstrate the coherence of the poetics
program, but they also suggest the variety of directions that a poetics of
Chinese cinema might take. Though the chapters are unified by historically
and theoretically informed analyses of particular films, genres, and oeu-

vres, they exemplify the multiple lines of inquiry available to the poetician.
Chinese-language cinema, partly by virtue of its festival exposure, is an
auteurist cinema, and several of the book’s chapters cast apposite empha-
sis on the stylistic traits of particular filmmakers (Hou Hsiao-hsien, Jia
Zhangke, Chen Kaige, Johnnie To, Wong Kar-wai, Fei Mu, and others).
At a more general level, the chapters canvass a range of genres and modes
of filmmaking, from melodrama and wuxia (swordplay) to detective fiction
and documentary. Still more broadly, several of the chapters consider the
relationship between the stylistic construction of particular films and the
industrial, economic, and sociohistorical forces bearing on that construc-
tion. The book’s chapters do not, of course, exhaust the possibilities of the
poetics program. But it is hoped that they demonstrate not only the virtues
of this research approach but also the intrinsic value of Chinese filmmaking
as an enduringly rewarding object of study.
As David Bordwell points out in the following chapter, the concepts of
poetics are not rigid or inflexible, and his own model of historical poetics
is but one framework among numerous possible alternatives. Bordwell
rehearses his poetics program before canvassing some ways in which it has
been refined and enriched by the study of Chinese-language cinema. He
makes his case by alighting on particular “middle-level” inquiries: the epi-
sodic tendencies of Hong Kong film as a potent creative alternative to clas-
sical Hollywood dramaturgy; the techniques of editing employed in Hong
Kong action cinema, their debts to other national styles, and their revival
and recasting of already existing norms; the confluence of factors shaping
the filmic style of Taiwanese directors such as Hou Hsiao-hsien; the pat-
terns of stylistic influence traversing Asian regional and national borders;
and the importance of individuals working within institutions, engaging
with norms and conventions, and cultivating personal styles. By means of
these research inquiries, Bordwell demonstrates how poetics both illumi-
nates and is illuminated by the films of the three Chinas.
Chris Berry explores the 1970s filmed versions of the Cultural
Revolution “model works,” redressing the critical neglect of these
highly politicized and affectively charged films. Focusing on the revo-
lutionary operas Taking Tiger Mountain by Strategy (1970) and Azalea
Mountain (1974), and the revolutionary ballet The Red Detachment
of Women (1971), Berry debunks the perception of the “model” films
as quintessentially Chinese, instead identifying within the films a syn-
thesis of indigenous and foreign elements, and of traditional and mod-
ern features. Not that the revolutionary model works were stylistically
10 G. BETTINSON
inseparable; as Berry demonstrates, the films shared certain aesthetic

principles, but Azalea Mountain employs a faster cutting rate, closer
framing, and more “immersive” strategies than its model predeces-
sors, the better to engage its viewer both politically and affectively.
Nevertheless, through a poetics analysis of the three films, Berry reveals
that each obeys the so-called Theory of Three Prominences, a politi-
cal theory governing cultural products during the Cultural Revolution
and which, Berry attests, informed the aesthetic design of the Cultural
Revolution model works in crucial ways.
Several of the chapters in the anthology converge on tradition as a
cornerstone of poetics. Traditions of various sorts serve a cardinal func-
tion in poetics analysis, for they provide a broad ground of norms and
conventions against which historical developments and deviations can be
weighed. Stylistic continuity and change come forward by reference to
pertinent traditions and backgrounds. In a chapter examining the influ-
ence of traditional Chinese painting on Chen Kaige’s early films, Peter Rist
evokes the paradoxical impulses of the young artist compelled, simultane-
ously, to emulate and transform established forms and practices. Rist sur-
veys the history of Chinese landscape and scroll painting; delineates this
tradition’s stylistic techniques; and limns the limited yet significant influ-
ence of classical landscape painting on the formative Chinese filmmakers,
such as Fei Mu. Not until the arrival of the Chinese Fifth Generation,
Rist argues, did Chinese cinema embrace—albeit in political and experi-
mental ways—the heritage of landscape painting. Rist perceives in Chen
Kaige’s Yellow Earth (1984) and King of the Children (1987) an urge to
innovate, manifested in Chen’s self-conscious (“modernist”) awareness of
the artistic traditions infusing the films. Rist demonstrates Chen’s debts
to—and deviations from—Chinese landscape painting by means of com-
parative formal analysis, buttressed by statistical breakdowns of the films’
ASLs and shot scales.
Precisely what is meant by a “traditional” Chinese film style anchors
James Udden’s comparative analysis of Fei Mu’s Spring in a Small Town
(1948) and its 2002 remake directed by Tian Zhuangzhuang. Udden con-
trasts the attitudes toward tradition articulated within the films themselves,
juxtaposing their advocacy of Confucian moral values. He also contends
that the thematic conservatism of Fei’s film belies the radicalism of its style,
the film’s “proto-modernist” aesthetic going largely unnoticed upon the
film’s release. Tian’s remake, however, fails to replicate Fei’s innovations
of staging, editing, and voice-over narration, instead virtually pastiching
an imagined traditional Chinese discourse. Among Tian’s putatively tradi-

tional resources is the long take device, which Udden notes is frequently
(mis)taken for a specifically Chinese trait. The chapter explores this and
other visual and aural strategies deployed by Fei, disclosing their radicalism
within 1940s world cinema, and arguing that they anticipate experimental
strategies (by the likes of Robert Bresson, Michelangelo Antonioni, and
Jean-Pierre Melville) that would come to define the modernist art cinema
of the 1950s and beyond.
The weight of tradition frames Emilie Yeh’s examination of Hou’s Café
Lumière (2003). Commissioned as a memorial to Yasujiro Ozu, Café
Lumière presented Hou with a unique artistic puzzle: how to mesh his
own authorial signature with that of the Japanese master? Here the artist
finds himself conscious of, and potentially overshadowed by, a venerable
forbear. However, Yeh suggests that Hou foregoes slavish imitation of
Ozuian stylistics, going so far as to subvert certain aspects of Ozu’s work
in order to differentiate himself. Hou’s stylistic departures from Ozu, Yeh
argues, are more instructive than the affinities; moreover, while many crit-
ics compare Café Lumière to Tokyo Story (1953), Yeh identifies a more
revealing—if lesser known—intertext in the Ozu canon: Tokyo Twilight
(1957). As Yeh’s comparative analysis demonstrates, Hou’s “rewriting”
of Tokyo Twilight offers a reflection not only on Hou’s stylistic proximity
to Ozu but also on Sino-Japanese cultural politics, ultimately pointing the
way to a new rapprochement between Taiwan and Japan.
I examine the fate of traditional modes of practice, as well as of local
storytelling norms, in my chapter on contemporary Hong Kong film-
making. I contest some widely held yet specious premises: first, that
the “Mainlandization” and “Hollywoodization” of Hong Kong cinema
eradicate local filmmaking practices and aesthetic norms; and second,
that the local routine of piecemeal script construction yields slapdash
plotting, and thus is inferior to the screenplay practices advocated in
Mainland China and Hollywood. I argue that not only have local work
routines endured in spite of institutional change, but that those practices
yield films of considerable complexity and ambition. In addition, I assim-
ilate several of these films to what I identify as a nascent “puzzle film”
trend in Hong Kong cinema. Disputing claims of a “post-Hong Kong
cinema,” my chapter draws on primary interviews with key participants
in detailing both the PRC coproduction system and the characteristic
script practices employed by Peter Chan, Johnnie To, and the Milkyway
Image studio.
12 G. BETTINSON
The poetics of digitally produced spectacle forms the basis of Song Lim’s
chapter. Focusing centrally on transnational martial arts ventures such
as The Grandmaster, Hero, and Crouching Tiger, Hidden Dragon, Lim
postulates a poetics of slowness—made possible by computer-generated
technology—that demotes the genre’s traditional emphasis on corporeal
action and stakes a claim for the cultural prestige conferred upon “slow
cinema” in contemporary global filmmaking. Lim contrasts the landmark
kung fu films of Bruce Lee with their latter-day effects-laden counterparts;
and he counterposes the genre’s traditionally “epic” scale with a “poetics
of smallness,” in which tiny objects (raindrops, beans) acquire a sensuous
and defamiliarizing force. Lim goes on to consider René Viénet’s situa-
tionist exercise Can Dialectics Break Bricks (1973) in order to indicate the
political potential of Chinese spectacle cinema; and he suggests that films
of this genre virtually efface the contributions of a specialized labor force,
subsuming the technical crew’s achievements to the auteur-poetician (Ang
Lee, Zhang Yimou, et al). For Lim, the films’ digitally upholstered action
scenes constitute nothing less than a new category of spectacle and incul-
cate a new kind of spectatorship characterized, primarily, by sensual plea-
sure. The genre’s traditional stress on speed and epic scale gives way to an
aesthetic of slowness and smallness.
Victor Fan contrasts the stylistic traits of 1950s Cantonese cinema
against the classical Hollywood style, considering whether Hong Kong
directors such as Lee Tit consciously recast the classic continuity system.
Fan begins by tracing the lineage of 1950s Cantonese cinema to the cen-
turies-old Cantonese Opera tradition, noting their shared dramatic and
narrational tendencies. By the 1950s, he suggests, classical Hollywood
norms had importantly modified Cantonese film dramaturgy; and yet cer-
tain Hong Kong films of this era deviated from the American style, to
particular aesthetic and political effect. Moreover, Fan argues, this nar-
rational deviation was actively desired by the Cantonese-speaking audi-
ence—but why? Fan investigates these matters through close formal
analysis of Lee Tit’s In the Face of Demolition (1953), a product of the
left-wing Hong Kong studio Union Film Enterprise. What emerges from
films of this ilk and era is what Fan calls a classical Cantonese style, distinct
from its Hollywood counterpart, which addresses the local audience’s
failed efforts at political agency, and provides recuperative narratives of
sociopolitical change. Such social ideals are dramatized not only through
narrative development but also, Fan demonstrates, through purposive
strategies of staging and editing.
In the final chapter, Rey Chow probes the aesthetics of the “real” in
documentary representations of China. Taking as a point of departure
Michelangelo Antonioni’s ethnographic work Chung Kuo/Cina (1972),
Chow gets to the heart of this film’s controversy by counterpointing dif-
ferent cultural attitudes toward the function and propriety of photogra-
phy itself. She draws on the work of Susan Sontag, Pierre Bourdieu, and
Roland Barthes to reveal the ways in which Antonioni’s well-intentioned
film inadvertently affronts its Chinese subjects and viewers; she investi-
gates the implications of a formal disparity between Antonioni’s objecti-
fying images and subjective voice-over narration; and she suggests that,
despite intentions to the contrary, Antonioni exoticizes Chinese culture.
Chow goes on to examine the status of documentary realism in the work
of Jia Zhangke, whose provocative uptake of the documentary mode in
films such as I Wish I Knew (2010) and 24 City (2008) muddies the dis-
tinction between fact and fiction. The poetics of “the real” in traditional
documentary, Chow contends, gives way in Jia’s aesthetic to “a new kind
of conceptual project,” one that envisions China from a wholly distinctive
perspective.
If the chapters in this anthology take up diverse lines of inquiry,
they all nonetheless epitomize what Bordwell (1995) calls middle-level
research; that is, they pursue modest theoretical and empirical objectives,
eschewing the Grand Theory that has often dominated the research field.
By espousing and undertaking a poetics of Chinese cinema, this book
endeavors to advance our knowledge of Chinese film aesthetics. Not least,
it seeks to affirm the artistry of particular Chinese movies, of the filmmak-
ers that made them, and of the polystylistic tradition of Chinese cinema
itself.
NOTES
1. Bordwell outlines his poetics model in Poetics of Cinema (2008), 11–56;
“Transcultural Spaces: Toward a Poetics of Chinese Film” (2001);
“Historical Poetics of Cinema” (1989); Making Meaning: Inference and
Rhetoric in the Interpretation of Cinema (Cambridge, MA; London: Harvard
University Press, 1989), 263–274.
2. See, for example, Todd Berliner, Hollywood Incoherent: Narration in
Seventies Cinema (2010); and Warren Buckland, Directed by Steven Spielberg:
Poetics of the Contemporary Hollywood Blockbuster (2006).
3. See David Bordwell’s Website on Cinema: Observations on Film Art. http://
www.davidbordwell.net/blog/
14 G. BETTINSON
4. Bordwell sets forth this research program in “Transcultural Spaces” (2001),

an essay reprinted and expanded in Sheldon Lu and Emilie Yeh (eds) (2005)
Chinese-Language Film: Historiography, Poetics, Politics, Honolulu:
University of Hawai‘i Press,141–162.
5. See Hjort (2010) and Bettinson (2015).
CHAPTER 2
Five Lessons from Stealth Poetics
David Bordwell
I can only be happy that poetics, as a perspective on studying films, has

attracted some vigorous practitioners and advocates. I wasn’t looking for
fellow travelers when I began exploring these ideas in the early 1980s. I
was just trying to understand films in what I thought were fresh and fruit-
ful ways. But since the results were frequently castigated, inaccurately, as
“formalism,” I felt the need to justify the approach on broader concep-
tual grounds. Hence my efforts to anchor particular projects in a wider
research program and a long-run research tradition.
That program constitutes only one choice among many. The general
concept of poetics has a long history, and its traditions are diverse. Initially
applied to literary works, it was eventually transferred to the systematic
study of other media. In many ways, it’s indebted to research traditions
in art history and music history. My version of poetics, as worked out in
relation to cinema, isn’t the only one possible.
But here goes with my version. I take poetics to be asking the key ques-
tion, According to what principles are artworks made? The artist may have
those principles consciously in mind (e.g., obeying the canons of the son-
net form) or may exercise them unconsciously. The principles bear upon
materials, form, and style. Materials, in my view, encompass the physical
D. Bordwell ()
University of Wisconsin-Madison, Madison, Wisconsin, USA

DOI 10.1057/978-1-137-55309-6_2
16 D. BORDWELL
stuff and tools of filmmaking (cameras, film stock, etc.) but they also
include subject matter and themes. These less tangible materials are taken
up by the artist and reworked through form and style.
Materials, forms, and styles answer to broader factors, notably norms.
These are the accepted (tacit or explicit) guidelines for creating the work.
Genre conventions form one example of norms, but there are norms of
narrative construction and stylistic patterning as well. Those norms, in
turn, have social standing; they are part of the institution within which the
artist works, and they may as well be salient for other institutions in the
broader culture. Norms are a pivot point between artistic agents and the
community in which they operate.
Similarly, the principles of artistic “making” shape audience uptake.
Not everything that audiences experience by means of an artwork is dic-
tated by it, but the design features of the work often solicit emotional and
intellectual responses from perceivers. The process of uptake itself oper-
ates within institutions as well, and the dynamic of materials/form/style
can be construed differently according to the perceiver’s institutional situ-
ation. And all of this takes place in history, during which principles, norms,
creative communities, and audiences can change. Put most abstractly,
analytical poetics scrutinizes the ways and means of particular films and
groups of films, while historical poetics seeks to explain how those quali-
ties came about. The one offers functional explanations, the other causal
explanations.
I think that poetics deals in “hollow” concepts. To ask about principles
is not to presuppose any particular content of them—say, the Oedipal
trajectory. To search for norms is not to assume ironclad rules; as noted in
Pirates of the Caribbean, norms are more like guidelines. Acting as a film
critic, the poetician has to be sensitive to how individual films actualize or
revise or reject norms (always remembering that any norm offers a menu
of weighted options). Grounding the argument in history, the poetician
will have to explore a wide amount of empirical material, ranging from
data about the film industry to trends in critical reception. The researcher
will need to be creative in constructing arguments about the most perti-
nent and proximate causes involved.
Whew! That is too condensed a summary, but it indicates some major
presuppositions that shape my research questions. If you want to put flesh
on the bones, you can consult a more detailed attempt to explain the
film poetics I propose.1 Early examples of these ideas in practice were The
Classical Hollywood Cinema (1985), Narration in the Fiction Film (1985),
FIVE LESSONS FROM STEALTH POETICS 17
Ozu and the Poetics of Cinema (1988), and Making Meaning (1989). In
these books, I argued for a historical poetics of cinema because it seemed
to me to offer unique scholarly advantages—particularly in contrast to
Lacanian psychoanalysis, Althusserian Marxism, and other theoretical
commitments of the period.
Apart from their conceptual and empirical shortcomings, these orien-
tations didn’t emphasize questions of cinema as an art. By contrast, the
poetics perspective impels the researcher to look and listen closely to films,
to chart the fluctuations of film form and style. In turn, historical poet-
ics seeks to embed that art fairly precisely in time and place, something
that again wasn’t on most theorists’ agenda. Questions promoted in this
framework also offer a way to unite theory and practice, to see how the art
of cinema may be tied to concrete creative activities of filmmakers. (This
is one reason the approach has attracted the interest of filmmakers.) And,
suitably extended, the poetics perspective suggests ways to understand
how films engage audiences.
The works I’ve mentioned were fairly upfront about defending this
approach. In other writing, I’ve practiced what might be called stealth
poetics—writing from that perspective without putting theory and
method to the forefront. This strategy is evident in short pieces, like
“Cognition and Comprehension: Viewing and Forgetting in Mildred
Pierce” and “Film Futures.”2 Stealth poetics is also at work in The
Cinema of Eisenstein (1993), which might have been called Eisenstein
and the Poetics of Cinema; On the History of Film Style (1997); Figures
Traced in Light: On Cinematic Staging (2005); The Way Hollywood
Tells It (2006); Planet Hong Kong: Popular Cinema and the Art of
Entertainment (2000, 2011); and Christopher Nolan: A Labyrinth of
Linkages (2013). Over the last decade, much of my writing on the web
at www.davidbordwell.net/blog is in the same vein: substantive ques-
tions are in the foreground, the poetics-based assumptions are in the
background.
Once you embark on a research program, there’s a temptation to see
it as a static framework, a big machine into which you will feed what you
encounter. This was indeed one of my objections to Grand Theory: that
films were “read” in light of what this or that thinker (usually French)
believed.3 The results were mostly predictable. In what follows, I hope to
sketch some ways in which my ideas about poetics have been expanded
and refined by encounter with specific material: the cinemas of the Three
Chinas.
18 D. BORDWELL
Because the concepts of poetics are flexible, they can be revised and
nuanced as one works through specific objects of study. If poetics has
helped me understand Chinese cinema, Chinese cinema has helped me
understand poetics—and made it unpredictable. Let me count some
ways.
NORMS AND FORMS
Mass-market theatrical cinema depends mostly on narrative forms.
Studying Hollywood cinema from the early 1910s to the present yields
a sense of storytelling norms with a great many conventions. Since the
early 1980s, Kristin Thompson and I have tried to spell them out. While
loose assemblage principles govern some Hollywood studio films, like the
revue musical that pulls together a variety of acts, most films rely on other
principles. Typically, we have quite tight causality overall, goal-oriented
protagonists, a rising curve of interest, a climax driven by a deadline, and
the resolution of the action. There are other classical strategies, such as
the use of “hooks” to weld scenes together and the compression of time
via montage sequences. We’ve argued that these and other principles of
narrative composition have formed the basis of filmmaking in Hollywood,
and many other places as well.4 Backing up that argument has involved
research on all eras of American and world filmmaking—a process that is,
no surprise, far from complete.
I reckoned that alternatives to classical construction were more “epi-
sodic,” but I hadn’t really thought through what that entailed. Working
on Hong Kong film forced me to think about episodic construction as not
simply a lack of classical rigor but a sharp, powerful alternative in its own
right.
The quick characterization would be “a cinema of set-pieces,” with
inevitable comparison to the musical. But through interviews and analysis,
I found that filmmakers worked out their own variant on episodic con-
struction via reel-by-reel plotting that padded the openings, deferred the
crucial conflict, and produced an extensive final phase across two reels.
(These discoveries also showed the virtue of studying films in 35 mm,
where the reel breaks are apparent.) Knock-on effects of this formal tradi-
tion included reliance on coincidence, something I hadn’t considered as
a positive force, and an emphasis on character revelation in the face of
changing situations (rather than on character development, purportedly
the goal of much Hollywood cinema).
Once I had a better sense of this local variant of episodic construction,

with its own menu of options, I was better equipped to characterize the
creative choices of particular filmmakers. The format clearly suited com-
edies and the revenge plots featured in so many wuxia pian, kung-fu films,
and urban cop thrillers. But I also learned that one could characterize
Wong Kar-wai’s work as a way of pushing episodic construction further,
creating distinct parts tethered to one character or another (Days of Being
Wild, 1990; Ashes of Time, 1994) or to different time frames (Chungking
Express, 1994; In the Mood for Love, 2000; 2046, 2004)—the whole film
becoming unified by pictorial and musical motifs. Alternatively, it became
apparent that thinking of actions as chunks could encourage the intricate
game of repetition that is common to films by Johnnie To Kei-fung and
Wai Ka-fai. Through analysis, I tried to show that Milkyway films, while
following the reel-by-reel format, filled the blank spaces with schemes,
counterplots, and doublings. The Mission (1999), for example, resolves
its main plot by the end of its seventh reel, leaving two more reels to go.
What would be the start of a climax in a Hollywood through-composed
line of action becomes a new plot, which winds up paralleling and clarify-
ing the old one and resolving itself in the final reel.
Finally, returning to my study of Hong Kong film after ten years, I
found evidence that a more Hollywood-inspired set of principles was
guiding some filmmakers. The outstanding example is the initial Infernal
Affairs entry (2002), but I found others. I didn’t have occasion to talk to
the screenwriters as I had before, but as an initial guess, I’d trace this trend
to a greater awareness of the three-act model in screenplay manuals and a
general sense that recent Hong Kong films had to compete with American
pictures. Perhaps also some screenwriters hope to enter the US film indus-
try, and a well-carpentered script can help them.5
In short, the conceptual distinction classical plotting/episodic plotting
got filled out with real content thanks to film analysis and inquiry into
craft practices of Hong Kong filmmaking. Just as we can find felicities
within the Hollywood norm, my conception of the episodic model gained
richness by intimate confrontation with an alternative tradition.
RULES—NO, GUIDELINES
How alternative is that tradition, though? Not absolute, of course: Hong
Kong stories for the most part are intelligible to people around the world,
and their mix of stars and genres are familiar parts of any film industry.
20 D. BORDWELL
One thing that studying Hong Kong cinema brought home to me were
the various ways we think about Hollywood and its “others.”
We might say, in a healthy affront to Hollywood hegemony, that other
national cinemas, especially those considered minor or marginal, are rad-
ically distinct. Each has its own aesthetic, stemming from unique local
traditions. My own research tends to suggest that on the dimensions con-
sidered by poetics, there is both convergence and divergence. Many nar-
rative principles carry across traditions. This is partly because Hollywood
filmmaking became a model for other countries in the early years of cin-
ema, when movies circulated quite freely around the world. The diver-
gence comes in ways in which filmmakers revise Hollywood principles or
replace them with others.
It’s likely that episodic plotting was common in the world’s popular
storytelling for centuries before cinema was invented. It might be the
“natural default” until other options, such as tight plotting, come to
prominence. From this angle, Hong Kong filmmakers would be preserv-
ing a perennial option quite self-consciously, for reasons of ease of produc-
tion and product differentiation. Still, there is a good deal of classicism
there: the kung-fu hero bent on revenge aims at a goal, and a hierarchy
of villains comes forward to block him from achieving it—with the final
showdown ruled by a deadline. The same basic pattern can be found in
many urban thrillers.
What, though, of film style? Again, at a general level, I’ve found that
Hollywood’s norms became very widespread. For example, rather than
thinking of Soviet montage as an equal alternative to classical editing,
it’s better to think of it as revising classical canons. Devices like cutting
in or back, shot/reverse shot, crosscutting two lines of action, matches
on movement—all were absorbed and reworked, sometimes in radical
ways, by young Soviet filmmakers. The Kuleshov effect was invented in
Hollywood. But it took Kuleshov and his peers to see how it could be
pushed to new expressive ends.
I learned much the same from studying Hong Kong editing, espe-
cially in scenes of combat. From the Shanghai cinema of the 1930s to the
Mainland and Hong Kong cinema of the postwar era, Chinese film edit-
ing was closely modeled on American continuity principles. By the 1960s,
after Hong Kong filmmakers had given up the free-for-all staging of the
early Wong Fei-hong films, they adopted a shrewd variant of constructive
editing. Cut from a fighter springing upward to his opponent looking
up and then to the first fighter soaring in the air; the three shots equal a
“weightless leap.” But once the fight was pulverized into several shots,
then it was possible for King Hu to play with various parameters: shaving
frames off each shot, sustaining the airborne phase through several shots,
dynamizing both leap and landing through unusual angles. Recognizing
that cinematic trickery was giving an impression of warrior prowess, he
decided to inflate that through editing that wouldn’t have been out of
place in an Eisenstein film.
Similarly, once filmmakers had mastered fine-grained editing, they
could exploit its rhythmic potential. During the 1960s, several studios
began developing a distinct approach to kung-fu and swordplay. Unlike
the more or less random tussling of fistfights in American films, Hong
Kong fights displayed a pattern of stasis—a flurry of action, itself fairly
rhythmic—and an instant of rest. I called this staccato pattern the pause/
burst/pause convention, and tried to show the great variations it could
create. What became more interesting was the way in which filmmakers
like Yuen Kuei carried this pattern down to the very editing of shots, cre-
ating complex sequences by intercutting fluid passages of movement with
shots that brake the action and prepare for the next burst. This dance-like
rhythm, derived from the practice of martial arts techniques, had strong
expressive possibilities. It could catch up the spectator’s body in a compul-
sive rhythm and accelerate the action for emotional effect. Again, having
the right tools helped me: I was able to count frames because I was work-
ing with 35 mm copies.6
I came to understand not only some specifics of Hong Kong action cin-
ema, and the ways it recruited film technique for engaging ends. I came as
well to understand that these filmmakers, untutored in film history, were
reviving practices explored by the Soviet filmmakers, above all Eisenstein.
He sought to create a montage that would galvanize the spectator through
appeal to his or her body. Few scholars had given credit to Hong Kong
action films, but I came to see them as being as important to the history
of editing as Soviet Montage.
Watching people flinch and clench before a Hong Kong film, I also
became convinced that Hollywood continuity editing was a very rich,
perhaps inexhaustible resource. Rather than invent comprehensive
alternatives, most filmmakers mined it for fresh possibilities. Far from
being a simple set of rules, continuity editing made every choice about
cutting bristle with a great many options—only some of which had
been realized by American cinema. The lesson: Rewarding poetics is
comparative.
22 D. BORDWELL
LONG TAKES, LONG LENSES

Researchers have long looked to technology as a proximate cause for the
ways films look and sound. We have associated the freewheeling look of
French New Wave films with the development of lightweight cameras and
faster film stock. Less obviously, the ease with which rapid editing caught
on in post-1960s, American cinema was facilitated by new tools, from
Scotch tape to digital storage. But it’s also clear that technology isn’t a
force unto itself. It has to be compatible with existing work routines and
division of labor, and its development will benefit when filmmakers can
articulate a specific problem they can’t yet solve.
One thing it’s easy to miss is the cascade of circumstances that can
shape a style. Hou Hsiao-hsien developed a long-take style that cultivated
subtle staging within a fixed frame. His early films show the beginnings of
this tendency, which gain in range and nuance in key moments of A Time
to Live and a Time to Die (1985), Dust in the Wind (1986), and City of
Sadness (1989). Hou explains his technique as simply the result of impro-
visation and letting the actors relax so that he can observe them from a
distance. But the precise flow of the performers’ movements, blocking
and revealing each other with minimal shifts, suggest that he was intui-
tively guiding them into very precise ensemble arrangements—perhaps
not through a detailed process but through something that felt spontane-
ous but was also strictly controlled.
This sort of precision staging has antecedents, most distantly in the
second decade of cinema, that period that gave us films like Ingeborg Holm
(1913) and the quietly virtuosic works of Louis Feuillade. It reemerged at
certain moments afterward, but usually in tandem with continuity editing.
What led Hou, with little interest in the history of cinema, to rediscover
techniques of staging that reach back so far?
It seemed to me that, first, the founding choice of the long take and
the fixed camera set limits from the outset. Some filmmakers make that
choice for its own sake, as a way of testing what they can do by eliminat-
ing the resources of editing. For others, it’s a forced choice, for reasons of
economy, speed, or the exigencies of location. For the earliest filmmakers,
it was simply the unquestioned premise of scenic construction; analytical
editing wasn’t a live option.
Whatever shapes the point of departure, if the camera is to be locked
down and several actors are to be present, the filmmaker faces the choice
of planting his characters in a lateral array (as Welles does in the famous
kitchen scene of The Magnificent Ambersons [1942]) or arranging them in

depth. If they’re arranged in depth, the likelihood is that they will have
to change positions in the course of the scene. That in turn obliges the
director to think about visibility—how to manage what we see and when
we see it. That leads to considering tactics like having actors turn away
from the camera, move to the periphery of the screen, or recede into the
distance: all ways of guiding attention. These tactics are part of directorial
craft of the 1910s and were, it’s clear, developed with great finesse before
receding in the editing-dominated eras that followed.
In this regard, the founding choice of the long static take creates
what engineers call “path dependence,” a commitment to limiting fur-
ther choices. But some choices remain open, and here’s where technol-
ogy becomes relevant. Hou started making films when certain norms
were in force, particularly anamorphic widescreen and location filming.
Commercial Taiwanese films of the 1970s and 1980s favored the 2.40
format, apparently because most theaters were equipped to show them
and they gave a certain cachet of quality. But to keep expenses down,
filmmakers shot on locations. City streets obliged filmmakers to use tele-
photo lenses to cover their performers. This constraint coaxed me to
think about problems of location shooting. The fastest way to execute a
location shot is from a distance, with a fixed camera, using long lenses.
That way, people wandering into the frame won’t be as disruptive as they
would be if your camera were close. They’re also less likely to notice the
filming.
The results, in contemporary-life Taiwanese romances and comedies
of the period, are many shots taken at a great distance. But one thing we
tend to forget about telephoto lenses is that they not only flatten pictorial
space but also make the playing space quite narrow. Actors have less room
to move horizontally. The widescreen format compensates for that cramp-
ing somewhat, but still, a director has to be fairly careful about staging
the action. The actors are crammed quite close together in the field of the
long lens, and they must not block each other.
I proposed that Hou began to master precision staging within these
long lens long takes. It seems plausible that the constraints of the industry
created favorable circumstances for precision staging. Location shooting
favored the telephoto lens, which obliged Hou to find creative ways to
fill a slice of space pictorially wide but physically narrow. Of course this
isn’t the whole story. Many other directors faced the same constraints
and didn’t pursue complex staging. I suggest that Hou hit upon it as a
24 D. BORDWELL
congenial way to work with nonactors and eventually realized he could

control our apprehension of the story more minutely this way.
In following years, he would complicate his precision staging by dark-
ening his compositions and moving the camera. These became further
ways of making the “just noticeable differences” in his characters’ behav-
ior artistically powerful. By the time he makes Flowers of Shanghai (1998),
the camera is arcing around a packed table, and the slightest shifts among
seated and walking characters become part of a rich pictorial flow. Hou
ceased to use the 2.40 format and pulled back from long lenses, but the
lessons learned while he deployed those tools were reapplied to later films.
What I learned from studying Hou’s style and seeking causal explana-
tions for it was something not evident from my initial understanding of
poetics. Forms and styles that have died out can be revived unwittingly by
filmmakers responding to historical pressures. Hou didn’t steep himself in
1910s cinema. Given certain choices he made, precision staging became
a plausible artistic resource independently of its prior uses. At bottom,
Hou’s style offers another argument against the linear model of artistic
development in cinema. Forms and styles once abandoned can reemerge
under very different circumstances—not through any Hegelian dialectic
of history, but through concrete, contingent choices artists make.7
ASIAN, MINIMALLY?
Another cluster of historical questions are raised by the ascent of Hou.
His signature works were made between 1983 and 1993. They thus run
parallel to the emergence of the new Hong Kong cinema and the Fifth
Generation in China. Other directors of the New Taiwanese Cinema of
the period did not rigorously cultivate his long-shot, long-take aesthetic
(although Edward Yang moved in that direction). We’re familiar enough
with such categories, as film history is often written in terms of national
schools like German Expressionism or Italian Neorealism.
But as I watched films from the 1990s, I wondered if another sort
of grouping made sense. I started to notice that Hou’s aesthetic choices
were appearing elsewhere in Asia a decade after his debut. I first noticed
the similarity in Kore-eda Hirokazu’s Maborosi no hikari (1995). But
there is a long tradition of static long-shot staging in Japanese cinema, of
which Mizoguchi was the most famous exponent, so I thought that Kore-
eda was harking back to that. (I was told by a Japanese friend, though,
that unfriendly Japanese critics attacked the film as a Hou facsimile.)
The mid-1990s brought international audiences a great many Asian film-

makers who seemed to be working along Hou’s lines. In Taiwan, there was
Tsai Ming-liang; in Japan, there was not only Kore-eda but also Aoyama
Shinji (Two Punks, 1996; Eureka, 2000) and Suwa Nobubiro (2/Duo,
1997; M/Other, 1999). Korea had Hong Sangsoo (The Day a Pig Fell in
the Well, 1996; The Power of Kangwon Province, 1998), while Thailand
had both Pen-ek Ratanaruang (6ixtynin9, 1999) and Apichatpong
Weerasethakul (Blissfully Yours, 2002; Tropical Malady, 2004). Even
China had its adherent: Jia Zhangke, particularly in Platform (2000) and
several films that followed. Japan’s Kitano Takeshi could also be included
in the group.
There are of course crucial differences among these filmmakers, but
they start from the long-take premise, usually in a fixed, fairly distant shot.
The staging might not be as dense and intricate as in a Hou film, and the
narrative structures might be far from Hou’s preferred ones—although as
he became a more elliptical storyteller, particularly from City of Sadness
on, younger filmmakers may also have been inspired by that side of his
work. And naturally some of these filmmakers have departed from these
beginnings; the development of Kore-eda and Jia toward other stylistic
options is apparent. Still, as a working hypothesis, I began to posit that
Hou’s work, at first apparently so singular, had become a model for others
in the region.
I called this trend “Asian minimalism,” and although in some essays
and a video lecture I’ve sketched some comparisons among the filmmak-
ers, the concept needs further testing. What these apparent stylistic affini-
ties taught me, however, is that we should be on the lookout for networks
of similarity across national borders.
Admittedly, any aesthetic approach to film could have homed in on
the affinities between Hou’s work and those of younger filmmakers in the
region. (Not that there are that many schools of stylistic analysis at play in
film studies.) But then someone would be tempted to associate the stylistic
properties with an “Asian view of the world,” characterized by contempla-
tion, meditation, and serenity. Such reflectionist ideas were long floated
about Ozu and Mizoguchi. But a perspective informed by poetics would
look for the more concrete causal circumstances within film culture that
could facilitate a pooling of artistic resources.
Those circumstances aren’t hard to find, in general. Today, filmmak-
ers can see each other’s works more easily than in past eras. Film festivals
bring together directors from all over the world—not only to watch films
26 D. BORDWELL
but also to meet, sit on panels, participate in marketing, and serve on

juries. (I’ve been told that Hou and Theo Angelopoulos learned of each
other’s works through Cannes.) Home video and satellite TV make a body
of work much easier to assimilate. And filmmakers are given many oppor-
tunities to visit other countries’ schools and cinematheques. In general,
the creative community of filmmakers hosts conditions for wide diffusion
of stylistic options, and access to a great many works created the possibility
of regional schools and international influences.
Even if we can make a case for Hou as a model for a regional trend, we
would have to be alert for ways that filmmakers differentiate themselves.
Tsai Ming-liang, for instance, quite early turned the style toward new sub-
jects and themes, often involving social anomie and homosexual eroticism;
he also displayed a sharp sense of humor, using long-take deep focus for
gags. Hong Sangsoo, using single-take scenes as blocks, built them into
parallel universe or split-viewpoint plots. While Hou by and large favored
somewhat diagonal compositions and cluttered or semi-opaque planes,
several of the other directors relied on more planimetric, frontal composi-
tions, and simpler spatial layouts. Again, if we posit that filmmakers work
within a community, we should expect divergence and even competition
among them. Once fresh stylistic territory is opened up, today’s film cul-
ture asks filmmakers to stake out part of the terrain as uniquely theirs.8
ALL TOGETHER NOW

The notion of a community of filmmakers brings me to the last lesson
I’ll consider. In working on American studio cinema, my coauthors and I
analyzed the tight interaction among sectors of the Hollywood film indus-
try. Within the fairly limited locale of Los Angeles and environs, there
could be rapid communication between various production sectors—film-
makers, technology firms, supply houses, and coordinating bodies like the
Academy. This interchange facilitated stylistic change through innovations
in sound recording, lighting, lenses, and the like.
Partly because the system we were studying was in the distant past,
I didn’t get a concrete sense of the give and take of Hollywood inter-
change. But visiting Hong Kong and interviewing critics, screenwriters,
directors, and martial arts choreographers brought vividly to life my air-
less abstractions about the creative community. Not that Hong Kong is
typical of all such artistic convergence points, but its relative smallness
and the frantic pace of its media industries gave a vantage point on a
restless, teeming film culture. In my 2000 book, I tried to convey the

rapid exchange among filmmakers, the way innovation creates copycats
and competitors, the way cycles pop up, throb along, and burn out. The
audience got caught up in this churn through unique institutions, like
the now-vanished midnight screening (a sort of early, small-scale version
of Twitter feedback).
Ten years later, I updated Planet Hong Kong to take into account
changes in the film culture after 1997. At that point, I was able to con-
sider the creative community from another angle. Given the rise of the
Mainland industry, what were filmmakers to do? I tried to reconstruct
their choice situation, basing my inferences on various survival strategies
that emerged in the wake of Hero (2002), China’s first bid to be a major
player in both regional and global cinema. The same dynamic of trial and
error, copy and competition, emerged. By focusing on four major direc-
tors—John Woo, Stephen Chow, Wong Kar-wai, and Johnnie To—I tried
to make their very different career decisions intelligible in the light of new
cultural, financial, and artistic opportunities.
By then the tight community I studied in 1994–1998 had changed.
I miss it. The biggest benefit of my Hong Kong research was the host
of friends I made and still have. My work there also convinced me even
more strongly that individuals matter in film history. Granted, they work
within institutions, they are prey to forces working behind their backs,
and the fruits of their efforts get appropriated in unpredictable ways.
Poetics can show us how these things happen. But ideology doesn’t
switch on the camera. Another reasonable task for a poetics of cinema
is to consider how creative individuals bring distinctive talents to a craft
tradition.
NOTES
1. “Poetics of Cinema,” in Poetics of Cinema (New York: Routledge, 2008),
11–55.
2. “Cognition and Comprehension: Viewing and Forgetting in Mildred
Pierce,” and “Film Futures,” in Poetics of Cinema, 135–150; 171–188.
3. Kristin Thompson offers, in her opening chapter of Breaking the Glass
Armor: Neoformalist Film Analysis (Princeton: Princeton University Press,
1988), a nuance: she points out that an approach can embrace many meth-
ods. What she calls Neoformalism took the Russian Formalists’ empirically
based theory as a guide. The approach asks you to spot intriguing things in
28 D. BORDWELL
a film. Guided by general principles of how films work, you canvas the ana-
lytical and explanatory methods on offer, and you try out the ones that seem
most illuminating.
4. See David Bordwell, Janet Staiger, and Kristin Thompson, The Classical
Hollywood Cinema: Film Style and Mode of Production to 1960 (New York:
Routledge, 1985); Kristin Thompson, Storytelling in the New Hollywood:
Analyzing Classical Narrative Technique (Cambridge: Harvard University
Press, 1999); David Bordwell, The Way Hollywood Tells It: Story and Style in
Modern Movies (Berkeley: University of California Press, 2006); and entries
on the blog “Observations on Film Art,” at www.davidbordwell.net/blog
5. For development of ideas sketched in this section, see Planet Hong Kong:
Popular Cinema and the Art of Entertainment, 2d ed. (Madison, WI:
Irvington Way Institute Press, 2011).
6. King Hu’s editing is treated in more detail in Planet Hong Kong, 160–165
and “Richness through Imperfection: King Hu and the Glimpse,” in Poetics
of Cinema, 413–430.
7. For more on ideas presented here, see Chap.6 of On the History of Film Style
(Cambridge: Harvard University Press, 1998) and Chap.5 of Figures Traced
in Light: On Cinematic Staging (Berkeley: University of California Press,
2005). See also “Transcultural Spaces: Toward a Poetics of Chinese Film,”
in Chinese-Language Film: Historiography, Poetics, Politics, ed. Sheldon
H. Lu and Emilie Yueh-yu Yeh, 141–162, and the 2015 online video lec-
ture, “Hou Hsiao-hsien: Constraints, Traditions and Trends,” at https://
vimeo.com/129943635
8. For further developments of these ideas, see “Beyond Asian Minimalism:
Hong Sangsoo’s Geometry Lesson,” in Hong Sangsoo, ed. Huh Moonyung
(Seoul: Seoul Selection, 2007), 19–29.
CHAPTER 3
Red Poetics: The Films of the Chinese

Cultural Revolution Revolutionary
Model Operas
Chris Berry
Tsui Hark’s three-dimensional (3D) action film, The Taking of Tiger

Mountain (智取威虎山), was a major Chinese-language box office hit
in 2014 and is listed as the tenth most popular Chinese film of all time
at the People’s Republic of China box office (http://www.cbooo.cn/
Alltimedomestic, accessed 30 September 2015). The film has a frame
story featuring a contemporary young man who stumbles across and
becomes fascinated by the 1970 Cultural Revolution-era film, Taking
Tiger Mountain by Strategy, which shares the same Chinese-language title
with Tsui’s film. It then proceeds to retell the core historical story of the
People’s Liberation Army rescuing the ordinary folk of Northeast China
from banditry during the power vacuum that developed there after the
Japanese invaders were driven out in 1945. However, it does so in the
contemporary action filmmaking mode, rather than the revolutionary
model opera mode of the 1970 version.
The frame story in The Taking of Tiger Mountain depends on us
believing that the young man has no idea what Taking Tiger Mountain
by Strategy is. Perhaps he may not have seen that particular film, but it is
C. Berry ( )
King’s College London, London, UK

DOI 10.1057/978-1-137-55309-6_3
30 C. BERRY
harder to believe that he would not recognize the imagery. Indeed, the
look of Taking Tiger Mountain by Strategy and the other so-called model
works—the yangbanxi (样板戏)—are instantly recognizable in China and
around the world as the aesthetic of the Cultural Revolution decade in
China (1966–1976), shared with posters (Evans and Donald 1999) and
other visual artworks.
After the end of the Cultural Revolution decade in 1976 and its rapid
repudiation during the Deng Xiaoping era that followed soon after, the
model works were dismissed. Kirk Denton (1987) undertook an early
semiotic analysis of Taking Tiger Mountain by Strategy, but after that, there
was a long gap in scholarly research on the model works. This neglect has
begun to change more recently. Xing Lu (2004) has examined the use of
political language in not only the slogans but also the posters and revolu-
tionary model operas of the period. Paul Clark has written a history of the
culture of the Cultural Revolution (2008), and Daniel Leese has written a
monograph on the Mao Cult (2011). More recently, Barbara Mittler has
produced a magisterial work which examines not only the practice of cul-
ture during the Cultural Revolution but also the memories of those who
took part (2012). Other work is going deeper into particular aspects of
Cultural Revolution culture, such as Gary Xu’s close analysis of the regular
feature films made when the industry was revived in the final years of the
decade (2013), and Yawen Ludden’s equally close analysis of the trans-
formation of Peking Opera into the revolutionary model operas (2013).
Despite this renewed interest in the culture of the Cultural Revolution,
so far there has been little work focused on the poetics of the film versions
of the model works, in the sense proposed by David Bordwell of cinematic
poetics as “the study of how films are put together and how, in determi-
nate contexts, they elicit particular effects” (1988: 1). Clark (2008) pays
a great deal of attention to the film versions, but is more focused on the
history of their production than how they work as film texts. Xu is focused
on the cinema of the Cultural Revolution decade, but on the feature films
and not the film versions of the model works.
This chapter begins the work of analyzing the poetics of the film ver-
sions. It focuses in particular on some of the best remembered and most
loved examples of the model works: the revolutionary operas Taking
Tiger Mountain by Strategy and Azalea Mountain (杜鹃山, 1974), and
the revolutionary ballet The Red Detachment of Women (红色娘子军,
1971). It undertakes the analysis of how the style of these films is put
together and how it works in four sections. First, it briefly introduces what
RED POETICS: THE FILMS OF THE CHINESE CULTURAL REVOLUTION... 31
is meant by the “model works.” Second, it considers the genealogy of the

model works aesthetic, pointing out that although these films look dis-
tinctly “Chinese” to foreign audiences, they are highly syncretic and mod-
ern hybrids. Indeed, their unique style can be considered as a response
to China’s effort to carve out its own unique revolutionary path at this
time. Third, it introduces the local and contemporary aesthetic doctrine
that could be considered as dictating how the elements composing these
hybrids should be combined. That doctrine was the “Theory of the Three
Prominences” (三突出), applied to all cultural output during the Cultural
Revolution era, including the cinema, and understanding the fundamental
principles it laid down is the first step in grasping the poetics of the model
works. And, fourth, it analyzes how the three prominences are achieved
in cinema, through the deployment of cinematography and editing, com-
bined with color, rhythm, music, pacing, montage, and other elements to
induce both a powerful embodied response and the taking up of a political
line. However, by way of a conclusion, it returns to Bordwell’s poetics to
argue that, because these two processes are articulated but not identical,
it is possible for audiences to remember and treasure their embodied plea-
sures today or to experience them for the first time, without necessarily
taking on board the politics attached to them, as appears to have been the
case with the character in the frame story of Tsui’s film.
THE “MODEL WORKS”

What are the “model works”? Answering this question raises two issues:
the concept of the model, and which particular works were designated as
model. Providing ideals for people to aspire to in this world is integral to
secular and modern cultures. If capitalist market culture uses advertising
to encourage citizens to imagine a process of self-transformation through
consumption, socialist revolutionary cultures used propaganda to prom-
ise transformation through participation in revolution and production. It
is no coincidence that during the Maoist era, there was no distinction
between propaganda, publicity, and advertising, with the same term (宣
传) covering all three meanings. Barbara Mittler (2012: 79) traces the
idea of using the model (样板) in culture back to Mao’s 1942 Talks at the
Yan’an Forum on Literature and Art, and Paul Clark (2008: 57) notes
how the term was deployed in the run-up to the Cultural Revolution to
designate the new revolutionary operas that were being developed at the
time.
32 C. BERRY
Efforts to reform the traditional Chinese opera repertoire had been

an ongoing part of the revolutionary process in culture long before the
Cultural Revolution itself began. The popularity of Chinese opera forms
with audiences meant they could not be easily discarded as “feudal trash.”
However, the process was slow, and Mao complained in 1963 that the
stage was still full of “emperors, kings, generals, chancellors, maidens, and
beauties” (He 1964, cited in Mittler 2012: 79). The development and
promotion of the revolutionary model works with modern settings was
designed to address this problem. On 28 November 1966, at the begin-
ning of the Cultural Revolution decade, the recognition of eight revolu-
tionary model works (革命样板戏) was announced (Clark 2008: 59) and
these were actively promoted during the May 1967 celebration of the
25th anniversary of Mao’s Talks at the Yan’an Forum on Literature and
Art, which had guided cultural policy ever since they were given in 1942
(Clark 2010: 177). The eight included five revolutionary Peking operas:
The Red Lantern (红灯记), Shajiabang (沙家浜), Taking Tiger Mountain
by Strategy, Raid on the White Tiger Regiment (奇袭白虎团), and On the
Docks (海港). The remaining three were composed of two revolutionary
ballets—The Red Detachment of Women, and The White-Haired Girl (白毛
女)—and a symphonic version of Shajiabang.
This early designation of eight model works is the origin of the com-
mon idea that they were the only things that could be seen during the
Cultural Revolution decade. However, Mittler (2012: 47) has shown that
the term “the eight model works” was in fact used very rarely during the
Cultural Revolution decade itself. Instead, it was seized upon and popular-
ized during the late 1970s, after the fall of the so-called Gang of Four and
as part of the repudiation of the Cultural Revolution.
Nonetheless, the idea has persisted, not only in popular memory but
even among scholars who should know better, that this was a decade of
“800 million people watching 8 shows (八亿人看八个戏)” (Clark 2008:
26). However, as Clark has clearly established, this was not true either in
terms of the general variety of culture available to audiences or in terms
of the number of model works. Taking the example of Tianjin, he shows
a wide variety of revolutionary works beyond the model works could
be seen during the National Day month of October 1966 (Clark 2010:
174–7). These included selected films from the 1950s and 1960s, and
these continued to be screened (Clark 2008: 150). Although imports from
both the liberal capitalist and “revisionist” Soviet bloc countries were not
screened, more than half the films in circulation were foreign films from
remaining “friendly” countries such as Albania, North Korea, Yugoslavia,

and Romania (Clark 2008: 150). As for the model works themselves,
a wide variety of new revolutionary stage performances was originated
in this period. So, it is not surprising that after the initial eight, others
were soon designated as “model works,” too (Clark 2008: 62–73). They
included Azalea Mountain, analyzed in further detail here.
GENEALOGY OF A HYBRID
Although the model works did not monopolize stage and screen dur-
ing the Cultural Revolution decade, they had a distinctive style and look
that is remembered and recognized not only in China but also around
the world. In accounting for this distinctive look, we must acknowledge
that the model works are hybrids, and the particular combination of ele-
ments making up their hybridity played a strong role in making them so
recognizable. As a first step toward understanding the poetics of the film
versions of the model works, what are these elements and where are they
drawn from?
For foreign viewers, the Chinese revolutionary model operas may seem
unutterably exotic and strange, and therefore they are sometimes mistak-
enly seen as essentially Chinese. The real situation is much more complex.
It is true that the revolutionary model operas are grounded in Peking
opera. However, Mao’s dissatisfaction with the continued presence of
emperors and beauties on the stage, cited above, indicates the kind of
changes that were being implemented to make opera modern and revolu-
tionary. Put simply, as in all other areas of life, after the establishment of
the People’s Republic in 1949, everything inherited from the “old soci-
ety” had to be assessed, and that which was “feudal” had to be eliminated.
Ever since the late nineteenth century, the desire to be modern and
the association of modernity with the West had led to various imports.
Marxist revolution, as a particular endorsed model of modernity, licensed
particular imports. At the same time, because the revolution was not only
a socialist but also a national revolution designed to throw off imperial-
ism, there was a desire to retain as much national culture as could be
seen to be progressive because of its association with the revolutionary
classes. Therefore, a process occurred of adopting those elements from
outside that could be considered modern and revolutionary and they
were Sinicized by their integration into the retained local elements. As a
result, the same revolutionary model operas that might look so exotic and
34 C. BERRY
Chinese to foreign viewers looked equally exotic to Chinese viewers at

the time, but for very different reasons—their innovations, often drawn
from foreign cultural origins, made them look modern and exotic.
This distinctive combination of Peking Opera traditions with new and
modern revolutionary elements can be seen, for example, in the opening
moments of Taking Tiger Mountain by Strategy. After the credits, there
is a dissolve into a white landscape, with a troupe of communist soldiers
in snow camouflage running out into a clearing from behind pine trees,
which cunningly disguise the point where wings to the proscenium stage
end. It is immediately clear that these characters are not from the feudal
but from the modern era, and that they are part of the “worker, peasant,
soldier” (工农兵) triumvirate of characters approved for leading roles in
socialist works because of their good class background.
Furthermore, they are not dressed in the elaborately stylized and con-
ventional costumes usually associated with Peking Opera, but instead in
what appear to be regular uniforms. Modernization means the adoption of
realism. Imported to the stage and into Chinese literature following inten-
sive contact with Western powers in the nineteenth century, realism was
also the only aesthetic mode approved in the Soviet Union and in Mao’s
Yan’an Talks, and was therefore associated with both socialism and being
modern (Anderson 1990; Huters 1993).
Not only the costumes are realist but so is the make-up. When the
soldiers who have run onto the stage come to a halt a few seconds after
running on stage, the film cuts to a close-up of the commander, who
is not made up with the sort of stylized mask we associate with tra-
ditional Peking Opera, but instead in a rather heavy but nevertheless
realist style that accentuates his features but does not cover them with
a design that diverges from them. This may be stage realism rather than
what is usually thought of as film realism, but it is realism nonetheless.
Barbara Mittler (2012: 57) rightly notes that within this realist frame,
there is some continuity from the Peking Opera tradition, with heroes
made up in glowing, warm, red hues, and villains in greenish pallor, as
we see later in the film in the villains’ lair on Tiger Mountain itself. This
somewhat exaggerated and type-based make-up can also be related to
the Maoist modification of socialist realism preferred after the break-
up with the Soviet Union: revolutionary romanticism and revolutionary
realism. This combination emphasized an idealized form of realism that
both revealed essential qualities and looked to the socialist future (Wang
2003).
The same adoption of this heightened realism is true of the stage itself.
Chinese opera is traditionally performed on a platform, around which the
audience gathers on three sides, but here we have a proscenium stage.
Minimal props—usually a table and a chair—are used on the traditional
opera stage, and what they signify depends on how they are used. But here
we have a painted backdrop that, albeit according to stage rather than
screen conventions, is realist, and props represent what they resemble.
Furthermore, when the dialogue begins, it is in contemporary spoken
Mandarin, rather than the classical language often used on stage. In prin-
ciple, this would help the political message get across.
However, having noted a host of changes implemented according to
the principles of bringing the content up-to-date, making it revolution-
ary, and following realist principles, it must be noted that not everything
has changed. In the same opening scene, when the soldiers need to enact
a journey through the snow and toward Tiger Mountain, they do so
running, leaping, and tumbling more or less on the spot in a series of
conventionalized movements, much like those used in Peking Opera to
communicate a journey to the audience. And, of course, they still sing
opera arias in Peking Opera style. But, aside from the arias, the music in
the film tends to supplement opera percussive punctuation of action with
full Western orchestral music and the use of leitmotifs associated with par-
ticular characters.
Not all the model works were revolutionary operas. Red Detachment
of Women is one of two revolutionary ballets in the initial tranche of
eight model works. Also set in the pre-1949 period, but this time in the
Southwest, on Hainan Island, it tells a story of a young woman rescued
from class exploitation who joins the communist army. Ballet is an import,
presumably regarded as politically acceptable because it was associated
with the Soviet Union rather than European court culture. However,
although this foreign origin is undeniable, a process of localization and
adaptation occurs in the development of revolutionary ballets like Red
Detachment of Women that means they are also hybrids.
This localization is most immediately evident in the Chinese narratives,
Chinese characters, and Chinese clothing and props of the revolution-
ary ballets. But it is also present in a wholesale transformation of ballet
movements and poses. The tradition of romantic love among princes and
princesses would be no more welcome on the revolutionary stage than
the opera emperors and beauties Mao despised. But, as Rosemary Roberts
(2008) has shown in her detailed analysis of gender in the revolutionary
36 C. BERRY
ballets, a whole new array of movements was developed for narratives in

which heroines were not tragic and delicate but revolutionary and defi-
ant. Although, as she argues, this does not reverse or overturn gender
hierarchy, with men still doing the lifts and jumps, the pas de deux in
Red Detachment of Women between the young woman and the man who
rescues her and later turns out to be the commander of the red detach-
ment itself is a good example of this transformation. Instead of individual
romantic passion, the dance becomes one of guidance and inspiration,
with the commander assisting the young woman, who has run away from
the evil landlord’s house, and ending with him supporting her arm as she
stands en pointe, while both of them gesture and gaze toward the revolu-
tionary future somewhere off-screen, their eyes glowing.
Furthermore, some of the new poses and moves in the revolutionary
ballet share much in common with Chinese opera. For example, at the
beginning of the film, the heroine appears chained to a pillar in the land-
lord’s dungeon, but she pulls at her chains, striking a defiant pose as she
stares off-screen. As Clare Sher Ling Eng points out (2009: 22), this is
the adoption from opera of the liangxiang (亮相) pose, which crystallizes
the emotions and attributes associated with a character. When the long-
awaited Party Representative (党代表) Ke Xiang (柯湘) finally makes her
appearance in Azalea Mountain, she is a prisoner being led to her execu-
tion, but, although also in chains, she strikes similarly defiant liangxiang
poses (Berry and Farquhar 2006: 62–4).
THE THREE PROMINENCES

If an examination of the model works as hybrids reveals the particular
mixture of traditional and new, and Chinese and foreign elements that
are combined in them, the first clue to how these elements are composed
can be found in the theory that guided all cultural production during the
Cultural Revolution decade. This is known as the “Theory of the Three
Prominences.” The first prominence concerns highlighting the positive
characters among all the characters in a work. Second, among the posi-
tive characters in a work, heroic ones should be given prominence. And,
third, the main heroic or central character should be highlighted most of
all. This theory was published in 1968 (Yu), and then cited repeatedly in
the years that followed.
As Clark (2008: 46) points out, not only was this theory developed
specifically in the context of the reform of opera and the production of
the model operas but also the concern with how to highlight the cen-
tral hero was evident in earlier discussions. However, it was applied much
more widely throughout the arts. For example, Laikwan Pang has related
the composition of portraits of Mao during the Cultural Revolution, in
which he was not only centered but also isolated from all those around
him, to the application of the theory of the three prominences (2012:
413–415). And Barbara Mittler (2012: 84) notes the application of the
principle of the prominences in the model works so that negative char-
acters “sing very few arias if any, and … they are accompanied only by
the lower, brassy sounding instruments of the orchestra or by low-range
Chinese instruments.”
In the cinematic versions of the model works, similar applications of
the three prominences can be seen. In Taking Tiger Mountain by Strategy,
the main hero Yang Zirong (杨子荣) has taken on the mission of infiltrat-
ing the bandits’ hideout up on Tiger Mountain. At 50 minutes into the
film, the scene of Yang’s arrival at their lair uses the theory of the three
prominences to put the spotlight on him—literally. After a fade up from
black, we are presented with a long shot of a few shadowy figures. The
camera pans right to the leader of the bandits sitting down on a chair on a
platform and wrapping a dark cloak around himself. The music is indeed
low notes and relatively quiet.
When Yang is called on, the camera cuts back over to the left, and two
bandits on each side of the frame point their bayonetted rifles toward the
entrance of the cave, shouting out loudly. Triumphal orchestral music,
with higher notes and a lot of brass instruments, plays a leitmotif associ-
ated with Yang to presage his arrival. And when he does appear, he is spot-
lit, throwing his coat back to reveal a bright white lining, his trademark
tiger-skin waistcoat, and a white neck scarf. The lighting, music, and the
bright colors all clearly demarcate him from the bandits. Where the ban-
dits have an unhealthy pale greenish pallor, Yang’s face is bright and shin-
ing, his wide and intense eyes outlined with black mascara.
The theory of the three prominences aims at the deployment of all the
elements in an artistic work to create clear delineation between positive
and negative characters, as well as among the different levels of positive
characters. Complex characters, or what in China were known as “middle
characters” (中间人物), had been advocated in the name of realism before
the Cultural Revolution (Hegel 1984: 208). The three prominences
eliminated them altogether. In these circumstances, the three prominences
operated largely through redundancy, so that all elements combined to the
38 C. BERRY
same effect, underlining each other ever more heavily. Therefore, in the
scene just examined, the spotlight on Yang Zirong combines redundantly
with the cut to an extreme close-up, the bold gesture of throwing his coat
open and holding his arms wide, the whiteness of the lining and his scarf,
his bright make-up, and the blast of brass instrument music to highlight
him as the main (and in this case only) hero in the scene.
The scene of Yang Zirong arriving in the bandits’ lair only involves a
hero and various villains. How does the theory of the three prominences
work to distinguish a main hero from lesser heroes and regular positive
characters? Ke Xiang’s first appearance in Azalea Mountain can help us to
understand. At first, the shots primarily feature her and the villainous local
government guards who are leading her to her execution. These shots
echo Yang’s appearance although there is no spotlight, because the scene
is taking place in daylight. But she appears at the top of the steps to the
government building (the yamen) when the doors are opened, dressed
in a brilliant white top, throwing her arms out, even though they are
chained, and striking a bold pose, much like Yang. Her top is stained with
red, presumably from her blood. As I have pointed out elsewhere (Berry
2012), red and white get used in the films based on the model works to
catch the eye, helping the heroic characters to stand out, whereas the vil-
lains sink into darker colors. Ke’s guards in this scene are dressed in black,
and whereas she faces the audience as she strikes her liangxiang pose,
they have their backs to us. She is individualized, they are an anonymous
group. Close-ups, percussion, and crescendos mark her poses.
However, although the distinction in the opening shots is between the
guards and Ke Xiang, the audience of the film knows from the narrative
leading up to this moment that the streets around the government office
are full of rebel militia who are eager to free her. Their leader would count
as a lesser hero, and most of them would be understood as generic posi-
tive characters. In this scene, Ke Xiang is the only character who strikes
a liangxiang pose. She is also the only character who sings, and the only
character in white. The scene develops as Ke comes down the steps into
the square, and, while she is still surrounded by black-clad guards, other
black-clad guards hold back the crowd on the left side of the frame.
Ke makes a speech to the crowd, and the film cuts to a close-up on the
militia captain, Lei Gang (雷刚), shouting out “Hear, hear!” The close-up
on him and his speaking role picks him out from the rest of the positive
characters. All the positive characters are dressed in relatively ordinary
clothes in mid-range colors such as browns and blues, but with a lighter
highlight in the form of, perhaps, a white shirt. This distinguishes them
from the black-uniformed guards, as well as from the brilliant white of Ke
Xiang. Among them, Lei Gang is distinguished by his red waistband and
another red cloth tied across his chest. Also, much like Yang Zirong in
Taking Tiger Mountain by Strategy, he wears an animal-skin waistcoat, in
this case a deerskin.
A stylized and highly acrobatic fight follows, in which Ke Xiang is freed.
At various moments, she and Lei Gang are close to each other, and the
camera frames them in two shots, in which they each occupy more or less
the same amount of screen space as they talk, and they are distinguished
from the rest of the positive characters. A slightly low camera angle is
deployed, making them seem grander and more heroic. But, once the
enemy has been trounced, the scene ends in a tableau. Ke Xiang stands on
a platform in the middle of the square, her arm thrust up triumphantly.
Lei Gang is slightly lower than her, and then the other militia fighters are
gathered around looking and pointing toward them. This arrangement
clearly communicates the triple hierarchy of the three prominences.
CINEMATIC POETICS OF THE MODEL WORKS

The analysis of the model works’ hybridity has given an understanding
of the elements composing them and making them distinctive, and the
analysis of the operation of the theory of the three prominences has shown
how those elements are combined and composed to signify the hierar-
chy of characters. But what remains to be further elaborated is how these
works operate to engage the viewer, not only in terms of clear political
signification but also emotionally in an embodied response. The previ-
ous section has already indicated the role of, for example, close-ups, in
accentuating liangxiang poses, or drawing the audience’s attention to the
primary heroes. By close analysis of the cinematic language of the films of
the model works, this section of the chapter attempts to extend that work
to begin to give a more concrete understanding of how cinematic poetics
work redundantly together with all the other elements in the films not
only to signify but also to engage the audience’s passions. Only in this way
can the long-lasting impact and appeal of these films be understood.
In taking up this approach to poetics as about embodied engagement as
well as meaning, this chapter engages with two directions in contemporary
scholarship. One is the broad and sustained commitment in Cinema
Studies and Cultural Studies to go beyond scholarship grounded in the
40 C. BERRY
investigation of meaning through semiotic analysis and to also attend to

the embodied experience of cinema before and beyond semiotic analysis.
This includes work on the Deleuzian idea of affect as bodily felt percep-
tion prior to cognitive processing (Massumi 2002); body genres (Williams
1991); phenomenology and the cinema (Sobchack 1991); revived inter-
est in the Eisensteinian idea of “attractions” (Gunning 1986); and much
more.
The second and more recent direction is the emphasis on passion in
Political Science. In specific regard to the Cultural Revolution, Michael
Dutton (2015) returns to the famous Rent Collection Courtyard group
of sculptures from the period and analyzes them as a “machine” for pro-
ducing not only meaning but also emotional, embodied responses on the
part of spectators, who were often moved to tears, outpourings of bitter
memories, and angry accusations. Dutton relates this mode of engage-
ment to Schmittian political theory and its emphasis on passion as crucial
to effective political mobilization. But in the context of this chapter, it is
noteworthy that Dutton’s work also effectively constitutes an investiga-
tion of the poetics of the Rent Collection Courtyard piece.
Relatively little attention has been paid to the cinematic poetics of the
model works so far. Perhaps this is because the general perception appears
to be that they are little more than recordings of stage works. Wang Qi
(2014: 32) writes that, “the model plays and films produce an extremely
flattened representational space, in which the characters have limited
depth as regards personality.” Although she is referring specifically to lack
of character complexity in this quote, it seems she also perceives the literal
space to be flattened: “Spatially, the mise-en-scène in the film reproduc-
tions follows the theatrical presentation and shows the stage on which the
performance takes place” (Wang 2014: 30).
Wang’s characterization of film versions of the model opera works is
correct, but it is also important to point out that the film versions do not
simply set the camera up in front of the stage and record the entire perfor-
mance in long shot and long take mode. Instead, I would argue that they
develop their own way of participating in the redundant orchestration of
elements, and that they do so in a manner that is designed to maximize
affective response on the part of the audience. Taking Tiger Mountain by
Strategy and The Red Detachment of Women were among the earliest of the
model works to be filmed, in 1970 and 1971, whereas Azalea Mountain
was filmed a few years later, in 1974. While the first two films have rela-
tively simple and indeed frontal cinematography with long takes, Azalea
Mountain has more dynamic cinematography, making the space much less
flat and with a speedier editing rhythm. Despite these differences, the lat-
ter is a more developed version of the strategies for filming these stage
performances that are already manifested in the earlier films rather than an
altogether new way of filming the model works.
The scene in which the detachment commander in The Red Detachment
of Women, Hong Changqing (洪常青), and his sidekick come across the
injured girl, Wu Qinghua (吴清华), can serve as an example of the early
cinematography and editing patterns in the film versions of the model
works. She has run away from the landlord, been caught, punished, and
left unconscious in an overnight rainstorm. The whole scene lasts approxi-
mately five minutes, and is divided into 18 shots. Two types of shot pre-
dominate: full shots that show the complete figures of the dancers as they
perform, and which are reframed to keep the dancers in the center of the
shot; and close-ups, which occur at moments of high drama when the
dancing stops and poses are struck. The full shots are mostly long takes,
and the close-ups much briefer. A shot breakdown is given in Table 3.1.
The logic of the cinematography and editing is subordinated to the
rhythms of the existing stage performance and the music, with the need
to show the dancing bodies determining that the full shot should be
used in many cases. But what also must be emphasized is how the cam-
era movement and the cutting to close-ups follow both the logic of the
three prominences and the logic of the dramatic passions the model works
seek to elicit. When the two men are dancing their process of checking
for enemies in shot 3, the camera tracks with Changqing, because he is
the hero and his sidekick is only a positive character. Similarly, the close-
ups are portioned up between the two heroic characters, Changqing and
Qinghua, and the sidekick does not get any.
But at least as important is the way the cinematography and editing
follow the emotional logic of events, encouraging the audience to feel the
passion of the characters and the situation. Shot 2 underlines the men’s
vigilance at the same time as it encourages us to feel Changqing’s heroic
passion when it moves in to a close-up. The close-up in shot 4 builds
on this logic when he sees something, and with shots 6 and 7, there is
the moment of exposure of the wounds. Robert Chi (2008) has written
at length about the importance of the dramatic exposure of Qinghua’s
wounds in the original feature film version of the Red Detachment story as
producing a somatic force that is felt by the audience. Here Changqing’s
shock and sympathy guide our response, and the close-up directs us
Table 3.1 Hong Changqing discovers Wu Qinghua in Red Detachment of Women
Shot no. Duration Size, angle Movement Music Narrative
42
1. 16 seconds Fade from Camera pulls back from A single note Hong Changqing and his sidekick enter
black to reveal night-time vegetation followed by a steady from the left, checking for danger as they
full shot and tracks right as but rapid percussion move through the forest at night
characters enter scene note on a cymbal,
C. BERRY
communicating
heightened attention
2. 7 seconds CU, low angle High notes, strings, Hong Changqing’s shining face, as he
urgent and dramatic stares past and over the camera, into the
middle distance
3. 1 minute, 3 Full shot, as 1 Pans with both men to The two men dance the process of
seconds keep them centered, searching the setting, relaxing as they
then follows Hong decide all is safe, putting their guns away,
when they go in and then turning their backs on the
different directions, camera and heading off
before returning to the
two of them together
4. 2 seconds CU Urgent low note of Hong turns round, noticing something
alarm
5. 20 seconds Full shot Camera reframes Discovering Wu Qinghua, Hong goes to
slightly to keep Wu assist her. As she comes to, she becomes
Qinghua in the center alarmed and dances fearfully between
of the shot them
6. 14 seconds Full shot, Cut reframes Wu and Move to close-up occurs as Hong sees the
tracking or Hong, each on one wounds on her arms and goes to staunch
zooming into side of image them with a handkerchief
a medium CU,
low angle
7. 52 seconds MCU, pulling Pulling back to full Close-up of Wu’s head looking away and
back to a full shot allows all figures then turning back to see Hong staring at
shot to be visible, and then her wounds. She grasps the men are not
reframes to center Wu her enemies, and then a pas de deux begins
when she dances along with Hong, in which we understand she is
narrating what has happened to her, and
then becoming distraught
8. 51 seconds MCU, low Tracking and moving The musical pace Close-up of Hong’s resolute expression,
angle, pulling with Hong to keep him becomes slightly before he goes and discusses something
out to full shot in the frame slower and more with his sidekick, and then they come and
steady, as Qinghua dance with Qinghua. This includes the
becomes more moment when Hong supports Qinghua as
assured she goes en pointe, and they gesture
toward the future beyond the camera,
somewhere off-screen to the right, toward
which she moves at the end of the shot
9. 9 seconds Full shot Hong has his back to us. Qinghua turns
back to look at him. He is holding
something out to her
10. 2 seconds ECU Two coins in his hands
11. 6 seconds CU Single string Qinghua looks down, then back up at
instrument playing a him, moved
high, poignant note
12. 2 seconds CU, low angle Hong gazes back at Qinghua
13 13 seconds Full shot Qinghua backs away from Hong en pointe,
and he offers the money for a second time
14. 6 seconds MCU Changqing offers the money for a third
time
15. 11 seconds MCU to CU Qinghua’s face, full of emotion, as she
moves forward to take the money
16. 15 seconds Full shot Camera tracks Qinghua takes the money, moves away
Qinghua’s movements from Changqing, then back to him, bows,
and then exits frame right
17. 8 seconds CU, low angle Changqing looks in the direction of
Qinghua’s departure
18. 10 seconds Full shot, Camera tracks Changqing moves to the left to rejoin his
RED POETICS: THE FILMS OF THE CHINESE CULTURAL REVOLUTION...
ending on CU, Changqing’s sidekick, they repeat the dance of checking

low angle, movements the environment, and the scene ends as
43
then fade to they freeze in a pose and the camera moves

black into a CU of Changqing’s radiant face
44 C. BERRY
toward the sight of the wounds as the trigger for his feelings, and then the
next close-up on Qinghua communicates the emotional depth of both her
suffering and her consequent resentment. The close-up at the beginning
of shot 8 communicates the moment that these feelings are transformed
into resolution on Changqing’s part.
The close-ups in shots 10, 11, 12, 14, and 15, cut back and forth
between Changqing and Qinghua, are the core and emotional high point
of the scene. They almost function as a shot and reverse shot structure,
except that the angles are closer to those of audience members looking left
and then right, following the logic of the drama, rather than those of the
characters themselves, with the exception of the final shot, which seems
almost from Changqing’s perspective. For audiences trained in Hollywood
drama, it is difficult to avoid seeing this as an erotically charged moment
between a man and a woman. Certainly there is passion here, but the
offering of the money represents material commitment on his part and the
receipt of it the beginnings of political commitment on her part. In this
way, the signified political meanings of the film are also felt by the audi-
ence, as the cinematic poetics of the scene redundantly combines with all
the other elements to try and produce an inescapable and overwhelming
passion.
A similar pattern of cinematic poetics can be found in Taking Tiger
Mountain by Strategy. In scenes of action involving many characters or
even lengthy speeches, long takes and full shots predominate, with track-
ing and reframing keeping the main hero in the center of the frame. Short
close-ups underline moments when poses are struck. In Azalea Mountain,
much the same logic continues to guide the way scenes are shot, suggest-
ing a consistent cinematic poetics for the film versions of the models. But
this time, the cutting is faster and the camera does not stay so consistently
back, as if with the audience in the front stalls. This more evolved style is
more dynamic.
This difference can be readily discerned if we compare the opening
scene of Azalea Mountain with the scene from Red Detachment of Women
analyzed above. Both scenes have a small number of characters and last
about five minutes. But, whereas the scene from Red Detachment of
Women has 18 shots over the whole five minutes, as Table 3.2 shows, the
first scene of Azalea Mountain uses 20 shots in just over two minutes.
This table represents the opening of the scene, in which Mother Du (杜妈
妈) lends the fugitive Lei Gang an axe to break his chains. As it goes on,
he discovers she is the mother of one of his fighters who has died. Having
broken the bitter news to her, he adopts her as his own mother, and she
in turn says her son might be dead, but she still has a grandson to join the
cause.
In both films, the cinematography and editing combine with the music,
acting, costumes, and so on in a pattern of redundant mutual reinforce-
ment. Also, the cinematography and editing are subordinated to the
requirements of the existing stage opera: when there are set pieces, arias,
or acrobatic performances, as in shot 6 in Table 3.2, the camera stays back
and maintains a full shot to display the performance uninterrupted.
However, not only is the cutting faster-paced in Azalea Mountain but
also there is an effort to use the camerawork to move the audience more
directly into the mise-en-scène, creating a more 3D and less flat experi-
ence. The cut from the first to second shot follows the standard classical
film logic of moving from a master shot into a tighter shot. But when we
cut to the third shot, at first, we do not know where it lies in relation to
the location shown so far and also beyond the first shot. This expands
our imagination of the location beyond a simple stage set. When the men
chasing Lei Gang race in from the lower left, we are able to link the spaces.
But then shot 4 cuts to a close-up of vegetation that could be anywhere
in the forest at night that we have been encouraged to imagine. Similarly,
although the narrative logic of Lei’s jump from shot 4 to the clearing in
shot 5 is clear, it is only with shot 6 that we get a master shot that enables
us to put these spaces together more precisely. Similarly, when Mother Du
comes up the mountain, the initial crosscutting between her and Lei Gang
does not feature a master shot. Not only are we taken deeper into the
mise-en-scène than a full shot in the metaphorical “front stalls” position
would take us but also the almost shot and reverse shot structure between
them combines with the percussion to make us feel the full tension of the
moment.
The same narrative logic of cruel experiences and treasuring up class-
based resentment, accompanied by the iconography of wounds, is found
in both Azalea Mountain and Red Detachment of Women. This different
style of cinematography and editing in Azalea Mountain compared to
Red Detachment of Women does not change the signification of the scenes
analyzed in terms of either narrative events or their political meaning.
However, it does work to create a more embodied and emotional experi-
ence for the audience in Azalea Mountain. Here, we can see that cin-
ematic poetics of the model opera film versions move the audience both
to understand and to feel.
Table 3.2 Mother Du frees Lei Gang in Azalea Mountain
46
Shot no. Duration Size, angle Movement Music Narrative
1. 11 seconds Fade from Cymbals, with much Night, handheld lanterns move from
C. BERRY
black to reveal shouting and dog barking deep left to foreground center, and off
LS, high angle to right
2. 3 seconds MS, high angle Percussion punctuating Darkly clad men with lanterns rush
action here and from lower left to upper right. One
throughout scene shouts “Look!” and they turn to look
toward lower right
3. 9 seconds LS, becoming A figure moves in the depths of the
CU dark. Lead pursuer enters frame from
bottom left, fires a couple of shots, and
sends men after him
4. 10 seconds CU Pans right to find Lei Gang parts the vegetation and looks
Lei Gang around. He sees something and leaps to
the right and down
5. 3 seconds CU Chains added to Lei Gang’s chained feet land in a
percussion clearing
6. 35 seconds LS Full range of instruments, Lei dances his indignation and
including wind and strings frustration, sitting to try and smash
chains with a rock
7. 2 seconds LS Mother Du coming up path on
mountain, notices something off-screen
right
8. 2 seconds MS Rapid low wood Lei hears something behind him and
percussion when Lei turns turns
communicates alarm
9. 2 seconds As 7 Rapid wood percussion of Mother Du notices something
alarm continues off-screen right
10. 4 seconds MCU Camera reframes Lei gets up and turns
to keep Lei
centered
11. 2 seconds MCU Du looks at him and then her gaze
drops
12. 1 second CU, high Lei’s shackled feet
angle
13 6 seconds As 11 Cymbal not punctuates her Du offers Lei her axe
move
14. 1 second CU Lei’s face
15. 3 seconds As 11, 13 Du offers Lei the axe again
16. 2 seconds MCU Lei understands, smiles, and moves
forward to receive the axe
17. 2 seconds As 11, 13, 15 Percussion pace picks up Lei moves into frame and takes axe
from Du
18. 6 seconds LS Lei’s smashing movements Lei moves to smash his chains in the
are marked percussively clearing, as Du moves to the
foreground, keeping a lookout, moving
out of frame to the right
19. 2 seconds MS Du keeping watch
20. 28 seconds LS, becoming Lei smashes his chains, then moves to
MCU foreground to return axe, which Du
takes in her left hand. She then offers
him a yam with her right hand, and he
is moved. He takes it, and they begin to
talk
RED POETICS: THE FILMS OF THE CHINESE CULTURAL REVOLUTION...
47
48 C. BERRY
CONCLUSION
In his book on the poetics of Ozu’s cinema, David Bordwell insists that
the features of Ozu’s work can only be accounted for by putting them
“into context,” making the discussion of “background” part of the work
of poetics. However, he goes on to write:
we can treat Ozu’s films as lying at the core of a set of concentric circles. We
cannot simply link the outermost circle – that is the broad and general fea-
tures of Japanese history and culture – to these films. The concentric circles
in between represent the more pertinent and concrete forces impinging on
the film – such forces as Ozu’s working situation, the film industry, and the
proximate historical circumstances of his milieu. (Bordwell 1988: 17)
This chapter has begun the process of examining the poetics of the films
of the Cultural Revolution model works. For Bordwell, his emphasis on
the “inner” circles of determination leads him to pay relatively little atten-
tion to broader social and cultural issues and to focus on the formal quali-
ties of Ozu’s works as developing in conversation both with the poetics of
other cinemas and also increasingly with his own earlier works.
In the case of the film versions of the model works, one would be
hard pushed to see their cinematic poetics as developed in an autonomous
realm of either industry or art form logics. All the published accounts of
the initiation and development of the model works in their various forms
attest to the fact that they were initiated through the cultural policies of
the Chinese Communist Party and its government, and that high officials,
including Mao’s wife Jiang Qing, took a very close interest in all aspects.
Paul Clark’s account (2008) is full of accounts about her directions and
advice, including, in the case of many film versions, demands for reshoot-
ing to reach her required standards.
However, although the production of the model works was strongly
shaped by politics, Bordwell’s emphasis on the autonomy of form may be
more useful in understanding the endurance of the model works films and
their imagery. Few if any audiences today go to see the stage performances
of the model works that are still put on in China or to buy the various con-
sumer items that make use of the iconography of the Cultural Revolution
in order to be filled with class hatred and pour out onto the streets seeking
out targets for their anti-bourgeois feelings. Precisely why they are inter-
ested in these works today requires further investigation. But this ability
to set aside the original political meanings of the model works suggests a
different type of autonomy; not autonomy in the realm of production but
autonomy in the realm of reception and consumption. This suggests that
new meanings get attached to the poetics of the model works, including
the film versions of the model works, by different audiences at different
times. Therefore, perhaps we can say that the very distinctive, innova-
tive, redundantly clear, and dynamic cinematic poetics of the model works
investigated here have a formal aesthetic appeal in their own right, aside
from or as well as the particular meanings that different audiences attach
to them.
CHAPTER 4
Renewal of Song Dynasty Landscape

Painting Aesthetics Combined
with a Contemplative Modernism
in the Early Work of Chen Kaige
Peter Rist
Surprisingly, an interest in the glories of Chinese art history was virtu-

ally absent from mainland Chinese film in the past. The Shanghai film-
makers of the 1930s and 1940s were involved in producing their own
modern brand of cinema that necessarily had to appeal to local audiences
through characterization and ideology—young people working together
against oppression through a kind of natural socialism—while emulating
the fast moving nature of Hollywood product. This model was followed
in the 1950s and 1960s by that of state-supported Communism, where
although the natural landscape would often appear, invariably it was used
as a heroic backdrop to revolutionary characters, ala Soviet Realism. It
wasn’t until after the Cultural Revolution that students of the state-run
Beijing Film Academy (BFA) had access to the full history of film, includ-
ing their own, as well as the great depth of China’s art history. Naturally
enthused by films of the numerous international ‘new waves’ of the
1960s, and moved by the richness of their specifically Chinese culture,
P. Rist ( )
Concordia University, Montreal, Canada

DOI 10.1057/978-1-137-55309-6_4
52 P. RIST
the filmmaking graduates were free for the very first time to regard the
past while seriously thinking of the future. By focusing on two of the first
three films directed by Chen Kaige—Yellow Earth (Huang tudi, 1984),
and King of the Children (Haizi wang, 1987)—I propose that Chen (and
his cinematographers) incorporates a keen understanding of the aesthetics
of Chinese landscape and narrative scroll painting combined with a need
to produce original, at times reflexively ‘modernist’ work. Another inter-
est here is that the ancient practice of Chinese painting was accompanied
by theory, to the extent that some theoretical principles of media—ink,
brush, silk, and paper—can be compared to aspects of twentieth century,
medium specific, high modernist art, where, say the representational
nature of painting, photography, and film becomes less important than
the medium itself. In a sense, then, I argue that these film works are
‘experimental’ both narratively and visually, through both Western and
Chinese interpretations.
To note that the tradition of landscape painting in China is ‘ancient’ is
a huge understatement. According to Yang Xin, the ‘use of brush and ink’
was developed during the Eastern Zhou dynasty (ca. 1100–256 B.C.) and
that ‘basic brush-made shapes have changed little since then’ (Yang 1997,
p. 1). Although not as old as figure painting, landscape painting became
highly developed at the end of the seventh century A.D. and the begin-
ning of the eighth, during the T’ang Dynasty, especially in the ‘green and
blue’ work of Li Ssu-hsün and his son, Li Chao-tao1 (Buhot 1967, p. 120).
During the Five Dynasties period (907–960), a Jiangnan landscape style
emerged around what is now Nanjing in the Southern Tang kingdom,
and copies of tenth-century hanging scrolls survive in the collection of the
Taipei museum as well as both a handscroll, The Xiao and Xiang Rivers
and a hanging scroll, Pavilion on the mountains of the immortals by the
Jiangnan master, Dong Yuan (d. 962).2
It was during the Northern period of the Song Dynasty (960–1127)
that the high mountain aspect of Tang-style landscape painting reached its
zenith with masterpieces such as Kuo Hsi’s (Guo Xi) Early Spring (1072),
the complexity of which is found in the ‘“curving” lines of mountains,
trees, and rocks’ as well as the atmospheric ‘naturalism’ and ‘life’ created
by ‘using blank areas of silk to suggest the penetrating clouds and mists’
(Ch’en 2003, p. 25).3 Other surviving masterpieces of the Northern
Sung are dominated by the verticality of a craggy mountain peak, coursed
by streams, dissected by mist and clouds, and dotted with trees; and, if
human or other figures are present, they are rendered minuscule in the
RENEWAL OF SONG DYNASTY LANDSCAPE PAINTING AESTHETICS COMBINED... 53
vast natural environment.4 Predominantly, these works are hanging scrolls,

a format which encourages a vertical scanning of the image, from land and
water to sky, or vice versa.
During the Southern Song dynasty period (1129–1279), one finds a
number of important developments in painting, including use of ink on
paper, rather than silk, and the perfection of the horizontal handscroll,
which would be gradually unraveled and ‘read’ from right to left to
reveal a theme or progression of some kind.5 A great surviving example
of landscape handscroll in the National Palace Museum, Taipei, is Pure
and Remote [Views of] Mountains and Streams (ca. 1200, 46.5 cm × 889
cm) by Hsia Kuei (Xia Gui, active 1190–1225) for which Watson declares
that ‘No better example can be found of perspective depth created by the
weight of black ink,’ and that a receding view into the work reveals a ‘uni-
form mist through which just enough of hill and coast penetrates to define
the total terrain’ (Watson 2000, p. 55). Landscape painting tended to
move away from the monumentality of the Five Dynasties and Northern
Sung periods to a quieter, more lyrical approach and from a harsh ‘real-
ism’ toward a more simplified style, where areas of empty space are promi-
nently included, and where there is a greater emphasis on depth.6 A good
example of these tendencies is found in Ma Yüan’s On a Mountain Path
in Spring (active before 1189–after 1225) (Fig. 4.1), where the mountain
itself is barely sketched, and where the left bottom corner of the frame is
filled with detail and one of the two male figures (larger than normal),
standing in the foreground, gazes up and right past delicate hanging wil-
low branches to a distant bird, alone in the top right quadrant except
for written verses of a poem: ‘wild blossoms dance when brushed by his
sleeves, the secluded birds cut short their song to avoid him’ (Ch’en 2003,
pp. 24–5). This painting exemplifies the bringing together of the different
art forms of the ‘Three Perfections’—poetry, calligraphy, and painting—in
Southern Sung court art.7
The oldest surviving written reference to painting, by Confucius, dates
from the sixth century B.C., and some Notes for a Landscape were written
by Ku K’ai-chih (Gu Kaizhi) in the fourth century A.D., while, arguably
the most significant critical text, The Six Techniques of Painting, was writ-
ten in the fifth century A.D., by Hsieh Ho (Xie He) (Lin 1967, p. 21,
pp. 27–9, pp. 34–5). The first two of these ‘techniques’ are judged to
be the most important, ‘creating a life-like tone and atmosphere (ch’i-
yün sheng-tung)’ and ‘building structure through brush-work.’ Dr. Lin
Yutang devotes a long discussion to the problems of translating Chinese
54 P. RIST
Fig. 4.1 Ma Yüan’s A Mountain Path in Spring
characters and writes of Hsieh Ho’s first technique: ‘As the Chinese use it
ch’i-yün is a bisyllabic word, a noun meaning tone and atmosphere; sheng-
tung is another bisyllabic word, an adjective, meaning fully alive, moving,
lifelike. The whole phrase means a “vital tone and atmosphere.” It sug-
gests a successful creation of tone and atmosphere that is moving and
alive, and by all Chinese criteria this tone and atmosphere, rather than
verisimilitude, is the goal of a painting’ (Lin 1967, p. 36). According to
Li’s translation, the other ‘techniques of painting’ are ‘third, depicting the
forms of things as they are; fourth, appropriate coloring; fifth, composi-
tion; and sixth, transcribing and copying’ (Lin 1967, p. 34). Alexander
Soper takes a slightly different approach, translating the first ‘condition’
of ‘(good) painting’ as ‘animation through spirit consonance,’ where ‘first
importance’ is ‘given to some quality that is never obtainable by technique
alone,’ and thus separating out artistry from craftsmanship (Sickman and
Soper 1961, p. 133). According to Sickman (aided by Soper’s transla-
tion), the ‘remaining five Principles are all concerned with the making
of a picture and involve technical procedures.’8 Clearly, the seeds of a
kind of ‘realism’ in the first, third, fourth, and fifth principles, a creative
‘impressionism’ in the first, second, and fifth principles, and, even, a ‘mod-
ernist’ regard for the medium itself—the properties of brush and ink—are
evident in the second principle. Along these lines, one could understand
that the Northern Song Dynasty featured a period of ‘realist’ landscape
painting, followed by an ‘impressionist’ tendency in the Southern Song
Dynasty, while a ‘modernist’ concern with the play of brush and ink on the
silk surface together with the texture of this medium and its scroll format
pervaded all.
If we return to look at In a Mountain Path in Spring, we can detect
a strong sense of what is now termed ‘intermediality’ (in the context of
twentieth-century postmodernism), where the ‘three perfections’ are
combined: the painter’s brush strokes are matched by the calligraphy in
the right-hand corner, which in turn links the words of the poet to the
subject of the painter. Perhaps even more remarkably, we can understand
the painting as an example of reflexivity, where we see the human form
of the scholar, who could have been drawn in the image of the artist.
In a smaller fan painting attributed to Ma Yüan, Gazing at a Waterfall,
the subjective figure of a scholar is looking down into an unclear space.
With the centrally framed mountain in the background typically obscured
in blank mist, a waterfall is clearly observed to the left of the painting
and away from the gaze of the scholar. One possible interpretation of
the painting is that the man may, in fact, be listening to the sound of the
waterfall rather than observing, perhaps persuading the observer of the
artwork to think of the sound of nature as well as its image, while possibly
alluding to a fourth art form: music. As Watson writes: ‘When persons are
introduced significantly into the scenes, which from the time of Li Tang
[Li T’ang, ca. 1050-after 1130] became increasingly the academy custom,
the theme is anthropocentric in a sense quite foreign to early tradition; the
feeling of the internal viewer of nature is prescribed and even a moment
of psychological perception indicated — the sound of the waterfall, the
glimpse of a bird’ (Watson 2000, p. 54).
In looking at Chinese Song Dynasty landscape paintings and narrative
handscrolls with Western eyes, it is easy to notice the differences from
European and North American works. Thus we notice multiple perspec-
tives and vantage points (and the absence of a single vanishing point), a
certain amount of flatness rather than sculptured depth, the smallness of
figures and objects, and vast areas of space containing nothing at all. As
valid as these perceptions of difference might be, it seems to me that the
tendencies for ancient Chinese art to be thought of as ‘exotically’ different
56 P. RIST
from Western norms (and hence ‘oriental’) and to depart from reality,
are emphasized too much. Clearly, the Chinese people have a tremen-
dous affinity with nature, which has always been reflected in their art. Of
Chinese landscape art, William Watson wrote in ‘Realism as Landscape’
that ‘However philosophized the meaning of landscape painting became,
its style never parted company completely with reality’ (Watson 1974,
p. 83). Arthur de Carle Sowerby went even further, in his book on nature
in Chinese art, where he illustrates his chapter entitled ‘Rocks, Mountain,
and Water’ (from the ‘Chinese name for a landscape, shan shui, literally
“Mountains and Water”’) with contemporaneous black and white photo-
graphs of Hua Shan (in Southeastern Shenshi [Shaanxi]) and the Huang
Shan or Yellow Mountains (in Southern Ahnwei [Anhui]), in comparison
with monochrome reproductions of classic landscape paintings, in order
to demonstrate how ‘realist’ these paintings of tree-speckled, craggy, mist-
laden peaks were (De Carle Sowerby 1940, p. 153, pp. 161–8). With
this penchant for the natural world together with the temporal dimension
enabled by the art of scroll painting, one would think that the invention
of cinema would be embraced in China, especially for its ability to reveal
the landscape, alive with movement.
During the silent film era, which lasted until 1936 in China, there was
very little aesthetic use made of the landscape and even less in the way
of allusion to landscape painting.9 Of course, with such a disastrous sur-
vival rate for silent films, it is impossible to make such a claim with any
real certainty. In 1995, Derek Elley translated into English and Italian
the only reliable filmography of pre-revolutionary Chinese film for the
years 1905–1937 (to the outbreak of the Sino-Japanese War, in July). He
claimed that of the approximately 1100 titles listed, less than five per cent
had survived (Elley 1995, p. 4).10 Many films have been rediscovered since
this time, including a number produced by the Lianhua Film Company
(United Photoplay Service). Together with Mingxing (Star), Lianhua was
the most notable of the Shanghai-based feature film production houses,
and of the 83 feature films (1930–37) listed by Elley, at least 30 are known
to be extant (36 per cent), one of the best survival rates for films of this
era in East Asia. Two of the most prominent Lianhua directors who con-
sistently shot large sections of their films on location were Sun Yu, who
was trained in the USA, and Cai Chusheng. Typically, a bucolic rural life
is contrasted with the corrupt, mechanized, fast-paced world of the city in
Sun’s films, and they often open in the countryside, on or near a river, for
example, Tianming (Daybreak, 1933), Huoshan Qingxue (Loving Blood of
the Volcano, 1932), and Xiao Wanyi (Small Toys, 1933). Although none of
the shots cover a large range of narrative incident, or are long takes, the
elegant camera movements in these films, directed by Sun Yu and shot by
Zhou Ke, can certainly be compared to the effect of unraveling a Chinese
horizontal scroll landscape painting.11
Whereas camera movement graced Sun Yu’s Lianhua films, Cai’s work
contained passages of what Catherine Yi-Yu Cho Woo says are linked with
the ‘soul of Chinese painting and poetry’ (Woo 1991, p. 22). However,
the Lianhua director who was most clearly conscious of ancient and tradi-
tional Chinese art and culture, and who was a relatively conservative figure
working in the left-leaning industry, was Fei Mu. He directed ‘nationalist,’
left-wing films such as Langshan diexue ji (Bloodbath on Wolf Mountain,
1937), but was best known for making films that espoused traditional
ethics, such as Tianlun (Song of China, 1935). This film was produced
by the right-leaning head of Lianhua, Luo Mingyou, in order to assuage
the Kuomintang (KMT) government. Only seven reels of its original 14
survive in an English and Chinese-titled version, with a traditional Chinese
musical soundtrack (by the Wei Chung Lo Orchestra), which had been
distributed in the USA by one Douglas MacLean.12
Fei Mu was only credited as Assistant Director (with Luo Mingyou
credited as Director), but it is fairly obvious from the visual evidence of
the extant Tianlun, with numerous low-angle shots of characters framed
against the sky, reminiscent of similar shots in Xiao cheng zhi chun (Spring
in a Small Town, 1948), that the real ‘director’ of the film was Fei.13
Whereas the titles clearly derive from Luo’s thinking—we are informed
in the film’s first title that ‘Filial piety’ is the ‘song of China’—the images,
which consistently mix the dynamism of Soviet and Hollywood film struc-
ture (high and low angles, rapid camera moves, shot/reverse shot nar-
rative scene construction) with Chinese aesthetic elements, are surely
attributable to Fei. The juxtaposition of country versus city in Tianlun is
even more skewed in favor of the former than in Sun’s films. Even before
the film’s credits, we witness an idyllic scene of a goatherd rescuing one of
his young flock. At least one scene suggests subtle links back to landscape
painting and ahead to the mastery of Spring in a Small Town. A younger
sister (played by Chen Yan Yan), who, unlike her city-loving brother, has
always shown loyalty to her parents, is contemplating leaving her home in
the countryside, and eloping. Her father has just told her that she doesn’t
have his permission to marry, and she is filmed from behind, in high angle,
seated on a riverbank. A reverse-angle cut is made to a frontal view of her,
58 P. RIST
with a tree in the background and the shadows of leaves flickering on the
ground. She seems to be contemplating her future and a cut is made to
a long shot view of the river. We see dwellings and trees in the far back-
ground, and a lonely figure, punting. We assume that this shot is seen
from the young woman’s perspective, although this isn’t necessarily the
case. As it develops, it becomes the longest take in the extant version of
Tianlun, with a large sailing barge gradually drifting into the film frame
and filling the left foreground space. It is a beautiful, reflective shot, invok-
ing the sense that this place is so peaceful, and the thought that she will
be missing the riches of a natural life if she chooses to leave it behind.14
Fei’s brave attempt at making a classically art-directed, contempla-
tive film under extremely difficult conditions during the ‘Orphan Island’
period in Shanghai, resulted in Confucius (1940), which has recently been
restored by the Hong Kong Film Archive.15 And with Spring in a Small
Town, he successfully combined the great traditions of Chinese landscape
painting in the Northern and Southern Song Dynasties which I discussed
earlier as being akin to ‘realism,’ ‘impressionism,’ and ‘modernism.’ We
can detect ‘realism’ in the film’s settings and performances, ‘impression-
ism’ in the creation of moods and in exterior (natural and architectural)
reflections of characters’ interior states, and to a ‘modernism,’ espe-
cially through the use of a female character’s ironic voiceover narration.
In Jim Udden’s book chapter, ‘In Search of Chinese Film Style(s) and
Technique(s),’ he claims that Spring in a Small Town is arguably the first
stylistically original Chinese film, especially in its use of voiceover narra-
tion, and the greatest of its time, whereas he does not find clearly dis-
tinctive ‘Chinese’ cinematic characteristics in other Shanghai-made films
of the 1930s and 1940s; for example, through the statistical analysis of
14 Chinese films, he shows us that there were no appreciable differences
between their average shot lengths (ASLs) and those of other films (Udden
2012, pp. 272–6; pp. 265–9).16
There are probably many reasons why the great historical tradition of
landscape painting is not prominently reflected in Chinese films during
the silent era and beyond.17 Indeed, the most likely explanation is that of
‘modernity.’ Cinema was regarded everywhere as the modern (‘seventh’)
art form of the twentieth century, and as a commercial medium of enter-
tainment. The technology of cinema was invented in France, Britain, and
the USA and was brought into China by entrepreneurs from these coun-
tries.18 Zhang Zhen provides a very interesting account of the situation
in Shanghai where the ‘time lag between early Euro-American cinema
and early Chinese cinema speaks certainly to the semicolonial nature of

Chinese modernity, especially with regard to “belated” technological
transfer and implementation. This temporal disparity, ironically, also sup-
plies testimonies to the persistence of early cinema not so much as a rigidly
defined aesthetic or period category, but as an emblem of modernity, or
rather multiple modernities’ (Zhang 2001, 235–6).19
After several attempts at starting real film education along the lines of
the Soviet model in Communist China, the BFA opened in June 1956 with
programs in directing, acting, and cinematography (Clark 1987, p. 63).
Everyone knows that the students who graduated in 1982 from the first
class to be admitted after the Cultural Revolution (in 1978) have been
labeled the ‘Fifth Generation.’ But we seem to have lost the understanding
that the label initially referred to a generation of film students rather than a
generation of filmmakers. As Klaus Eder wrote in 1993, they were the ‘5th
generation of film-makers having been educated since 1949’ (Eder 1993,
p. 8). In an interview with Eder, Li Shaohong, one of the female directing
students who graduated in 1982, suggested that ‘The 4th generation had
already entered film school in the 1960s and had already graduated before
the Cultural Revolution,’ and that the ‘3rd generation, were the students
brought up by Soviet experts in the 1950s’ (Li 1993, p. 100). There are
different interpretations of what constitutes the first three generations,
but there is general agreement that the ‘Fourth Generation’ constituted
those students who graduated from the BFA just before the outbreak of
the Cultural Revolution and its subsequent closure.20
According to Ni Zhen, who was a professor of Art Direction and Film
Theory at the BFA from 1980 to 2000, the core courses for students in the
Directing Department in 1956 ‘were the “Art of Film Direction” (which
included acting classes), “Film Montage Theory,” and “Screenwriting,”
and they were designed after consulting Soviet teaching materials’ (Ni
2002, p. 56). Ni asserts that all classes followed this model and that the
same core system was revived in 1978 (Ni 2002, p. 56). Indeed it is clear
that People’s Republic of China (PRC) filmmaking from 1949 until the
Cultural Revolution was dominated by ‘socialist realism,’ and that, typically
when exterior scenes were filmed away from the studio, the Chinese land-
scape functioned as a natural backdrop to the heroic foreground action.
For example, Yingjin Zhang writes, ‘A glorification of the eight female
martyrs who fought Japanese invaders in northeastern China, Daughters
of China (Zhonghua nüer, dir. Ling Zifeng, 1949; b/w) is shot largely on
60 P. RIST
location in forests and establishes a rugged look of cinematic realism and a

fresh documentary aesthetic’ (emphasis in original) (Zhang 2004, p. 193).
Arguably, the most prominent genre for the deliberate, aesthetic
use of the natural landscape was the ethnic minority film. The Uighur,
northwestern-set, black and white, Visitor on Ice Mountain (Bingshan
shang de laike, dir. Zhao Xinshui, 1963) uses the Pamir mountain back-
drop as a frontier/barrier and threshold to spying and political intrigue.
Perhaps the most picturesque of these films is the Bai ethnic musical
romance, shot in rich color, Five Golden Flowers (Wuduo jinhua, dir. Wang
Jiayi, 1959), which brings a horseman from the countryside to find his
beloved Jinhua, a party official in Dali, Yunnan province.21 In the first
half of the film, there are many panning and tracking establishing shots
that view the Cangshan mountains behind beautiful Lake Er Hai, and
that could have been included to deliberately evoke handscroll painting.22
Nevertheless, the closest connections between classical landscape painting
and Chinese films of the first 17 years of the PRC can be found in the
work of the Shanghai Animation Film Studio, especially the short, ink
wash films directed by Te Wei: Where is Mama (Xiao ke dou zhao ma ma,
1960) and The Cowherd’s Flute (Mu di, aka, ‘Buffalo Boy and the Flute,’
co-dir. Qian Jiajun, 1963). Where is Mama was inspired by the twentieth-
century water color painter, Qi Baishi, who was in turn inspired by the
Ming dynasty paintings of Bada Shanren, while The Cowherd’s Flute was
based on the work of Li Keran, a follower of Qi who was famous for ink
wash paintings of water buffalo and who combined Western oil painting
techniques into his landscape work.23
Although there is now a growing recognition that ‘Fourth Generation’
Chinese filmmakers were experimenting with narrative incident, structure,
and style, when the film studios reopened after the Cultural Revolution,
the appearance of Yellow Earth (Huang tudi, 1984) at the Hong Kong
International Film Festival on 12 April 1985 announced the arrival of a
completely new kind of Chinese film, and made stars of its director, Chen
Kaige and cinematographer, Zhang Yimou.24 It is very unlikely that clas-
sical Chinese art, including landscape painting has ever been on the cur-
riculum of the BFA for budding directors and cinematographers.25 But,
in an interview with Klaus Eder at the Cannes International Film Festival
in 1993, Chen noted that they were the first ‘generation’ to ‘break away’
from Soviet-inspired ‘communist realism’ and that they were ‘controlled
by this ideal that we must make visually beautiful films,’ as well as their
being ‘influenced by Chinese classic art, by the way Chinese artists did
things—painting, poems, some classic novels’ (Chen 1993, p. 92). Many

people have commented on these connections with the past, and in her
book on Yellow Earth, Bonnie S. McDougall includes a long discussion
on the making of the film that refers to numerous interviews with the
filmmakers, wherein she states that in general, Chen and Zhang chose the
‘difficult approach of bending new techniques to express new ideas in the
context of a traditional society depicted in its own aesthetic’ (McDougall
1991, p. 28).26 More specifically she notes that the harsh, Loess Plateau
and Yellow River landscape locations in Shaanxi were chosen for their
essential Chinese historical and cultural significance as well as for their
‘hard beauty,’ and that the ‘composition of the picture frame, in regard
to such things as the compression of the landscape into one surface, with
virtually no perspective and with human figures appearing as marginal,
is based on traditional Chinese painting … rather than contemporary
Western film’ (McDougall 1991, p. 27; p. 39).
A comprehensive analysis of the connections between Yellow Earth and
painting was written by Chris Berry and Mary Ann Farquhar for a book
entitled Cinematic Landscapes. In it, they argue that ‘Like landscape paint-
ing, Yellow Earth emphasizes the natural over the human world, imag-
ery over narrative, and symbolism over (socialist) realism.’ They compare
cinematography to brushwork, the use of light and color to the use of
ink, and in recognizing a similar concentration on composition in both
painting and film media, they state that ‘we have the three key principles
in the theories of traditional Chinese painting’ (Berry and Farquhar 1994,
p. 85). However, Berry and Farquhar focus their visual comparisons of the
film with paintings of the contemporary, post-1949 Chang’an school that
Zhang Yimou referred to specifically as his inspiration, and interpret the
‘composition of the shots’ as embodying the ‘same idea (yi) of maternal
warmth and intimacy which informed the camerawork, color, and light-
ing’ (Berry and Farquhar 1994, p. 85; p. 95). Thus, whereas it was clear
from Chen Kaige’s interviews that he wanted both coldness or hardness
and warmth to emanate from the yellow earth, and whereas it is also clear
that many aspects of the film were made deliberately ambiguous, in part
to avoid censorship, it seems to me that to insist on ‘the most important
aspect of classical aesthetics in the film’ being the ‘makers’ idea of its mean-
ing’ is questionable (Berry and Farquhar 1994, p. 97). Indeed, in another
book chapter, where Berry is investigating the viewing subject of Chinese
films of the 1980s, he clearly states that Yellow Earth is such a strategi-
cally elusive film that it can be ‘read’ by viewers in so many different ways,
62 P. RIST
including it being a ‘didactic’ work ‘along the lines of the classical main-
land cinema,’ and that few other early Fifth Generation ‘texts’ employ
‘ambiguity and elusiveness to the extent that Yellow Earth does’ (Berry
1994, p. 102).
As an indication of the film’s ambiguity, the principal female character,
Cuiqiao (played by Xue Bai), a teenager who aspires to a better life, and
wishes to join the Eighth Route Army in Yan’an (Yenan, in 1939), is pre-
sumably drowned in the Yellow River, when she tries to cross it at night
in a small boat. We all assume she drowns, but we don’t actually see this.
She disappears. Offscreen, she is singing a communist song, taught her by
the soldier, Gu Qing (Wang Xueqi), which abruptly ends when the camera
tilts up from the river to the moon and her brother, standing on the shore,
calls out to her on two occasions. On the second of these calls, the fast-
moving Yellow River is shown and a series of five dissolves reveals the river
from night to day to night, and to a second dawn. Presumably, to avoid
censorship, even the budding communist, Cuiqiao’s death was rendered
ambiguously, and the reasons for it are a mystery. Did she commit suicide,
or was her attempt foolhardy, and her presumed drowning accidental? Or
as Rayns asks, was her ‘disappearance’ a ‘symbolic’ act, suggested by the
fact that her ‘singing voice [was] cut-off in mid-syllable—the syllable in
question being “Comm”’ (Rayns 1989, p. 30).
According to the Cinemetrics database, Yellow Earth has an ASL of 9.7
sec.27 This figure is matched by my calculation of Bonnie S. McDougall’s
statistics—9.74 sec.—achieved through dividing the length of the film
print she analyzed, 2385 meters [87 min. 10 sec. @ 24 fps], by 537 shots.28
In fact, there is not too much of a difference between these figures and
other films of the period. David Bordwell notes that ‘most mainstream
[Hollywood] films had ASLs between 5 and 7 seconds’ in the 1980s,
while the Cinemetrics database shows a number of US-made ‘art films’
from 1984 having a longer than 9.7 sec. ASL, including Wim Wenders’
Paris, Texas 12 sec., Woody Allen’s Broadway Danny Rose, 14–15 sec.,
and John Cassavetes’ Love Streams, 15.6 sec. (Bordwell 2006, p. 122).
What is of more interest in terms of relating Yellow Earth to landscape
painting aesthetics is the consideration of the distance from the camera to
the human subjects and the percentage of the film’s running time devoted
to different shot scales. Because of the variegated systems used in measure-
ment of shot scale, I have decided to combine Extreme (or Big) Close Ups
(ECU) together with Close Ups (CU) and Medium Close Ups (MCU)
as ‘close shots’; Medium Shots (MS) with Medium Long Shots (MLS,
or plans américains) as ‘medium shots’; and Full Shots (FS), Long Shots
(LS), and Very (or Extreme) Long Shots (ELS) as ‘long shots.’29 I am also
aware that detail shots of objects, animals, or even parts of the landscape
could have been considered by the analysts in the first grouping of more
closely framed shots, and no distinction here is made between shots con-
taining human characters, or not—something that could have been very
useful.
According to my understanding of the Cinemetrics ‘advanced break-
down,’ 45.5 per cent of Yellow Earth’s shots are ‘close,’ 13 per cent
‘medium,’ and 41.5 per cent ‘long,’ with my calculations of McDougall’s
scales resulting in 48 per cent close, 16.5 per cent medium, and 36.5
per cent long. The Cinemetrics table allows for calculations to be made
also on running times, with 25 min. of close shots, 17.4 min. of MS,
and 41.9 min. of LS, fully 50 per cent of the running time of the film, a
remarkable statistic, I feel. Barry Salt has allowed Cinemetrics to publish
his extensive statistical analyses on their website, and, although he hasn’t
provided many shot scale breakdowns for contemporary films, it is pos-
sible to make some comparisons with Yellow Earth. I have chosen to com-
pare seven US films, of which two, Tim Burton’s Beetlejuice and Woody
Allen’s Broadway Danny Rose are from 1984, and another Allen-directed
film, Purple Rose of Cairo is from 1985, along with three of the many titles
from 1979 that Salt has analyzed: John Badham’s Dracula, Mark Rydell’s
The Rose, and John Huston’s Wise Blood as well as one more recent film,
Jonathan Demme’s Silence of the Lambs (1991). Of these, only one film
has fewer than 50 per cent of the shots being ‘close’—Broadway Danny
Rose: 46 per cent—while this same film is the only one with more than
19 per cent of its shots being ‘long’ (24.5 per cent). Not surprisingly, the
film with the highest percentage of close shots, and lowest percentage of
LS is the most recent—Silence of the Lambs: 77 per cent and 9 per cent,
respectively.30
Chen Kaige’s second feature film, also shot by Zhang Yimou and also
produced by the Guangxi Film Studio, was The Big Parade (Da Yuebing,
1986). As interesting visually as Yellow Earth, with its widescreen, cin-
emascope frame providing wide, ELS exterior views contrasting dynamic,
tight, and claustral interior scenes of all-male military interaction (both
positive and negative), The Big Parade is much more closely related to the
European, modernist avant-garde of the 1920s than to Chinese, tradi-
tional landscape painting.31 Like Yellow Earth, The Big Parade is both dia-
lectical and ambiguous; where, for example, in the beginning, the formally
64 P. RIST
rigorous cinemascope compositions can be perceived as being beautiful,

they are eventually understood as showing the monotonous cruelty of
army life.
Chen Kaige’s third feature film, King of the Children (Haizi wang,
1987) was produced at Xi’an studio, by Wu Tianming, the man who was
already regarded as the person most responsible for promoting young
filmmakers, in particular Fifth Generation graduates of the BFA, and who
was able to fund Xi’an’s ‘art’ films (yishu pian) and ‘exploratory’ films
(tansuo pian) by producing commercially successful films (Rayns 1989,
p. 19). As a director at Xi’an, Wu had championed shooting on location
with his first solo effort, The River Without Buoys (Meiyou hangbiao de
heliu, 1983) being shot in the Hunan countryside (and containing many
impressive landscape shots) and this became a feature of his role as the
head of the studio, to which he was elected in 1983. Key examples of these
films include In the Wild Mountains (Ye shan, dir., Yan Xueshu, 1985)
shot in Shaanxi province not far from the studio, the extremely experi-
mental The Horse Thief (Daoma Zei, dir. Tian Zhuangzhuang, 1986), shot
in Tibet, Gansu, and Qinghai, and the most commercially successful film
of the group, Red Sorghum (Hong Gaoliang, dir. Zhang Yimou, 1987),
unusually set in eastern China, in Shandong province.
With Zhang being given the chance to direct Red Sorghum, another
of the Fifth Generation cinematographers who had moved with them to
Xi’an, Gu Changwei, led Chen’s crew to the southwestern and subtropical
countryside of Yunnan province to shoot King of the Children. Together,
Chen and Gu fashioned probably the final great Fifth Generation film, and
arguably the most stylistically complex and deliberately ambiguous film of
this Chinese new wave. There are at least three scholarly, close analyses of
King of the Children in English, two of which place the film in the context
of ‘modernism’: Rey Chow’s and Xudong Zhang’s (Chow 1991; Zhang
1997). In criticizing Chow’s interpretation of the film, Zhang stresses that
it is ‘complexities and contradictions that keep alive an alternative mod-
ernism and make it productive in a moment of ambiguity’ (Zhang 1997,
p. 284). In recognizing the numerous landscape shots and the structural
use of blankness, Zhang focuses on the open subjectivity of the teacher
protagonist, Lao Gan (Lao Gar, ‘skinny,’ or ‘bean pole,’ played by Xie
Yuan), looking ‘for an object that will return his inquiring gaze—his stu-
dent, the cowherd, the smiling peasant, or the rolling mountains’ (Zhang
1997, p. 301). Both writers recognize that King of the Children is an
extremely autobiographical work for Chen Kaige, who was ‘sent down’
during the Cultural Revolution as a zhiqing (educated youth) to be re-

educated by working on the land, in his case, cutting bamboo and rub-
ber trees for two years or more in Xishuangbanna, the Dai Autonomous
Prefecture at the most tropical southwestern corner of Yunnan.32 While
still in Yunnan, Chen met the writer, Ah Cheng, who had been a teacher
there, and who later completed the novella on which King of the Children
is based.
The third analysis, written by An Jingfu, is titled, ‘The Pain of a Half
Taoist: Taoist principles, Chinese landscape Painting, and King of the
Children.’ An goes back to the Zhou Dynasty philosophical writings
of Lao Tzu (Laozi, 571–531 B.C.) and Chuang Tzu (Zhuang Zhou,
369–287 B.C.) for his definition of the Tao as the ‘way of nature and of
human life,’ and argues that their ‘teaching’ persuaded ‘people to pre-
serve their inborn nature by practicing inaction and by living in emptiness,
silence and purity’ (An 1994, p. 118; p. 119). For An, when Lao Gar
‘looks into the distance,’ of the landscape, he senses the ‘power of the
Tao,’ while the other part of his character, the Confucian half believed that
there was some hope in the profession of teaching (An 1994, pp. 122–3;
pp. 120–1). An suggests that the ‘whole picture resembles a Chinese land-
scape painting, and the slow pace of the film reinforces this impression’
(An 1994, p. 122). After stating that ‘there are more than one hundred
landscape shots,’ defined in the notes as ‘shots in which mountains and
trees appear in the background, whether they are hidden or obvious,’ An
regards the ‘stillness’ and ‘emptiness’ of these shots, and how the human
figures ‘often appear small,’ while they and the landscape are often ‘bathed
in mist’ (An 1994, p. 122; note 8, p. 125).33
According to my own statistical analysis, there are 364 shots in
King of the Children. When the titles are excised, the running time of
the film is 105 min. approx., giving an ASL of 17.6 sec. As a reminder,
the Cinemetrics ASL of Yellow Earth was 9.7 sec., and four other films
directed by Chen have received the Cinemetrics treatment. Regarded by
date, the ASL drops; gradually to Farewell My Concubine (1993): 9.2 sec.,
and then, drastically: ‘One Hundred Flowers: Hidden Deep,’ segment of
Ten Minutes Older: The Trumpet (2002): 5.5 sec.; The Promise (2005): 5.3
sec.; Sacrifice (Zhao shi gu er, 2010): 4.8 sec. I suspect that Life on a String
(Bian zou bian chang, 1991), Chen’s beautiful, but baroque and less
ambiguous goodbye to classical Chinese and Fifth Generation yishu pian
aesthetics, may have reversed his trend toward a more conventional, narra-
tive editing approach. Like Yellow Earth, a large percentage of scenes were
66 P. RIST
filmed on exterior locations, and after inspecting the opening 30 min. of

the film, I calculate that over 30 per cent of the shots are ELS. However
the ASL of this first section is 7.7 sec., representing a faster cutting rate,
even, than Farewell My Concubine.
My shot scale analysis of King of the Children resulted in 165.5 LS,
73.5 MS, and 125 close shots. Of the LS, no fewer than 69 are ELS, with
a running time of over 25 min., and 71 are LS, with a running time of
28 min. Indeed, all of the ‘long shots’ combined represent over 60 per
cent of the film’s length. Unlike Yellow Earth, the minority of King of the
Children’s shots were filmed outside. However, the school classrooms are
open structures with low walls, no windows or doors and thatched roofs,
so that many of the daytime interior shots also view the exterior courtyard.
Other daytime interiors are often lit with natural light, and nighttime inte-
riors are extremely dark, and contain discreet candle lighting. In part, this
approach is reminiscent of the 1930’s films of Jean Renoir where a con-
scious attempt was made to link interior and exterior spaces in the guise
of a humanitarian ‘realism,’ and we can also observe a recognition of the
ancient historical Chinese cultural appreciation of nature.34 On the other
hand we can understand the encroachment of nature as a criticism of the
lack of development of agricultural communities and domestic technolo-
gies under state communism. Interior spaces are sometimes more empty
than exterior ones; the little furniture that exists is prone to breakage,
and tree stumps are used as tables and chairs. This last item is interest-
ing because it is an indication of primitive ingenuity as well as a sign of
deforestation.
I counted 146 exterior shots, only 40 per cent of the total, although
more than 100 of these are ‘long shots,’ of which 66 are ELS scale, with
a running time of 24.5 min., an ASL of 22.25 sec. Although the ASL
of the first half of the film is almost identical to that of the second half,
and although this split also applies to the occurrence of exterior LS/ELS,
landscape shots—34 in the first 53 min., with an ASL of 22.4 sec.—there
are big differences in compositional and editing style between sequences.35
In the first four scenes inside the classroom where Lao Gar learns that the
children don’t have textbooks, and where he is initiated into the system
of them copying text that he writes on the chalkboard, there are a total
of 65 shots lasting 495 sec., an ASL of 7.6 sec., with the editing being
much faster than for the film as a whole. These scenes are broken down
by analytical editing, à la Hollywood, with a number of reverse-angle cuts,
between teacher and students, where the camera is often angled between
45° and 60° to the subject, and with a variety of shot scales: 19 LS, 13
MS, and 33 close shots (65 per cent), a far greater proportion than for any
other scenes in the film.
The opening two shots of King of the Children provide a more repre-
sentative introduction to the film’s predominant style. The opening, pre-
credit shot following a brief title card is a 60 sec., ELS time lapse and
dissolving view, revealing, out of the frame-filling fog, a diagonal pathway
leading to a school on the top of a hill, with a mountain range behind in
the far distance. The shot ends with a sunset and a fade to black. The mist,
the passing shadows cast by unseen clouds and the position of the camera,
perpendicular to the school, together with the mountainous terrain itself,
combine to create a sense of unraveling a mystery, while simultaneously
recalling Chinese landscape painting—distanced view, blank areas of the
frame, apparent flatness or multiple perspectives. We don’t see any human
figures, but the soundtrack recognizes their offscreen presence with cow
bells, a musical human cry, and an explosive percussive noise, perhaps rep-
resenting the chopping of bamboo.
The second shot is even more mysterious, and longer: 117 sec. The
camera regards a dark, interior space, at 90° to a wall. Natural light shines
through a doorway at the extreme right edge of the frame, while a long
bamboo pipe is being smoked by a barely visible man at the left side of
the frame. We gradually understand the smoker to be the ‘captain’ of a
Red Guard rural work unit and Lao Gar is introduced through his voice
and his shadow on the floor as he stands, and then squats, offscreen in the
doorway. His arm enters the frame initially to pick up packets of matches
and cigarettes thrown by the boss onto the floor and later to collect a sheet
of paper sitting atop a tree stump inside the room, presumably inform-
ing of his teaching assignment. Of course, this is not a ‘landscape’ shot,
but the blankness of the wall that is foregrounded by Lao Gar’s cigarette
smoke matches the misty emptiness of the opening shot. In addition, we
can certainly compare the flatness of the perspective here with a choice of
‘modernist’ European filmmakers of the 1960s, in particular, Jean-Luc
Godard, while the smoke that appears white in the light from the doorway
is even reflexive of cinema projection itself, when cigarette smoke would
illuminate the projector’s beam. Later in the film, there are recurring exte-
rior perpendicular ELS views of a school house where the thatched straw
roof fills the top half of the frame and hangs so low that Lao Gar and his
students’ upper bodies are visually fragmented by it. When the teacher
first arrives at the school, there is a perpendicular view of his dormitory
68 P. RIST
window frame, seen from the outside with crude, children’s graffiti sur-
rounding it on the bamboo wall. Lao Gar sits down inside, behind the
window in MS and drinks from a gourd, while human figures (probably
children), blurred by the telephoto lens dash past the camera in front
of the wall. This shot (#50) lasts for 55 sec. It reminds us of the credits
sequence where the background is a page from an apparently old book,
and where the credits for the film’s cast and crew appear through a series
of dissolves on the left, while the title of the film is shown in vertical
Chinese characters in a torn frame on the right. During the credits, the
offscreen sound of chalk on a blackboard is heard. Thus we had been pre-
pared for a treatise on pedagogy, the oldness of which was suggested by
the tattered nature of the page. With shot 50, we are given both an image
of twentieth-century modernism, with its flatness and fragmentation, and
a reflection on painting with the frame-within-a-frame composition, and
the black ink appearance of the drawings.
After a seven-minute introduction to all of the zhiqing gathered in their
dormitory, including the extremely disheveled Lao Gar, we witness his
trek to the school accompanied by his friend Lao Hei. There are 15 shots
in this three-and-a-half minute sequence (ASL, 14.2 sec.), 13 of which are
ELS framings. In three of these shots there are no human figures visible
and in the ten others the two men are very small in the frame. Matching
the film’s opening shot, the scene begins in mist, but the aura of mystery
is continued through images of strange-looking trees and bushes, burnt
stumps, and the recurrence of fog and mist. There is camera movement in
five of the shots, sometimes used to follow the figures, but in two striking
cases, where the men are not seen, the camera pans right past a foggy, tree-
filled landscape, for its full duration (11 and 15 sec., respectively). Because
of the forested, misty, mountainous landscape that is being viewed, because
there are no human figures in the shots, and because the lateral tracking
movement is so deliberate, it is as if we are unraveling a handscroll.
In the third shot, we view Lao Gar and Lao Hei from behind, with
their lower bodies obscured by the bottom frame edge (not unlike a
Michelangelo Antonioni or Godard framing), walking into the lush green,
misty forest background along a path. The next shot reveals the moun-
tainous nature of the terrain, and we struggle to find the tiny figures of
the men as they walk up from behind some fallen logs, toward the cam-
era. In another apparent reverse angle we find the men perched on the
downslope of a hill. The camera views them in very high angle showing a
mist-shrouded river in the far distance. This is the first instance where we
see the men, already fragmented, dropping below the frame, at the end of
the 19-second shot, momentarily eliminating their figures. Through the
next five shots, the sound of wood chopping intensifies and is now mixed
with birdsong. The third of these is an apparent reverse-angle reaction
shot of the two men, framed against a blank sky. They both have puzzled
looks on their faces; they turn around and with their backs now to the
camera, they walk into the blank background void accompanied by a slight
pan to the left. We can only imagine what they have been observing. The
12th shot shows the head and shoulders of a young cowherd from behind
in low angle, and holds until his walking away from the camera causes him
to drop below the frame, leaving a view of trees and a misty sky in the
frame. With a similar background and low angle on the two young men,
framed in MS, the implication is that the next shot is yet another reverse-
angle reaction shot. They don’t move, and again look puzzled.
There are a number of significant features of this initial landscape
sequence. It illustrates a journey from the zhiqing base camp to the school
where Lao Gar is going to teach, and this journey is reminiscent of the
subjects of Sung Dynasty paintings such as Travelling Among Mountains
and Streams (c.1000). The consistent use of reverse-angle cutting on
movement suggests that the journey is linear, although the discontinuous
nature of the depicted landscapes suggests otherwise. The two apparent
reaction shots of Lao Gar and Lao Hei are interesting for a number of
reasons. They are reminiscent of low-angle shots from ‘socialist realist’
works of propaganda where the protagonists are heroically framed against
an open sky (the first) or a picturesque landscape (the second), and of
PRC commissioned posters.36 I am persuaded (because of the use of the
low-angle camera) that Chen and Gu chose these shots to remind Chinese
audiences of the near-past and to be critical of communist propaganda,
in part, because the context is ambiguous. After all, how better to coun-
ter rhetorical propaganda than by providing ambiguous images that resist
interpretation. In the first instance, we don’t know what they are looking
at and in the second, we can’t know what they are thinking. Throughout
the film, Lao Gar is seen in similar poses, sometimes framed against the
(tilted) chalkboard and straw ceiling, while his consistently chaotic, messy
appearance is the antithesis of the ‘hero’ image. (Occasionally, one or
more of the children are also filmed in low-angle interior shots.)
I find the situation where the two male characters are either absent from
the shot, or can barely be seen, gradually emerging into and disappearing
out of the bottom of the frame especially interesting because it makes the
70 P. RIST
viewer’s task a struggle to see the protagonists. We have to work to see

them, and we are also encouraged to contemplate the image. It goes with-
out saying that the act of regarding a painting, any painting, is different
from that of watching a film. When we look at a painting we ‘contemplate’
the image. We can scan it with our eyes, and move closer or farther away,
but the painting remains static. Some paintings, especially large murals, are
more complex and contain more intricate detail than others and require a
longer period of concentration to appreciate fully, and I argue that this is
the case with many Chinese hanging scroll landscape paintings that require
considerable vertical scanning. Of course the handscroll is a case apart, and
this form which involves actual movement—unraveling with the hand—
and whose appreciation must also be time based, has often been noted as
a precursor to lateral tracking or panning shots in film.37
We hardly ever get a clear impression of the mountainous background,
and the general overcast appearance of the sequence is both reminiscent
of the blank spaces of Southern Sung paintings and suggestive of mysteri-
ous surroundings. I think the filmmakers’ intention here was for the film’s
audience to already be questioning what is being seen and heard. In short,
our experience of watching the film should be a learning experience, and
one through which we won’t necessarily find the answers. During the
classroom scenes, Lao Gar has no choice but to have the students copy the
characters he writes on the board, but he eventually manages to encour-
age their creativity in writing little compositions, and reads two of them
to the class. In the first part of this scene, that begins at the 59 min. mark,
the teacher’s approach is matched by the creativity of both the style of the
film—shot composition and montage—and of the students themselves.
While Lao Gar is reading the first, very short account of ‘Going to
School,’ and while, through laughter, the teacher explains that it is ‘very well
written,’ there is a series of 11 shots, beginning with a detail shot of the carv-
ings of two heads on a crossbeam, with ‘tongues of red paper fluttering in
the breeze.’38 There are five other, strange detail shots, one of which shows
two fingers, with faces painted on them pushing up through holes in a desk-
top and another of a boy’s hands rolling a piece of paper like a ball of wool.
Two of the other detail shots contain camera movement: a track along the
front of the teacher’s desk revealing crudely drawn graffiti, ‘a weird sea, land
and air battle scene; a teacher holding a cane and yelling “Copy! Copy!;” a
female teacher with her hair sticking out,’ and a crane up a column contain-
ing a vertical row of red painted, small footprints (Chen and Wan 1989,
p. 98). However primitive these images might be, they suggest artistic
potential in the children, and in combining calligraphy with figure painting

and carving, especially where the tracking shot exists as a strangely playful
pastiche of a handscroll, we are given naive echoes of classical Chinese art.
The imaginative editing here is also reminiscent of Sergei Eisenstein’s ‘intel-
lectual montage’ in Octyabr (October, 1928), where, in the Winter Palace
sequences, although it appears the director was using non- or extra-diegetic
inserts, he actually filmed many objects inside the palace (or objects that
were obtained from the palace) to create associational montages.39 It is as if
Chen and Gu found examples of student doodling in an actual schoolroom
(at the Mengxing State Farm Middle School, Mengla, Yunnan) and decided
to highlight them here in the guise of criticizing learning by rote. Indeed,
extending the connection with October, this King of the Children sequence
is one of many that could be termed ‘dialectical’ in a very open sense, rather
than in a Marxist (and Eisensteinian) closed sense, wherein a finite synthesis
must follow a series of antithetical edited shots.
Not long after this sequence, in an exterior scene, the students are on
an expedition to chop bamboo for repairing the classroom roofs. Of the 25
shots, 11 are ELS in scale, and in one of them (#246), there is a slow, 360°
pan right, lasting the full 75 sec. duration of the shot from Lao Gar and
the students in the far distant background walking alongside a river on the
left, away from them past trees in the mist to Wang Fu (Yang Xuewen) and
then following the river to pick up the group as they near the camera. In an
attempt to win a dictionary, Wang Fu has been felling trees overnight, and
his isolation in the frame represents his separate status. In part, the arcing
camera also sketches this situation, while continuing to convey a sense of mys-
tery to the environment and echoing the narrative potential of a handscroll.
Throughout the film, the offscreen sounds of tree felling together with the
off and onscreen presence of fires build a textural recognition of the perils of
slash and burn agriculture, which for Chen Kaige was surely an autobiograph-
ical memory. When he had returned to Xishuangbanna to scout locations for
King of the Children in the autumn of 1986, he revisited his work station and
along the route found that rubber trees now occupied many of the slopes, but
that the great primary forest that he had known had disappeared for ever.40
Surprisingly, perhaps, Chen’s recollections of working in the Yunnan tropi-
cal rainforest were tinged with nostalgia. Now outside their classroom, the
children’s learning encompasses the natural world, while the film continues
to dialectically test the audience’s apprehension of what constitutes education
and how film can stylistically ‘educate’ while connecting the distant past of
landscape painting with the recent past of the zhiqing work experience.
72 P. RIST
After the very first ‘landscape’ scene, there are at least seven other exte-
rior scenes that contain shots where Lao Gar, or his best student Wang
Fu, are isolated in the frame, and where the frame is empty for part of the
shot’s duration, usually because of the character walking down a slope or
off the frame. Invariably, there is a period of time spent when Lao Gar is
motionless, either facing the camera or away from it looking into the land-
scape background. In the first case, we wonder, as with shot 28, what he is
looking at, and as with shot 33, what he is thinking. In the shots where the
teacher or the student have their backs to the camera, and sometimes gaze
into the landscape, I sense a really strong connection being made with
Southern Song paintings, where the foreground figure was often a reclu-
sive scholar/painter, or an upper-class figure. I have already suggested
that a painting such as Ma Yüan’s On a Mountain Path in Spring is hence
‘reflexive,’ and King of the Children goes even further in this direction. We
have an allusion to Southern Song painting, which leads to questioning
the status of the human figure. As stated earlier, Lao Gar is anything but
a ‘hero,’ and thus, such shots contradict the representation of leaders of
the revolution in ‘socialist realist’ films, and now here we can understand
the difference between court figures of ancient classical painting, and the
‘king’ of the children, who is closely related to the peasantry.
In the film’s concluding scene, Lao Gar, who has been released from his
duties for not sticking to the curriculum, is framed once more looking out
from the schoolyard plateau into a mountainous background, obscured
by mountain mist. With his back to the camera, he walks down and away
from the school and is next seen on the strange hill of burnt tree stumps
seen in the film’s first exterior journey scene. Bonnie S. McDougall calls
this location ‘Easter Island’ in recognition of the mysterious nature of the
sculptures there and how these burnt remnants of trees, seen in a series
of detail shots, resemble human and animal figures (Chen and Wan 1989,
p. 119). In addition, the mysterious young, presumably illiterate cowherd,
also found in that early scene, is seen from behind, peeing, probably to
give his animals access to salt. This scene is perhaps a perversion of a clas-
sical landscape image with the ‘scholar’ Lao Gar’s gaze into the landscape,
more mystified than ever. But perhaps he is moved by the magical tree
figures rising out of natural vegetation, and impressed by the cowherd’s
creativity, which is matched by his own creative construction of a new
Chinese character for Cow/Piss, that he had invented earlier for the stu-
dents and which is reprieved at the very end of the film, chalked on the
blackboard with a fine ray of sunlight crossing it (#362).
NOTES
1. According to Wu Hung (1997, p. 65), it is possible that a vertical hand-
scroll in the collection of the Palace Museum, Taipei, entitled, Sailing
Boats and a Riverside Mansion may be attributed to Li Ssu-hsün. Michael
Sullivan (1980) is somewhat skeptical about the importance of Li father
and son, but traces the development back much further and provides
numerous examples of surviving seventh-century Sui Dynasty cave paint-
ings including landscapes, at Tunhuang, especially pp. 111–112. The old-
est surviving hanging scroll painting on silk dates from the Western Han
Dynasty, about 180 B.C. (Sullivan 1973, pp. 45–6).
2. See The Pride of China (2007, p. 29, p. 83); Watson (2000, pp. 1−3).
3. See also Hsü (1999), and Watson (2000, pp. 11−12).
4. See Ch’en (2003, pp. 24–9); and Sickman and Soper (1961, pp. 203–14).
5. Sabine Hesemann discusses the horizontal scroll as it was developed earlier
in the Southern style, exemplified by the work of Dong Yuan (e.g., The
Xiao and the Xiang) who worked for the court of Nanjing (937–962),
where the land was relatively flat. According to Hesemann, he was ‘a man
of the south [and], created an effect of great scope in his landscape com-
positions,’ ‘China: The Song Period and the Aesthetics of Simplicity,’
(Hesemann 1999, p. 145).
6. See Hesemann (1999, pp. 155−6). See also Cahill (1997, p. 8).
7. See Treager (1997 [1980], p. 124); Cooper and Cooper (1997, p. 83);
and Ch’en (2003, p. 35).
8. Soper’s translation of the other five principles is as follows: The second is
‘structural method in use of the brush.’ The third is ‘fidelity to the object
in portraying forms.’ The fourth is ‘conformity to kind in applying colors.’
The fifth is ‘proper planning in placing (of elements).’ The sixth is ‘trans-
mission (of the experience of the past) in making copies’ (Sickman and
Soper 1961, p. 133).
9. See my essay, ‘The Presence (and Absence) of Landscape in Silent East
Asian Films’ (Rist 2006a).
10. The original filmography was published as Volume One of Zongguo diany-
ing fazhan shi (The History of the Development of Chinese Cinema) in 1963
and republished in 1980 after the Cultural Revolution.
11. For more on visual style in the Lianhua films, see my essay, ‘Visual Style in
the Silent Films made by the Lianhua Film Company [United Photoplay
Service] in Shanghai: 1931−35,’ (Rist 2006b).
12. The version of the US release that exists in China, and that was published by
the China Film Archive, Beijing on VCD is only five reels of 35 mm in length,
but a longer, seven reel version was discovered recently as a 16 mm print, with
a musical soundtrack, and with a running time of approximately 67 min.
Hopefully, an original Chinese release version will eventually be rediscovered.
74 P. RIST
13. In discussing Fei Mu’s films and writings at some length in a book on
Chinese cinema before 1949, Jubin Hu stated that his ‘interpretation of
Spring in a Small Town is that while Fei Mu accepted new ideas about
culture and ethics he also wanted to preserve Chinese tradition to a certain
degree’ (Hu 2003, p. 177). Elsewhere, Hu discusses Fei Mu’s writings on
Chinese ‘national’ film style, where, as early as 1941, he had argued for
such a style to, ideally, combine ‘tradition and modernity’ (p. 182).
14. For more on style in Tianlun, see, ‘The Presence (and Absence) of
Landscape in Silent Chinese Films’ (Rist 2007).
15. During the Japanese occupation of Shanghai, from 1937 until the out-
break of the Pacific front of World War II in December 1941, it was pos-
sible to make films in the foreign concessions (the ‘orphan island’). On the
occasion of the film’s restoration and public screening in Hong Kong, a
little booklet was published that included a bilingual essay, ‘History and
Aesthetics,’ written by archivist/historian Wong Ain-Ling (Wong 2009,
pp. 4−9).
16. He also shows that with two other stylistic features considered by scholars
to be typically ‘Chinese—medium shot scale and camera movement com-
bined with flatness—the statistical evidence doesn’t support these claims’
(Udden 2012, pp. 269−72).
17. For example, in commenting on a neighboring cinema, David Bordwell
said that ‘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?’ (Bordwell 2005, p. 98).
18. The Lumière Brothers introduced the cinema to China in 1896. See
Robertson (1991 [1980], p. 3).
19. See also, Laikwan Pang, ‘Shanghai Films of the 1930s’ (Pang 2011, p. 58).
20. In the very first book written in English on the ‘Fifth Generation,’ and one
that remains a first-rate introduction to the subject, Tony Rayns wrote,
‘No one seems fully clear how the nickname [Fifth Generation] was arrived
at. Some take it to mean that the film-makers were the fifth distinct group
to graduate from the Film Academy, which has a periodic rather than an
annual intake. But simple mathematics makes that explanation unlikely: the
Academy was founded in 1956, closed between 1966 and 1978, and offers
its students four-year courses’ (Rayns 1989, p. 16). Rayns has recently
provided a detailed discussion of the five generations (Rayns 2014, p. 16).
21. The narrative is driven by a comedy of errors, where the male visitor/hero
Ah Peng finds the wrong Bai woman named Jinhua on four occasions
before finally finding his loved one—hence the number ‘Five’ in the title.
There is also an interesting reflexive dimension to Five Golden Flowers,
where two artists working for the Changchun Film Studio have come to
Yunnan to record folk songs, and paint landscapes and figures.
22. Another example is Third Sister Liu (Liu sanjie, dir. Su Li, 1960, b/w).
Yingjin Zhang argues that this film ‘represents an attempt to approximate
the viewing experience afforded by the traditional hand-scroll painting’
(Zhang 2004, note 4, p. 304).
23. See Ehrlich and Jin (2001, pp. 10−11).
24. See, for example, Shelly Kraicer, ‘Rediscovering the Fourth Generation’
(Kraicer 2008, p. 30). For the reception of Yellow Earth, see Rayns (1989),
where he writes, ‘The screening was received with something like collective
rapture, and the post-film discussion stretched long past its time limit’
(p. 1); and, ‘The torrid enthusiasm of the Hong Kong audience was
repeated when Yellow Earth had its western première at the Edinburgh and
Locarno festivals four months later’ (p. 2).
25. According to the English-language website of the BFA, the Directing
department core includes, ‘Drama, Performing, Video and Audio lan-
guage, Directing Art [Films?], Documentary Film Theory and Making,’
while the listing of 13 other ‘basic courses’ includes ‘Analysis of Art Works’
and ‘General Introduction of Art,’ which may or may not include Chinese
classical landscape painting examples. http://www.bfa.edu.cn/eng/2012-
12/17/content_57148.htm
26. See, for example, Silbergeld (1999a, pp. 43−7); Yau (1991, pp. 64−5).
27. This figure was submitted by Zhang Zizhao on 27 October 2010, the full
breakdown of the advanced mode being available at http://www.cinemet-
rics.lv/movie.php?movie_ID=16882
28. My calculation uses 27.36 meters of 35 mm film being equal to one min-
ute of running time taken from one of the very useful calculation tables
presented in Cherchi Usai (2000, p. 174). McDougall counts the end titles
as 11 shots, whereas I consider them as one shot only, although I main-
tained the count of the opening titles as six shots (Zhang 1991).
29. The Cinemetrics measurement of shot scale is based on Barry Salt’s system,
illustrated at http://www.cinemetrics.lv/salt.php, where BCU (Big Close
Up) frames the head only, CU frames the head and shoulders, MCU is
from the waist up (although the image shows from the stomach or chest-
up), MS, ‘includes from just below the hip to above the head of upright
actors’ MLS, ‘shows the body from the knees-upwards,’ and, although not
defined by Salt, FS shows the full body in the shot, LS, shows the body
filling only half of the frame, while VLS (Very Long Shot) shows the actor
small in the frame. For my own measurements of shot scale, I have tended
to use the same scheme for closely framed shots (with ECU, instead of
BCU), but I usually employ ‘MS’ more widely (to include Salt’s MS and
MLS), while I reserve MLS for full body shots (FS). I use ELS (‘Extreme’)
in place of VLS. McDougall uses ECU, CU, MC, MS, LS, and ELS, with-
out defining any of the scales as Salt does. See Zhang (1991 p. 174)
30. See Cinemetrics http://www.cinemetrics.lv/satltdb.php
76 P. RIST
31. I am thinking here, especially of Fernand Léger’s (and Dudley Murphy’s)

Ballet mécanique (France, 1924) that comically connects human behavior
with machines and critiques the world of advertising, but also of celebra-
tions of mechanization and the modernity of cinema such as Walter
Ruttmann’s Berlin; Symphony of a Great City (Berlin; Die Sinfonie der
Grosstadt, Germany, 1927) and Dziga Vertov’s Man with a Movie Camera
(Chelovek s kino-apparatom, USSR, 1929), and abstract animation, includ-
ing Oscar Fischinger’s German cigarette commercials of the 1930s.
32. Tony Rayns provides an excellent overview of the connections between
Mao’s zhiqing system of re-education and the Fifth Generation, especially,
pp. 2−15 (Rayns 1989). The best account of Chen’s experiences in Yunnan
is available in French, not English; see Chen (2001 [1989]).
33. The endnote 8 continues, ‘Most of these [landscape shots] are LS, although
sometimes MLS, MS, or even close-ups are inserted. In all cases, the func-
tion of the natural background remains constant.’
34. I have found no specific reference to Chen Kaige being influenced by
Renoir. Ni Zhen refers to the BFA students being shown ‘Chinese classical
cinema, Hollywood film of the thirties and forties, and the Soviet classics’
(Ni 2002, p. 94). Later, he notes that Japanese films were imported into
China in ‘the late seventies,’ and that ‘Filmmakers such as Bergman,
Resnais, Godard, Truffaut, and Antonioni, and writers such as Kafka,
Sartre, Camus, Wolfe, Garcia Marquez, Faulkner, Bellow, and Hemingway
all became objects of intense fascination and half-comprehending worship’
(p. 98).
35. The only commercially available DVD of King of the Children is published
by Guangzhou Beauty Culture Communication Co. Ltd. in China, #DE
1619-26, and it is a very poor quality version: poor resolution, with color
shifts, bad subtitling, incorrect masking giving an aspect ratio of 2:1; and
worst of all, reels seven to nine are inverted, complicating segmentation
and close statistical analysis.
36. See, for example, ‘Comrades-in-arms’ (Zhanyou, 1977), ‘I contribute pre-
cious deposits to the mother country’ (Wo wei zuguo xian baozang, 1979),
‘Sternly attack criminal activities’ (Yanli daji xingshi fanzui buodong,
1983), and ‘Comrades Zhou Enlai and Deng Yingchao’ (Zhou Enlai he
Deng Yingchao tongzhi, 1983), in Landsberger (2001 [1995], p. 40,
p. 142, p. 160, p. 103). See also model paintings such as Chairman Mao
Goes to Anyuan (Mao zhuxi qu Anyuan, 1967) by Liu Chunyua (born
1944), and especially, in contrast to figures in the landscape of Yellow
Earth, Chen Yifei’s (1946−2005) Eulogy of the Yellow River (Huanghe,
1972), in Chiu and Zheng (2008, 121, pp. 36−7; detail, xii).
37. See, for example, the writing on the relationship between Mizoguchi
Kenji’s long take and camera movement style of the 1930s and 1940s and
e-makimono Japanese scroll painting (Sato 1994, pp. 170−1); (Burch

1979, pp. 228−34).
38. The quotations are taken from McDougall’s translations in Chen and Wan
(1989, pp. 98−9).
39. See, for example, Taylor (2002, p. 38). See also, Bordwell (1993, p. 85):
‘The toys, cutlery, plates, glassware, and other luxury goods are evidently
in the palace, but they are filmed in such intense close-ups or framed
against such neutral backgrounds that they seem to hover in a purely sym-
bolic space.’
40. ‘Le long de chemin, le paysage s’était profondément modifié. Les pentes,
de part et d’autre de la route, était maintenant occupées par des plantations
d’hévéas…. Mais la grande forêt primaire que j’avais connue, elle, a disparu
à jamais’ (Chen 2001, p. 196).
CHAPTER 5
Poetics of Two Springs: Fei Mu versus

Tian Zhuangzhuang
James Udden
In the inaugural issue in 1993 of the Chinese journal, Film Art (Dianying
Yishu), an eight-page article on Spring in a Small Town argues that Fei
Mu’s film reflects an “Eastern” cinematic language distinct from either
Hollywood or Western art cinema (Ying 1993). This is a reminder of how
difficult it is to discuss Chinese cinema without some reference to tradition,
or without some search for cultural specificity through cinema. However,
with Fei Mu’s Spring in a Small Town (1948) in particular—now com-
monly referred to as the greatest Chinese film of all time—tradition seems
an impossible issue to avoid. References to tradition and culture in Fei Mu
are not misguided, to be sure. Even his own daughter, Fei Mingyi, stresses
how her father was imbued with numerous traditional forms such as poetry
and theater. Yet strikingly, she adds how Spring in a Small Town in particu-
lar represents a deeper reflection on the role of tradition immediately after
the eight-year war with Japan and the civil war that was still raging when it
was made (Fei 1996: 9). In other words, this particular film is very much
about the uncertain fate of tradition in the present day of the late 1940s,
as it is about tradition per se.
That Spring in a Small Town centers on the fate of a cultural tradition is
only half the story, however. Thematically, this film covers familiar territory
J. Udden ()
Gettysburg College, Gettysburg, Pennsylvania, USA

DOI 10.1057/978-1-137-55309-6_5
80 J. UDDEN
with its exploration of moral conflicts within a Confucian milieu; stylisti-

cally, however, it is beholden to no tradition whatsoever—neither East nor
West. In the context of the late 1940s, this film’s style represents a radical
transformation of cinematic language even outside of China.
This becomes manifestly clear when comparing the 1948 original to the
2002 remake by Tian Zhuangzhuang. Side by side, Fei Mu’s version is the
more forward-looking of the two, sitting on the cutting edge of a burgeon-
ing post-war cinematic modernism normally associated with European
and Japanese cinema of the late 1940s and early 1950s. By contrast, Tian’s
version is almost backward-looking, as if he reimagines Fei Mu’s classic
along the lines of what many assume this film to be, rather than what it
truly is. This is especially evident in how much the revised version relies
on the long take compared to the original. Yet in doing so, three salient
traits of the 1948 version are absent in the remake: Fei Mu’s unusual
tendencies in staging; Fei Mu’s intricate editing strategies; and most of
all, Fei Mu’s radical experimentation with the voiceover. With these three
traits alone, Fei Mu deserves to be discussed alongside contemporaries he
likely never knew of, most of all Jean-Pierre Melville, Robert Bresson, and
Michelangelo Antonioni. A seeming paradox emerges: thematically speak-
ing, Fei Mu is a more “traditional” director than Tian Zhuangzhuang; but
stylistically speaking, the former is the proto-modernist, while the latter
follows a more imaginary ideal of what a “traditional” Chinese film puta-
tively ought to look and sound like. That these crypto-modernist traits
were missed by nearly everyone at the time was the result of the liminal
state of Chinese cinema in the late 1940s just before China became the
People’s Republic of China.
TWO SPRINGS; TWO TRADITIONS

There is no denying that compared to most of his Chinese contempo-
raries, Fei Mu was a director steeped in tradition. Carolyn Fitzgerald has
provided a thorough overview of how deeply Fei Mu drew from tradi-
tional Chinese lyrical poetry, theater, and landscape painting. She argues,
for example, that in Spring in a Small Town, Fei Mu uses a technique from
Chinese poetry of “juxtaposing images in parallel couplets to produce
subtle and evocative contrasts” (Fitzgerald 2013: 174–75). She further
documents an anecdote from the screenwriter of the film, Li Tianji, that
Fei Mu had recited to him a famous poem called “A Spring Scene” (by
Su Shi, ca. 1094–1097) that “expresses themes of melancholy springtime,
POETICS OF TWO SPRINGS: FEI MU VERSUS TIAN ZHUANGZHUANG 81
unrequited love, and sadness over the passage of time” (212), all traits
quite evident in the film.
By comparison, Tian Zhuangzhuang as a whole displays more ambiva-
lence toward tradition, something not uncommon for Fifth Generation
directors who often explored how much tradition may have been respon-
sible for the calamities that beset China after 1949. How the new version
ends compared to the original bears this out. In both versions, Zhang
Zhichen, the guest, leaves the house of his friend, Dai Liyan, after nearly
having an affair with Dai’s wife, Yuwen, who had once been Zhichen’s
youthful love before he had left for the war. But the final two shots of
Tian’s newer version implies that Yuwen is still caught in the same predica-
ment of a loveless marriage that she now solemnly accepts as her rightful
duty: in the penultimate shot, she has returned to her room to embroider
once more in desperate solitude; the final shot shows the empty wall of
the small town, still in ruins due to the war. In Fei Mu’s version, by con-
trast, Yuwen is outdoors walking along that same wall, looking at Zhichen
departing from afar. Yet soon enough Liyan joins her at the wall, and she
points at something in the distance, an implication at least, that something
has been resolved for the better despite her clear love for Zhichen. By all
appearances, the original has a happier, more Confucian ending compared
to the remake.
Yet things are hardly as simple as that. Fei Mu is clearly a Chinese
filmmaker imbued with Confucian values, yet Spring in a Small Town
shows a marked difference from any of his previous works. As a whole,
Confucianism is a didactic tradition, but this film is anything but didactic.
What is striking about Fei Mu’s classic is the utter lack of a villain that
smacks more of 1930’s Renoir than traditional Confucian drama. Zhichen
may be a Western-trained doctor, but he does not represent the “cor-
rupting” values of the West as one would expect: he is sympathetically
portrayed as a good man in love with someone he could not marry years
earlier due to family objections and the war. Likewise, Yuwen is anything
but a tart villainess, but is treated with equal sympathy in this film as a
woman caught in a profoundly stifling marriage, yet someone who also
recognizes that her husband, Liyan, is a good person despite his illness
and temper. Liyan in turn eventually realizes the two are in love, and even
comes to believe they would be better off with each other if he were not
in the way. None of these are pure Confucian archetypes; all three char-
acters are fully fleshed out human beings; all “have their reasons” to echo
the famous line from Renoir’s Rules of the Game. It was precisely Fei Mu’s
82 J. UDDEN
seeming inability to condemn anyone in this film that would soon get him
in so much trouble with the new communist regime that would take over
in 1949. That utter lack of moralizing, that indelible sense of sympathy for
all parties concerned, including those from the “wrong” classes, is largely
what relegated this film to near oblivion until it was rediscovered in the
1980s.
That Spring in a Small Town is such a classic has much to do with when
it was made, and not just by whom. This film was produced during the
latter days of a heated, bitter, and devastating civil war, which itself came
on the heels of an eight-year total war with Japan that for many Chinese
was the equivalent of a holocaust. Fitzgerald suggests this new-fangled
realism versus the former didacticism of Fei Mu is due to an old society
now lying in ruins around him. Liyan himself is not the Confucian hero
of old in Fei Mu’s earlier works, but just an ordinary man like Fei Mu
himself (186). That the total devastation of the war looms especially large
over the original becomes more manifest when compared to the remake.
In the original, Liyan is introduced as a man sitting helplessly among
those ruins, at one point seemingly trying to rebuild it piece by piece, and
yet looking futile in doing so. Among the ruins, the servant Old Huang
asks Liyan, “Don’t we now have peace?” Liyan looks ahead and says,
“Peace?” Then he quietly scoffs. In the remake, that brief response and
significant scoff are conspicuously absent. The long, traumatic experience
of constant warfare evidentially had deepened Fei Mu’s art profoundly:
Spring in a Small Town was borne from the wellspring of abysmal cul-
tural, historical, and even existential angst. Moreover, the recurring motif
of the wall casts a shadow even over that supposedly “happy” ending:
throughout the film, those decrepit walls repeatedly serve as a constant
reminder of the uncertain present-day conditions in which this film takes
place. Moreover, they still lay in ruins as the film ends; nothing had been
rebuilt. As Olaf Möller puts it, “Traditional values are still observed in this
culture all in ruins; the inconsolable feeling is that there’s no future for
these mores” (Möller 2011: 21). Fei Mu no doubt remained true to his
Confucian values to the bitter end, yet it appears he wavered when it came
to judging how well those values could adapt to new, uncertain, festering
historical conditions.
If the 2002 remake seems markedly removed from the historical cir-
cumstances within which the original was made, then we can presume
there are significant stylistic differences as well. But which of the two
versions is more “traditional” in style? The answer depends on how one
defines tradition stylistically speaking as opposed to thematically speaking.

Often people are led astray by one particular technique when it comes to
Chinese cinema: the long take. The long take is arguably one of the most
misunderstood cinematic techniques in history. For starters, many assume
the long take achieves certain effects along the lines laid out by Bazin,
namely realism and ambiguity that is true to both the temporal and spatial
relations of the world. In truth, however, there has never been unanimous
agreement about what long takes actually achieve. Theo Angelopoulos,
one of the more famed long take directors, for example, denies that his
long takes have anything to do with realism (Durgnat 1990: 44). Pier
Paolo Pasolini argued that montage is objective, while the long takes are
intrinsically subjective (Pasolini 1980). It seems better to view the long
take as merely a blank canvas of sorts with little intrinsic meaning in and
of itself: it is not the “size” (i.e. duration) that matters, but what is done
within the boundaries of the long take that carries true weight. One only
has to look at polar opposites of Russian Ark (Alexandr Sokurov 2002)
versus Gravity (Alfonso Cuaron 2013) to realize how true that is, that in
fact the long take by itself tells us very little.
Of course, there is a claim sometimes made that somehow the long
take is a specifically “Chinese” trait. As I have argued elsewhere, however,
there is scant evidence to back up this claim. The average shot lengths
for Chinese films in the 1930s and 1940s do not exceed the average shot
lengths found elsewhere in the world at the time. On average, there were
more long takes in Chinese films from the 1940s than the 1930s, but
this was no less true elsewhere, including in Hollywood. Moreover, there
are numerous contemporary filmmakers whose long takes on average had
already far exceeded Fei Mu in Spring in a Small Town, including not
only Kenji Mizoguchi in Japan but also Jean Renoir, Otto Preminger, and
shockingly, even Alfred Hitchcock in the late 1940s (Udden, 2012). Even
if one shifts the discussion from sheer duration to more the “quality” of
the long takes, problems still remain. The best known claims in this regard
were by Lin Niantgong, who argued that Chinese long takes are indicated
by an ever roaming camera (jing you) akin to viewing Chinese landscape
paintings where the human subjects are decentered. He likewise argues for
a Chinese tendency he terms as “montage within the long take,” further
adding that the Chinese were doing this long before any European film-
maker picked up on this possibility in the 1960s (Lin 1991: 7–35, 41–49).
There are several reasons to question both of these claims as well, since
they are based on rather rigid and dubious definitions of both long takes
84 J. UDDEN
and montage: for example, a roaming camera within long takes is largely a
global norm both past and present.
For the sake of argument, however, let us assume that Lin is correct
on both counts: which of these two versions of Spring more resembles
Lin’s description of a “traditional” Chinese style? Without a doubt, it
is the 2002 remake by Tian Zhuangzhuang. The Fei Mu original aver-
ages just under 25 seconds per shot; Tian’s remake averages nearly three
times that amount, a truly long take film by both yesterday’s and today’s
global standards. Unlike Fei Mu’s version, the bulk of the scenes in
Tian’s version are plan sequences (meaning one continuous take for one
scene). As a result, they do somewhat resemble having “montage within
long takes,” something not unexpected when a scene is not broken up
into multiple shots captured from multiple camera setups. The most
telling instance is the drinking scene during the celebration of the 16th
birthday of Liyan’s little sister. In a long take lasting just over six min-
utes, all five characters in the film are present at once. This is a pivotal
moment in the narrative since midway through the revelry a seemingly
innocent remark is dropped by Zhichen about Yuwen: “I never could
outdrink her!” For Liyan, this confirmed that the two do have a history,
and maybe still have feelings for each other. Liyan suddenly becomes
very taciturn and withdrawn. This long take underscores the central con-
flict in this film between passion and propriety, and hence we have some-
thing like montage (i.e. conflict) within a long take. The stark contrast
between the happy drinking games by the others versus the suddenly
sullen face of Liyan in the background between them is made palpable,
and nobody (save for the viewers) notices when he quietly leaves. Slow
arcs and pans are employed here to follow the subtle shifts in position
by various characters at various moments. Nevertheless, this long take is
rather derivative, in part because the cinematographer was none other
than Mark Lee, who had already done much more daringly dense long
takes for Hou Hsiao-hsien. This particular scene even resembles a pared
down, simplified version of similar scenes in Flowers of Shanghai (1998),
since the lighting is primarily motivated by two oil lamps on the table
awash a golden glow. To wit, it is almost as if Tian Zhuangzhuang took
Lin Niantong’s ideas about long takes and camera movement literally
and decided this was an appropriate homage to Fei Mu. We shall soon
see, however, how radically different is Fei’s version of this same scene
in 1948.
THE TRADITIONAL MODERNIST

The actual story of Spring in a Small Town is rather simple, the very stuff
of melodramas everywhere. Two friends love the same woman, but the
woman truly loves the one she is not married to. One needs no specific
cultural training to understand the basic conflict that ensues. With these
two versions, then, the real issue becomes the moment by moment ren-
dering of this rather modest narrative. In doing so, one realizes how over-
rated the long take can sometimes be. Tian Zhuangzhuang’s version far
exceeds the original in terms of shot duration, as we have just seen, but it
pales in comparison when it comes to the indelible poetic qualities of the
original. This is largely because the originality of the 1948 version lies not
in the use of long takes per se but in three other aspects: the staging of
the actors at certain moments; the dynamic use of editing in certain key
scenes; and the voiceover of Yuwen, the one trait most studies of this film
recognize. Tian’s unrelenting pursuit of the long take came at the expense
of these three indelible traits. As a result, the 2002 remake is bleached of
much of the depth, the richness, and the subtlety of Fei Mu’s classic. It is
a more heavy-handed film instead, making the drama almost too obvious
at points. For example, unlike in the original, in the remake, Liyan goes
outside and audibly weeps right after the drinking game mentioned above.
Back inside soon thereafter, Zhichen’s drinking clearly gets out of hand in
a more blatant, almost violent, way. By comparison, Fei Mu exercised the
greatest restraint in his version, de-dramatizing the high-pitched emotions
to such an extent as to ratchet up the underlying tension and uncertainty
about what these characters might actually be thinking—or might do—
resulting in a more evocative quality overall.
Fei Mu achieved this remarkable overall effect by not aping any of the
traditions available to him in the late 1940s. He certainly remained a
Confucian filmmaker, albeit now one tempered dramatically by the bitter
experience of perpetual warfare. Nevertheless, even if he was Confucian,
this did not mean he was rigidly anti-Western in every possible way. Fei
Mu clearly opposed the May 4 ideals of rejecting all things “traditional”
in favor of all things “modern” (i.e. “Western”), but this did not translate
into a pursuit of a “traditional” cinematic style. In fact, Fei Mu was not
only conversant in several Western languages; he was also well versed in
the Western film theories of the time. Unlike most of his contemporaries
in China, he was less interested in imitation as in the exploration of film
language. Most commentators today recognize how “modernist” this film
86 J. UDDEN
seems, but some still attribute this to Chinese culture. As Susan Daruvala
states: “the film transcends its ostensibly conservative narrative surface and
through its cinematic techniques becomes a truly modernist work. It is, in
fact, these aspects that seem the most closely related to Chinese aesthetics”
(Daruvala 2007: 175). However, Chinese aesthetics is only a partial expla-
nation at best, and perhaps only for the voiceover. Just as Bresson was a
very Catholic filmmaker but not one necessarily in pursuit of a “Catholic”
cinematic style, so was Fei Mu a Confucian filmmaker open to any possi-
bilities the medium itself offered him. What he came up with in 1948 was
almost without precedent. Let us begin with the staging.
Staging
Staging is indeed a delicate art, one that only a handful of directors in
history have ever truly mastered. To date, however, few have ever even
considered how radical Fei Mu’s staging strategies in this film actually are
for the time. (Only Mizoguchi’s works in the 1930s were more arguably
radical in this regard.) Tian’s 2002 version does display some competent
staging (e.g. the drinking scene), but as already noted, it hardly goes as
far as his contemporaries such as Hou Hsiao-hsien. In Spring in a Small
Town, on the other hand, some strikingly oblique staging techniques are
employed that would later be identified with Michelangelo Antonioni. Fei
Mu could not have possibly known who Antonioni was in 1948, since the
Italian director did not complete his first feature-length, fictional film until
1950 (Chronicle of a Love Affair, another film involving a love triangle of
sorts). It is extremely unlikely that Antonioni saw Fei’s film either, since
it barely registered at the time within China, let alone anywhere else. In
other words, both directors came up with similar solutions in complete
isolation from each other.
In the case of Spring in a Small Town it would seem that staging should
be relatively simple, since there are only five actors total that appear in the
film (plus one chicken), something done largely for budgetary reasons.
Yet throughout Fei maintains a strong sense of mystery about what these
characters are truly feeling, and what they might actually do. Indirect
staging is one of the key causes of this effect. At times, we see characters in
oblique angles that do not allow a full registering of their facial emotions.
Sometimes characters have their backs completely to the camera, enhancing
the guesswork required of any viewer, something that would eventually
become an Antonioni trademark. Fei Mu couples this strategy with other
stylistic traits that further obscure the hidden depths of these characters,
most of all the night scenes in Zhichen’s room which are often captured in
extreme chiaroscuro lighting schemes that are motivated by the frequent
electrical blackouts, another subtle reminder of the troubled larger con-
text these characters inhabit.
The first clear instance of this staging motif occurs right after the opening
credits: a long shot shows Yuwen wandering along the wall from directly
behind. Later, when she is embroidering and Old Huang announces the
arrival of a guest named Mr. Zhang, she is shown in a mostly oblique pro-
file. The camera then tracks in slowly as her voiceover suggests it could be
him, but it also could not be him. Right after Zhichen’s arrival, a scene
occurs indoors where Little Sister is singing a song for everyone else, yet
Zhichen cannot keep his eyes off of Yuwen. Yuwen, however, first has her
back to Zhichen, after which her back is completely to the camera, leaving
us no clues whether she is reciprocating his look or not. After Zhichen
settles in his room, Old Huang brings him an orchid, a gift from Yuwen:
there is a cut-in to the orchid placed on the table, but the two shots
sandwiching that cut-in both show Zhichen from behind, obscuring his
reaction as well.
The most telling moment, however, occurs the first time Zhichen and
Yuwen are alone at the wall to talk things over. Zhichen asks her: “If I
asked you to leave with me, would you just say ‘whatever’?” She answers:
“Seriously?” She then walks in front of him and then gives a brief, enig-
matic smile until her back is to the camera. He slowly walks behind and
thus places his back to the camera as well. Zhichen then picks up a stray
brick from the ruined wall and throws it into the river below. She sees this
and extends her arm to him. Then he grabs her, and she suddenly stops.
He then gently grabs her arm only and they slowly walk further away from
the camera. A dissolve to the next shot reveals them in an extreme long
shot along a path with a bamboo fence on the right side. Once again, we
can only see them from behind as they walk further away from the camera
in a scene done with no sound whatsoever. They briefly step apart from
each other, looking at each other; then slowly they come back together as
a chicken walks by in the foreground. She then grabs his arm. For reasons
unclear since no sound is heard, suddenly they stop. She separates herself
and runs off ahead of Zhichen, her back to the camera once again. He
then begins to chase her as the image fades to black. There is no way of
telling what words, or even glances, or even hints, were being exchanged
at that moment. Fei suggests much, but reveals very little.
88 J. UDDEN
This use of oblique staging is used at select moments in this film, but
there are other staging strategies in play as well. Fei Mu staged the drinking
scene in a much different manner than Tian, largely because Fei employed
more editing. One medium close-up of a slightly inebriated Yuwen shows
her moving over toward Liyan and looking at him, only to then push him
away lightly as she decides to drink with Zhichen instead. The next shot is
a medium shot of all three as she commences a new drinking game. Then
there is a cut-in to a concerned Liyan as he realizes what is really going
on. As he sits down, the camera pans right back to Yuwen in a medium
close-up. Liyan then returns to the frame as he walks slowly and dejectedly
behind her, and the camera pans right to follow him. In this case, the edit-
ing and the staging are intricately synchronized. Despite the fact there is
no long take in this scene, Fei Mu renders it in a much more delicate and
poetic fashion than Tian Zhuangzhuang due to the other stylistic choices
he made. This includes not only staging but also editing.
Editing
The pivotal drinking scene calls attention to another aspect of style in
Spring in a Small Town that has been largely overlooked: editing. As did
most directors in the late 1940s anywhere, Fei Mu completely abides by
the well-established rules of continuity. This means this film displays none
of the fragmentation that will return with a vengeance in the late 1950s
and 1960s among the various new waves around the world. Fei Mu does
use more dissolves than was the norm at the time, and for very calculated,
poetic effect. Yet he also displays real dexterity and imagination within
the supposedly rigid confines of continuity editing even when only using
straight cuts. Take this drinking scene: Tian Zhuangzhuang captured this
scene in a single long take over six minutes long as seen above; Fei Mu
did this same scene in just under three-and-a-half minutes and nine shots.
Yet as seen above, Fei Mu actually gained something with the variety
of shot scales that he used to emphasize key moments, most of all the
closer shot scales toward the end, which both hide and reveal at the same
time. The first four shots of this scene are fairly standard when it comes
to continuity editing, but the fifth shot, the medium close-up following
Yuwen, allows us to register her unexpected and significant little shove of
Liyan. Likewise, the seventh shot described above follows Liyan instead as
he realizes what is going on. The eighth shot isolates Yuwen and Zhichen
happily continuing their drinking game, indicating they have no idea that
Liyan has caught on. The last shot shows Liyan in a medium shot with his
back to the camera which exemplifies how staging and editing are work-
ing in tandem again. He slowly sits down and grabs toward his heart in
an understated manner. The end result is a much more richly delineated
scene, the “conventional” editing notwithstanding.
One of the oldest (and global) editing conventions is crosscutting,
namely the cutting back and forth between two lines of action in two
separate spaces, usually implying they are occurring simultaneously. Fei
Mu uses this particular technique for very calculated effect on the first
night of Zhichen’s arrival. After Yuwen has provided Zhichen with some
extra bedding and they discuss Liyan’s health, the siren goes off indicat-
ing that the electricity is about to be cut off. She turns on the other light
and lights a candle as well. They both then sit down. The next shot is a
medium close-up of Liyan in his own room lying on bed, opening his eyes
and looking ahead. One can surmise that the siren simply has woken him
up, but one can also speculate that he somehow suspects something is
going on in the other room, even though Yuwen and Liyan normally sleep
in separate quarters. The next shot returns to Zhichen’s room from the
same camera setup as before, and the two are still seated without saying
much of anything, and Yuwen is looking away. Right after the electric-
ity cuts off, there is a cut back once more to Liyan whose eyes are now
at least half closed. Then we return once more to Zhichen’s room, only
now Yuwen is looking intently Zhichen’s way, unlike before. He suddenly
begins to laugh, but she breaks down crying instead. He tries to comfort
her, but she merely signals for him to leave her alone and she decides to
leave. Using one of the oldest editing tricks in the book, Fei Mu in a very
understated manner expresses the subtle depths of despair and longing
that are emerging. Multiple viewings still do not make clear the exact pur-
pose of those two inserted shots of Liyan lying in bed, yet it is impossible
to imagine this scene without them, providing this moment with an added
albeit indefinable texture.
Still nowhere is editing more ingeniously employed in this film than
the remarkable scene of an outing on a boat that occurs roughly a third
of the way in. This particular scene underscores how different these two
versions actually are. Tian captures this scene in one long take of roughly a
minute-and-a-half in length; Fei’s version is slightly longer at 1 minute and
40 seconds, yet he does this in fifteen shots instead of one, all conjoined
via straight cuts. In Tian’s version, Yuwen and Liyan are sitting side by
side as husband and wife, with Zhichen behind, and Little Sister in front
90 J. UDDEN
singing; in Fei’s version, Yuwen is seated behind both Liyan and Little
Sister who instead are sitting side by side in the front. Just behind Yuwen,
seated alone, is Zhichen who is standing as he steers. This provides ample
opportunities for stolen glances between the two secret former lovers that
never occur in Tian’s version. However, only the ninth shot even suggests
that the two indeed are looking at each other at the same time, and only
for a split second. In the third shot, a medium shot of Zhichen shows him
looking Yuwen’s way very briefly, but the subsequent shot from the front
does provide no clue as to whether she was looking back. In the fifth shot,
we see a medium shot of Yuwen as she looks back Zhichen’s way, then
looks down in front of her. The camera then tilts up and pans left back to
Zhichen, but he is only looking straight ahead, not at her. The 14th shot,
a long shot from the front, suggests that Yuwen may have just ended yet
another brief look Zhichen’s way, but he is still looking elsewhere, not at
her. Fei Mu enriches this scene by inserting four “empty” shots of water
rushing by, adding a poetic touch: the first and last shots of this scene
are isolated shots of the waves in the river rushing by; the seventh shot is
another inserted shot of the water as well, while the 11th shot is a strik-
ingly brief shot of an oar thrusting through the same water. For a scene
that is so quickly edited compared to the film as a whole, the end result is
nevertheless richly evocative and suggestive, a near-perfect balancing act
of the carefree obliviousness of two people in front versus the understated
yet palpable tension of two people in the rear. Without the editing in this
case, the lyricism would have been lost. Indeed, that is exactly what hap-
pens in the remake.
Voiceover
If the crucial roles played by staging and editing have been largely over-
looked in previous discussions of Fei Mu’s 1948 original, the same can-
not be said of the voiceover. What is particularly noteworthy about this
voiceover is how it can be seen as both traditional and ultra-modern at
the same time. Both Carolyn Fitzgerald and Susan Daruvala convincingly
source many techniques in this film back to a particular genre of Chinese
poetry known as “boudoir poetry” from the tenth to the thirteenth cen-
tury. These poems often involve a woman alone in her room, longing for a
lost lover, often watching a candle as it “melts into tears.” Daruvala notes
that all of the visual motifs in this film, whether it be the moon, the orchid,
or the candle can be found in this sort of poetry (175). Fitzgerald builds
on that idea to suggest the voiceover is also derived from this same strand
of traditional poetry (196).
On the other hand, Nick Browne notes how often the voiceovers
are used as “structural and compositional resources” in several modern
European films such as Bresson’s Diary of a Country Priest (Browne 1980:
233). Moreover, he notes how in Bresson’s case there is a complex rela-
tionship not only between sound and image due to the voiceover but also
between what is past and what is present: “The ‘pastness’ of the voice-over
narration is qualified, to the advantage of the fiction of presentness of
action, by including, on the level of the narrated story, the act of writing
the diary” (235). Once again, for the same reasons, we cannot source his
staging techniques to Antonioni, Fei Mu could not possibly have been
inspired by Bresson’s classic to utilize the voiceover in this way: Diary of
a Country Priest came out in 1950, two years after Spring. Nor could he
have found inspiration in another innovative use of the voiceover from
Europe at that time, namely Jean-Pierre Melville’s Silence of the Sea, which
did not come out until 1949. Clearly, Fei came up with this technique
on his own, independent of the other two, just as Melville and Bresson
had done so independently of him. Nevertheless, this reveals how much
Fei Mu was on the cutting edge of a burgeoning trend found in post-war
world cinema on multiple fronts.
Many filmmakers have now long explored the temporal ambiguities
that the voiceover offers, not to mention the ambiguous divide between
what is diegetic versus what is non-diegetic. (Terrence Malick is one recent
example of a director fully committed to such ambiguities.) Yet Fei Mu
is arguably one of the first directors in history to fully recognize these
possibilities. Carolyn Fitzgerald makes an astute observation by how at
times Yuwen’s voiceover points out things she as a character could not
have known: for example, while she was outside with Zhichen, Liyan goes
to her room alone, and the voiceover notes it is the “first time he had
been there in years”; or when Zhichen is arriving, she describes somewhat
redundantly in her voiceover: “He came from the train station. He walked
through the city wall.” Fitzgerald further points out the profound tem-
poral ambiguities sometimes lost in the English subtitles, since Chinese
is uninflected, unlike English (192–194). This suggests that Chinese as
a language might have offered Fei some new ways of thinking about cin-
ematic language in regard to the voiceover. Yuwen is both narrator and
character; she inhabits a liminal state, not just somewhere between the past
and the present but somewhere in between the diegetic and non-diegetic
92 J. UDDEN
worlds. Sometimes she is speaking for herself, expressing the depths of

her despair in the present; sometimes she is speaking strictly for our ben-
efit about things she as a character could not possibly know. Tradition,
whether poetry or even language itself, provided Fei Mu an opportunity
to create something that was ahead of its time. At the same time, however,
Fei Mu was inspired by tradition, but he was not constricted by it; he was
not in search of a perfect analogue of boudoir poetry in cinema, because
he knew full well that no such thing exists. Instead, Fei Mu searched for
the right mixture of cinematic techniques from a modern art form that
happen to express what he loved from his own tradition. The voiceover
was one of those resources.
The end result is not something familiar and classical in the Hollywood
sense of the time, but something very modernist instead. There are some
astonishing moments in that voiceover beyond what was just described.
From the very first sentence spoken about passing life every day in a town
that never changes, the voiceover displays a breathy, deeply haunting, and
utterly desolate quality. For a culture whose dramatic traditions emphasize
social roles rather than the depths of individual psychology, this voiceover
is unprecedented for how deeply it penetrates the mind of a solitary female
character. Early on, for example, as Yuwen’s voiceover questions how sick
Liyan really is, she then adds: “I lack the courage to die, he lacks the cour-
age to live.” Perhaps such a line can be sourced to “boudoir poetry,” but
it seems this is just as likely a very modern expression of “post”-war angst
in a Chinese setting.
What is most overlooked about this device, however, is a deeper pur-
pose for this voiceover that is structural in nature. Tian’s remake is lon-
ger than Fei’s original by roughly 20 minutes, yet it begins right away
with Zhichen’s arrival, completely eliding the desperate psychological
backdrop for Yuwen that the original took pains to establish for the first
11 minutes. During that opening section, it as if there is nothing Yuwen
will not reveal to us. Then suddenly some key shifts begin the moment
Zhichen arrives. In the first shot, where we see Zhichen carrying luggage
and walking along a path, the voiceover says, “Who knew that person had
arrived?” Later, just over 17 minutes into the film, she finally goes out
and approaches him for the first time. At that very moment, an unsettling
shift occurs in the voiceover: “He really doesn’t know I married Liyan.
Why did you come? For what possible reason could you have come? How
can I even face you?” Most commentators have noticed this dramatic shift
from third to second person. Yet this marks an even more profound shift
in the film overall. After spilling to us via her voiceover the utter depths
of her domestic despair for the first ten-plus minutes, from that point
on Yuwen’s voiceover becomes mostly perfunctory and merely descrip-
tive. Some commentary thereafter merely provides a modicum of tempo-
ral orientation: “The morning of the second day;” “The morning of the
third day;” “The second Sunday, the ninth day after his arrival.” Some
are completely redundant: “We went rowing on the river. Little Sister
sang.” Only at brief moments are any somewhat deeper psychological
facts revealed, such as when her voiceover indicates she regretted saying to
Zhichen something she had never contemplated before, “unless he dies.”
Even then, her immediate facial expression provides sufficient evidence
of her own horror for saying such a thing. What the voiceover does not
provide, however, are any clear clues as to which way she is leaning after
she sees Zhichen and recognizes fully their predicament. Profound psy-
chological explicitness had now shifted to an equally profound psychologi-
cal concealment instead. Moreover, this large-scale strategy is dovetailed
with the other techniques mentioned earlier to achieve that same effect,
namely editing, staging, and sometimes lighting. What should have been
trite melodrama, is instead transformed into a profoundly poetic work
replete with mystery, a seamless amalgam of the traditional and the mod-
ernist, an almost perfect even if inadvertent last testament of a sensitive yet
open-minded Confucian artist, one deeply troubled by the world he faced
in China in 1948.
CONCLUSION: THE LOST OPPORTUNITY

Fei Mu died not long after completing Spring in a Small Town, aged
only 44. In all likelihood, he never became aware that he was part of a
larger, global trend, a post-war revival of cinematic modernism that had
largely been in eclipse during the 1930s and the war. Fei Mu was not
somehow “ahead” of his time even though he predated by a year or two
techniques we normally associate with the early Antonioni, Bresson, and
Melville; rather, like these others, he was a product of his time. It seems
no coincidence that the first heyday of cinematic modernism occurred in
the 1920s after World War I, while the second heyday of the 1950s and
1960s came on the heels of World War II. In both cases, the years follow-
ing a traumatic and total war apparently spawned two distinct generations
of filmmakers outside of Hollywood, who collectively searched for a new
cinematic language in face of a brave, new, and most uncertain world.
94 J. UDDEN
(Even in Hollywood this was somewhat true, albeit within much more
restricted commercial parameters.) For the latter generation, the baby
steps of post-war modernism were first being taken by Italian neorealism
just as Fei Mu was making Spring. Later this would be fully developed
by directors who became known as part of “art cinema” or sundry “new
waves” scattered across the globe starting in France. Fei Mu was part of
this early, larger exploration across the globe, working alongside not only
Antonioni, Bresson, Melville, but even Kurosawa in his one-off experi-
ment, Rashomon (1950), which would in time become the very epitome—
if not the very definition—of the most universal trait of modernist/art
cinema, namely ambiguity.
But Fei Mu was different than these other notables because he was
Chinese. Fei Mu was not just a product of his time; he was also a victim of
his time. Unlike his contemporaries in Italy, France, and Japan, he made
this film in the throes of a devastating civil war. Moreover, unlike these
other contemporaries, there were virtually no portals for Chinese filmmak-
ers outside of China. Venice famously introduced Kurosawa and Japanese
cinema to the rest of the world in 1951 when it showed Rashomon and it
won the Golden Lion, opening the floodgates for Japanese cinema there-
after. There was no similar gate for China at the time. Instead, after the
takeover of Mao’s communists in 1949, Chinese cinema would be cut
off from the rest of the world until the 1980s with the Fifth Generation,
including Tian Zhuangzhuang, who collectively came to admire their
own lost cinematic master, Fei Mu. Before that, however, Fei Mu became
persona non grata in his own country, and remained unknown outside
of China. Today, he still is for the most part. Today, few recognize that
what Fei accomplished cinematically is the equivalent of what Melville,
Kurosawa, Antonioni, Bresson, and the neorealists had accomplished
around the same time. Indeed, given the hostile conditions in which he
operated, one could even argue his accomplishment was the greatest of
them all.
Spring in a Small Town suggests one other thing: there is no contradic-
tion between “tradition” and “modernism.” This is not to suggest that
they are perfectly compatible, but that they can be independent variables.
In terms of story, this film is about as clichéd as they come; in terms
of its cinematic rendering, however, it is audaciously and breathtakingly
original for its time. By comparison, Tian’s Springtime in a Small Town
seems rather old-fashioned, even backward-looking. It uses long takes,
but long takes are now largely an art cinema cliché, and few can employ
them with the originality of Bela Tarr, Hou Hsiao-hsien, or Jia Zhangke.
This is not to denigrate Tian’s remake—it is a very honorable homage,
well made even if not very original. The problem is that he is remaking
what is considered in Chinese-speaking circles today the greatest Chinese
film ever made, yet he failed to recognize how truly great it is or why.
He failed to replicate the deft strategies in both staging and editing; and
remarkably he decided to leave out the voiceover. Yet the original is not
merely arguably the greatest Chinese film ever made, it is arguably one of
the great overlooked films of its time period. Belatedly, the Chinese have
rightfully reclaimed their lost cinematic master. Now it is time for the rest
of the world to recognize Fei Mu for what he truly is: one of the most
overlooked cinematic masters in history.
CHAPTER 6
Remaking Ozu: Hou Hsiao-hsien’s Café

Lumière
Emilie Yueh-yu Yeh
CAFÉ LUMIÈRE PRISM

To commemorate the 100th birthday of Ozu Yasujiro (1903–1963), the
Japanese film studio Shochiku invited Hou Hsiao-hsien to Tokyo to make
a film to honor Japan’s most venerated director. This brought Hou to one
of the countries where he first built an international reputation (Hasumi
2008). Where then did he begin his dialogue with Ozu, whose style for a
long time was assumed to be the prototype for Hou’s work? How to make
a ‘tribute film’ that also bears the signature of Hou Hsiao-hsien?
As the commission must be a Japanese language picture, shot in Tokyo,
Hou was expected to remake Ozu’s best-known work, Tokyo Story (1953).
Did Hou remake Ozu as Yamada Yoji, another Japanese veteran, did in
his Tokyo Family (2013)? An unlikely story. Hou goes against the grain,
making only tangential reference to his target source Tokyo Monogatari.
He deliberately bypasses the aging couple as a point of departure, revers-
ing the sentimentality of familial disintegration. In remaking Ozu, Hou
turned to the couple’s daughter, a young rebellious writer whose daily
routines set the stage for Hou’s own Tokyo story.
Hou began the story idea with a friend, his long-term Japanese inter-
preter, Osaka Fumiko. Osaka likes to work in coffee shops and Hou’s
E.Y. Yeh ()

Hong Kong Baptist University, Hong Kong, China

DOI 10.1057/978-1-137-55309-6_6
98 E.Y. YEH
visits to Tokyo were often filled by intervals in a variety of cafés (Lan

2004). Thus the Japanese title Kohi jiko, ‘café light’ and its approxi-
mate Chinese title Kafei shiguang, ‘coffee time.’ Japanese kanji 珈
琲—kohi—is used to indicate coffee, with 時光—jiko—representing a
Chinese phrase for time. Here Hou makes a compound title with both
familiarity and distance to both his Japanese and Chinese audiences.
As a Japanese word for coffee, the characters of kohi (珈琲) will not
seem correct to Chinese viewers. At best, it would appear a misnomer.
The correct Chinese phrase for coffee is 咖啡, pronounced as kafei, that
transliterates the English term. And jiko is not a Japanese word either.
So a Chinese audience would have doubts in reading the film’s title
while a Japanese audience would need to surmise the meaning of jiko as
something between time and light, comparing the two kanjis, ji (time)
and ko (light). Hou uses this intercultural title to create ambiguity, déjà
vu, familiarity infused with hesitation.
Café Lumière (2003) is thus conceived as an intercultural translation,
intended as a multi-coded work, with its production background, trans-
national implication, and authorial imprints. The multiple valences extend
further by a direct reading of its international title, literally, the Lumière
Café, taking us straight back to the centenary of cinema. They pair the
Ozu/Hou doublet with the Lumière brothers, the legendary founding
fathers of cinema, and a café, the site where the first public cinema exhibi-
tion took place, in Paris of late 1895. ‘Café Lumière,’ coffee shop under
light, is thus more than a handy reference to a personal experience and
an intercultural wordplay. It is also a historical conceit, evoking the gene-
alogy of cinema (which is now the aging media of the new century) by
offering a retrospective on historical phases. Recall the scene that initiates
the film. A low angle shot with a tram rolling across a gray sky choked
with power lines, utility poles, and rooftops. It is ghostly, not clear what
time of day this is, and on the soundtrack we hear only the moving train
itself. In its unhurried yet confident course, the train sounds its brakes, as
if arriving at its destination, leaving the cinematographic terrain moving
offscreen.
Café Lumière thus opens with a tram arriving at a local station, as
if to rehearse the primordial moments of the cinema. Later we learn
that this is likely to be the vehicle that carries the protagonist back
to her Tokyo apartment at the beginning of the day. This trundling
car mobilizes a Deleuzian ‘crystal image,’ a time machine carrying a
series of sounds, tracks, and stories that converge and jostle with one
REMAKING OZU: HOU HSIAO-HSIEN’S CAFÉ LUMIÈRE 99
another. As the film unfolds, we will soon witness the image of time, a
prism carried by trains and tracks crisscrossing in Tokyo’s vast transport
network.
Café Lumière centers on Tokyo counterparts, their interests, idiosyncra-
sies, and ambiguous relations. Yoko (陽子, sun child) and Hajime (肇, to
begin) are good friends, and they are unconventional. Yoko is a freelance
writer who occasionally works as a Japanese teacher in Taiwan. Like Osaka
Fumiko, she likes to work in coffee shops. At the beginning, Yoko is writ-
ing a book on the legendary Taiwanese–Japanese composer Jiang Wenye
who was active in 1930s and 1940s. Meanwhile, she is pregnant with the
child of her Taiwanese boyfriend whom she has no intention of marrying.
Hajime runs a bookstore and when he’s not attending to his business, he
records the sounds of trains running through the arteries of metropolitan
Tokyo. Yoko too spends much time on trains. The two young people’s
paths crisscross in the intricate network of the city’s mass transit system.
Although they seem to lead completely different lives, with their distinct
interests and inspirations, they are connected, knowingly or unknowingly,
intentionally or accidentally, through the passing trains. Café Lumière on
its contemporary facade thus mobilizes several overlapping conjunctures,
each of which—the Lumière brothers, arriving trains, Jiang Wenye, Ozu,
and the structure of Tokyo’s arteries—mark the nodal points of Hou’s
commission.
On the face of it, Café Lumière embodies a new page in Hou Hsiao-
hsien’s career. It is Hou’s first foreign and commissioned work: his first
attempt to address film history and the first picture he shot completely
outside Taiwan, outside his comfort zone. With its novelty, Café Lumière
has been appraised from a variety of perspectives and in multiple takes
(Yue 2008; Hasumi 2008; Udden 2009; Lin 2009; Sing 2010; Chang
2011; Wada-Marciano 2012). In Taiwan, critical attention to Hou had
declined since Good Men, Good Women in the late 1990s. But when Café
Lumière was released, it prompted a number of substantial treatments.
Taiwan scholar Lin Wenchi (2009) uncovers the mise en abyme inside
Café Lumière, outlining its mirroring effects that assimilates Yoko’s state
of mind to Hajime’s electronic sketches and Maurice Sendak’s Outside
Over There, a story about stolen babies. Lin attends to the intertextual
specificities of the film and redirects our attention to other textual dupli-
cations in Hou’s oeuvre. Another scholar Chang Hsiao-hung (2011,
pp. 9–10) situated the film in Gilles Deleuze’s theory and argued for the
film’s realization of ‘a new philosophy of becoming in-between the latent
100 E.Y. YEH
and the manifest, the virtual and the actual.’ Chang focuses on the edit-
ing device of fades used in the beginning and closing of the film. Instead
of treating fades as a set of formal properties indicating the passage of
time, Chang transcribes the fades-in and -out as an affective switch that
reconfigures Tokyo as a ‘city-body.’ Here Chang calls attention to Tokyo’s
public transport as an organic entity that awakens repressed desires and
mobilizes unspoken relations, despite its impersonal, functional design.
In advancing this city-body image, Chang elaborated on the perceived
‘obliqueness’ (Yeh 2001, p. 69) of Hou’s drama in Daoist ‘blandness’
(dan)—a neutralization of sounds and fury into a state of indifference, and
transcendence. Chang says that in this blandness, the liminality of visible
and indiscernible, or actual and virtual is illuminated. What is normally
opposed, conceptually and categorically, can be merged, exchanged, and
plunged into undecidability.
By contrast, Sing Song-yong (2010) views Café Lumière as Asian cin-
ema’s answer to the idea of ‘cinema revisited,’ an emergent genre cat-
egory and practice in rewriting film history. Taking issue with adaptation
study as a rewrite of a priori work, Sing suggests looking at the remake as
an uncanny ‘phantom-effect’ of the deceased, the lingering spirit of cin-
ema past. The spectral phantom-effect is nonetheless double binding. On
the one hand, citing film classics is a tactic for contemporary filmmakers
to shelter the cinema regime beset by digital imaging proliferation, the
overflow of electronic simulacra. Yet a revisit to a bygone era, meanwhile,
perpetuates and sustains problematic influences of the cinema ‘fathers.’
Cinema revisited is thereby a possible trade-off, a new bottle containing
old, sometimes bitter sediments.
These three articles are important additions to the literature on Hou
Hsiao-hsien. They provide new thoughts and methods to grasp Hou’s
representation of the millennial life, routinely carried by communica-
tion technologies such as cellphones, computers, audiovisual recordings,
and public transport. While respecting their arguments, I take a holistic
look at Café Lumière, examining its rendition of film history and the
cultural politics between Japan and Taiwan. In paying homage to Ozu,
Hou evokes Jiang Wenye as a ‘fellow traveler,’ who accompanies him in
his pursuit of a Taiwanese ‘Tokyo story’ in the new century. Jiang is a
contemporary of Ozu but never received due recognition and respect
throughout his life or posthumously. Hou’s shrewd recovery of Jiang
Wenye is where we can anchor the politics of a Taiwanese–Japanese
coproduction. Instead of shying away from a textual comparison between
Hou’s ‘remake’ and its possible host, I take a straightforward approach,

by way of segmentation, to note the textual correlatives, parallels, and
intergenerational reverberation, and to reach an understanding of Hou’s
design in interweaving the life of a Tokyo woman with the history of
cinema and Sino-Japanese cultural politics. Time and history in Café
Lumière are codependent on each other. By tracing the diegetic time of
Café Lumiere, we find Yoko’s schedule is not as ‘bland’ (Chang 2011,
pp. 20–1) as it might have appeared. Instead, her ‘uneventful’ daily activ-
ities, when mapped out chronologically, reveal epistemological clues in
bringing to light Yoko’s agenda. That is, by examining her ostensibly
aimless daily routines, we come to a revelation of a young woman’s desire
in the Tokyo megacity. And with these clues we can advance to the analy-
sis of the Hou/Ozu/Jiang relation, which offers points of entry into
Hou’s revision of cinema and history.
CAFÉ LUMIÈRE SEGMENTATION

Day One
Scene 1: Introducing Yoko. At home, on the phone with someone about

returning from Taiwan, but soon departing for Takasaki. Landlady
interrupts, accepts gift of sweets. Back on the phone, Yoko shares ‘weird
dream’ about a melting baby. [Title credit]
Scene 2: Tram ride, followed by JR train. On the train, Yoko inspects a
pocket watch.
Scene 3: At bookstore, sees Hajime. He gives her music CDs that she
ordered, and happily receives pocket watch from Yoko, for his birthday.
The watch is a souvenir celebrating Taiwan railways. They listen to the
music [Jiang’s Maggio Suite] and discuss Jiang Wenye, a musician who
frequented a Tokyo coffee shop called Dat. Hajime provides informa-
tion on changelings, re: Yoko’s dream.
Scene 4: Yoko checks a locker at the station. Re-packs. [Jiang music
continues.]
Scene 5: Yoko falls asleep on the train.
Scene 6: Rural Yoshii station near Takasaki. Met at station by Father.
Scene 7: At home, Mother in kitchen. Yoko requests beef potato stew for
dinner. She falls asleep.
Scene 8: At night, Yoko turns on kitchen light, looking for food, and tells
Mother she’s pregnant.
102 E.Y. YEH
Day Two
Scene 1: Grave cleaning. Passing trains. Rural landscape.

Scene 2: Family eats noodles.
Scene 3: At home, family rests. Yoko leaves house by bike. Parents at
home, Mother discusses Yoko’s future, scolds Father.
Scene 4: Raining. At country station, Yoko asks about a cat. [Jiang’s
Formosa Dance]
Day Three
Scene 1: Erika coffee shop [Formosa Dance]. Yoko makes appointment

with Madame Koh Nobu, Jiang Wenye’s Japanese widow. Phone call
from Hajime. From Nonchan Yoko hears amusing gossip about Hajime.
Scene 2: Yoko leaves Erika. Bookstore. Hajime gives her ‘Outside’ picture
book. She treats him to Erika delivery coffee.
Scene 3: On the train, Yoko inspects the book.
Scene 4: Home. She reads Sendak story, Outside, Over There, in English.
Scene 5: Late night, during thunderstorm, she calls Hajime, explaining
her recollection of birth mother.
Day Four
Scene 1: Tomaru bookstore, Koenji. Owner knows nothing of Jiang

Wenye. Yoko takes a picture.
Scene 2: A call from Hajime, appointment at Ochanomizu station.
Scene 3: Yoko falls ill en route, in Shinjuku.
Scene 4: Resting on platform, Yoko calls Hajime.
Scene 5: Ochanomizu converging trains. On platform below, Hajime
records.
Scene 6: At Yurakucho station, Hajime learns Yoko is pregnant.
Scene 7: In a café, studying old map of Tokyo, seeking Jiang’s Dat café.
Directions provided by café manager. [Jiang’s Three Dances fades in].
Scene 8: Yoko and Hajime at the site of Dat café. [Jiang’s Bagatelles].
Yoko takes a picture.
Scene 9: Yoko arrives home [music fades out]. Phone call from Mother
about their visit to Tokyo the next day. She requests tasty beef potato
stew.
Scene 10: Yoko asleep, feeling ill. Hajime visits, cooks for her and shows
her his train-womb graphic. [Jiang’s Bagatelles].
Day Five
Scene 1: Café Erika. Yoko calls Hajime, leaving him a voicemail.

Scene 2: A different coffee shop. Yoko with Madame Koh, admires her
scrapbook, hears stories of Jiang Wenye.
Scene 3: Yoko leaves coffee shop, takes train. [Jiang’s Formosa Dance] As
trains pass, we see Hajime, but Yoko doesn’t.
Scene 4: Yoko meets her father. They arrive at her flat, with Mother.
Yoko enjoys beef potato stew brought from home. Father feeds her his
potatoes.
Scene 5: Borrowing sake from landlady, receiving delivery sushi. Yoko tells
parents about her Taiwanese boyfriend and his family. Father drinks.
Day Six
Scene 1: Yoko returns sushi boxes. Finds Nonchan, who tells her Hajime
is out recording.
Scene 2: Train ride downtown. Yoko falls asleep.
Scene 3: Hajime enters the train and sees Yoko. He comes over.
Scene 4: They both ride the train. Hajime continues recording.
Scene 5: Both exit the train, pause on platform.
Scene 6: Ochanomizu converging trains. [Closing song]
YOKO’S SIX DAYS

Following the segmentation, we see that the film records Yoko’s six days
since her arrival in Tokyo. In these six days, we find that her daily activi-
ties are filled with two recurring and interwoven threads: reunions and
the Jiang Wenye research. What connects these two threads is her friend
Hajime.
The day after her return to Tokyo, Yoko re-establishes her relations and
at the same time, continues with her research. Gift exchanges carry this
forward. In scene one, Yoko calls Hajime and gives a typical souvenir (a
pack of pineapple sweets) to her landlady. In scene three at Hajime’s book-
store, she receives Jiang Wenye’s music recordings that Hajime sources
and reciprocates with a birthday present, a Taiwan pocket watch. In this
exchange, we see the first indication of the film’s Japanese-Chinese title—
time (jiko, shiguang), presented in a timepiece and a gift to a friend. Here
‘time’ is given to Hajime. It is not just a piece of time offered in goodwill;
it also reminds us that, from the initial phone call to the bookstore visit,
104 E.Y. YEH
Yoko’s ‘first’ time after arriving back in Tokyo is to be spent with Hajime.
Yoko’s offering of her time to Hajime reveals that he occupies a key place
in her mind. Only after seeing Hajime, Yoko returns to her parents for
another reunion.
The theme of reunion and research continues on the second day. Yoko
and her parents visit the family grave. In the afternoon, she revisits her
old haunts, listening to the rain while Jiang Wenye’s music enters non-
diegetically, which transports her back to Tokyo to her regular café, Erika.
Though at home with her family, Yoko’s mind is elsewhere. The music of
Jiang Wenye might indicate her preoccupation with research; it may also
link to the recordings given by Hajime the day before. Here research and
Hajime converge and intertwine.
Day three repeats the first day: Yoko reunites with Hajime and the
two exchange gifts, again. Hajime gives her a book that relates to her
dream about stolen babies—Outside Over There by Maurice Sendak—and
she treats him to a pot of coffee. Coffee, from the film’s title, makes its
first appearance. Being pregnant, Yoko drinks only milk in coffee shops
and coffee does not really have a place in the story until it is presented
as another gift from Yoko to Hajime. With the second gift, coffee and
time are reunited, and their union takes place in Hajime. Late at night
Yoko calls Hajime about her dream. Day three thus begins and ends with
Hajime. To this point, Café Lumière emerges as a love story, though
implicit and understated.
On day four, Yoko once more reunites with Hajime, who keeps her
company in researching Jiang’s old traces in Tokyo. It begins with Yoko’s
visit to Tomaru, an old bookstore frequented by Jiang in the 1930s. She
then meets with Hajime who has found an important clue for her proj-
ect. She reveals to him her pregnancy. Together they discover the site of
Jiang’s favorite café, Dat. Later that day, Hajime checks up on her and
they share his computer drawing, as if offering a response to Sendak. In
the picture book, a baby is spirited away to a weird, uncertain ‘over there.’
In Hajime’s artwork, the baby is floating, cradled in a web, or womb, of
ersatz JR trains. On this day, Yoko deepens her research and her relation-
ship with Hajime.
Yoko’s fifth day is again filled with reunion and research. The day’s
major events include a meeting with Madame Koh and hosting her parents
visiting Tokyo. Madame Koh shows Yoko her family photo album where
we have glimpses of the versatile and cosmopolitan Jiang Wenye in Japan.
Meanwhile, Yoko tries to reach Hajime but fails, as if the day is not com-
plete without spending some time with Hajime. Providentially they briefly
cross paths on parallel trains. Without knowing it, Yoko and Hajime do
meet. This fateful, serendipitous connection comes to fruition on the last
day. The day opens with Yoko looking for Hajime and they finally catch up
with each other, on the train. End of story.
Reunion and Jiang Wenye occupy Yoko’s schedule in these five days.
In this period, Yoko has two reunions with her parents, still, Hajime is
the person whom she actually wishes to see. That he is the centerpiece
in Yoko’s time is fully illuminated when she gives him first the pocket
watch and second, a coffee. The gifts are both tangible and intangible for
they signify trust, companionship, and bonding. Furthermore, Hajime’s
importance is magnified by his involvement in her project, as he is useful
in providing key information. Without the project on Jiang, the meet-
ings with Hajime would not have been so frequent. Research on Jiang is
also deepened by the result of reunions with Hajime. Things brought by
Hajime—a music disc, an old map and a picture book that reminds her of
motherhood—all of these objects serve as memoir involuntaire, allowing
Yoko to remember her childhood and to excavate a forgotten figure, a
concealed history.
Hence, reunion and Jiang Wenye research have a mutual connection,
constituting each other. For instance, the research trip taken by Yoko and
Hajime together ends with an intimate sharing. That day after finding the
old Dat site, Hajime appears at Yoko’s place to look after her, after she falls
ill. He cooks for her and then shows her his artwork comprised of spiral
digital images of trains that indicate the Tokyo transit system. Within this
picture is a self-portrait—a floating, fetus figure enclosed in a red environ-
ment and wired by a headphone and a pocket watch. This uncanny image
simulates an umbilical cord connecting fetus to a womb-like space satu-
rated with staggered trains. Trains appear like mother’s body, but within
this body, Hajime is alone and seems lonely, desperate. Yoko comments on
the sad eyes of the self-portrait. Hajime agrees.
The scene is pivotal—a culmination of their friendship and suggestive
of a growing bond between the two. From the beginning of the film,
Hajime’s presence is indicated in the phone conversation, though implicit.
The night when Yoko comes to grasp her dream, she calls Hajime. Once
more her friend’s importance is noted, though not explicitly depicted,
let alone melodramatically. Hajime is all along a helpmate, but always
with enryo, restraint (Richie 1974, pp. 155–6). Hours after he learns of
her condition, he arrives and shares his picture of a child inside a womb,
106 E.Y. YEH
responding to the news confided by Yoko earlier. Here, Hajime sends a

consoling message to Yoko. Evidently, he too feels like a lonely child inside
the tangled machinery of railways, routinized though with an organic
rhythm of its own. Here, Hajime is both friend and guardian to Yoko. He
helps her trace the steps of Jiang Wenye in a Tokyo that no longer exists;
he also helps her cope with an uncertain future by showing understanding
of her maternal anxiety. With this exchange of feelings, Yoko is no longer
alone as she finds a companion in her journey to motherhood. Like an
odd couple, Yoko and Hajime share their secrets, idiosyncrasies, fears, and
hopes (Fig. 6.1).
So we see that her first action the following day, before the crucial
meeting with Madame Koh, is to call Hajime. She wishes to meet but
does not have time. We then see their brief crossing and though they fail
to make contact with each other by voice or sight, they do ‘actually’ meet,
unintentionally, by the accidental design of the city’s complex railway. In
other words, though they can’t see each other because they are not look-
ing for each other, they are seen together, spatiotemporally, because of the
bond they have, and the sentiments they share.
These six days of Yoko are not loosely structured, nor are they as ‘bland’
or casual as they seem. Instead, these days present a tissue of entwined
contexts and implications, carrying specific agendas of Yoko’s pursuit of
Fig. 6.1 Yoko and Hajime meet up inside the train, unexpectedly
security, friendship, and love, perhaps. The segmentation illuminates the

film’s schematic mapping of a contemporary Tokyo woman with transna-
tional network, mobility, and above all, freedom. She pursues her interests
and chooses her own biopower, with or without the sanction of her family
or society.
INTO THE DARK: TOKYO TWILIGHT

How does the segmentation help us come to terms with the Hou/Ozu
relation? Most writings on Café Lumière use Tokyo Story as a convenient
peg on which Hou’s tribute to Ozu could be hung. Possibly because of
the limited knowledge of Ozu’s oeuvre, more likely driven by publicity,
most writers clutch at the obvious. They find first, the 180-degree axis
violation in shooting Yoko’s parents’ living room. We see the shot taken
from the living room, showing her stepmother cooking in the kitchen.
This is followed by its reverse shot taken from the kitchen, showing Yoko
lying on the tatami in the living room (Udden 2009, p. 173). Second, the
rice wine borrowing alludes to Noriko, the stepdaughter, borrowing food
and drink for her in-laws staying with her (Udden 2009, p. 173). Third,
parents from the countryside with nowhere to stay avail themselves of
children’s cramped space (Yue 2008).
On the making of Café Lumière, the French documentary, Metro
Lumière, presents a segment comparison between the film and Ozu’s
Equinox Flower and An Autumn Afternoon, based on the premise of the
daughter’s marriage. One review on the film’s DVD release mentions the
comparable situation of the marriage conundrum of the daughter in Ozu’s
Late Spring and Late Autumn (Schager 2005). Yoko, like Noriko in Late
Spring and Ayako in Late Autumn, has reservations about the sanctioned
path for a woman to follow in adulthood. Their refusal of marriage thus
becomes the conflict in need of a resolution.
Tokyo Story, Late Spring, An Autumn Afternoon, and Equinox Flower
are among the finest works by Ozu. Motifs in these films, such as fam-
ily dissolution, interchangeability of roles and recurring situations, the
irrevocable time passages, and so on, could be found in Café Lumière.
Nonetheless, the similarities tend to be provisional, derived from a general
take on certain situations, like the melodrama of the unmarried daugh-
ter and her worrying parents. In fact, Yoko is nothing like Noriko (Late
Spring) or Ayako (Late Autumn), whose concerns are leaving parents
unattended once they themselves get married. Filial obligation is behind
108 E.Y. YEH
the decision of the Ozu daughters. Does this circumstance apply to Yoko
in Café Lumière? By no means. Yoko is more like an indulged child with
demands than a filial daughter anxious to look after her parents. Even in
Tokyo, Yoko is fed by her parents visiting from the country. Yoko’s out-
of-wedlock pregnancy is the issue. Her parents wonder if she can afford to
be a single mother, since they do not have the means to support her. This
is certainly not the case in Late Spring or Late Autumn, thereby making
the comparison weak and unsustainable.
While offering their Hou/Ozu associations, many critics concede that
Café is far from being an Ozu remake. Indeed, it is difficult to compare
the two. How then can we better substantiate the Hou/Ozu relation
beyond the perfunctory? Instead of looking for similarities, I suggest that
we examine places where Hou’s rendition is notably different from Ozu.
Following this, we find that Café Lumière rewrites an Ozu film never men-
tioned in the literature. That film is Tokyo Twilight (Shochiku 1957).
As the title suggests, the film tells a gloomy story set in Tokyo. It cen-
ters on a difficult household presided over by single father Sugiyama,
whose wife Kisako left him and two young daughters with his subordinate
years ago. Sugiyama’s elder daughter Takako returns home with her baby
girl following a rift with her alcoholic husband. Meanwhile, the younger
daughter Akiko is pregnant and has been searching in vain for her boy-
friend Kenji. Kisako has returned to Tokyo and runs a mahjong parlor
frequented by Kenji’s friends. In the coming days, Akiko goes through a
police interrogation, an abortion, and a dreadful reunion with Kisako who
wants to know the daughters she left behind. Distressed by the boyfriend,
and shocked by her mother’s illicit affair that split the family apart, Akiko
is ‘accidentally’ killed at a train crossing. The mother who seeks in vain
Takako’s forgiveness eventually leaves Tokyo and her daughters for good.
Like Café Lumière, Tokyo Twilight (henceforth Twilight) is about Tokyo
and about light. Against the luminosity in Café’s Tokyo, Twilight portrays
the city’s gradual dissolution at day’s end, into the dark and its hidden
secrets and shame. It was Ozu’s last black-and-white film. The light in
Twilight is so dismal that it closes with the death of the younger daughter.
Akiko’s plight begins with the discovery of her condition, which she with-
holds from her family. Knowing Kenji will not help her, she chooses to
terminate the pregnancy. Brokenness, disillusionment, and defeat define
Akiko’s state, underscored by the film’s dusky photography and nocturnal
settings, in the depth of winter. A good portion of the film takes place
late at night, inside seedy bars and cafés where Akiko hopelessly searches
or waits for Kenji. At one point, Kenji stands her up at a coffee shop. She
waits. As the night deepens, the café gradually becomes menacing. In an
unusual sequence, Ozu cuts to the bizarre faces of the patrons sitting near
her. These faces either leer at the saddened Akiko or sink into an abyss
of confusion. One of these faces turns out to be a plainclothes cop. He
takes Akiko away and books her at the police station. Akiko’s pregnancy
thus brings her a series of lapses associated with downtrodden places—a
gambling den in a decrepit area, a remote gynecological clinic that rids
working girls of their problems, a murky café patrolled by the police, and
a late-night police station populated with alcoholics, perverts, and petty
thieves. All of these are punitive measures to portray Akiko as a deflated
character.
As if the disillusionment toward Kenji is not enough for Akiko to bear,
the film renders her abortion yet another chilling humiliation. Akiko
undergoes the procedure alone and is taken as a bargirl. Back home, Ozu
stages an excessively melodramatic scene to push the wounded, dejected
Akiko further into despair by having her confronted by Michiko, her tod-
dler niece. This begins with a couple of shots showing Takako attending
to her domestic chores while Michiko plays in the corridor. The corridor
is filled with sunlight coming through the window. Then Akiko returns
home; she rests at the entrance that leads to the corridor where Michiko
plays. Takako notes her weakness, rushing out of the frame to make up
a bed for her. The film then cuts to two consecutive shot/reverse shots
featuring the two characters left in the corridor—Michiko and her suffer-
ing aunt. In Ozu’s typical cutting, in the first shot we see Michiko playing
and looking at her aunt and the reverse shot is the dark back of Akiko who
also looks back at her niece. The second set of shots changes to a close-up
of Akiko’s face, covered by shadows. Baby girl Michiko with her innocent
look faces the shadowy Akiko who has just aborted her future baby. Akiko
can no longer hold her emotion, burying her sobs with her hands. In
the bedroom, Takako tells Akiko about the visit from their aunt earlier
that day with the photos of two young men for her marriage prospect.
Akiko coldly replies that she cannot and will not marry. Immediately after
Takako leaves, Akiko sinks into an emotional slough, out of shame and
guilt about her abortion (Figs. 6.2 and 6.3).
Café Lumière, on the other hand, uses abundant light to safeguard
Yoko and the life she is carrying. Hou Hsiao-hsien and cinematographer
Mark Li Ping-bing follow closely the visual motif, time, and light, sug-
gested by the film’s title and the name of Yoko, child of the sun. The result
110 E.Y. YEH
Fig. 6.2 Michiko greets auntie Akiko (Tokyo Twilight, 1957)
Fig. 6.3 Akiko buries her sobs in her hands (Tokyo Twilight, 1957)
is omnipresent sunlight on the streets of Tokyo and glassy reflections on

windows, coffee shops, and bookstores where Yoko spends her time. Even
on Yoko’s train rides, we see so much bouncing reflection that the frame
becomes a cascade of onrushing images, dizzying, colliding, abstract. The
torrent of light translates lucidity and simplicity in a time of change. Yoko’s
life is rushing forward, with its own steady rhythm and plan.
In Tokyo, Yoko has her own place, a flat she does not share with oth-
ers. She is also comfortable in occupying a space in public, either in coffee
shops, on the trains or sidewalks. For Yoko the boundary segregating the
‘feminine’ from the ‘masculine’ space is dissolving. It is difficult to see
a gendered spatial layout here (domestic as feminine and public as mas-
culine). Her workstation is often inside a coffee shop; she walks the city
of Tokyo to conduct research, treating the city as her field of investiga-
tion. Nor do we experience an absolute binary of the domestic as con-
finement against the public as liberty. Coffee shops and trains are just as
habitable as home. The built environment for Yoko is resolutely different
from that associated with Akiko. Yoko constantly de-territorializes and
re-territorializes to chart new spaces; Akiko hovers, circles, and ends on a
deathbed: ‘I don’t want to die’ are her last words.
Yoko’s decision to be a single mother is not only supported by the
bright summer light, but also the warmth she receives from people around
her: parents who cook her favorite beef potato stew, coffee shopkeepers
that bring her milk, and above all, Hajime who helps her any way he can.
Yoko is the blessed child of the sun. She is protective of the child she car-
ries, but does not withhold its existence from her parents or Hajime.
Tokyo Twilight has been considered one of Ozu’s darkest films because
of its melodramatic excess and its rare noir façade—dark streets, seedy
café, bars, gambling parlor, and police station (Bordwell 1988, p. 340).
Ozu too, after feuding with his screenwriter, considered the film a failure
(Li and Shu 2003, p. 150). It is an anomaly in light of Ozu’s controlled
features of classic home drama—the perennial problem of aging parents,
departing daughters, dissolution of traditional family. But Ozu also com-
plained that most people did not read the film correctly. This film is about
the father’s household eventually resuming its order and peace, after the
death of his younger daughter and the departure of his elder daughter and
former wife. Resumed stasis concludes with Takako returning home to try
to work out her marriage and Kisako leaving without any reconciliation
with her estranged family. Indeed, the film’s last scene opens with a sunny
morning when the father begins his daily routine; the house without the
112 E.Y. YEH
daughters is now refreshed with lots of light. Ozu said that subplots about
the daughters were mere embellishment: ‘although the film had been said
to be about a young woman’s transgression, for me, the emphasis was first
and foremost on Ryu Chishu’s life – how a man whose wife has deserted
him would cope... as for the younger generation, it merely served as a
parallel. However most people only had eyes for what was intended as
embellishment to the main theme’ (Li and Shu 2003, p. 150).
Despite Ozu’s claim that Tokyo Twilight is a ‘father’s picture,’ the rep-
rimand given deviant women is harsh—Akiko with her life and Kisako in
her exile to the far north. Twilight is a rare noir picture in Ozu’s postwar
career, a response to Nikkatsu’s popular taiyo zoku, the sun-tribe youth
genre created by Ishihara Shintaro in the early 1950s. Ozu’s portrait of
disruptive youth is a challenge for the father to endure, a respectable
banker. Such a view is not shared by Hou, however. In Café Lumière, the
father chooses to remain silent and supportive toward his daughter’s deci-
sion. Hou’s depiction of a puzzled, yet empathetic father reflects a posi-
tive and liberating depiction of the mobility and freedom enjoyed by the
young woman opting for an unconventional path.
Mobility in Twilight is treated as irresponsibility and destructiveness.
Sugiyama’s wartime mobility seems to bear the blame for the breakdown
of the family. It was during Sugiyama’s secondment in Korea that his wife
committed adultery at home. When he returns, Kisako elopes with her
lover to Manchuria. Wartime mobility enabled by Japanese imperialism
and licentious youth’s postwar liberty are presented as hindrances to a
traditional, well-functioning family. Mobility for women especially results
in a definite internment and domesticity. The film’s ending seems to sug-
gest that women with freedom bring about terrible consequences; it is
bad for them and their children too. Akiko’s constant travel between her
affluent neighborhood Zoshigaya and Gotanda, the lower part of Tokyo
populated by cheap tenements, bars, cafés, and mahjong parlors eventually
ends with a dark closure. Takako heeds the warning and decides to return
to her husband. What choice does she have?
Café’s pro-life and open-endedness is thus contrary to Twilight’s abor-
tion, death, finality, and awful resumption of order. Here Hou’s tribute
to Ozu is not revisiting or spectral haunting (Sing 2010, p. 31), but con-
version of a tragedy into a feminist, pro-choice quest. The Tokyo gamine
treasures her choices, her freedom, her transnational mobility, and linguis-
tic, intercultural competency. Though Hou was once criticized for being
insensitive to women’s struggle against the backdrop of political turmoil
or social dislocation (Mi 1991; Liao 1989), autonomous female figures

have always been present in Hou’s pictures; from his earliest Cheerful
Wind to Flight of the Red Balloon, there is no shortage of strong women
who prefer to follow their own convictions (Shen 2013).
Yoko in Café Lumière reprises Hou’s previous heroines. Her trans-
national movement is not just a personal hobby, but a response to a
historical calling. Histories related to Japan and its former colonies in
East Asia are often dark and difficult to face, as in Tokyo Twilight. As
a postwar picture, Twilight re-examines the aftermath of the war and
holds an equivocal view on human vulnerability, in that trust and loyalty
are subject to circumstances. Café meanwhile attends to history as a well
of inspiration and resources. Yoko’s cosmopolitan disposition therefore
takes her into the gaps in imperial history and seeks to repair the histori-
cal connection between Japan and Taiwan. In the next section, we will
look at how this mobility prompts a new understanding of the Taiwan–
Japan relation and Hou’s own position in the long list of Taiwanese
talent residing in Japan.
REUNION WITH JIANG WENYE: A GLIMPSE

OF THE EMPIRE’S EDGE
Yoko’s major task in those six days is to trace the fading steps of Jiang
Wenye, a composer who lived in Taiwan, Japan, and China. It is also
by revisiting Jiang that Yoko is able to form a bond with Hajime, who
becomes her surrogate family in Tokyo. Here reunion binds research and
research promotes reunion. Who is Jiang Wenye and why was he writ-
ten into this story? Jiang Wenye or Koh Bunya (1910–1983) was an
obscure figure in Sino-Japan history until the early 1980s when censor-
ship on Taiwan nationals in Communist China was relaxed. Jiang Wenye
was born in Taiwan, grew up in Southern China, was educated in Japan,
and became a prominent composer in Japanese-controlled North China
in the late 1930s. In 1945, when the Pacific War ended, the Taiwanese–
Japanese composer Jiang was tried in Beijing for wartime collaboration
and served time. After his release, he continued to stay in Beijing, contrib-
uting to building the new curriculum of music education. Like many of his
peers, Jiang was purged during the anti-rightist campaign of 1957–1958.
Condemned as a counter-revolutionary during the Cultural Revolution,
Jiang was sent down to a labor camp for reform in the 1970s. Jiang died a
few years after his rehabilitation in 1978.
114 E.Y. YEH
In many ways, Jiang Wenye’s life embodies the gray areas between
colonial Taiwan and her ruler Japan, between the two warring states of
Japan and China, and between Communist China and Nationalist Taiwan.
Jiang came from a wealthy family whose cross-straits shipping business
in early twentieth century brought him a cosmopolitan lifestyle blended
with local, regional, and international cultures. Jiang was raised Catholic
in the expatriate quarters of Xiamen, a southern treaty port populated
with British, Portuguese, and Spanish residents. He knew Western music
and was a talented baritone at a young age. This prepared him to break
into the Japanese music establishment when he chose to give up his train-
ing as an engineer to become a musician. Without formal pedigree, he
managed to practice music professionally in the guarded and conservative
circle of classical music. In advancing his composition, Jiang was inspired
by the American Russian musician Alexander Tcherepnin (1898–1977),
who encouraged Jiang to seek the roots of music that was closer to home
in East Asia. Jiang’s Formosan Dance, a piano piece comprised of motifs
from Taiwan’s aboriginal music, won a prize at the 1936 Berlin Olympic
Games. The piece is heard on day two and day five, in Café Lumière.
So when an invitation to teach music in Beijing came, Jiang happily
accepted the offer. Like many Taiwanese enticed to pursue their ambitions
in China, Jiang’s work there helped advance Japan’s ambition to unify
Asia into an anti-Western bloc. This was the so-called Greater East Asia
Co-prosperity Sphere (Dai Towa Kyoeiken). When the war ended, these
Taiwanese in China, including Jiang, were arrested and tried for treason
for their assistance to Japan, their adopted country. After serving time,
many of these people chose to stay on in China. They were all buried in
the postwar Chinese historical narrative for the identities they once held,
and their work for the opposing side.
In Beijing, Jiang was immersed in the ritual music of Confucius as well
as provincial folk melodies. Jiang’s involvement in the production of the
Greater East Asian culture brought him professional satisfaction and finan-
cial rewards that he would not have had if he had stayed behind in Japan.
War against China provided endless inspirations and raw materials and
the Japanese imperial army awarded him with exposure and recognition.
Between 1938 and 1942, orchestras in Tokyo played the music Jiang sent
back from Beijing. Meanwhile, he was commissioned to write propaganda
songs to be broadcast in the puppet state of Manchukuo (Tamura 2007).
In addition, he composed scores for five ‘national policy’ films to advance
the agenda of the imperial army (Lin 2005, pp. 102–6). These works
include documentary films on Japan’s conquest of two capitals: Nanjing

(dir. Akimoto Takeshi 1938) and Beijing (dir. Kamei Fumio 1938), and
three feature films: Path to the Peace of East Asia (Toyo heiwa no michi, dir.
Suzuki Shigeyoshi 1938), Daughter of the Earth (dir. Li Pingqian 1940),
and Hot Wind (Neppu, dir. Yamamoto Satsuo 1943). These films were
major productions directly or indirectly funded by the Japanese military
(Tsuji 1998, p. 58; Misawa 2012, p. 224). Like many Japanese responding
to the call of the Emperor, Jiang spent his most productive years in assist-
ing Japan’s aggression against China. It was no small wonder the Chinese
authority arrested him when Japan was defeated and surrendered.
Why is Jiang Wenye invoked in a story about twenty-first-century Tokyo?
To what extent does Jiang’s story testify to art’s dependency on imperial
power and identity politics? Like Hou who had to adapt himself to the new
space in Tokyo, Jiang Wenye responded to opportunities available during
the Second World War. Though there is hardly any similarity between the
lives of the two, Hou’s resurrection of Jiang has at least three intentions.
First, to redeem Jiang’s place in history, the first ‘Japanese’ to win a major
award in an international music contest. Although Jiang was rehabilitated
in Taiwan and China, he remains virtually unknown in Japan. Second, to
show the genealogy of Taiwanese art in contemporary settings, that Hou is
not the first nor the only one, but one of several, including a gifted musi-
cian who was forgotten, buried in postwar disavowal. Taiwanese/Chinese
talents active in postwar Japan include singer Teresa Deng Lijun, esteemed
chess player Lin Kaihou, and Sadaharu Oh, a celebrated pro baseball player
and manager of the Yomiuri Giants. These figures once sought professional
achievements in Japan and opportunities, fortune unavailable in a fragile
island state like Taiwan during that period. They became household names
in Japan, and Deng even grew to be a popular voice all over East Asia. It
is important therefore, for Hou, to remind the world of Jiang Wenye, a
forsaken figure fallen into the cracks of history. Finally, to resurrect Jiang is
to invite his spirit to walk with Hou as a fellow traveler in his exploration
and discovery of Tokyo’s past and present.
To do so, Hou brings Jiang’s music into the soundscape of Yoko’s
journey during those six days. And the Jiang Wenye moments are, save
the closing credits song, the only times the soundtrack privileges music.
Besides these, the film’s soundtrack has only ambient city sounds and dia-
logue. On these occasions, Jiang’s music is also used as a bridge for spatial
transition and a sonic vehicle carrying Yoko’s thoughts and movement.
On day one, scene three, in the bookstore, Yoko gives a piece of time to
116 E.Y. YEH
Hajime in exchange for two compact discs of Jiang’s music sourced by

Hajime. They listen to one, along with Hajime’s dog. This is Maggio Suite
(May Suite 1935), which plays over Yoko’s account of Jiang and lingers on
the soundtrack when Yoko goes to the train station, preparing to depart
for the countryside. Maggio Suite has two roles: as a sound bridge for a
smooth, plausible transition, and also like a benign angelic figure on air.
It is Yoko’s sonic companion from diegetic (bookstore) to non-diegetic
audition (train station), as she moves from place to place, city to coun-
try. Following Yoko home, on day two, scene 4, Jiang’s Formosa Dance
resurfaces during her visit to childhood haunts, on a rainy afternoon, then
transporting her back to Tokyo, settling her in at the sunlit Erika Café,
while talking to Madame Koh on the phone. On day four, scene 7, just
before Yoko finds Dat spot, Three Dances (1936, op. 7) fades in. At the
Dat site, the soundtrack changes to Bagatelles, op 9, while she takes a
picture. Jiang Wenye does not yet disappear. Later that day, Hajime pays a
visit, and another piece from Bagatelles op. 9 is heard when Yoko appreci-
ates Hajime’s digital self-portrait (scene 10). Here Jiang’s music seems to
provide a support of the enhanced relationship of these characters. Finally,
on day five, scene 3, after the meeting with Madame Koh, as Yoko walks
into sunlight, Formosa Dance follows her, passing some tall willow trees,
waving gently in the wind (Fig. 6.4).
Fig. 6.4 Madame Koh shows Yoko photos of Jiang Wenye (aka Koh Bunya
1910–1983)
Hou’s placement of Jiang Wenye in the film’s diegesis is more than

a historical rehabilitation. Jiang is invited back, less to proclaim his lost
reputation than to perform, with Hou, a collaborative film-work between
Taiwan and Japan. Despite being a commissioned work, Hou’s interest
lies not in paying a literal tribute to a Japanese master, but in juxtaposing
himself with other masters, including the Taiwan-Japanese musician Jiang,
an Ozu contemporary. This advances both the textual and historical depth
of Café Lumière, through Hou’s double, simultaneous citation of masters
of different kinds. Compared to Ozu, the citation of Jiang is made more
explicit and pronounced. Jiang’s performance in this contemporary film is
twofold. First, it continues the conservation effect of Taiwan’s renowned
artist Li Tianlu, in The Puppetmaster, in that Jiang’s music is re-animated
in contemporary scenes, albeit brief and discreet. Second, it calls atten-
tion to Jiang’s involvement in film history despite its dark affiliation with
Japanese imperialism. And, given that wartime record, there is no better,
other way to ‘rehabilitate’ Jiang by inviting his best-known compositions
to play in a contemporary rendition of the city where his talent bloomed,
but never had the chance to return. Jiang Wenye interludes are an exqui-
site, ingenious movement added in Hou’s juxtaposition with Ozu and
Japanese cinema.
There are key parallels between Hou and Jiang, their liminal cultural
place, international reputation, and vulnerability to political contin-
gency. Reunion and research, the two agendas in Yoko’s six days, light
up a forsaken, forgotten figure in Sino-Japanese history. Yoko’s research
illuminates an opaque spot in the past and aligns Hou with Jiang in the
passage of Japanese and Taiwanese art. Indeed, Hou and Jiang are both
like passengers on a historical railway. This alignment and alliance is for-
ward looking, suggesting a new Sino-Japan relation that actively looks to
recognition and reconciliation, leaving disavowal behind.
INTO THE LIGHT: A CODA

Like a tourist, Yoko sees the sights of Tokyo. She takes pictures, reads,
writes, and asks for directions. She interviews Jiang’s widow. She looks
for Hajime, a lot. She is not lost, though she often has to rest, and picks
her way through places of waiting, meeting, and transit. Tokyo’s vastness
is mapped by the number of train lines announced, shown, and imagined.
Every station seems a portal to another trajectory, beckoning toward dis-
tinct terminals. Viewers familiar with the city will have their own memories
118 E.Y. YEH
of various stations, lines, and transit points. Yamanote, Chuo, Sobu, and
other major commuter lines appear and reappear. Green line Yamanote
seems prominent, because it runs in a never-ending loop through central
Tokyo, and even reappears in Hajime’s artwork. Hou and cinematogra-
pher Lee Ping-bing were shooting on the fly, because they did not get per-
mits from the railways. While in the trains and stations they capture many
people just going on their way, alone, smoking a cigarette, gawping at
the actors Asano Tadanobu and Yo Hitoto, or hurrying with their foreign
children in tow. Everyone is on the move, on their various ways. Yoko’s
peregrinations make her part of a larger, diffuse entity, with anonymous
figures as metonymy for her and she for them.
Café Lumière pauses on intermediate, transient spaces where Yoko’s
routines and quests alight. On the trains, there is so much reflection from
the windows that the frame becomes a rushing image cascade, dizzying,
colliding, abstract. In both these visions of transit spaces and glassy reflec-
tions, the setting is paradigmatic. That is, Yoko’s story is set aside, sug-
gesting other crisscrossing lines of action and agents who happened there
and then. Peoples’ private lines of action take place simultaneously across
the complex public transport lines. This invites contemplation of simul-
taneous, intersecting pursuits. Hasumi writes that ‘the constantly pass-
ing trains seem to have been transformed into something other than a
mode of transportation,’ unlike those of Hou Hsiao-hsien or Lumière’s
La Ciotat locomotive (Hasumi 2008, p. 193). Could that ‘something’ be
a contemplation of time, chance, and ephemerality?
Within the reflective, multifaceted prism of Café Lumière, coffee shop
under light, we see an overspill of light that penetrates the organs of the
city, the minds of its dwellers, a crepuscular film of Ozu, and the colonial
blemish of Taiwan and Japan. Making a straight, clear-cut dissection opens
an aperture on the textual details that carry the filmmaker’s centenary
musings on film, history, and politics.
CHAPTER 7
Hong Kong Puzzle Films: The

Persistence of Tradition
Gary Bettinson
Since the worldwide success of Crouching Tiger, Hidden Dragon in 2000,

globalizing forces from two different sources are said to have diluted
the local flavor of Hong Kong cinema. On the one hand, Hong Kong
filmmakers have faced increasing pressure to ‘Hollywoodize’ their long-
standing production practices and storytelling strategies. On the other
hand, the pressure to ‘Mainlandize’ Hong Kong movies has intensified in
concert with an industrial shift toward cross-border coproduction. These
two pressures are not entirely distinct—to Mainlandize is in part to adopt
Hollywood-style strategies already assimilated by the Mainland film indus-
try. Nevertheless, critics argue that both pathways endanger the cultural
and aesthetic distinctiveness of Hong Kong cinema. These global forces,
it is argued, augur the abandonment of local production practices, the
dissolution of an indigenous aesthetic ‘flavor,’ and the suppression of cul-
tural and political expression.
The tendency toward Mainlandization gained impetus with the
Closer Economic Partnership Arrangement (CEPA), a free trade agree-
ment signed by Hong Kong and the People’s Republic of China (PRC)
in 2003. CEPA, in turn, triggered a sharp increase in joint ventures
produced under the aegis of the State Administration of Radio, Film
G. Bettinson ()
Lancaster University, Lancaster, England

DOI 10.1057/978-1-137-55309-6_7
120 G. BETTINSON
and Television (SARFT).1 Among its various mandates, SARFT makes

the preproduction screenplay a prerequisite for shooting, a practice
alien to traditional Hong Kong filmmaking. Similarly, the pressure to
emulate Hollywood film production revolves around the pre-composed
screenplay. Industry executives and commentators, anxious to revive the
region’s ailing cinema, have urged Hong Kong filmmakers to imitate
Hollywood script construction, to ‘correct the traditional tendency to
undervalue the importance of scripts,’ and to adopt ‘good storytell-
ing techniques’ from Hollywood movies.2 Local critics adduce Infernal
Affairs (2002) as an exemplar. Mounted from a full-fledged script
(Marchetti 2007, p. 178), the film displays a level of narrational com-
plexity rare in Hong Kong cinema (and moreover it became a worldwide
commercial success).
Implicit in these script-oriented imperatives, however, is the fal-
lacy that Hong Kong cinema’s traditional storytelling practices—most
saliently, the local custom of shooting without a complete script—divert
the filmmaker from canonical script structure, yielding simplistic and
structurally inferior plotting. By contrast, the Hollywood-style, SARFT-
prescribed method—that of the preproduction screenplay—typically
leads to ‘well-told’ storytelling. This chapter takes issue with these
assumptions. I argue that local storytelling practices have endured in
spite of state regulations, that those practices are hardly inimical to
narrational complexity and sophistication, that a ‘local flavor’ persists
even in Hong Kong–China coproductions (which themselves bear
traces of Hollywoodization), and that a strain of local puzzle filmmak-
ing coalesced in the CEPA era, indebted to Hollywood yet not wholly
Hollywoodized. This cluster of claims testifies to the adaptability, as well
as the continuity, of Hong Kong cinema’s distinctive and durable work
routines.
For many critics, the effacement of local characteristics from Hong
Kong cinema is most apparent in the jingoistic military costume dramas
jointly produced by Hong Kong and China in the post-CEPA years. The
Myth (2005), A Battle of Wits (2006), The Warlords (2007), An Empress
and the Warriors (2008), and Saving General Yang (2013) embody
Chinese nationalist fervor at the expense of Hong Kong elements. The
purported erasure of cultural heritage prompts critic Bono Lee (2012,
p. 193) to posit the emergence of a ‘post-Hong Kong cinema,’ in which
the trademark features of Hong Kong cinema survive only as structuring
HONG KONG PUZZLE FILMS: THE PERSISTENCE OF TRADITION 121
absences. According to Esther Yau (2015, p. 17), ‘the Cantonese film
legacies and local Hong Kong stories that gave [Hong Kong] cinema its
reputation remained largely absent from the “coproduction films” (he pai
pian).’ Critics regard the rise of China coproductions and the concomi-
tant threat to Hong Kong scriptwriting practices and storytelling norms
as symptoms of a wider Sinofication of Hong Kong culture. For Esther
Cheung (2015, p. 96), at stake is nothing less than ‘the ontological defini-
tion of Hong Kong cinema,’ a claim echoed by Lee’s notion of post-Hong
Kong cinema.
Still another source of cultural erasure, critics maintain, is the emergent
Hong Kong puzzle cinema. As Bordwell (2006a, b) points out, ‘intricate
plotting and well-earned twists [had] never been strong points of local
cinema,’ but Hong Kong puzzle films teem with such elements, resem-
bling the virtuosic plotting of Hollywood productions such as Shutter
Island (2010), Black Swan (2010), and The Usual Suspects (1995). At first
glance, such films testify to the Hollywoodization of Hong Kong movies.
And yet, I shall argue presently, the Hong Kong puzzle film maintains a
long-standing tradition of local storytelling (as such, it is not really a ‘new’
trend at all). More broadly, I refute the existence of a post-Hong Kong
cinema. In what follows, I examine three contemporary Hong Kong
films—a defiantly ‘local’ policier (Mad Detective [Johnnie To/Wai Ka-fai,
2007]) and two China coproductions (Peter Chan’s Wu Xia [2011] and
Johnnie To’s Blind Detective [2013]). All three films motivate narrational
complexity by embellishing a hoary crime genre trope: an overzealous
detective hero, endowed with a heightened intuitive capacity to empa-
thize with the criminal, uses his inexplicable psychometry to solve the
narrative crime.3 These complexly plotted movies, moreover, are subsum-
able to Hong Kong’s nascent puzzle-film trend. Along with other puzzle-
centered dramas such as In the Mood for Love (2000), Infernal Affairs,
and 2046 (2004), they are apt to jumble story chronology, conflate objec-
tive and subjective reality, furnish unreliable flashbacks and untrustworthy
narrators, spring deus ex machinas, and foreground other narrationally
restricted devices. So much narrative experimentation, I shall argue,
springs partly from the films’ industrial and creative modes of production.
By way of illustration, I outline the work methods employed by Peter
Chan, Johnnie To, and Wai Ka-fai; and I examine the three films’ nar-
rational stratagems, highlighting both their complexity and their embodi-
ment of local traditions.
122 G. BETTINSON
PUZZLES AND PRACTICES: PETER CHAN

AND WU XIA (2011)
A veteran of the pre-handover Hong Kong cinema, Peter Chan was

weaned on local scriptwriting practices. He inherited the local custom
of ‘script-by-brainstorming,’4 a process of collective creation that, under
Chan’s auspices, results in a draft shooting script. Several rewrites may
be undertaken prior to filming, but even the formalized shooting script
remains but a loose prototype of narrative action. As per Hong Kong
tradition, the story coalesces in piecemeal fashion throughout the filming
and postproduction phases.5 ‘Like all creative processes in Hong Kong, we
write as we go,’ Chan says. ‘So we don’t really know exactly how the sto-
ry’s going to turn out.’6 Before the 1997 handover, the custom of over-
seas presales seldom necessitated full script approval; foreign distributors
settled for a story outline and a cast of bankable stars. However, if ‘piece-
meal plotting’ (Bordwell 2000, p. 183) held sway in the pre-CEPA era,
Chan’s later films such as Wu Xia, The Warlords, and American Dreams
in China (2013)—all mounted as China coproductions—were required
by SARFT to be fully scripted in advance of shooting. In accordance with
SARFT regulations, these scripts underwent a process of official review,
enabling SARFT not only to impose pre-emptive censorship, but also (as
critic Lisa Leung points out) to ensure an acceptable level of ‘quality’
(2013, p. 121).7 After shooting and editing, the films were subjected to a
further mandatory round of SARFT censorship.
How did Chan adapt deep-rooted local practices to the straitened
demands of PRC coproduction? Wu Xia’s script genesis testifies to the
artistic and industrial pressures that shape a coproduction’s story con-
struction. For the film’s main plotline, Chan revived the genre topos of a
master swordsman who, retired from the martial world and disguised as a
common villager, is coerced back into the jianghu (martial underworld).
By Chan’s own admission a ‘typical Chinese wuxia storyline,’ this prem-
ise has formed the basis of numerous Shaw Brothers wuxia pian. (Chan
took The New One-Armed Swordsman [Zhang Che 1971] as a model.)
Embracing a well-worn genre situation, however, intensified the artistic
and commercial pressure of novelty: ‘How do we make a martial-arts
movie that is different? That was the first and foremost consideration,
even before we had thought up the characters and story details.’ Along
with screenwriter Aubrey Lam, Chan chose to amplify the realistic impli-
cations of actual kung fu combat, investing scenes of physical action with
forensic plausibility. He also sought to complicate filmic narration, the

better to defamiliarize a formulaic plotline. The final script submitted to
SARFT contained both these elements, as well as an implicit aim—again in
the quest for novelty—to modify leading man Donnie Yen’s star persona.
As often, SARFT’s script censors excised aspects of the script judged
politically contentious. No longer could the script’s villainous characters
be depicted as ethnic Tanguts, for fear of inflaming peoples in Northwest
China (particularly Tibetans, whose heritage closely intersects with Tangut
history). In addition, scenes of physical violence had to be truncated and
sanitized. (In the absence of a movie ratings system in the PRC, films must
be designated suitable for all viewers; hence graphically violent scenes are
forbidden.) Only once Chan had acceded to the censors’ demands was Wu
Xia put into production. Yet despite the forced adoption of working prac-
tices quite divorced from those of Hong Kong filmmaking—including the
creation of a full-bodied preproduction script—Chan’s de facto shooting
method ensured that local filmmaking practices survived by stealth. His
strategy fused pragmatism and sidelong rebellion:
Because the script needs to be censored before you make it into a movie,
people assume that the script is thereafter written in stone. But that’s not
true. Script censorship and postproduction censorship are two different pro-
cesses. SARFT executives can make us censor as much as they want during
the script-approval process, but that doesn’t mean we have to shoot their
version of the script. The film is going to be censored again anyway, at the
distribution-approval stage. So, I still shoot the things they asked me to
remove from the script.
Contravening SARFT’s ordinance, Chan filmed the banned scenes of

unflinching violence, and downplayed rather than expunged the negative
portrayal of the Tanguts: ‘This element is still present in the finished film,
but it had to be sort of hidden, because China is very nervous about
offending different ethnic groups.’ Typically, a different set of censors
than those at the preproduction stage assess the completed film, and Chan
finds that ‘things censored at the script-approval stage [often] pass the
censors at the postproduction censorship stage’—largely, he says, because
the final assembly makes cogent the banned element’s narrative logic and
affective force.8 Such tactics discredit the notion of post-Hong Kong cin-
ema, which presupposes that local filmmakers capitulate to PRC-imposed
ideology. Rather, directors such as Chan navigate the coproduction system
124 G. BETTINSON
through their strategies of plotting and shooting, finding tactics of resis-

tance against Mainland cultural imperialism.
The coproduction system enforces unfamiliar practices, but Chan relied
on tried-and-proven Hong Kong methods during shooting. Scripted
scenes were revised each day throughout filming. Even combat scenes,
containing minimal dialogue, would be reconceived to accommodate on-
set invention, as when Donnie Yen (the film’s star and action director)
improvised stylized fight choreography between takes. This experimental,
trial-and-error mode of practice encouraged Chan to probe the possibili-
ties of the script’s narrational complexity, and ultimately accounts for Wu
Xia’s puzzle-centered dramaturgy. While the SARFT-approved script pro-
vided Chan a stable architecture, the exploratory shooting method fos-
tered collaborative interplay and narrative experimentation.
What narrational strategies have Chan’s working methods engen-
dered? A brief plot synopsis of Wu Xia will be useful here. Set in the early
Republican era, Wu Xia centers on Jinxi (Donnie Yen), a peaceful villager
suspected of murdering two bandits while working at the local paper mill.
The opening plot phase depicts the bandits’ attempt to ransack the mill,
Jinxi’s brave attempt to ward them off, and the apparent stroke of blind
luck that enables him to defeat them. This theft situation will be replayed
later in the plot but in ways that contradict our initial understanding. In
the first rendition of the crime, the narration presents the violent skir-
mish as objective truth; nothing indicates that the narrated action may be
unreliable. But subsequent replays ambiguate the initial event by shifting
among different characters’ perspectives, somewhat akin to Kurosawa’s
Rashomon (1950).9
Investigating the bandits’ death, Detective Xu (Takeshi Kaneshiro)
attempts to reconstruct the crime, and thereby initiates the narration’s
first replay of the incident. Far from providing a straight iteration of the
robbery, Detective Xu’s version recasts fundamental details of the event.
In Xu’s replay, Jinxi kills the bandits not accidentally, but intentionally
and surreptitiously, by means of furtive kung fu maneuvers. Since Xu did
not witness the crime firsthand, the viewer grasps this subjective replay
as hypothetical, hence corrigible: Xu’s suppositions may be flawed or
even wholly erroneous. Though we do not yet know it, the narration’s
misdirection rests on our default tendency to accept seemingly objective
action as trustworthy. Deviously, the narration’s apparently objective pre-
sentation of the crime—which does not initially arouse our skepticism—is
to be proven false, while the subjectively motivated replay—which we are
invited to treat skeptically—will eventually be validated. For a substantial

portion of the plot, Wu Xia equivocates on these incompatible scenarios.
Is Detective Xu’s counterfactual hypothesis valid? Or did the crime occur
as Jinxi attests and as initially presented by the narration? More broadly,
is Jinxi the reluctant hero he at first appears to be, or does he conceal
another identity behind a benevolent charade? It is this central ambiguity,
prolonged and nurtured by an equivocating and untrustworthy narration,
that sustains narrative interest for much of the film’s plot.
Triggering our distrust, Wu Xia throws doubt on Detective Xu’s cred-
ibility. Traditionally in detective fiction, we presuppose on the part of the
detective sagacity and sharp deductive reasoning. But Xu is quickly estab-
lished as an eccentric and possibly schizophrenic figure. This psychological
eccentricity motivates a peculiar quirk of the narration. In Xu’s tentative
replay of the robbery, he imaginatively inserts himself amid the ferment;
thus the flashback incorporates shots of Xu scrutinizing the crime as an
invisible observer (Fig. 7.1). This narrational gambit is to be grasped as a
symptom of Xu’s split personality. Across the film, Xu’s doppelgänger—his
‘rational’ self—frequently manifests at his side, coaxing him to forgo emo-
tion in favor of brute logic. In short, the narrative presses us to construe
the misanthropic detective as fallible. Peter Chan’s narration here cues a
crude but compelling hypothesis: Xu may be outlandishly embellishing
Fig. 7.1 Wu Xia: Detective Xu (Takeshi Kaneshiro, left) imaginatively ‘wit-

nesses’ the paper-mill skirmish
126 G. BETTINSON
Jinxi’s account of the robbery, perhaps as a result of an incoherent psyche

and a zealous devotion to the law.
As it puts Xu’s deductions in doubt, the narration seeks to promote
Jinxi’s account—that is, the first version of the robbery presented by the
narration. Crucial here is what Meir Sternberg calls the primacy effect, the
set of pregnant first impressions generated by the text (Sternberg 1978,
p. 94). In the opening reel, the narration situates Jinxi at the heart of
tranquil domesticity. Only in retrospect does the viewer grasp these scenes
as structural misdirection. By introducing Jinxi as a familial and benign
figure, the narration makes Detective Xu’s allegations against him hard
to countenance; the potency of the primacy effect prevents us from easily
reconciling our first impression of Jinxi with the traits of a savage killer.
As a function of the narration, the primacy effect in Wu Xia plants false
impressions from the start, and continues to shape our hypotheses even
after Detective Xu furnishes his counterfactual version of the robbery.
Also at play, however, is the competent viewer’s awareness of Donnie
Yen’s martial-arts persona. In this context, the primacy effect acquires spe-
cial emphasis, the opening scenes crucial in subduing Yen’s star image. As
Peter Chan observes:
Donnie Yen has a contract with the audience. The audience knows he always
plays an action hero, but we had to make him seem like an ordinary peasant.
So it’s a little bit harder to play that game of deception with the audience,
because they have a preconception about who he is and the roles he usually
plays.
Even a viewer not acquainted with Yen’s action persona may surmise, on
the basis of genre conventions, that Yen’s protagonist will eventually be
unmasked as a martial-arts maven. On the other hand, viewers are accus-
tomed to ‘star vehicles’ designed to radically subvert a pre-established star
image; indeed, Yen himself has since defied and travestied his action-hero
persona in local New Year comedies such as An Inspector Calls (2015),
All’s Well Ends Well 2012 (2012), and All’s Well Ends Well 2011 (2011).
Notwithstanding the varying levels of extrafilmic competency among the
audience, the film ultimately presupposes either a spectator unversed in
Yen’s previous work, or one willing to suspend prior knowledge of Yen’s
persona and play along with Chan’s ‘game of deception.’ In either case,
Wu Xia’s distribution of knowledge is engineered to conjure and sustain
ambiguity around Yen’s placid villager.
As corollary, we cannot wholly discount Detective Xu’s theory. Wu

Xia’s narration arouses a hesitation between the incompatible premises
posited by Xu and Jinxi. It does so partly by toggling between their respec-
tive points of view. As noted, certain flashbacks, or replays, are subjectively
anchored to Detective Xu. Certain other flashbacks are focalized around
Jinxi. Cunningly, some of Jinxi’s flashbacks are reliable while others are
not, and the narration doesn’t overtly distinguish the trustworthy flash-
back from the false one. (The most salient instance of this is Jinxi’s spe-
cious recounting of the robbery.) Further, whereas the narration renders
Detective Xu’s subjective thoughts transparent, it strategically withholds
subjective access to Jinxi. Thus, Jinxi remains subjectively opaque, at least
until the narration discloses his secret past; Jinxi’s memories of his own
martial adventures and barbaric history are withheld from us by a selec-
tively restricted narration, which thereby preserves the central enigma of
his identity.10
Following the robbery scene and its variant replays, Wu Xia again tele-
graphs the potential duplicity of its narration. Intending to expose Jinxi’s
true identity, Xu throws him off a bridge; Jinxi is suspended perilously on
the spindly branches of a tree before plummeting to the river below (Fig.
7.2). Incredibly, Jinxi survives the fall unscathed. As in the initial robbery
scene, the action here is presented as objectively true. And once more,
Detective Xu challenges our comprehension of the action by advancing
a counterfactual scenario, depicted by the visual narration, in which Jinxi
breaks his own fall by summoning internal chi energy. Here again the
Fig. 7.2 Spindly branches—or perhaps internal chi energy—rescue Jinxi (Donnie
Yen) from certain death in Wu Xia
128 G. BETTINSON
more tentative account, furnished by Detective Xu, will eventually be

validated, and the initial ‘objective’ version discredited, exposing the nar-
ration’s deceptiveness. When the narration finally disambiguates Jinxi’s
identity—disclosing his hidden past as a master swordsman—the viewer is
obliged to recall and reinterpret the prior action. Now the first version of
the paper-mill theft, and of Jinxi’s plight on the bridge, is to be construed
as false. Conversely, Detective Xu’s outré replays are validated, affirming
his genius as a detective. By means of this gap-filling strategy, Wu Xia’s
narration resolves our mutually exclusive hypotheses, and satisfies our
long-aroused curiosity as to Jinxi’s true identity.
Once the plot disambiguates Jinxi’s identity, the film’s final act swerves
onto familiar swordplay territory, and the narration becomes less auda-
cious. Until that fairly late plot stage, the narration flaunts the strata-
gems of puzzle cinema: it self-consciously disarrays the viewer’s epistemic
knowledge (is Jinxi an imposter?), ontological knowledge (does the nar-
ration present actions that aren’t real?), and real-world knowledge (can
our default inferences be trusted?). It also triggers curiosity hypotheses
(perhaps the robbery incident did not occur as we initially thought) and
suspense hypotheses (the narration may eventually reveal Jinxi to be an
impostor), keeping them simultaneously in play for much of the film. But
why does Wu Xia temper its narrational duplicity in the plot’s latter sec-
tion? ‘The audience is waiting for Donnie Yen to be “Donnie Yen” the
action superstar, and display his fighting ability,’ Chan states. ‘I felt that
we’d gone as far as we could at that point in the story without testing the
audience’s patience any longer.’ Here, then, another kind of knowledge
structure—in this case, intertextual—mediates and molds Chan’s narra-
tional construction. As already noted, the competent viewer’s knowledge
of Yen’s persona shapes her or his hypothesis-formation, fueling the prem-
ise that Jinxi is secretly a martial-arts adept. In other words, this fund
of intertextual knowledge generates expectations concerning the present
film’s narrative development. Just as important, these intertextual expecta-
tions were anticipated by Chan during production and crucially influenced
his suspense-building strategies; hence he opts to reveal Jinxi’s true iden-
tity before the film’s final act rather than, say, at the climax.11 The audi-
ence’s intertextual expectations, therefore, constitute a kind of external
pressure impinging on the film’s aesthetic construction.12 Once the enigma
of Jinxi’s identity is resolved, the raison d’être for narrational complexity
is dissipated. Thereafter, Wu Xia can stage a narrationally straightforward
denouement, showcasing Yen’s fighting prowess and thereby fulfilling

audience desire.
The kind of intricate narration on display in Wu Xia is not sui generis
in Hong Kong cinema. A post-CEPA surge in cross-border coproduc-
tions may have sharpened local filmmakers’ attention to script matters,
but Wu Xia’s complex storytelling cannot be wholly attributed to joint
ventures and SARFT script demands. Other influential factors are at
play. For instance, Peter Chan had already begun probing narrative form
(albeit it in a coproduction context) in films such as Perhaps Love (2005),
whose labyrinthine plot boldly reworks the backstage-musical genre.13 Wu
Xia therefore extends Chan’s authorial engagement with esoteric plot-
ting. Then there is the international puzzle-film trend, which gathered
momentum in the USA subsequent to The Sixth Sense and Fight Club
(both 1999), and spread to South Korea (Il Mare [2000]; Oldboy [2003]),
Mainland China (Suzhou River [2000]; Hero [2002]; Mystery [2012]),
and other East Asian cinemas. Ever alert to transnational market trends,
Chan assimilated Wu Xia to this pan-Asian and global vogue for complex
storytelling. At the same time, however, he dovetailed the film with a dis-
tinctively local tendency for experimental storytelling.
Most influential upon Wu Xia’s narration is a heritage of locally pro-
duced martial-arts mystery and detective films, notably from the Shaw
Brothers studio. The late-1970s spawned pertinent titles directed by Chor
Yuen (Murder Plot [1979]; Heroes Shed No Tears [1980]), Sun Chun (The
Avenging Eagle [1978]), and Zhang Che (The Brave Archer [1977]; Life
Gamble [1979]). Typically adapted from Chinese literary sources, these
films anticipate contemporary puzzle cinema—most saliently, they flaunt
the narrational restrictedness, elaborate plot imbroglios, abrupt volte-face,
and surprise revelations ingredient to the genre. By the end of the 1970s,
the fashion for dense and duplicitous plotting was a discriminable trend
within Hong Kong cinema (if not quite the dominant narrative mode),
even permeating the local independent sector—examples include Flaming
Swords (aka Strife for Mastery [1977]) and Blooded Treasury Fight (1979).
Wu Xia harks back to this flurry of martial-arts mystery films (with which
Chan had grown conversant in his youth14). Here, then, is one source
of the film’s localism: Wu Xia revives a strong tradition of narration-
ally ambitious Cantonese cinema. At the same time, the film invokes the
Mandarin-language swordplay movies produced by Shaw Brothers in the
1960s and early 1970s, chiefly by means of allusion (evoking Zhang Che’s
The New One-Armed Swordsman) and intertextual casting (Shaw veterans
130 G. BETTINSON
Jimmy Wang Yu and Kara Hui),15 reminding us that the Hong Kong and
Mainland film industries had been commercially interlinked long before
1997.
Wu Xia’s complex narration, then, is not explicable solely by refer-
ence to the post-CEPA era of preproduction scripting. Nor does the film’s
complexity attest to the Sinofication of Hong Kong cinema, for complex
narration is not a primary or exclusive trait of Mainland movies. Rather,
Hong Kong’s popular cinema has always intermittently explored the archi-
tectural and cognitive appeals of ludic narration and intricate plotting.
For all its superficial symptoms of Mainlandization—its official coproduc-
tion status, its historical Mainland setting, its Mandarin-language sound
track—Wu Xia is a film firmly rooted in Cantonese filmmaking tradition.
The example of Chan and Wu Xia is not an isolated case, and illustrates
that, despite claims to the contrary, the China coproduction system has
not nullified Hong Kong’s long-standing tradition of piecemeal story con-
struction. It is perhaps unsurprising that predominantly local filmmakers,
such as Ann Hui and Pang Ho-cheung, cleave to Hong Kong production
practices during their occasional excursions into China coproduction ter-
ritory (Hui’s The Golden Era [2014]; Pang’s Love in the Buff [2012]). But
even Peter Chan, ostensibly the most Mainlandized of Hong Kong’s film-
makers, adapts local work routines to PRC production constraints.
Localism also survives and flourishes in the films of Johnnie To,
Wai Ka-fai, and the Milkyway Image studio. Here again I take ‘Hong
Kong localism’ to encompass narrationally demanding storytelling.16
Traditionally, Hong Kong’s commercial cinema has been identified with a
signature set of storytelling norms: episodic plotting; frequent attractions
(a chase, a gag, a fierce skirmish); cohesion devices such as motifs, par-
allelism, chance, and coincidence; brazenly sentimental situations; tonal
incoherence; a disregard for character change; and unpredictable endings
(Bordwell 2000, pp. 178–198). This set of narrative features is, I think,
partly what critics mean when they speak of a distinctive ‘local flavor’ char-
acterizing Hong Kong films; and it is this traditional Hong Kong aesthetic
that is at stake in the post-handover era of China–Hong Kong coproduc-
tions. Critics, however, have counterpointed this aesthetic to the kind of
canonical, ‘well-told’ story construction sanctioned by SARFT. From this
angle, Hong Kong cinema’s traditional narrative principles—episodicity,
tonal ruptures, an absence of character arcs, and so forth—run counter to
‘quality’ Hollywood-style storytelling. By extension, the local filmmaker’s
traditional work practices are perceived as deficient, for they engender a
less than robust narrative architecture. Thus these practices and aesthetic
qualities are hardly apt to yield storytelling techniques of any sophistica-
tion or complexity. Yet, as I aim to demonstrate in the rest of this chapter,
the work routines employed by To, Wai, and their Milkyway colleagues
not only preserve local practical and storytelling norms, they also foster
narrational innovation and intricacy.
SCREENWRITING PRACTICES AT MILKYWAY IMAGE

In studio-era Hollywood, the screenwriter might not meet the direc-
tor, much less be invited to the set. Even today Hollywood screenwriters
often are not recruited beyond the preproduction phase. Much the same
obtains with PRC coproductions: the writer’s function typically ends upon
SARFT approval of the preproduction screenplay. At Milkyway, however,
the role of the screenwriter extends from preproduction to postproduc-
tion, even encompassing prerelease marketing and promotion. (Members
of the writing team frequently compose promotional synopses and mar-
keting taglines.) Moreover, writers are sometimes engaged on two films
at once: thanks to conflicting shooting schedules and cast availability, the
scriptwriters fulfill postproduction duties while at work on the next film’s
preproduction. The role of the Milkyway screenwriter, then, is both all-
encompassing and integral to every phase of production.
Some Milkyway scripts are signed by a single screenwriter, but
many of the studio’s films are accredited to a small ensemble of writ-
ers. Occasionally a script is attributed to ‘The Milkyway Creative Team,’
which consists of an interchangeable cadre of writers. By the late 1990s,
Johnnie To and Wai Ka-fai had established a production system adopted
from Television Broadcast Limited (TVB)17: script duties would be man-
aged by a supervisor, with a second tier of house writers (including Yau
Nai-hoi, Au Kin-yee, and Yip Tin-shing) assigned to particular projects.
Presiding over story development is Wai Ka-fai—the ‘gatekeeper of the
script,’ as one staff writer puts it. Despite the studio’s hierarchal staff
structure, an openly collaborative ethos typifies the work environment.
Division of labor is loosely orchestrated, while the kind of specialization
fostered in Hollywood—where, for example, a writer skilled in charac-
ter dialogue may be enlisted to furnish punchy one-liners—is forfeited
in favor of a collective, free-flowing interplay of ideas. From this method
spring certain advantages. Not least, the close-knit collaborators function
as a mutual sounding board, sharpening story details before submitting
132 G. BETTINSON
the plot scenario for studio and investor approval. An added efficiency
stems from the familiarity of colleagues reunited across successive projects.
How is the Milkyway screenplay developed? Two parameters typically
guide story construction at the outset: a prearranged theatrical release
date; and an a priori cast to whom principal roles must be tailored. The
release deadline is usually ordained by the film’s financiers, and broadly cir-
cumscribes the kind of narrative and milieu that can be feasibly designed,
manufactured, and filmed. (Hence, granted only a 22-day schedule, Wai
conceived Fantasia [2004] as a midrange comedy rather than as, say, an
effects-driven space opera.) Just as financiers determine a deadline, they
assert a degree of control over casting. Media Asia, for instance, invested
in Life Without Principle (2011) on condition that director To cast its cli-
ent, singer Denise Ho, in a primary role. In such cases, the task of ‘star
development’ befalls Milkyway’s writing staff. (Peter Chan pursues a simi-
lar task by casting Donnie Yen in Wu Xia, in this case refreshing an extant
film-star persona.) Given these prerequisites, the writers’ story construc-
tion is, as Milkyway producer Shan Ding puts it, ‘made to order’—the
narrative is designed to meet certain bespoke specifications.18 But Wai and
his team approach these constraints not so much as creative hindrances as
liberating mediations forcing them to innovate.
With a handful of actors and a deadline in place, the creative team
weighs up potential subject matter. Typically, though not always, the ini-
tial story concept is conceived by Wai. Sensitive to market demands, he
seeks to capitalize on recent box-office hits; at the same time, he fixates
on ways to invigorate generic formula, while also contriving scenarios
intended to deepen the personas of recurring stars (such as Andy Lau and
Lau Ching-wan). Once the film’s premise is agreed, Wai dispatches his
staff to research the basic subject matter, the results of which are woven
into a two-page plot treatment. Wai’s team then submits this treatment
to Johnnie To, upon whose approval the project rests. If To sanctions the
plot synopsis—and sometimes he is apt to demand revisions—the writing
staff constructs a scene breakdown, mapping out the macrostructure of
the plot. (Traditionally, Hong Kong scenarists disregarded the structural
felicities of ‘acts’ and ‘turning points’ [Bordwell 2000, p. 122], but the
Milkyway writers—in a rare departure from local custom—instinctively
parse their plot outlines into three distinct acts.) Devoid of dialogue, the
skeletal scene breakdown specifies each scene’s location, characters, and
fundamental events. It is subsequently distributed to the studio’s produc-
tion departments (costume and wardrobe, art design, practical stunts, and
so forth), much as the shooting script in studio-era Hollywood served as

the basis of preproduction planning. At Milkyway, however—even in the
post-CEPA era—no screenplay exists prior to shooting. Indeed, the term
‘shooting script’ is a misnomer; only at the end of postproduction does a
script fully coalesce. In lieu of a preproduction screenplay, the schematic
scene breakdown is an important document for Milkyway’s production
departments, but it serves only to limn the general story action rather than
to operate as an airtight plot outline or an ironclad blueprint for shooting.
These practices, de rigueur in Hong Kong cinema, proved controversial
when Milkyway embarked on China coproductions in the early 2000s. To
the writing staff, SARFT’s policy of a full-blown preproduction script was
anathema, and circumventing it required the guile of the studio’s produc-
ers. One solution was to submit bogus screenplays to SARFT. Freelance
writers were commissioned—for a fee of HK$5000—to produce fake
shooting scripts satisfying SARFT criteria.19 Meanwhile, the studio’s house
writers secretly prepared the de facto synopsis and scene breakdown, as
per studio custom. Milkyway producers abandoned this ploy once SARFT
executives grew wise to their subterfuge.20
For Milkyway films that are not subject to SARFT’s script regimen (i.e.
those that are not mounted as China coproductions), Wai’s staff writers
compose scenes day by day as filming progresses. As prompted by the
shooting schedule, the writers produce a formal script page complete with
character dialogue, but only for the scene or scenes being shot that day.
Typically, the script page is not finalized (and therefore not distributed to
the director, leading actors, and crew) until the day the scene is shot.21 (It
is not the case that Wai withholds plot details from the actors for strategic
effect, as is customary practice for Woody Allen or Ken Loach; rather, a
fully fleshed-out plotline is simply not available during shooting.22) For the
films that he directs, Johnnie To adopts a steady routine. During filming,
one of Wai’s writers remains on set with To, supplying last-minute scene
revisions at To’s behest; concurrently, Wai and his colleagues compose
the script page for the next day’s shoot. At the end of each shooting day,
the writing team reviews the dailies, discusses To’s script notes, and fine-
tunes the next scene to be shot. From day to day, the Milkyway process of
screenplay composition is perpetually piecemeal, on-the-fly, and mercurial.
These facile working methods enable the writers to shape characteriza-
tion and performance. Their presence on set during shooting, as writer
Au Kin-yee affirms, enables them ‘to observe the actors giving their per-
formance, and allows us to feed the actors material that we think they can
134 G. BETTINSON
deliver well, or that will surprise the audience.’ For instance, the female
protagonist of My Left Eye Sees Ghosts (2002) had at first been conceived
by Wai as a ‘tough lady’; but, Au recalls, ‘after shooting for a few days we
realized it’s hard for [actor] Sammi Cheng to play tough; she’s better play-
ing someone who needs help from others, so we changed it and reshot the
scenes.’ During the filming of Needing You… (2000), the writers modi-
fied Cheng’s heroine to better exploit the actor’s naturally ‘eccentric’ fig-
ure movements; similarly, Blind Detective yokes its narrative situations to
Cheng’s elastic physicality. If Hollywood screenwriters are seldom per-
mitted on set, the Milkyway writers’ proximity to the shooting process
crystallizes star personas, characterization, and story action. Likewise the
ritual of reviewing dailies: ‘When we watch the dailies every day,’ says Au,
‘It not only helps us to observe how well the actor is delivering the mate-
rial, but also gives us inspiration when creating the scenes to be shot later
in the schedule.’
Postproduction editing offers a final opportunity to alter story particu-
lars and dialogue. Though Hong Kong filmmakers adopted direct sound
recording in the late 1990s, Milkyway filmmakers still favor postsynchro-
nization23; consequently, dialogue revisions can be implemented late into
the dubbing phase, effecting a local custom of ‘postproduction plotting’
(Bettinson 2014). The postproduction phase also generates foreign-
language tracks—and alternative cuts of the film—for overseas markets,
and here too Wai’s staff is on hand to rewrite dialogue and redistribute
scenes. It is only at the end of the postsynching process that something
resembling ‘the screenplay’—until now a piecemeal aggregation of scene-
length pages—can plausibly be compiled. In numerous ways, then, screen-
play composition at Milkyway Image differs sharply from Hollywood
practice. If the main role of the Hollywood screenwriter is fulfilled prior
to shooting, the Milkyway scriptwriter’s contribution far exceeds the pre-
paratory phase, and it consists of a relatively variegated array of tasks.
Mad Detective, signed by both To and Wai, encapsulates these local
practices. Conceived as a commemorative film to mark a decade of
Milkyway output, Mad Detective found its genesis in characteristic points
of departure: an attempt to evoke a former success (in this case, To and
Wai’s Running on Karma [2003]); a desire for artistic novelty (the
paradoxical effort to depart from the ‘Milkyway style,’ which, according
to producer Shan Ding, had by 2007 become ‘clichéd’); and a preliminary
cast of established and prospective stars, some of whom were Milkyway
regulars (Lau Ching-wan), and others that were imposed on the project
as conditions of finance (newcomer Andy On, a client of investor China

Star). The impulse to innovate gained impetus from the lead actor’s per-
sona: ‘Lau Ching-wan had already made so many movies with Milkyway,’
notes Au. ‘For us writers, it was a question of, How do we make sure that
he doesn’t repeat his previous characters? And so we set out to create a
very different kind of protagonist for him.’ As with Running on Karma,
the Milkyway scenarists created an eccentric cop protagonist gifted with
‘special abilities,’ including an acute capacity for empathy and an ability to
reenact crime situations with startling accuracy.
These psychological traits, in turn, motivated some elliptical and com-
plicated narrational tactics. Indeed, Wai’s initial two-page synopsis—which
thematized schizophrenia, and posited Lau’s role as but one manifestation
of several distinct identities—was deemed too byzantine for mass con-
sumption. ‘The film’s original working title was Who is Who?’ states Ding.
‘And that summed up the problem: Johnnie To couldn’t follow the story,
none of us could figure out who was who; the scenario was so confusing.’
When the film’s distributor balked at Wai’s knotty synopsis, To ordered a
rewrite. An amended draft evinced greater clarity, but it remained deliri-
ously offbeat; hence To and his investors reconceived Mad Detective as a
niche ‘festival movie,’ effectively ceding the Mainland and international
markets.24 Lacking a preproduction script, Wai’s creative team sculpted
the overall plot throughout shooting and editing, constantly monitoring
its intelligibility. If the dailies proved confusing, the problematic scene
would be rewritten for greater legibility, and reshot later in production.
Finally, Au Kin-yee attests, Mad Detective’s theatrical cut differed sharply
not only from Wai’s initial plot synopsis but also—thanks to a drastic reor-
dering of scenes during the assembly phase—from the story manifested at
the end of filming.25
As an indigenous production, Mad Detective exemplifies local filmmak-
ing practices, including that of postproduction plotting. It also embodies
the aesthetic norms of pre-handover Hong Kong cinema (tonal detours,
visceral set pieces, quirky heroes, gleeful vulgarity), norms that are alleg-
edly effaced in Hong Kong–China coproductions. At the same time, the
case of Mad Detective reminds us that local filmmaking is not exempt from
production pressures. Domestic ventures hardly guarantee unbridled
artistic freedom; financiers, distributors, and market tastes impinge on the
story-construction process just as surely as SARFT censors curb creativity
on the coproduction model.26 Other familiar forces beset purely local pro-
ductions too—the industrial pressure to balance recondite plotting with
136 G. BETTINSON
comprehensibility, for instance, and the obligation to juggle novelty with

familiar norms and schemas. As Au Kin-yee maintains:
We were reminded from the start [of Mad Detective] about market demand.
Everybody expects a Johnnie To film to be a genre police action movie, and
so, for distribution reasons – no matter how crazy we made our detective
protagonist – we had to try to keep Mad Detective within the context of a
typical police crime story. (Bettinson 2016: 43)
If domestic Milkyway product such as Mad Detective epitomizes a dis-

tinctive Hong Kong flavor—and does so, incidentally, while maintaining
a high level of narrative sophistication—what becomes of this local flavor
when the Milkyway studio embarks on China coproductions? Here again,
the Hong Kong filmmaker does not automatically forfeit local storytelling
principles. Nor is narrational complexity sacrificed on the altar of SARFT’s
straitened censorship. The following analysis of To’s Blind Detective—a
Hong Kong–PRC joint venture—aims to demonstrate that, contrary to
standard belief, some Hong Kong filmmakers have found it possible to
assimilate local norms and practices to the China coproduction system.
Moreover, they have done so without compromising narrative experimen-
tation and ingenuity.
A COMPLEX CASE: BLIND DETECTIVE (JOHNNIE TO,

2013)
Blind Detective’s eponymous protagonist—Johnston, played by Andy
Lau—possesses the psychological ability to solve crimes through role-play,
reenactment, and leaps of imagination, aided by a strong empathic iden-
tification with the criminal and victim. As in Wu Xia and Mad Detective,
the crime-solver’s psychological idiosyncrasies justify the narration’s play-
fulness. Other similarities link the three films. For instance, the detective
protagonist mentally infiltrates the crime situation—in Blind Detective, the
historical murders of several schoolgirls—and, paradoxically, the narration
depicts the sleuth retroactively investigating the crime as it unfolds.
Also like Wu Xia, Blind Detective furnishes reenactments marked as
provisional. Each successive reenactment of the principal murder revises
those that precede it, and signals a new phase in the detective’s investiga-
tion. The effect is akin to what David Bordwell calls ‘multiple-draft nar-
ratives,’ whereby the narration parcels out variants on the initial situation
(in this instance, a murder), with each new version presented as a more
plausible, more complete revision of the last (Bordwell 2008, p. 184).
As Johnston modifies his deductions, the narration correspondingly sup-
plies counterfactual replays (or drafts) of the murder incident. That these
replays are subjectively motivated means that the viewer’s grasp of the
crime coalesces only in stages, coterminous with the detective’s investiga-
tive progress.
In this respect, Blind Detective echoes Wu Xia, but To’s film adopts
its own narrational stratagems. Unlike Wu Xia, Blind Detective furnishes
no preliminary, apparently objective rendition of the crime against which
to contrast the ensuing counterfactuals. Thus, the film does not hood-
wink the viewer as Wu Xia does; from the start, we are aware that each
successive replay is tentative and likely to be modified at a later juncture.
Moreover, whereas Wu Xia passes off its initial depiction of the paper-mill
robbery as objective truth, Blind Detective is far less deceptive, explicitly
identifying subjective replays as subjective by assigning them a discrete
visual tone. Each replay is drenched in blue-tinted hues and saturated
top lighting, marking a stylistic departure from the film’s objective action
(Fig. 7.3). Prima facie, then, the film’s narration does nothing to arouse
our skepticism. Unlike Wu Xia, it signals subjective action instantly and
overtly; and it establishes from the outset that the replays are not definitive
accounts of the crime, but rather ongoing stages in an investigation, hence
subject to revision.
Fig. 7.3 Blind Detective: a harsh lighting scheme and distinctive color palette
denote subjective action as Johnston (Andy Lau) investigates a crime
138 G. BETTINSON
Having established its offbeat premise, Blind Detective elaborates its

replay gambit in challenging ways. Subsequently, Johnston not only envi-
sions himself amid the past action (as does Detective Xu in Wu Xia); he also
interacts with the key suspects and victims, gathering clues that advance
his investigation. Granted, this constitutes an oblique form of causal moti-
vation, but To motivates it thoroughly by the protagonist’s bizarre psy-
chology.27 As already noted, the initial replays are subjectively anchored
to Johnston, motivated as symptoms of his intuitive brilliance. But as the
plot unfolds, Johnston trains his sidekick, Goldie (Sammi Cheng), in the
ways of imaginative reasoning, and soon the narration furnishes replays
focalized around her subjective imagination. Here again the narration pos-
its a revised draft of the murder situation—once more, explicitly marked
as subjective action by an oversaturated, blue-toned visual palette—but,
as with previous replays, the reliability of this hypothetical reenactment
remains moot. By now Blind Detective has established its intrinsic norm
of subjecting the central crime situation to frequent revision, and each
successive replay leaves unresolved gaps in the mystery. Consequently,
the viewer expects subsequent replays to refine, redress, or repudiate the
preceding ones, and awaits a definitive reenactment that will conclusively
disambiguate the crime.
In detective fiction, of course, it is common for the narration to provide
partial flashbacks to the criminal act. At the climax, the traditional detec-
tive plot will reveal the missing information, ‘filling in’ the flashback and
resolving the enigma. Sometimes the narration hasn’t signaled that the
initial flashback is incomplete, thus the late-arriving exposition generates
surprise. Blind Detective does not furnish flashbacks exactly, but rather
provisional and possible replays of prior events (for instance, the multiple
drafts of the murder) and speculative extrapolations of the murder situ-
ation (as when Johnston imaginatively simulates an interrogation of the
prime suspect). While the possible replays often get invalidated, the imagi-
native extrapolations obliquely nudge the detective closer to the enigma’s
solution. As in Wu Xia, the sleuth’s methods elicit the spectator’s doubt.
Does Johnston truly possess outlandish intuitive gifts, as he believes?
Unlike classical detectives, Johnston repeatedly errs in his guesswork. (‘I
guess I’m wrong again,’ he intones near the film’s end.) As the narration
continually discredits Johnston’s hypothesized replays, it throws his pro-
ficiency as a detective and his reliability as a focalizing agent into doubt.
More broadly, the viewer’s skepticism toward Johnston and Goldie fosters
narrative unpredictability—a formal convention of Hong Kong cinema
that, according to the Mainlandization thesis sketched earlier, is imperiled

by the PRC coproduction model.
The climax marks the apogee of the film’s narrational ploys. Johnston
deduces the identity of a missing schoolgirl once known as Minnie. Now
an adult and nine-months pregnant, Minnie has spent the last few years
in hiding, assuming an alias and altering her physical appearance by cos-
metic surgery. Her husband, a restaurant chef, has earlier been intro-
duced as Johnston’s chief murder suspect, and the plot has also detailed
the husband’s habitual adultery. At the climax, Johnston visits the res-
taurant and confronts Minnie about her furtive past. Suddenly disturbed
by an unsettling sensation at his feet, the sightless detective intuits that
he is standing in a stream of blood. This inference is seemingly corrobo-
rated by the visual narration, the camera revealing the butchered corpses
of Minnie’s husband and his lover nearby. As Johnston attempts to flee,
Minnie exclaims: ‘My waters have broken – help!’ At this point, the viewer
(along with the blind detective) must redress the initial impression that
Johnston has stumbled upon a fresh murder scene. Indeed, the narra-
tion now encourages the inference that it is amnionic fluid, not blood,
that Johnston detects underfoot, and moreover, that the image of dead
lovers—neither of whom is now in view—was false. After all, the film’s
narration has set forth unreliable action before, and much of the viewer’s
activity has been spent amending false inferences. The inconclusive action
here comes forward as one more narrational feint, one that augments the
possibility that Johnston is an untrustworthy focalizing agent and hence a
fallible, inept crime-solver.
The narration, however, abruptly delivers another deus ex machina.
As Johnston tends to the pregnant Minnie, he suddenly recoils in pain:
Minnie thrusts a butcher knife into his stomach, stabbing him frenziedly.
What, we are led to ask, is the reality status of the action now? Is Johnston
truly in mortal danger, or is this yet another of the narration’s specious
images? Is Minnie on the cusp of childbirth? Do the corpses of the adulter-
ous lovers languish nearby, if indeed they really are dead? As the spectator
negotiates this blinding array of hypotheses, Johnnie To sustains narrative
ambiguity by refusing to cut back to the spot previously occupied by the
lovers’ corpses. By denying the spectator a clear purchase on the dramatic
space, To maintains uncertainty about the action’s reality status. Only
when Goldie arrives at the restaurant and tackles Minnie does the visual
narration allow us to survey the action in toto. At last the actual state
of affairs is clarified: Johnston is severely wounded by Minnie’s assault;
140 G. BETTINSON
the lovers are dead; and Minnie is giving birth. (Thus the scene confirms
two antithetical hypotheses: the floor is soaked with blood and amnionic
fluid.) The narration’s sudden omniscience elicits what cognitivists call
spectatorial ‘insight’ (Berliner 2013, p. 201), that pleasurable ‘eureka’
moment in which the viewer gains mastery of the narrative situation.
In this brief climactic action, To’s narration ambiguates the drama
and deftly throws the viewer’s hypotheses into rapid flux. Our oscillat-
ing hypotheses are triggered by several tactics. First, Johnston’s physical
defect—his lack of sight—fosters not only his own bewilderment but also
the viewer’s uncertainty, since throughout Blind Detective the narration
closely aligns us with Johnston’s inferences. Second, the visual narration’s
strategic restrictedness withholds images that would permit the viewer a
confident grasp of the scene’s actual events. Finally, the foregoing story
events indelibly shape the viewer’s equivocal uptake at the climax; by this
late plot stage, we have thoroughly internalized the film’s pattern of sup-
plying unreliable action. Consequently, we entertain the prospect that
aspects of the climax may be untrustworthy and subject to correction.
The narration has utterly discombobulated us, but it has done so with-
out violating the detective convention of fair play. The initial shot of the
dead couple—which for a short interval we reinterpret as narrative mis-
direction—is visually consistent with the other shots in the scene. That
is, the shot does not display the blue-tinctured, harshly-lit visual scheme
afforded all the scenes ‘discredited’ as subjective or false. As such, the
spectator ought not to have doubted the veracity of the climactic images
(including the initial images of the blood-spattered floor and the lovers’
flaccid bodies). In sum, the climactic narration ingeniously manages to
misdirect us, send our hypotheses into a tailspin, and sustain the stylistic
intrinsic norm established near the film’s start (viz., presenting unreliable
action in a discrete visual register). If the climax leaves us reeling, To fur-
nishes a brief coda that strives to restore equilibrium: several years on from
the bloodbath, Johnston, Goldie, and Minnie’s orphaned child find stabil-
ity as a boisterous yet loving makeshift family.
RECONSIDERATIONS
What conclusions may be drawn from the foregoing analyses? Certainly
our major case studies testify to Hong Kong cinema’s artistic innova-
tions, but they also compel us to reconsider some deep-rooted critical
assumptions. Chief among these is the alleged demise of Hong Kong’s
traditional filmmaking practices. As we have seen, local filmmakers have

not so much acquiesced to SARFT mandates as assimilated long-standing
methods to new production circumstances. Granted, Chan and To—as
commercially successful, senior industry figures—may wield a degree of
bargaining power with SARFT that is not available to most other Hong
Kong directors. Still, the case remains that both Chan and To are account-
able to SARFT policies and procedures, that they conjure ways to defy
the censors, and that they resort, more or less openly, to locally estab-
lished methods of production. These production methods, moreover, are
capable of yielding plots no less adventurous than those wrought from the
Hollywood-style preproduction screenplay.
The foregoing analyses belie this cinema’s reputation for slapdash story-
telling. When critics characterize Hong Kong cinema as narratively crude,
they invariably base this contention upon local craft practices: Hong Kong
films, they suggest, are simply produced too haphazardly, with too little
detailed preparation, to deliver well-told stories in the Hollywood mold.
Yet this premise generates a teleological fallacy, for the same work routines
that engender a structurally anarchic film such as All’s Well End’s Well
(1992) also yield the tight-woven Wu Xia and Mad Detective. Further,
this deterministic premise—which holds that an apparently loose and
casual scripting method yields loose and casual plotting—undervalues
both the narrative proficiency of Hong Kong scriptwriters and the nar-
rational rigor of certain Hong Kong films. By contrast, I have sought to
indicate that while local filmmaking practices do shape a film’s aesthetic,
that influence is not inevitable or absolute: a range of plot types, more or
less complex or straightforward, may spring from the industry’s standard-
ized work routines.
The rise of the Hong Kong puzzle film, moreover, hardly signals the
abandonment of local traditions. This trend, I argue, is not wholly or even
chiefly attributable to the CEPA trade agreement and the post-handover
rise of Mainland coproductions. Though it dovetails with an international
vogue for complex storytelling, the puzzle-film tendency elaborates a local
tradition of narrative experimentation. Antecedents exist in the Cantonese
melodramas of the 1950s, the wuxia and kung fu extravaganzas of the
1960s and 1970s, certain films of the Hong Kong New Wave, and the
1990s art cinema of Stanley Kwan (Center Stage [1991]) and Wong
Kar-wai (Ashes of Time [1994]). In other words, narrational dynamism
has always been a hallmark of this simultaneously ‘crowd-pleasing’ and
142 G. BETTINSON
‘richly…artful’ cinema (Bordwell 2000, p. 2), and its sources are both
indigenous and international.
Most broadly, I dispute claims that the characteristic Hong Kong aes-
thetic—that which some critics refer to as local flavor—hovers on the
brink of extinction. As critics have noted, a stridently indigenous strain of
Hong Kong cinema has emerged in the post-CEPA period. Suffused with
local atmosphere, these films run the gamut from assertions of carnal out-
rageousness (Vulgaria [2012], Golden Chickensss [2014]) and exercises in
nostalgia (Echoes of the Rainbow [2010], Gallants [2010], Young Bruce
Lee [2010]), to films stressing geographical specificity (Crossing Hennessy
[2010], Big Blue Lake [2011], Firestorm [2013], Aberdeen [2014], Dot
2 Dot [2014], Kung Fu Killer [2014]28) and local textures and rituals
(Ann Hui’s The Way We Are [2008], Fruit Chan’s The Midnight After
[2014]).29 Such films can be understood within the specific context of pro-
democratic political uprisings such as the Umbrella Movement, of which
the reputed Mainlandization of Hong Kong cinema is but one concern.
Perhaps surprisingly, however, the traditional Hong Kong aesthetic
has weathered even the China coproduction system. As Bordwell (2013)
points out, the coproduced Blind Detective ‘yields something like a Hong
Kong comedy of the 1980s,’ replete with plot digressions, oblique causal-
ity, antic physical humor, Cantonese vernacular, and a tone that veers from
gross-out vulgarity to sober pathos. In addition, the ‘false bottoms’ of
the climactic scene’s narration—distressing the reliability of the detective’s
experience—delivers the unpredictable closure typical of Hong Kong’s
‘golden age’ movies. From this perspective, the heralding of a post-Hong
Kong cinema is premature. The poetic traditions of this malleable yet
remarkably robust cinema rumble on.
I am grateful to Peter Ho-Sun Chan, Au Kin-yee, Shan Ding, and Joey

O’Bryan for providing background information for this chapter.
NOTES
1. SARFT, Mainland China’s state censorship body, supervises the country’s
film, television, and radio industries. Its branches include the Film Bureau,
whose function includes the censorship of all films released in the Mainland.
2. See Chan (2009) ‘Policies for a Sustainable Development of Hong Kong
Film Industry.’ Public Policy Digest (July). http://www.ugc.edu.hk/rgc/
ppd1/eng/04.htm. Retrieved 5 August 2015.
3. More recently, NBC’s Hannibal offers an updating of this narrative

schema. For his part, Peter Chan cites the contemporary US television
series CSI and House as influences on Wu Xia. Author interview with Peter
Chan, 3 April 2014, Hong Kong.
4. This work method became systematized by local studios such as Golden
Harvest and Cinema City in the 1980s (Bordwell 2000, p. 121; p. 172).
5. Notes one of Chan’s scenarists: ‘The typical way our company operates is
to make revisions to the screenplay while we shoot the film – it’s been like
this since He’s a Woman, She’s a Man. We’d be discussing the script and
changing the lines on set every day’ (quoted in Li 2012, p. 86).
6. Author interview with Peter Chan, 3 April 2014, Hong Kong. All subse-
quent quotes attributed to Chan derive from this interview.
7. According to Kimburley Wing-yee Choi and Steve Fore (2015, p. 146),
SARFT liberalized the preproduction script mandate in 2004, thereafter
requiring only a plot synopsis. However, Hong Kong filmmakers still today
submit fully composed preproduction scripts to the state authorities.
8. See for further information Bettinson 2016, pp. 44–45.
9. This plot outline and the analysis that follows pertain to the Hong Kong
and Mainland Chinese version of the film released under the title Wu Xia.
An alternative international cut—reedited and distributed by The Weinstein
Company, and retitled Dragon—alters the film’s original plot structure in
various ways. For a comparative analysis of Dragon and Wu Xia, see
Bettinson 2016.
10. Slipperier still, the flashbacks are apt to slide between narrators. A flashback
might be launched by Jinxi, but completed by Xu. In one scene, Jinxi’s
flashback is hijacked by Xu, skewing the narration’s replay toward the
detective’s own speculative version of events. Further narrational complex-
ity springs from the referential nature of the flashbacks assigned to Xu.
Whereas some of the narration’s replays are to be grasped as Xu’s conjec-
tures, others are intended as Xu’s unequivocal memory of actual past
events. Then there is the narration’s propensity to ambiguate not only the
past (as in the conflicting accounts of the robbery) but the narrative future
as well. On occasion, a flash-forward will present a possible future which
then bleeds into an actualized event; hence the narration both suppresses
and flaunts its capacity not only to skip over significant portions of story
time, but to radically disarray the viewer’s knowledge.
11. In the Weinstein Company’s reedited version (Dragon), the definitive rev-
elation of Jinxi’s identity arrives sooner than in Peter Chan’s Director’s
Cut, thanks to the deletion and redistribution of scenes from the film’s first
half.
12. There is an economic explanation too: as a mainstream commercial ven-
ture, Wu Xia cannot risk a resolution wholly reveling in Rashomon-esque
ambiguity.
144 G. BETTINSON
13. For insightful analyses of Perhaps Love, see Stephen Teo (2008), Vivian Lee
(2009), and G. Andrew Stuckey (2014).
14. ‘My first influence [as a child] was Zhang Che’s movies,’ notes Chan. ‘The
Chor Yuen detective movies arrived when I was in my teens – I probably
watched every single one of them.’
15. Like Wu Xia, Chan’s previous China coproduction—The Warlords—finds
a source in a 1970s swordplay saga signed by Zhang Che (The Blood
Brothers [1973]).
16. This is not to claim that all Hong Kong films construct demanding narra-
tions, only that some Hong Kong films do, and, moreover, that the local
tradition of filmmaking is not inimical to sophisticated storytelling.
17. Both To and Wai cut their teeth in the Hong Kong television industry,
directing and writing mini-series and made-for-TV movies at TVB during
the 1970s.
18. Author interview with Shan Ding, 5 April 2014, Hong Kong. All subse-
quent quotes attributed to Shan Ding derive from this interview.
19. Outlandish as it seems, this gambit is not anomalous even for purely local
productions. Other Hong Kong filmmakers employ fake scripts in order to
attract financiers; see Szeto and Chen (2013, p. 13).
20. Another gambit occurs when shooting commences, though only in the
case of local Hong Kong productions for which there is no strict release
date already imposed (such as Mad Detective, or Soi Cheang’s Accident
[2009]). ‘The standard procedure [at the start of a Milkyway production],’
states Shan Ding, ‘is to shoot for a couple of days and then shut down
production for a month or so.’ This hiatus enables Wai’s staff to finesse the
film’s story premise, and reassures investors that production is underway.
Other studio departments also undertake further planning during this
interval. The filmmakers may then use these revisions to persuade finan-
ciers to pour further capital into the production.
21. From To’s perspective, this method limits the prospect of ‘creative interfer-
ence’ from stars and their business representatives, enabling him to ‘pro-
tect the story’ and retain overall artistic control. Author interview with Au
Kin-yee, 5 April 2014, Hong Kong—all subsequent quotes attributed to
Au Kin-yee derive from this interview.
22. Johnnie To discusses this aspect of his process in Ingham 2009, p,136.
23. For a detailed account of Milkyway’s sound-design strategies, see Bettinson
2013.
24. In fact, Mad Detective became moderately successful in Western markets.
Budgeted at less than HK $5m, the film also proved domestically profit-
able, generating revenues of HK $12m at the Hong Kong box office.
25. Such was the necessity for postproduction plotting that the editing phase
consumed three months of a nine-month production schedule, a fairly
extensive period by local standards.
26. Financier Media Asia rejected the initial treatment for Milkyway’s locally
produced Motorway (2012), necessitating substantial revisions by writer
Joey O’Bryan. Author interview with Joey O’Bryan, 20 April, 2013.
27. Thanks to the Milkyway writing staff, Blind Detective’s crime and romance
plotlines are deftly interlinked. In one subjective replay, a murdered school-
girl counsels Johnston on his romantic tribulations.
28. As if in defiance of Mainland encroachment, the closing credits of Kung Fu
Killer pay tribute to an assembly line of local industry figures ‘for uphold-
ing the fine tradition of Hong Kong action cinema.’
29. Interestingly, a number of these ostensibly local productions were part-
funded by Mainland Chinese companies; see Cheung 2015, p. 57.
CHAPTER 8
Can Poetics Break Bricks?
Song Hwee Lim
In 1973, a martial arts film made in the previous year in Hong Kong
underwent a somewhat unexpected transformation. Adopting a method
known as détournement, the situationist film-maker René Viénet sub-
stituted the original dialogue in The Crush/Tangshou taiquandao (Doo
Kwang Gee/Tu Guangqi, 1972) with an ‘anarcho-Marxist reading in
French of the entire image-track as though the film were an allegory of
class struggle between “proletarians” and “bureaucrats”’ (Morris 2004,
p. 182). The result, Can Dialectics Break Bricks?/La dialectique peut-elle
casser des briques? (1973; hereafter Dialectics), exemplifies the ‘situationist
strategy of diverting elements of affirmative bourgeois culture to revolu-
tionary ends’ (McDonough 2004, pp. xiii–xiv). However, as Meaghan
Morris notes, ‘particular instances of détournement can date very quickly’
and, in the case of Dialectics, ‘the joke does wear thin over the duration of
a full-length film,’ though the ‘image-track soon asserts its power’ because
‘the story is exciting and easy to follow’ (2004, p. 182).
The title of Dialectics is clearly inspired by the opening scene of the
original film, which features a martial arts training session that takes place,
as the détourned voiceover puts it, ‘early one chilly morning in a coun-
try where the ideology is particularly cold.’ The film’s image track wastes
S.H. Lim ()

The Chinese University of Hong Kong, Hong Kong, China

DOI 10.1057/978-1-137-55309-6_8
148 S.H. LIM
no time in showcasing the corporeal spectacle of brick breaking by flying

kicks and head butts; even a boy late for training can perform the same
feat with his fist and feet. Notwithstanding the dialectical intention lay-
ered over the soundtrack, the image track never ceases to remind us of
the presence of a Chinese kung fu film whose main attraction remains the
spectacular performance of bodily skills. This foregrounding of ‘corpo-
real authenticity’ (Hunt 2003, p. 29) is symptomatic of the time of the
film’s production in the 1970s when the kung fu genre enjoyed popular
globalization, thanks, in no small part, to the entrance of the symbolic
dragon that is Bruce Lee. Since then, the genre has witnessed an increas-
ing reliance on technology, from the use of wirework (hence the coinage
of the term ‘wire-fu’) by non-action stars (Hunt 2003, pp. 23–4) to the
more recent digital turn to ‘cyber-fu’ (p. 197). As Leon Hunt asks in his
book on Chinese kung fu stars, in the age of The Matrix (Wachowski and
Wachowski 1999), ‘[d]id Jackie Chan and Jet Li achieve global “pres-
ence” just as the body disappeared under the onslaught of the digital?’
(p. 18). Indeed, corporeal spectacle risks being subsumed by technologi-
cal spectacle in what Vivian Lee terms the ‘new cinematic aesthetic’ (2007,
p. 13), a ‘“digital imaginary” [that] displaces the exotic (Chinese) tradi-
tion into the familiar realm of digital technology, resulting in a hybrid,
multiply coded, culturally ambiguous, and therefore transnational visual
medium for global consumption’ (p. 10).
Two further shifts between the 1970s and the present period are note-
worthy in relation to the globalization of the Chinese martial arts genre
(which, for the discussion in this chapter, will include both wuxia and
kung fu films) as well as the genre’s digital turn. The first shift correlates
to China’s political clout. Bruce Lee’s emergence bore the memory of
China’s humiliation at the hands of the West and Japan over the previous
century or so, thus validating Lee’s smashing of the plaque of the ‘Sick
Man of Asia’ in Fist of Fury/Jingwumen (Lo Wei 1972). Enter the new
millennium, however, the dragon that symbolizes China (particularly the
People’s Republic of China, or PRC) no longer crouches or hides but is
leaping confidently onto the world stage. The staggering rise of China as
a global economic giant forms the backdrop to the worldwide success of
Ang Lee’s Crouching Tiger, Hidden Dragon/Wohu canglong (2000; here-
after CTHD) and the opening ceremony of the 2008 Beijing Olympics
masterminded by Zhang Yimou, which featured martial arts in the perfor-
mance. Martial arts, therefore, is indexical of a transformation from shame
to pride, an imaginary of Chineseness and nationalism that has found a
CAN POETICS BREAK BRICKS? 149
new iteration in the current PRC leader Xi Jinping’s slogan of ‘China

Dream.’ Moreover, if Bruce Lee’s display of corporeal spectacle represents
a form of visible (on-screen) physical labor that embodies the last (if not
the only) resort of the economically poor,1 the digital spectacle in the
post-CTHD films, like the Chinese calligraphy and painting executed on
a cutting-edge LED scroll in the Beijing Olympics opening ceremony,
parades an invisible (behind-the-scene) technological labor that can be
bought with money—capital that can be liberally spent to domesticate
Mr. Science even when the will to court Mr. Democracy has not always
been there.
The second shift across the two periods concerns the cultural status of the
cinematic genre. Conventional kung fu films can be regarded as belonging
to the category of body genres (alongside pornography, the horror film,
and melodrama), marked typically by ‘“excesses” of spectacle and displays
of primal, even infantile emotions’ that result in their low ‘cultural esteem’
(Williams 1991, p. 3)—cue Bruce Lee’s primal scream. By contrast, direc-
tors who have been drawn to making martial arts films in the new millen-
nium tend to be the arthouse type with an international reputation, from
Zhang Yimou (Hero/Yingxiong, 2002; House of Flying Daggers/Shimian
maifu, 2004; and others) and Chen Kaige (The Promise/Wuji, 2005; The
Monk Comes Down the Mountain/Daoshi xiashan, 2015) to Wong Kar-
wai (The Grandmaster/Yidai zongshi, 2013) and Hou Hsiao-hsien (The
Assassin/Nie Yinniang, 2015), though a couple of commercial and main-
stream directors have also jumped on the bandwagon (Feng Xiaogang’s
The Banquet/Yeyan, 2006; Stephen Chow’s Kung Fu Hustle/Gongfu,
2004).2 In particular, the use of bullet time to make visible small virtual
objects engenders a new kind of digital spectacle that eschews epic scale,
fancy camera movement, loud sonic effect and fast speed for a poetics of
smallness, stillness, silence, and slowness. Buttressed by the brand names
of auteurs and high production value, this digital aesthetic rescues the
notion of the spectacle from its low cultural status through the staging of
virtual objects in slow motion at a time when a slow temporality has been
enjoying high cultural prestige within world cinema.3 As a result, a rou-
tinely debased popular genre predicated on fast corporeal spectacle now
gains critical credence among the hallowed halls of arthouse cinema in the
form of digital slowness—or spectacle as poetics.
Digital technology, politics, and poetics, therefore, meet at the junc-
ture of the new millennium to herald a new relationship to labor, where
the accolades of the auteurs, the physical prowess of kung fu stars, the
150 S.H. LIM
choreography of wire-fu, and expensive visual effects created by technical

personnel (not to mention machines and software) compete for attention
in high-value productions enabled by transnational capital. If the act of
breaking bricks was allegorized in Dialectics, as the voiceover states right
from the start, to advocate ‘the termination of the exploiters,’ and if cin-
ema was regarded by the situationists as lending itself ‘particularly well
to studying the present as a historical problem, to dismantling processes
of reification’ (Viénet 2004, p. 184), how can we understand the various
forms of labor brought about and reconfigured by the use of digital tech-
nology in filmmaking, especially as they relate to the economy of exploita-
tion? To put it differently, to what extent does this new digital aesthetic
challenge the long-standing hierarchy between notions of poetics and
spectacle? As the spectacle of the human body breaking bricks is displaced
by the poetics of virtual objects flying in bullet time, what is the labor of
the martial arts film and in what ways is it amenable to political critique
and utopian imagination?
Because the revival of the martial arts genre in Chinese cinemas coin-
cided with the use of digital technology (particularly computer-generated
images, or CGI) in global filmmaking, I want to propose in this chapter
that the use of digital technology has implications for the notions of both
spectacle and poetics. I contend that post-CTHD martial arts films deploy
digitality to redefine the meaning and role of spectacle to the extent that
the genre attains a high cultural status through the poetics of slowness. In
what follows, I will first delineate the different notions of spectacle and
poetics to provide a clearer working definition of the two terms. I will then
analyze the use of bullet time in Zhang Yimou’s House of Flying Daggers
and Hero to demonstrate how the different forms of labor (of the director,
the actor, the choreographer, the cinematographer, and visual effects per-
sonnel) coalesce via the virtual objects created digitally and, in the event,
dismantle the hierarchy between spectacle and poetics. Finally, I will draw
on Wong Kar-wai’s The Grandmaster (whose diegesis is based on the life of
Ip Man, who famously taught Bruce Lee the wing chun school of martial
arts) to argue for a reconstituted notion of (labor) force that demands a
sacrifice in order to engender a new politics of aesthetics.4 If conventional
martial arts films had depended on actors’ ability to physically break bricks
to perform bodily spectacle, this chapter asks, in post-CTHD martial arts
films in which digital technology generates virtual objects in slow motion,
can poetics break bricks?
THE POETICS OF SPECTACLE

The antithesis between poetics and spectacle has long been established;
the relationship between the two is hierarchical but not mutually exclu-
sive. In Poetics, Aristotle (1996, p. 3) defines ‘the art of poetry’ (or poet-
ics) as ‘the effect which each species of poetry has and the correct way
to construct plots if the composition is to be of high quality.’ Among
these ‘species of poetry’ (the phrase is closer to what we now call genres),
Aristotle specifically discusses the notion of spectacle in relation to tragedy
and its performance. For Aristotle, tragedy has six component parts which
determine its quality: plot, character, diction, reasoning, spectacle, and
lyric poetry (1996, p. 11). Of these components, plot is ‘the soul of trag-
edy’ and character is second in importance, whereas spectacle ‘is attrac-
tive, but is very inartistic and is least germane to the art of poetry’ (1996,
pp. 12–13). In Aristotle’s reckoning, therefore, poetics is not simply about
the mechanics of constructing a piece of work but, more importantly, an
effect that would engender a high quality in the work. Within this opera-
tion, spectacle is the lowest ranked, for ‘the effect of tragedy is not depen-
dent on performance and actors; also the art of the property-manager has
more relevance to the production of visual effects than does that of the
poets’ (1996, p. 13).
Aristotle’s view on spectacle is worth examining in more detail not only
because it is placed at the very bottom rung in terms of artistry but also
because its relation to plot parallels the tension between spectacle and nar-
rative in film theory. Indeed, what is remarkable about Aristotle’s appraisal
is not so much that plot is ‘preferable’ to spectacle ‘for the evocation of
fear and pity’ in tragedy, but rather that plot alone is sufficient to achieve
catharsis: one only needs to hear the plot—‘even without seeing [the
play]’—to shudder and feel pity at what happens (1996, p. 22). Spectacle,
in this regard, is utterly superfluous. It is staged during theatrical produc-
tion, an extra-textual layering not unlike digital visual effects created dur-
ing postproduction in the filmmaking process that adds little or nothing
to the plot, which remains ‘the source’ of tragedy (1996, p. 12). The
hierarchy between poetics and spectacle is firmly maintained, as the labor
intensive and ‘decorative’ process of staging spectacle is the job of the
lowly property-manager, unbefitting of the high and mighty poet.5
While Aristotle’s account acknowledges the division of labor and the
role of individuals assigned to execute each task, his exclusion of spectacle
from the effect of the plot needs to be questioned. Whether for a theater
152 S.H. LIM
performance in Aristotle’s time or for a filmmaking process in the history

of cinema, it is incredible to claim that plot alone suffices in achieving its
effect. Hearing the plot of Oedipus the King (Aristotle’s example) does not
prepare an audience for the mise-en-scène of actor, action, lighting, color,
costume, set, props, music, and so on that makes each theatrical staging
or cinematic adaptation of Sophocles’ play unique and different. In fact,
what leaves a lasting impression could well be the spectacle rather than the
plot, for example, the drum beats accompanying the first fight scene set on
the rooftops in CTHD or the richness of colors saturating Hero. Spectacle
is one of the means through which plot elements are presented. It is not
superfluous to the plot, but rather central to its revelation.
The exclusion of spectacle from plot, however, has also been prevalent
in film theory. In an article that aims to cross the great divide of spec-
tacle versus narrative, Aylish Wood summarizes two positions in existing
analyses of spectacular cinema. The first position proposes that ‘spectacle
interrupts narrative progression,’ drawing attention to itself ‘as surface,
an excess that distracts from temporal flow’ (2002, pp. 371–2). This
position echoes that of Aristotle insofar as spectacle is effectively brack-
eted from the narrative (such as song-and-dance numbers in the musi-
cal) and deemed as an attraction or a distraction to the extent that ‘the
plot becomes simply a pretext for the real raison d’être, the parade of
splendors’ (Bordwell 2000, p. 178). The second position argues that spec-
tacle ‘enhances the effect of narrative’ (Wood 2002, p. 371; emphasis in
original) so that contemporary spectacular cinema ‘combines spectacular
elements with narrative elements and integrates them within the experi-
ence of the film’ (2002, p. 373). It is the second position that I will take
in this chapter, and I will elaborate in the next section how this position
not only dissolves the boundary between narrative and spectacle but also
dismantles the Aristotelian hierarchy between poet/director (poetics) and
property-manager/visual effects personnel (spectacle) through an analysis
of the use of bullet time in recent Chinese martial arts films.
The hierarchy between poet/director and property-manager/visual
effects personnel, and that between the effects their roles are designed
to produce (poetics vs. spectacle) have survived from Aristotle’s time to
the present day. Indeed, Aristotle’s thesis written in fourth century BCE
foreshadows a mid-twentieth-century French notion of film authorship
that privileges the director over his collaborators. In an essay first pub-
lished in 1948, Alexandre Astruc introduced the idea of la caméra-stylo
for the ‘film-maker/author’ who ‘writes with his camera as a writer writes
with his pen’ (Astruc 2009, p. 35), a notion inspired by literary writing
produced by an individual and mapped onto the collective practice that is
cinema. In a famous article published in 1954, François Truffaut (1966)
built on this notion to make a distinction between auteurs who ‘often
write their dialogue’ and, in some cases, ‘invent the stories they direct’
and metteurs-en-scène who merely execute the filming of the script (Cook
2007, p. 390). Today, even when directors are not directly responsible for
the plot of their films as screenplay writers, they are often designated sole
authorship of their films and occupy the top rung of a production system
that includes collaborators such as visual effects personnel.
While I consider the use of digital technology as marking a paradigm
shift in the mode of spectacular cinematic production, this hierarchy has
been maintained because the labor of digital work goes unacknowledged
and because directors tend to jealously guard their precious status. Not
unlike early cinema audiences who went to exhibitions ‘to see machines
demonstrated’ (Gunning 1986, p. 66), audiences who encounter today
the new forms of cinematic technological innovations (from 3D to IMAX)
typically reserve their awe and wonder for the novelty of the technol-
ogy itself rather than recognize the labor of the personnel operating the
machines. Moreover, directors can be dismissive of—or they may disavow
the role of—digital technology in filmmaking. Quentin Tarantino, who
engaged Yuen Woo-ping (CTHD’s choreographer) on both his Kill Bill
films (2003, 2004), apparently expressed disdain toward all that ‘CGI
bullshit’ employed in films such as The Matrix trilogy by the Wachowski
Brothers (1999, 2003, 2003), which also featured Yuen as stunt chore-
ographer (North 2005, p. 58). In interviews, both Zhang Yimou and his
cinematographer Christopher Doyle ‘downplay[ed] the role of high-end
visual effects in Hero’ (Farquhar 2010, p. 185). Besides, industry prac-
tice dictates that visual effects personnel creating the Aristotelian lowest-
ranked spectacle must remain anonymous (or, at best, a name among a
long list of end-credit names rolling at great speed), while the director is
prominently proclaimed as the author of the film, from the textual (open-
ing credits) to the extra-textual (publicity posters and junkets). According
to Mary Farquhar, the visual effects specialists working on Hero were
contracted through international studios, but the names of those studios
are not listed in the credits; instead, the names of the 65 people respon-
sible for visual effects appear almost at the end of the credits, ‘just before
“wardrobe, hair and makeup”’ (2010, p. 185). Even the role of the visual
effects supervisor for Hero was merely to present the storyboards and
154 S.H. LIM
concept art to the director, whose input would be ‘to determine whether
the design fitted into his vision of the film’ (N.A. 2003, qtd. in Farquhar
2010, p. 195 n. 6). In the final analysis, it is the director who possesses
overall vision (poetics), claims authorship, grants interviews, and makes
decisions about the visual effects (spectacle) of the film.
THE POETICS OF SLOWNESS

How, then, has digital technology facilitated a paradigm shift in the
staging of spectacle in post-CTHD martial arts films, and how does this
unsettle the Aristotelian hierarchy? I propose that digitality makes its mark
in three ways, and I will demonstrate these aspects via an examination
of selected shots in Zhang Yimou’s Hero and House of Flying Daggers,
released, respectively, two and four years after CTHD. Firstly, the exten-
sive use of bullet time, which is rooted in the tradition of slow motion
(North 2005, p. 54), introduces a poetics of slowness to a genre better
known for its fast fists of fury. In this, I depart from Vivian Lee’s read-
ing of the two abovementioned Zhang Yimou films, in which she con-
cludes that the use of digital technology results in ‘a multilayered visual
pattern that emphasizes not only speed but also the staging of special [sic.]
effects as the films’ most spectacular spectacle’ (2007, p. 25; emphasis
in original).6 Secondly, the slowness enabled by bullet time reformulates
the notion of spectacle as a kind of poetics not just through its stretched
temporality but also because many of the virtual objects in Zhang’s films
are miniscule rather than epic in scale. While Zhang’s films also deploy
what Kristen Whissel (2010) calls the ‘digital multitude’ in the form of,
say, innumerable soldiers firing uncountable arrows at Jet Li’s character in
Hero, my focus here is on objects that downplay size and speed in favor
of smallness and slowness.7 Finally, as I shall explain below, the manner in
which these small virtual objects appear within the narratives makes visible
the labor of visual effects personnel to the extent that this usually unseen
form of labor demands to be acknowledged—on-screen. As such, this new
visibility of labor possesses the potential to undo the hierarchy between
poetics and spectacle, and to provide a critique about the unequal power
relation inherent in collaborative artistic practices.
It is worth noting from the outset that what constitutes a spectacle at
a certain historical juncture is dependent upon what an audience has or
has not previously encountered. For example, in the early days of cin-
ema, ‘[w]hat today seems innocuous – a shot of some workers coming out
of a factory, say – then seemed magical, a spectacular demonstration of

the miracles of science more exciting than the bullet photography in The
Matrix is for us today’ (Arroyo 2000, p. vii). Thus, we need to ask what
specific aspect in Zhang’s staging of visual effects is spectacular, what is
new in his use of bullet time for the martial arts genre, and in what sense
can the notion of spectacle be understood instead as poetics. In fact, using
slow-motion photography to make visible the movement of small beings
and objects (from insects to raindrops) is commonplace in science and
nature documentaries.8 Yet I would contend that what is new in Zhang’s
use of bullet time is the role these small objects play in the narratives so
that visual spectacle is not exterior to plot and is turned into poetics.
Let us first consider a more conventional use of bullet time. In House
of Flying Daggers, a famous scene set in the Peony Pavilion stages an Echo
Game in which a blind dancer (Zhang Ziyi) responds to sounds of beans
hitting a circle of drums by flicking her long sleeves to hit the drums in the
right order, only for the game to turn into a fight scene between the dancer
and the captain of the police (Andy Lau) who flicks the beans. Within this
exquisitely choreographed sequence is the bullet-time motion of flying
beans, first with one bean hitting one drum, followed by a doubling of the
number of drums hit by a single bean, then a quadrupling. When a whole
bowl of beans is eventually scattered, Zhang holds her posture in stillness
while virtual beans circle her like orbiting comets (Fig. 8.1).
Fig. 8.1 Zhang Ziyi holds her posture in stillness while virtual beans circle her like
orbiting comets in House of Flying Daggers. Copyright Beijing New Picture Film
Co., China Film Co-Production Corporation, Edko Films, Elite Group Enterprises,
Zhang Yimou Studio
156 S.H. LIM
This bullet-time scene exemplifies a poetics of slowness, especially in

contrast to the extended spectacle that immediately follows it. After two
shots of beans falling in slow motion to the ground, the dancer flicks her
sleeves to hit the circle of drums but the performance of high jumping and
somersaults is closer to martial arts than to dance (hence the use of a body
double).9 This extended sequence is pure spectacle for its corporeal skills,
fast editing to rhythmic drum sounds, multiple camera angles, and lavish
colors of the costume, prop, and set. On the other hand, the two bullet-
time shots of Zhang holding her posture while numerous beans encircle
her inject a sense of slowness through the stillness of the camera as well as
that of the performer, the slow motion of virtual beans, and the relative
silence of the soundtrack (soft sounds of beans hitting drums). Moreover,
the smallness of the virtual objects lends a spatial dimension to bullet-time
technology, for the latter not only slows things down in time but also ren-
ders small things visible in space. Bullet time, therefore, makes possible the
visibility of virtual small objects through both time and space and, in the
process, enables the scaling down (and simultaneous upgrading in status)
of spectacle as poetics.
In their smallness, slowness, stillness (the camera typically holds still
while virtual objects are flying), and silence (flying objects are often
accompanied only by the sound of their flight), these objects transform
the staging of fighting spectacle in martial arts films into moments of tem-
poral suspension and contemplation worthy of the poetics of art cinema.
Moreover, in these two films by Zhang Yimou, bullet-time technology
challenges the Aristotelian hierarchy because these small virtual objects are
foregrounded to the extent that they invariably intervene in any martial
exchange between opponents. These objects become the third character
between opponents whose martial exchanges must now include them, be
it as weapon (daggers) or prop (beans). The visibility of these small objects,
in turn, undermines notions of the spectacular and the epic, which are
inextricably bound to the category of the popular, a lowly cultural marker
deemed as ‘only entertainment’ and ‘the opposite of serious works of art
designed to change perceptions, push aesthetic boundaries and challenge
our views’ (Arroyo 2000, p. vii).10 Since these small objects are only visible
with the aid of bullet time, their appearance in fight sequences brings a
poetics of slowness to the otherwise high-speed corporeal spectacle.
A prime example of the central role granted to small virtual objects is
the first fight sequence in Hero, which features the protagonist Nameless
(Jet Li) and an opponent Sky (Donnie Yen). The climax of the sequence
is staged as a mind game in which the two fighters, filmed in color, stand
still with their eyes shut, while their alter egos, filmed in black-and-white,
carry on a fight which showcases the actors’ high-speed bodily skills, invis-
ible wirework, and visual effects, accompanied by a tune played by a blind
old man on his zither. The mind game ends when the strings on the zither
break and, following a brief moment of silence, the film returns to its
default color mode to enact a final scene in which Nameless kills Sky, this
time accompanied by an extra-diegetic score and sound effects.
What is most remarkable about this final scene is the privileging of
small virtual objects whose intervention between fighting opponents now
takes on a haptic dimension. If the virtual beans in the abovementioned
scene in House of Flying Daggers merely encircle the dancer but do not
touch her, Nameless in Hero must come into physical contact with virtual
objects before executing the deathblow. As Nameless charges toward his
opponent, a close-up of his face in profile shows him breaking through six
strings of raindrops, splashing the droplets as his face comes into contact
with them (Fig. 8.2). The next shot reprises this hapticity but this time
showing Nameless’ sword piercing through the rain before it reaches the
opponent three shots later.
Here the digital spectacle of virtual raindrops is no longer an extra-
textual layering or bracketed from the narrative, but becomes a plot ele-
ment in itself, occupying as crucial a role as Jet Li’s character. In narrative
Fig. 8.2 Close-up of Jet Li’s face in profile as he breaks through six strings of
raindrops while charging toward his opponent in Hero. Copyright Beijing New
Picture Film Co., China Film Co-Production Corporation, Elite Group Enterprises,
Sil-Metropole Organisation, Zhang Yimou Studio
158 S.H. LIM
terms, the appearance of virtual raindrops is seemingly superfluous since

it does not have any bearing on the plot of the film. The creation of these
small objects, however, generates new effects in martial arts films so that
our attention is not just on the fighters’ corporeal skills or their mid-air
flight enabled by wirework. Rather, as Nameless’ face smashes and splashes
those virtual raindrops, we are faced with a new category of spectacle
(which we may now call poetics) that grants visibility to smallness and
hapticity through slowness. Moreover, this digital aesthetic reformulates
the relationship between spectacle and narrative/plot insofar as the former
is now indistinguishable from the latter. To put it differently, the plot in
this instance of digital intervention is not ‘Nameless kills Sky’ but rather
‘Nameless encounters raindrops whilst killing Sky.’
More importantly, as slowness begets visibility and generates hapticity,
it also interpellates a new spectatorial practice that recognizes the finger-
print of the virtual raindrops as that belonging not to the actor, director,
choreographer, or cinematographer but unmistakably to the visual effects
personnel and their digital machines. Indeed, in this shot the labor of
the actor—whose action has been choreographed, directed, and shot—
encounters, quite literally, the labor of the visual effects personnel as well
as the labor of the digital machines, all rendered visible on-screen in the
form of splashed raindrops. The hierarchy between all these forms of labor
is dismantled in this digital remediation because none of them would have
existed without the others since the shot is a composite of the various types
of expertise that each contributes. No more talk of ‘CGI bullshit’ because
CGI has transformed spectacle into poetics. This, I would argue, is the real
impact of post-CTHD martial arts films: a reformulation of spectacle as
poetics that also raises questions about the politics of poetics. To address
the latter aspect I will draw on Wong Kar-wai’s The Grandmaster to inter-
rogate notions of sacrifice and (labor) force that could engender a new
politics of aesthetics, an aesthetics whose ethos is not violence but peace,
and whose practice is premised precisely on not breaking bricks.
THE POLITICS OF POETICS

In my book Tsai Ming-liang and a Cinema of Slowness, I propose that
the use of long takes and the narrative trope of waiting in a body of con-
temporary films comprise aesthetic acts that (following Jacques Rancière)
‘promote new modes of temporal experience, new ways of seeing, and
new subjectivities that are politically committed to an ethos of slowness’
(Lim 2014, p. 33). In his book On Slowness: Toward an Aesthetic of the

Contemporary, Lutz Koepnick examines the use of slow-motion photog-
raphy in order to tease out the stakes of this technology. With Neo from
The Matrix in mind, Koepnick argues that ‘slow motion today has largely
come to encode and celebrate a world and cinematic enterprise in which
neither the random nor the irreversible may ever thwart a hero’s inten-
tional actions’ (2014, pp. 176–7). Drawing on the films of Tom Tykwer
as counterpoint, Koepnick lays down the stakes of what Rancière (2006)
would call ‘the politics of aesthetics’:
What is at stake is to embrace the special effect that is cinema as a lens not
only allowing us to see the world differently, but restitching the fabrics of
what can be thought and done. What is at stake is to explore the indetermi-
nacy of aesthetic experience as an effigy holding up the claims of subjectivity
and sensory experience in spite of their ever increasing dematerialization.
Rather than lionize heroic self-determination while at the same time over-
powering the spectator, then, slow motion in Tykwer invites protagonists
and viewers alike to probe the promises of a world in which we can safely
and playfully yield to what exceeds our control (2014, p. 181).
To re-view the first fight scene in Hero in Koepnick’s terms, the virtual
raindrops may not thwart the hero’s intentional action to kill his opponent,
but the spectacular slow-motion image of the hero’s face being splashed
by these small objects en route to his big mission does present us with a
new poetics, one in which six strings of virtual raindrops serve as a kind of
obstacle course through which the hero must pass. The presence of rain
(albeit man-made here) is a force of nature beyond our control. To bor-
row Koepnick’s description of Tykwer’s slow-motion images of free fall,
these virtual raindrops ‘cause protagonists and viewers alike to experience
the present – aesthetically, as it were – as a site of disjunction that suspends
any direct relationship between certain actions and their desired effects,
between intentional movement and anticipated destinations’ (Koepnick
2014, p. 180). In the films of Tykwer and Zhang, the human agents’
intentional actions are invariably mediated by the poetics of slow-motion
photography to the extent that their physical force must be thwarted,
decelerated, and sacrificed for a new politics of slow aesthetics.
This new politics of aesthetics finds a specific configuration in Wong
Kar-wai’s The Grandmaster in relation to the trope of brick breaking.
Despite the film title’s ostensible reference to Ip Man (played by Tony
160 S.H. LIM
Leung Chiu-wai), the narrative grants equal space to another plot thread
which features Zhang Ziyi as Gong Er and the revenge she seeks for her
father’s death at the hands of his disciple, Ma San (Zhang Jin). In fact, it
is the duel between Gong and Ma on a railway platform with high-speed
passing trains that has been described as the film’s ‘most spectacular set
piece’ and ‘one of the few concessions to CGI in a film that relies largely
on the in-camera magic of 35mm’ (Kermode 2014). The duel runs for a
substantial duration of approximately six minutes (around one-twentieth
of the film’s total running time),11 from which I want to highlight two
shots to illustrate how this film exemplifies the transformation of the spec-
tacle of breaking bricks into a poetics of slowness.
Both shots are drawn from the start of the duel. In the first instance,
Gong has been forced to back onto a pillar and when Ma attempts to
punch her, she dodges and he hits the pillar instead. There is a close-up of
Ma’s fist hitting the pillar, followed by another close-up of its impact on
the other side of the pillar: a nut that bolts metal strips on the pillar shakes,
but it remains intact while tiny snowflakes splash from above the strips
(Fig. 8.3). This is immediately followed by Gong’s counterattack, hitting
the bench on which Ma has been forced to sit and from which he dodges
Gong’s assault. There is now a close-up of both of Gong’s palms hitting
the wooden planks of the bench, sending snowflakes into the air, followed
by another close-up of the impact of Gong’s palms on the other side of the
Fig. 8.3 A nut that bolts a metal strip on the pillar shakes but remains intact while
tiny snowflakes splash from above the strip in The Grandmaster. Copyright Block 2
Pictures, Jet Tone Films, Sil-Metropole Organisation, Bona International Film
Group
planks: a loosening of a screw alongside its bracket, the screw then jump-
ing halfway out of the hole but descending back into it (Fig. 8.4).
Like the bullet-time shots in Zhang Yimou’s two films discussed above,
these two shots in The Grandmaster are rendered in slow motion and they
privilege small virtual objects. The staging of these moments of spectacle-
as-poetics, I would argue, is categorically different from a preponderance
to create the digital multitude in many films to ‘dramatize their pro-
tagonists’ relationships to sudden, often apocalyptic, historical change’
(Whissel 2010, p. 91). In those films, ‘in order for the protagonists to
thwart the multitude’s rapid movements and the dreaded change it rep-
resents, everything must happen at an accelerated pace’ (2010, p. 96). By
contrast, in The Grandmaster the historical change facing the protagonists
(the ownership and passing on of martial arts traditions) is dramatized as
duels between two characters only, and aided by a couple of virtual objects
in slow motion, thus differing from the digital multitude in both pace and
scale.
More importantly, Wong’s film reconfigures the trope of brick break-
ing in martial arts films, mobilizing the virtual objects to underline that
this new digital aesthetic is exactly about not breaking things. This is a
recurring trope in The Grandmaster, especially in fight scenes between key
protagonists.12 Indeed, the film could hardly bring itself to break a piece of
cake—never mind a brick—in the fight scene between Ip Man and Gong’s
Fig. 8.4 A screw loosens alongside its bracket, with the screw then jumping half-
way out of the hole before descending back into it in The Grandmaster. Copyright
Block 2 Pictures, Jet Tone Films, Sil-Metropole Organisation, Bona International
Film Group
162 S.H. LIM
father, which is staged like a tai chi exercise with its slow movements and
circling trajectories.13 Later, in the fight scene between Ip Man and Gong
Er, the deal the former proposes is that he will admit defeat if anything
is broken in the exquisite brothel where the duel takes place. As such,
the protagonists’ response to historical change is at once more peaceful
and more ambitious. Peaceful, because the emphasis is not on physical
force itself but rather on its control, hence the relative non-violence in the
staging of confrontations: in the duel between Gong and Ma the human
agents are shown literally to be pulling no punches, but these punches
do not break the brick-and-mortar of the pillar or even wooden planks.14
Rather, the protagonists’ control (and transference) of force is visual-
ized by its impact on small virtual objects in slow motion.15 Ambitious,
because, as Ip Man replies to the suggestion by Gong’s father that the
nation (unlike martial arts) does not have a distinction between north and
south, why limit one’s vision to the nation when the cake could represent
the whole world (and not just the martial arts world)? This approach of
non-violence and global ambition is what we understand today as ‘soft
power’ (Nye 2004).
This poetics—and its politics—fundamentally restructures the spec-
tacular tradition of the Chinese martial arts genre, including a belief in
sacrifice in the form of human labor for the attainment of the physical
prowess to break bricks. Sacrifice, as Rey Chow argues in an insightful
essay in which she identifies mimesis as its ‘conceptual double or con-
joined twin’ (2006, p. 132), often presents itself as a tempting notion
‘subscribed and adhered to by the victims and their community, as an
inalienable part of their belief’ (p. 134). This sacrificial logic has found a
popular expression (a popularity not necessarily measured by box-office
intake but rather qualified by structure of feeling) in early Zhang Yimou
films (such as To Live/Huozhe, 1994), in which endurance (itself a form
of sacrifice) equates to being (Chow 1996). Whether showcasing the suf-
fering of ordinary people during the Cultural Revolution or the sacrifice
one must make to perfect kung fu skills, mimesis is the preferred mode of
representation for the staging of violence as spectacle, evidenced by the
sweat, blood, and tears the protagonists must parade.16
The staging of small virtual objects, however, is not always amenable
to the logic of sacrifice. While digital technology is used to mimetically
represent the nuts and bolts in the two CGI shots in The Grandmaster dis-
cussed above, these virtual objects upstage the notion of human sacrifice
if only because the physical ability to break bricks is no longer an essential
criterion for the choice of actors in the new digital cinematic aesthetic. To
put it another way, the sacrificial logic is here reformulated so that it is not
predicated upon force but the control of force, not upon destruction but
upon preservation of objects, not upon violence but upon peace, not upon
corporeal spectacle but upon digital poetics, not upon on-screen labor but
upon off-screen labor, and, finally, not upon sacrifice but upon play.
For a situationist such as Fourier, ‘sacrifice itself is the thing to repudi-
ate’ and the sanctification of hardship in the Catholic religion should be
accompanied by a sanctification of sensual pleasure (Wark 2013, p. 62). As
digital technology becomes more and more adept at generating non- and
post-human forms, we are increasingly invited, in these post-CTHD mar-
tial arts films, to relish the sensual pleasure of small virtual objects shown
in slow motion, displacing at once the human agents and their corporeal
skills gained through sacrifice. Like the technique of détournement, these
phantom objects are a game in which poetics does not break bricks, not
because it cannot, but because it chooses not to. If ‘[e]verything is at
stake, but the world is still a game’ (Wark 2013, p. 17), the stake that is
sacrificed in this new digital aesthetic is the notion of human sacrifice itself,
rooted in the corporeal spectacle of kung fu fighting. This digital aesthetic
raises instead the stakes of slowness and smallness in order to present us
with new sensual pleasures, painstakingly created behind the scenes by a
labor force whose sacrifice can no longer be subsumed under the pedestal
of the director-poet, a dialectical power dynamic that must now disappear
under the façade of these post-CTHD martial arts films.
Acknowledgments Thanks to Song-yong Sing, Adrian Martin, and McKenzie
Wark for pointing me to sources on the situationists; to Rey Chow, Paul Bowman,
and Jessica Ka Yee Chan for their helpful feedback on a draft of the chapter; and to
the editors of the volume for their careful reading and suggestions.
NOTES
1. As Slavoj Žižek suggests (here paraphrased by Paul Bowman), the ‘strong
appeal of martial arts films in ghettos across the world over was initially
class-based’ because ‘[t]hose who have nothing […] have only their bod-
ies, only their discipline, only their desire’ (Bowman 2013, p. 174).
2. It must be qualified that the boundary between arthouse and mainstream
is not always clear-cut, and it can be argued that directors such as Zhang
Yimou and Chen Kaige started as arthouse directors but have become
more mainstream over the years, whereas an arthouse director in Hong
164 S.H. LIM
Kong (like Wong Kar-wai) still faces commercial pressure to cast stars in
the lead roles in a way that arthouse directors in Taiwan do not.
3. On the consumption and cultural prestige of slow cinema, see Lim (2014)
and Schoonover (2012).
4. Today the fascination with Bruce Lee is as strong as ever: a five-year (from
2013 to 2018) exhibition on his life and art is currently running at the
Hong Kong Heritage Museum; and books on him, academic (Bowman
2010, 2013) or otherwise, continue to be published. This fascination has
been extended to his former master Ip Man, who has recently become the
subject of no fewer than six films: Ip Man/Ye Wen (Wilson Yip 2008), Ip
Man 2/Ye Wen 2 (Wilson Yip 2010), The Legend is Born: Ip Man/Ye Wen
qianzhuan (Herman Yau 2010), Ip Man: The Final Fight/Ye Wen: Zhongji
yizhan (Herman Yau 2013), The Grandmaster (Wong Kar-wai 2013), and
Ip Man 3/Ye Wen 3 (Wilson Yip 2015).
5. My use of the term ‘decorative’ echoes Andrew Darley’s discussion of
visual digital genres (such as spectacle cinema, music video, and computer
games) that are deemed as ‘“lesser” forms of art or culture’ since they ‘tend
greatly to play up form, style, surface, artifice, spectacle and sensation, and
they dilute meaning and encourage intellectual quiescence’ (2000, p. 6).
6. I believe Vivian Lee is here referring to visual effects (created in postpro-
duction) rather than special effects (generated in-camera).
7. I thank Jia Tan and Jessica Chan for bringing Whissel’s work to my
attention.
8. I thank Rey Chow for reminding me of this.
9. In the DVD (Pathé P-SGB P916301001) commentary which features
Zhang Yimou and Zhang Ziyi in conversation, the director disclosed that
he engaged a dance choreographer six months in advance for the choreog-
raphy of this sequence. However, he decided at the last minute before
shooting to adopt an action rather than dance approach to the sequence,
and recalled the martial arts choreographer Ching Siu-tung from his vaca-
tion in Hong Kong to shoot the sequence.
10. The coupling of action cinema and spectacle is evident in the book title,
Action/Spectacle Cinema (Arroyo 2000), and the coupling of epic film and
spectacle can be seen in the book title, Epics, Spectacles, and Blockbusters: A
Hollywood History (Hall and Neale 2010).
11. Note that there are three versions of the film with varying lengths. The one
released in East Asia and to which I refer runs for 130 minutes, whereas the
US version is 108 minutes because, as Wong explained to an audience in Los
Angeles, ‘[w]e have an obligation to give the picture [to The Weinstein Co. for
U.S. release] within two hours, so we have to create a shorter version’ (Appelo
2014). The version submitted to the Berlin film festival in 2013 was 120 min-
utes long (see https://www.berlinale.de/en/archiv/jahresarchive/2013/02_
programm_2013/02_Filmdatenblatt_2013_20137779.php#tab=filmStills;
date accessed 7 October 2015).
12. It is only in the scene introducing a third protagonist, The Razor
(Yixiantian, played by Chang Chen), that the tiles and cement are smashed
when The Razor flings an opponent onto the pillar at the end of a fight
sequence set outdoors in the rain.
13. My argument here is inspired by Paul Bowman’s suggestion of the scene’s
‘tai chi aesthetic’ in his feedback to a draft of the chapter. A character
watching the duel between Ip Man and Gong’s father also comments upon
it using a tai chi analogy.
14. I highlight ‘relative’ here because Gong still needs to beat Ma to his feet in
order to regain ownership of her father’s legacy.
15. In fact, physical force—its restraining control rather than brute execu-
tion—is a recurring theme in all Ip Man films. I thank Paul Bowman for
highlighting this to me and for this reading of control and transference of
force.
16. In this we return to Linda Williams’ notion of body genres, in which
ecstasy is shown in pornography, horror, and melodrama by, respectively,
ejaculation, blood, and tears (1991, p. 9).
CHAPTER 9
Poetics of Parapraxis and Reeducation:

The Hong Kong Cantonese
Cinema in the 1950s
Victor Fan
Is there such a thing as a poetics of Hong Kong Cantonese cinema in

the 1950s? Like many enthusiasts, critics, and historians of the Cantonese
film, I began asking this question—at first unknowingly—when I grew
up in Hong Kong in the 1970s. At that time, the two Cantonese televi-
sion channels, TVB Jade and Rediffusion’s Chinese Service, showed their
respective collections of Cantonese films from the 1950s and 1960s under
the title Yueyu changpian (Jyutjyu coengpin or Cantonese-language fea-
tures). This title is often derogatorily, or at times, affectionately, dubbed as
Yueyu canpian (Jyutjyu caanpin or worn-out Cantonese films). For many
people, the pleasure of watching a Jyutjyu coengpin requires a taste for a
set of stylistic traits that are distinct from classical Hollywood cinema and
from the Hong Kong Cantonese films made after the 1960s. In other
words, there is an unspoken assumption that these films demonstrate a
stylistic consistency in the deployment of their narrational devices and
audiovisual elements.
Yet such claims of stylistic distinctiveness, so common in social con-
versations in the past and social media today, are often impressionistic.
V. Fan ()
King’s College London, London, England

DOI 10.1057/978-1-137-55309-6_9
168 V. FAN
In spite of the vibrancy of the studies of the Cantonese film, not many
scholars devoted their research to rethink whether the Cantonese film is
best considered a ‘classical’ system. Nonetheless, making such a claim is
not easy, and it begs a number of questions. First, very few scholars today,
as David Bordwell et al. (1985, pp. 13–15) once did, have the resources
to choose a hundred random samples from the Hong Kong Film Archive
and conduct a detailed formal analysis of each film. Second, even Bordwell
would agree that each film in the classical system instantiates a set of dif-
ferences from a structural norm, rather than a set of rules that determine
the syntactic organisation of each film’s formal elements (Bordwell 1985,
pp. 1–26; Bordwell et al. 1985, pp. 4–6). Third, historically, Hong Kong
Cantonese cinema in the 1950s does not have one unified narrational
style. Rather, there are stylistic conventions specific to individual studios,
political positions (leftwing or rightwing), and genres. Last but not least,
it is not easy to explain why so many Cantonese filmmakers and specta-
tors, who were very familiar with the classical style of narration in both
Hollywood cinema and Mandarin cinema, developed a stylistic paradigm
that defied or violated those expectations.
Answering all of these questions would be beyond the scope of this
chapter. Instead, I focus on one single film, Weilou chunxiao [Ngailau
ceonhiu or In the Face of Demolition, Lee Tit 1953], one that exemplifies
the Hong Kong Cantonese leftwing cinema in the 1950s, or more specifi-
cally, the stylistic paradigm of the Zhonglian dianying gongsi (Zunglyun
dinjing gungsi or Union Film Enterprise, 1953–64), a studio that domi-
nated the market and cultural imagination of the Cantonese cinematic dis-
course in Hong Kong and the Chinese diaspora in Southeast Asia, North
America, and Europe between 1953 and 1964 (Zhou 2011, pp. 9–18).
By exemplification, I do not mean that the film serves as a prototype of
all Union films; rather, it is best understood as a site where stylistic dif-
ferences can be traced and reexamined, from which further studies of
the poetics of Hong Kong Cantonese cinema in the 1950s can be devel-
oped. Here, I argue that In the Face of Demolition shows a poetics dis-
tinct from the classical Hollywood style. However, such distinctiveness
is neither the result of the Cantonese filmmakers’ attempt to modify the
classical Hollywood system, nor the fact that the Cantonese viewers use
a cognitive map different from their Mandarin and Euro-American coun-
terparts. Rather, based on Miriam Hansen’s (1999, pp. 59–77) argument,
I propose that the Hong Kong Cantonese cinema in the 1950s was a
public sphere where contesting notions of political affiliation, socioeco-
POETICS OF PARAPRAXIS AND REEDUCATION: THE HONG KONG CANTONESE... 169
nomic values, and aesthetic sensibilities were negotiated, and the resulting
classical Cantonese style actively serves as a medium of negotiation. More
specifically, the Union films in the 1950s can be seen as a form of para-
praxis (Freudian slip), defined by Thomas Elsaesser (2009, pp. 190–92)
as the ‘failure of performance and performance of failure’. In the 1950s,
leftwing intellectuals and cultural producers have failed—in the sense that
they chose to stay in Hong Kong instead of returning ‘home’ to construct
‘New China’—to fight against capitalism, and the closing of the border
between Hong Kong and Shenzhen (Shumchun) also made it impossible
for them to return ‘home’ without risking their status quo. In this light,
these films perform the failure of the leftwing filmmakers and audience
to take any active political agency. They offer narratives in which intel-
lectuals are reeducated and reintegrated into the masses, thus offering the
spectators a ‘second chance’ to rehearse the possibility of having a political
agency to activate sociopolitical changes.
FROM THE CANTONESE THEATRE TO THE CANTONESE

STYLE OF NARRATION
The unique narrational style developed in Hong Kong Cantonese cinema
in the 1950s can be traced back to the zhengbenxi (zingbunhei or feature
play) of the Yueju (Jyutkek or Cantonese theatre, more popularly known
as the Cantonese opera), a dramatic format developed between the fif-
teenth and the seventeenth centuries from the Kunqu and other regional
theatrical forms. During local or familial celebrations and worships, theat-
rical troupes would be hired to perform for hours or even for days. Hence,
unlike a classical Greek drama or an Elizabethan play, a Cantonese xi (hei
or play) is not organised around a unified space and time. Rather, it con-
sists of a series of ben (bun, acts, or numbers), each featuring a dramatic
action and showcasing a particular role (in theatrical terms) or voice (in
musical terms) called hangdang (hongdong) (Chan 2007, pp. 155–7; Fan
2015, p. 155; Lai 2010, pp. 34–51; Mak 1940, p. 17; Ouyang 1954,
p. 150). By the eighteenth century, most one-evening performances fol-
lowed a programme called da paichang shiba ben (daai paaicoeng sapbaat
bun or 18 acts in a grand programme). These 18 acts do not necessarily
form a dramatic whole; even if they do, individual acts can remain episodic
without any clear indications of causal, spatial, and temporal relationships
(Fan 2015: p. 172; Lai 2010: pp. 40–8).
170 V. FAN
After the institutionalisation of the Jingju (Peking opera) in 1790,

and the emergence of the xiyuan (heiyuan or playhouse) in trading ports
such as Guangzhou (Canton), Xiamen (Amoy), Foshan (Fatshan), and
Hong Kong in the eighteenth and nineteenth centuries, the Cantonese
theatre began to borrow dramatic concepts such as qi (hei or exposi-
tion), cheng (sing or development), zhuan (zyun or dramatic twist), and
he (hap or denouement) in the organisation of acts. In 1906, a group
of Chinese students in Tokyo including Li Shutong (1880–1942), Zeng
Xiaoju (1873–1936), and Ouyang Yuqian (1889–1962) established the
Haruyanagisha (Chunlishe), performing and advocating the shingeki
(xinju or new play, also known as the wenming xi or civilised play). A
new play often takes key acts from a European or American drama, and
the actors would perform (represent) and illustrate (present) the narrative
through a series of acts. In the Canton-Hong Kong region, performers
added singing to these acts as early as 1907, thus incorporating the for-
mat of the ‘new’ play into the Cantonese theatre. With the emergence of
the classical Hollywood narrational style in American cinema, Cantonese
playwrights began to pay attention to problems such as causal, spatial,
and temporal continuity and coherence in the construction of their plays
(Fan 2015: pp. 154–5; Liu 2012: pp. 20–145; Lai 2010: pp. 172–3 and
pp. 176–7).
Early Cantonese films were in fact produced by Cantonese theatre
troupes commissioned by playhouses. A more music-oriented version of
a play would be released on stage; a more dialogue-heavy version would
be released on screen; and eventually, the music would be released on the
phonograph. By the 1930s, while the cinema and the theatre became two
separate institutions and cultural practices in Shanghai, they were seen as
one single media environment in the Canton-Hong Kong region (Fan
2015, p. 157; Lai 2010, pp. 188–9). Hence, while Shanghai cinema began
to develop its own narrational paradigm under the influence of classical
Hollywood, Soviet montage, German expressionism, and French impres-
sionism, Hong Kong cinema remained highly ‘theatrical’ (Bao 2015,
pp. 153–264).
Today, only two Cantonese films from the 1930s survive: Nüxing zhi
guang [Neoising zi gwong or Light of Women, Ko Lei-heng 1937] and
Nanguo zimeihua [Naamgwok zimuifaa or Twin Sisters of the South, Lai
Ban 1939]. The syuzhets of both films are made up of episodes or acts.
Each act is causally open, but such openness is not necessarily answered or
closed immediately in the subsequent act. Rather, each open thread is con-
nected to a larger question or enigma that cannot be resolved unless the

larger problem of social inequity can be addressed. For example, in Twin
Sisters, two twin sisters brought up in two different social classes and eco-
nomic environments, Siu Tip and Tai Tip (Wu Tip-ying), fall in love with
Cantonese musician Lee Chung-ching (Ng Cho-fan). Siu Tip, brought up
as a country girl, is first engaged with Chung-ching, but then she has to
escape from the countryside to the city in order to evade an arranged mar-
riage. When Chung-ching begins to lose hope that he would find Siu Tip
again, he meets her twin sister Tai Tip, brought up by her wealthy foster
parents in Hong Kong, and falls in love with her. This love triangle cannot
be resolved via Chung-ching’s personal choice. Rather, by the end of the
film, Tai Tip leaves Hong Kong and her wealth and joins a theatre troupe
in the country to fight against the Japanese, and Siu Tip abandons her pas-
sive country upbringing and fights for her own future by joining Chung-
ching not as a dependent, but as a modern and progressive companion.
In these two films, each scene is considered a self-unified act that is
conveyed largely in long shots, three-quarter shots, and medium shots.
Actors are strategically positioned around the center of the frame, with
the protagonist of the scene and their antagonist forming an axis. The
dialectical relationship in each dramatic beat is emphasised either by deep
staging, or by means of using the tracking shot or pan, rather than edit-
ing. Neither of these two films employs the classical Hollywood shot/
reverse shot structure. Instead, characters would alternate their positions
in the frame, sometimes with camera movements, that would emulate the
effect of the shot/reverse shots. Furthermore, musical sequences are usu-
ally presentational, either with very minimal editing to cut from a long
shot to a medium shot in order to bring the spectators closer to the per-
former, or with a cutaway shot that shows a listener or an audience. For
example, in Twin Sisters, Tai Tip performs a Cantonese theatrical act on
stage. The scene is preceded by a placard indicating that what follows is a
xizhongxi (heizunghei or a play within a play). It then cuts to a long shot
of the auditorium with the theatrical act on stage. Once Tai Tip begins to
sing, the film cuts to a long shot of the stage itself and stays there for the
entire introduction of the aria. Then, when Tai Tip’s emotion begins to
escalate and when a second aria begins, the film cuts to a medium shot of
her singing directly to the camera. As film historian Po Fung points out,
early Cantonese cinema often includes musical numbers that unapologeti-
cally lift the audience from the narrative, and the mise-en-scène of the film
172 V. FAN
would remind the viewers that they are watching the actual performer—as
opposed to the character—singing to them directly (Fan 2015, p. 171).
Many Shanghai filmmakers and critics considered the Hong Kong
Cantonese cinema chuzhi lanzao (coarsely produced and excessively made),
primarily because these films speak vernacular (not literary) Cantonese, and
at first glance, they are merely filmed theatre (Anon 1937, no page; Fan
2015, p. 160; Chiksan 2009, no page). Yet some of these stylistic devices
were used into the 1950s despite the fact that many Cantonese filmmakers
were indeed trained in Shanghai and wrote critical essays on European film
theory and Hollywood cinema. Meanwhile, Cantonese film viewers would
often go to the movie theatres to watch a Cantonese feature in conjunc-
tion with a Hollywood or Mandarin feature. If one follows Bordwell and
Thompson through and through, the Cantonese audience seemed to be
culturally conditioned to employ a different cognitive map to understand
the Cantonese film. Alternatively, as I will demonstrate, some of these nar-
rational devices have the effect of negotiating some of the social, cultural,
and political affects and sensibilities specific to the viewers in the Canton-
Hong Kong region. As director Hu Peng (Wu Pang, 1909–2000) recalled,
when he brought films in the 1940s to the countryside in the Guangxi
(Gwangsi) region to entertain the Cantonese migrant workers, they often
preferred Cantonese rather than Mandarin productions, despite the fact
that the Mandarin films often had higher production values and cinematic
techniques. For writer Song Wanli (Sung Man-lei), the Cantonese films
aroused in these viewers a Guangdong jingshen (Gwongdung zingsan or
spirit of Guangdong or Gwangtung) as though they were these spectators’
lived realities re-presented in their own language (Sung 1938: no page).
Meanwhile, the Cantonese films produced in the immediate years after
the Pacific War (1941–45) by Grandview studio in San Francisco, with
the Chinese-American communities in the USA as their primary audience,
follow the classical Hollywood style of narration faithfully.
THE UNION FILM ENTERPRISE AND ITS PARAPRACTICAL

POETICS
Most Cantonese filmmakers left Hong Kong during the Pacific War
and joined local theatre troupes to perform patriotic plays or itinerant
projection teams to show newsreels or national defence features in the
countryside. Between 1941 and 1950, leftwing filmmakers led by direc-
tor, screenwriter, actor, and theorist Lu Dun (Lo Duen, 1911–2000) and
actor and producer Zhang Ying (Cheung Ying, 1919–84) made blatantly
pro-Communist films including Cihen mianmian wu jueqi [Cihan min-
min mou zyutkei or Everlasting Regret, Lo Duen 1948] and Zhujiang lei
[Zyugong leoi or Dawn Must Come, Wong Wai-yat 1950] in an attempt to
encourage the wartime émigrés from Mainland China to return to their
hometowns to contribute their efforts to the revolution, or to organise
mass movements in Hong Kong against the colonial government. Their
efforts stopped around 1950 and 1951, when it was apparent that the
Chinese Communist Party (CCP) had no intention of taking over the
sovereignty of Hong Kong, and when a group of union leaders in the
film industry were deported to the Mainland under the silent consent of
the Beijing government. After that, the new head of the Xinhua News
Agency, Liao Chengzhi (1908–1983), stopped funding socialist films in
Hong Kong. Instead, he offered a startup fund to film producer Liao
Yiyuan (1920–2002) to invest money in buying movie theatre chains and
distribution companies, thus allowing independent left-leaning producers
to finance and distribute their films without any state intervention (Chung
2004 [2007], pp. 111–19).
Even though the Union Film Enterprise is always discussed as the pro-
totypical leftwing film studio, it was not part of Liao Yiyuan’s distribution
and production circuit. The company was set up in 1953 as a coopera-
tive between actors, screenwriters, and directors including Ng Cho-fan
(1911–1993), Cheung Ying, Bai Yan (Pak Yin, 1920–1987), Huang Manli
(Mary Wong 1913–1998), Lo Duen, Wu Pang, and many familiar names
in the industry (Zhou 2011, pp. 9–18). Although Cheung Ying and Lo
Duen had connections with Liao Yiyuan, the company’s chief executive Ng
Cho-fan declined any interaction with the CCP. In fact, Union’s indepen-
dent status allowed their films to be distributed to pro-Kuomintang (KMT
or Nationalist Party) distributors in Southeast Asia and North America,
including the Shaw Brothers and Cathay (Law 2011, p. 44). Although
Union produced only 64 features in its 11 years of existence, its members
opened independent film companies under its auspices to produce Union-
styled films (Zhou 2011, p. 15). For example, Cheung Ying produced
and directed Diqihao siji [Daicathou sigei or Driver No. 7] in 1958 with
his own company Overseas Chinese Films, with the aim of promoting the
achievement of ‘New China’ in the North American and Southeast Asian
markets. Meanwhile, Lo Duen also directed and produced films for the
Sun Luen film company, which received funding from Liao Yiyuan.
174 V. FAN
Although Union films belong to a variety of genres, the studio is best

known for making petit urbanite films, with plots developed around an
ensemble of lower-middle-class and working-class characters. In addition,
their films also show a deliberate effort to give narrative agency to margin-
alised women, including dance hostesses and home workers. These char-
acters can be roughly grouped into: (1) émigrés from Mainland China
who settled down in Hong Kong before or during the Sino-Japanese War
(1937–1945), including drivers, day labourers, and seamen (who were
the most unionised groups in Hong Kong during the 1950s), intellectuals
and educators, dance hostesses, and home workers, (2) new immigrants
from the Mainland who settled down in Hong Kong after 1949, includ-
ing factory workers and house helpers, (3) entrepreneurs from Shanghai
including manufacturers, company managers, accountants, and real-estate
investors. Many of these plots are set in tenement buildings that are rented
to a baozu [baaozou or rent contractor], who subdivides each floor with
wooden panels and curtains into makeshift units and sublets them to indi-
vidual tenants. This is a living arrangement easily recognisable by many
lower-middle-class and working-class viewers in Hong Kong during the
1950s; it also allows characters from all walks of life to interact with each
other and condenses social, political, and economic conflicts in a densely
populated and claustrophobic space. In addition, the rent contractor, who
always appears to be an evil character who demands money from those
who struggle for their living, is in fact constantly under the pressure of the
landlord or developer, thus allowing the viewers to trace all the social ills
to the ‘landowner’ as the ultimate symbolic substitute of capitalism.
That being said, none of these petit urbanite films can be considered
anti-capitalist. As Shek Kei (1966, p. 11) pointed out, most Hong Kong
leftwing films leave the capitalist system untouched. In fact, these films
merely suggest that through hard labour, communal loyalty, and indi-
vidual perseverance, these lower-class characters can overcome monetary
issues and demonstrate their strength against the upper middle class. In
these films, the Canton region is still fondly called or remembered as one’s
xiangxia (hoenghaa or hometown or home village), and the fact that one
cannot cross the border to return to one’s hometown is effectively erased.
By the same token, mobility between Hong Kong and ‘home’ is either
actively denied (e.g. nobody travels or expresses any intention to return
‘home’), normalised (e.g. some characters can somehow travel between
Canton and Hong Kong as though border control did not exist), or
retarded and delayed until a film’s denouement. For example, in the end-
ing of Xilu Xiang [Sailou Coeng or The Kid, Fung Fung, 1950], the father
walks his children along the railroad back to their ‘home’ as an imaginary
resolution to the social ills and poverty that the family suffered.
The lack of critical sensibility against capitalism and the deliberate
amnesia of the Mainland–Hong Kong divide can be attributed to colo-
nial censorship, which stipulated that cinema may not instigate politically
sensitive subject matter or express any blatant political position for fear
of generating ‘uncomfortable feelings’ in the ‘neighboring country’ (i.e.
China). However, these unique features of the Hong Kong Cantonese cin-
ema can also be understood as a form of parapraxis. As Thomas Elsaesser
(2009, pp. 190–2) argues, the ‘para’ in parapraxis suggests that each failed
praxis both instantiates a failure and actively performs the failure in order
to address, mediate, and reconstruct a deeper trauma. For example, if I
promise you a cup of coffee, and I bring you a glass of whisky instead, I
fail to perform what I promised to offer. At the same time, I am also per-
forming the failure to perform a repressed desire that had long or recently
traumatised me (e.g. I have always wanted to have whisky with my father,
but he always denied me the opportunity). As I mentioned in the begin-
ning of this chapter, in the 1950s, leftwing intellectuals and cultural pro-
ducers have failed—in the sense that they somehow chose to stay in Hong
Kong instead of returning ‘home’ to construct ‘New China’—to fight
against capitalism, and the closing of the border between Hong Kong
and Shumchun also made it impossible for them to return ‘home’ with-
out risking their status quo. These films allow them to re-experience the
affects of their failed performance. Yet by performing their collective sense
of failure, the leftwing filmmakers and the left-leaning audience regained a
sense of sociopolitical agency.
IN THE FACE OF DEMOLITION: A NARRATIVE

OF REEDUCATION
In the Face of Demolition is based on a Cantonese-language novel written

by Zhang Wangyun (Cheung Mong-wan, 1960 [1930]), Renhai leihen
[Janhoi leoihan or Traces of tears in the sea of humans], serialised in the
Hong Kong newspaper Dazhong bao [Tai Chung Pou] in the 1930s. In
fact, director Li Tie (Lee Tit, 1913–1997) adapted the novel as a film
under the same title in 1940. The original novel is about a young writer
Chow Ping moving into a living unit in a tenement building in Lan Kwai
176 V. FAN
Fong, a poor alley in Central (known today as the most vibrant bar district
in the city), and witnesses how his housemates struggle with their lives.
The 1953 film is by no means a faithful adaptation of the literary work.
It merely borrows the novel’s settings, plot elements, and character types
and migrates them to postwar Hong Kong.
As critic Kong Tong (1953, no page) argued, In the Face of Demolition
centers around Law Ming (Cheung Ying), a young teacher who moves
into an old building in Fai Fu Alley (literally, an alley where everybody
gets rich quickly). Law Ming falls in love with one of his housemates Pak
Ying (Tsi Lo-lin), a dance hostess, but soon he loses his job. Pak Ying
introduces Law Ming to a man who claims himself to be a newspaper edi-
tor, yet he turns out to be a fake. Angered by Pak Ying and frustrated by
life, Law Ming asks his uncle, the landlord of his building, to give him a
job as a rent collector. Nonetheless, once Law Ming assumes his position,
his uncle presses him to collect all the rent within three days, because the
building turns out to be due for demolition. Not being able to tell his
housemates about this, Law Ming pressures his housemates to pay their
rent as though he were some greedy and abusive figure. In order to pay
his rent, a poor housemate Tam Yi Suk (Wong Cho-san) sells his blood.
When the doctor refuses to take more of his blood, Yi Suk takes up a
job as a day labourer and breaks his back. In a stormy evening, Yi Suk
passes away. While all the housemates are worrying about getting money
for his funeral, the wife of Law Ming’s best housemate Leung Wai (Ng
Cho-fan) suffers from obstructed labour and must be sent to the hospital.
Law Ming goes to his uncle, demands that the building be repaired, quits
his job, and asks for his last month’s pay. He then runs to the hospital to
give his salary to Wai. Not only that, since his blood type matches Mrs.
Leung’s, Law Ming donates his blood to save her. Eventually, Park Ying
forgives Law and the film ends with the housemates promising that they
will always help each other: ‘All for One; One for All!’
As Kong Tong pointed out, the film is by no means driven by the goals
and obstacles of Law Ming alone. In fact, Law seems to lack any goal in
life besides having a passing fancy in the first half of the film to become a
writer. His aimlessness in life is matched with his temperamental manner
(as Ying points out, he is like a thermometer: ‘warm and passionate when
everything goes well, and cold and mean when things go against his will’)
and his penchant to ‘save face’, a negative prototype of a middle-class
intellectual. In this sense, Kong Tong (1953, no page) argues that the
romance between Law Ming and Pak Ying is a highly developed subplot
that was put there to satisfy the increasing demand from the audience for
a Hollywood-styled narrative development.
The syuzhet of In the Face of Demolition is indeed highly unusual from
the classical Hollywood perspective. Unlike a classical Hollywood film, in
which the opening sequence is about the disturbance of a status quo, In
the Face of Demolition reverses this process by presenting a status quo that
has already been disturbed. In fact, this opening sequence, lasting as long
as 14 minutes of screen time, condenses and rehearses the overall narrative
by allowing the viewers to see how all the key characters gradually come
together as a collective. In addition, as Lam Nin-tung (1978, p. 4; Fan
2015, p. 168) argues, the organisation of the narration’s devices, especially
its camerawork and editing, is based on the principle of social dialectics:
• The extensive use of the long shot and medium long shot to preserve
the integrity of the dramatic space.
• The use of deep-focus composition to create dramatic tension
between characters within a unified dramatic space.
The opening sequence condenses a constellation of signs that can be

used as a user’s manual for reading the rest of the film (Elsaesser 2012,
pp. 115–17). The camera then tilts down to a plaque that reads ‘Fai Fu
Alley’ (again, an alley where everyone gets rich quickly). It then continues
to tilt down until it shows an arch opening onto the housing complex. The
plaque ‘Fai Fu Alley’ in fact condenses the problem that disturbs the status
quo of the house: the principle of ‘everyone getting rich quickly’ under
capitalism. Later on in the film, Law Ming and Wai both describe Fai Fu
Alley as a rotten house that leaks everywhere, and that someone may get
killed if the landlord fails to repair it. Eventually, the landlord ignores the
problem, as the building is due for demolition. Yet before he knocks it
down, the building collapses in a typhoon and kills Manager Wong (Lo
Duen), a smalltime moneylender who used to be employed in a big trading
company. The plaque Fai Fu Alley therefore sets up a metonym between
the building and the capitalist society as a whole. Nonetheless, when the
building finally collapses, it kills neither the working-class collective who
manages to leave the house in time, nor the landlord who chooses to
ignore the problem. Rather, it kills Manager Wong, an opportunist who is
both a victim and an exploiter under the capitalist system. If we push this
idea further, the Cantonese title Ngailou cheonhiu, literally, the dawn of
spring in a condemned building, puts the emphasis on the possibility of
178 V. FAN
retaining, renewing, and overcoming the condemned sociopolitical posi-

tions, rather than taking an active agency to change it.
A set of dialectical relationships is then established through camerawork
and editing. When the camera continues to tilt down, a rickshaw comes
into the frame from the right and enters the arch. The camera dollies
slowly to the right, away from the arch; it then cuts to a medium long shot
of the rickshaw stopping in front of a building. Law Ming gets off the rick-
shaw and instructs the driver to take his belongings up to his new place.
As Law Ming knocks on the door of the apartment, which is located on
the second floor, the film cuts to a long shot inside, and we see Manager
Wong walking down the stairs from the third floor to open the door for
him. As Law Ming enters, he steps on Wong’s toes, thus suggesting that
the arrival of Law Ming will stir up a series of problems that are going to
‘step on Manager Wong’s toes’.
From this point on, the film begins to convey the story by means of
montage. In fact, the entire opening segment consists of 108 shots that
compartmentalise individual characters in their own individuated spaces,
and editing juxtaposes these discrete spaces against each other. For
instance, when Law Ming walks up the stairs, the film alternates between
medium shots and medium close-ups of Law Ming and Manager Wong.
Law Ming is often shot from a high angle, and Wong is often seen from
a low angle, thus foregrounding the sense of superiority Manager Wong
has over Law. Meanwhile, Wong is rarely seen together in the same shot
as Law. The only exception is a low angle shot of Wong flirting with Fong
in the foreground as they stand at the bottom of the staircase, while Law
stands in the background on the top of the staircase paying the rickshaw
driver, which verticalises the dialectical relationship between Law Ming
and Wong by juxtaposing them in the same shot (see Fig. 9.1).
Besides the dialectics between Law Ming and Manager Wong, this
opening sequence also establishes a dialectic between Law Ming and Yuk
Fong. As Law Ming gets to the top of the stairs, he accidentally kicks over
the shrine of the menguan (mungwun or door-keeping god). When Yuk
Fong finds out, she warns Law Ming that he is going to have stomach
ache that evening as a divine punishment. In this sense, Law Ming as an
educated man is juxtaposed against Yuk Fong, a new immigrant from ‘the
country’, who is still bound by feudalistic superstitions. Eventually, Fong
will be raped by her cousin-in-law and become his concubine: a ‘victim’
of feudalism.
Fig. 9.1 Verticalisation of the social dialectics in In the Face of Demolition
Nonetheless, far from being a progressive educator, Law Ming is soon

revealed to be an intruder. After Law Ming first meets Yuk Fong, he asks
her to look for the rent contractor Sam Ku. The film cuts to the kitchen,
where Sam Ku is doing laundry. Up until this point, and in fact, through-
out the entire film, there is no establishing shot giving the viewers a sense
of the overall topography of the building. Rather, the film relies on the
spectators’ familiarity with the generic layout of a tenement building,
either through their lived experience, or through other films belonging to
the same genre. After a brief introduction to Sam Ku’s husband Pat Sin
(Ko Lo-chuen), who performs kung fu with his frighteningly loud voice
(a comic relief), the film cuts to the room of Yi Suk, where Sam Ku asks
him and his family to move in order to vacate the space for Law Ming. Yi
Suk is cooperative, though Sam Ku is impatient. In this sub-segment, the
film alternates between a three-quarter shot of the entire room with Yi Suk
and his family, and medium shots to medium close-ups of Sam Ku, thus
foregrounding the dialectical relationship between the rent contractor and
the tenants by separating them into two spaces.
When Yi Suk eventually moves out of the room, he meets Law Ming
in a claustrophobic corridor. Here, in a long shot, Law apologises for
making Yi Suk move. Yi Suk then blames the problem on money, an issue
180 V. FAN
that will recur in Yi Suk’s personal trajectory. From this point on, the
film begins to frame characters that occupy contesting social positions in
the same shot, and the dialectical relationships among them are conveyed
through staging. Strategically, this technique allows the film to pack more
and more characters into the same frame as they begin to develop a sense
of in-group loyalty. As this sub-segment continues, for example, Yi Suk
and his family, Sam Ku and Pat Sin gather together around the staircase
as Yi Suk begs Sam Ku to rent them the space under the staircase, where
they can set up a bed. The space, however, has been occupied by Manager
Wong for Yuk Fong. Yuk Fong joins the group and agrees to offer that
space to Yi Suk. The wife of Manager Wong then appears and scolds Yuk
Fong for giving up her bed. Here, the film alternates between the group
and Mrs. Wong, thus allowing the viewers to sense that an opposition has
now developed between the moneylender’s wife and the rest of the ten-
ants. As the argument escalates, Wai returns home from work and joins the
rest of the group. Meanwhile, Pak Ying also appears on top of her panel
and helps negotiate. The tenants convince Sam Ku to rent the space to Yi
Suk, although Yi Suk reveals to Wai that he does not have a bed anymore.
Eventually, Law Ming steps up, joins the group and offers his bed to Yi
Suk and his family. The entire segment eventually ends with a long shot of
the whole group celebrating their solidarity with the film’s slogan: ‘All for
One; One for All!’ (see Fig. 9.2).
Fig. 9.2 Development of in-group solidarity through gradually packing dialecti-

cally conflicting characters into a single long shot
After this moment, Wai introduces Law Ming to all the housemates.
However, a special emphasis is given to the relationship between Wai and
Law Ming, and then Law Ming and Pak Ying. Wai originally takes Law
Ming as a bully. However, when Law Ming agrees to offer his bed to Yi
Suk, Wai becomes deferential. In a series of shot/reverse shots, Wai asks
Law his name. Out of respect, Law asks Wai to introduce himself first.
Wai tells Law Ming his name, and calls himself a ‘taxi driver’, thus fore-
grounding his social class as a unionised worker. Law Ming then intro-
duces himself as a teacher, which inspires Wai’s admiration. Later on,
when Wai introduces Law Ming to Pak Ying, Wai tries to describe the
character ‘Ying’ to Law Ming. However, while he remembers ‘Pak’ as in
‘white’, he cannot remember how to write ‘Ying’ (as in ‘crystal’). When
Law Ming steps in by saying, ‘Ying as in crystal’, Wai utters, ‘Please shake
hands, you intellectuals.’ The weighted introduction between Law Ming
and Pak Ying is important not only because their romantic subplot will be
the first plotline that will be developed, but also because they are the two
educated ones that stand out from the rest of the tenants. In fact, as Wai
later criticises over and over again in the film, Law Ming and Pak Ying’s
problem is their pride, that they want to ‘save face’. In some ways, the film
is about how the solidarity and in-group loyalty of the tenants eventually
reeducate Law Ming, and to a lesser extent, Pak Ying, to become coopera-
tive members of the masses.
To a certain degree, In the Face of Demolition is an ensemble piece
interwoven by the trajectories of several key characters. I do not use the
term ‘plot’ here, as most spectators would come out of the film not nec-
essarily remembering a ‘story’. Rather, they would remember a set of
characters—or prototypes—whom they may recognise among their fel-
low petit urbanites, and who make decisions with moral characters that
they may remember and emulate, or detest. Besides Law Ming, the film
features Wai prominently as a down-to-earth taxi driver who always makes
quick, effective, and morally fair decisions in times of crisis and hardship.
In the entire film, Wai does not have a goal, and he overcomes his difficul-
ties with hard work (e.g. when he loses his job, he becomes a day labourer;
when he needs money for his wife, he goes out and looks for it from his
friends). He serves as a moral compass for the entire household. Yi Suk
is almost kicked out of the house in the beginning because he fails to pay
rent, and twice in the film, he declares that he would persevere in order
to make money, pay for his debts, and bring up his children. In this sense,
he does have an objective, but not in the form of any personal determina-
182 V. FAN
tion. Rather, he is driven to take action because of the social inequity that
entraps him. If one looks at all the rest of the characters—Manager Wong
and his wife who only care about money, Yuk Fong, the kindhearted
cousin-in-law of Wong who is being raped by Wong, and Sam Ku, who is
a prototype of a rent contractor who qishan pa’e (heisin paangok or bullies
the kind and fears the bigger bullies)—none of them has what classical
Hollywood cinema would call an individual goal.
Based on the observation of film scholar Lam Nin-tung, I have argued
elsewhere that the Hong Kong Cantonese cinema has
a narrative structure that is not based on a cause-and-effect chain triggered

by a character’s want and obstacles. Rather, the film is constructed as a
series of chronological events, each being caused by the larger socio-political
conditions that leave the characters with no political agency to change the
situation. Meanwhile, the seriousness of the effects of social oppression
would escalate until the film calls for the audience’s own reflection upon
their actual sociopolitical conditions (Fan 2015, p. 169).
However, without any goal or objective, the film still works upon what
Viktor Sklovsky would call narrative retardation—the deferral and dis-
placement of the key narrative goal. The narrative question is not how
these key characters achieve or fail to achieve their individual objectives.
Rather, one is drawn to follow the narrative steps in order to find out what
it takes for these housemates to realise that they have a collective goal. If
a Bildungsroman is about a girl or boy being educated through facts of
life into a self-made bourgeois individual, the Cantonese film in the 1950s
is about a group of individuals, who are already weathered by the hard-
ship of life, reeducating themselves as a community that can persevere and
grow under capitalism. In short, these individuals have already failed to
perform; but by performing and rehearsing their individual failures, the
spectators began to develop a sense of affinity with these individuals and
achieve a form of in-group loyalty with the marginalised classes.
CONCLUSION
In the Face of Demolition demonstrates a poetics distinct from the classical
Hollywood style. Given that a director like Lee Tit was very familiar with the
Hollywood narrational paradigm, a case can be made that the Cantonese
filmmakers in the 1950s consciously modified the classical Hollywood sys-
tem. Yet, this cannot explain why the Hong Kong Cantonese-speaking
audience, equally well-versed in the classical Hollywood style, demanded
or welcomed such modification in the first place. In this sense, Miriam
Hansen’s suggestion of seeing the cinema as a public sphere provides a
better understanding of why Cantonese cinema developed such a distinc-
tive narrational style. The Union films, I argue, can be considered as a
form of parapraxis (Freudian slip). These films actively perform the failure
of the leftwing intellectuals to perform in a colonial space where they had
no active sociopolitical agency, and their collective failure to even con-
sider returning ‘home’ and reconstruct the ‘national’ space. These films
offer narratives in which intellectuals are reeducated and reintegrated into
the masses, thus offering the spectators a ‘second chance’ to rehearse the
possibility of having a political agency to activate social change. In the
Face of Demolition lays out the dialectical relationships among workers,
intellectuals, and opportunists who are in fact all victims of capitalism.
These failed figures are dialectically juxtaposed in the film through cam-
erawork and editing, yet the film never offers any imaginary resolution to
their problems besides generating a sense of in-group loyalty and mutual
dependency among them. Their problems are actively displaced into the
abstract, yet affectively appealing slogan, ‘All for One; One for All!’ In
a sense, this memorable slogan is the emblem of the leftwing intellectu-
als’ collective failure. However, by performing such failure and rehearsing
their collective trauma, Cantonese cinema, or more specifically, the Union
films, kept alive a social ideal that was rapidly dwindling in postwar Hong
Kong.
CHAPTER 10
China as Documentary: Some Basic

Questions (Inspired by Michelangelo
Antonioni and Jia Zhangke)
Rey Chow
A couple of years ago I was invited to speak in a small forum on Chinese

cinema, of all places in the world, in Vienna. During the discussions, I
noticed an interesting divide between those who were supposedly ‘native
informants’ and those who had come to Chinese cinema as foreign
observers. We could be looking at the same images, but the ways that we
talked about them were usually different, or presumed to be different by
both sides. On the side of the foreign observers, a tone of hesitancy and
uncertainty, a need to defer judgement and to respect the more authori-
tative views of the natives was in the air. On the side of the native infor-
mants, the predominant tone was a mixture of anxiety and paranoia and,
most of all, apprehension about the possibility of being orientalised and
misunderstood.
The gap between these positions is what interests me and where I
would locate the future of a visual research area such as Chinese cinema
in the age of hypermediality. Paradoxically, this widening gap is the result
of globalisation, when increasingly close contact between the two sides
has not only become constant, but is virtually unavoidable. As bearers of
R. Chow ( )
Duke University, Durham, NC, USA

DOI 10.1057/978-1-137-55309-6_10
186 R. CHOW
specific attitudes and perspectives, so-called native informants and foreign

observers are partners in a long historical relationship, the complexity of
which lies not in either position alone but rather in their entanglement.
Ultimately, the two positions may be thought of as two fictional charac-
ters. Much like the manner in which Jacques Lacan has described the situ-
ation between men and women—that there is no sexual relation—perhaps
a similar hypothesis can also be ventured in this case, namely, that in the
end there is no relation between the native informant and the foreign
observer? Yet, like men and women, native informant and foreign observer
are forever thrown together and must coexist, collaborate and communi-
cate as though there were a relation somewhere. How to proceed?
Personally, what I have found most challenging when writing about
Chinese cinema is exactly this fantasy relationship (or non-relationship)
between the two characters, and in particular the implicit moral judgement
that tends to shape and colour it. This implicit judgement is part of the
epistemic burden called ‘China and the West’, in which the two sides have
been linked by a lack of equality since the mid-nineteenth century, when
China found itself to be the inferior partner who had to accept defeat as
well as to strengthen, reform and update itself. In the post-colonial world,
this epistemic burden, typically borne by those on the side of the colo-
nised, has fundamentally structured the way that images continue to be
produced and consumed, so much so that even when faced with the same
images, those who identify with the native informant typically see very dif-
ferent things from those who identify with the foreign observer.
This gap between native informant and foreign observer is perhaps
most pronounced when the images involved are of a documentary nature,
whereby the claim to the real that has been associated with the photo-
graphic medium since its inception becomes part of the conceptual prob-
lem of what counts as the documented content.1 As mutuality, reciprocity,
agreement and consent—all those ideals of civilised sharing—seem increas-
ingly sensitive and elusive in the course of global interactions, how to
talk cross-culturally about seemingly straightforward-looking images has
become something of a formidable task. Often those imagistic elements
that readily seem to convey a sense of the real are exactly the places where
the identity of the image itself becomes contentious rather than secure. If,
as has been meticulously discussed by scholars of photography, the photo-
graphic image’s claim to reality has become indeterminate, paradoxically,
as the technical capacities of the medium itself become perfected,2 a visual
encounter across cultures can only complicate the situation further because,
CHINA AS DOCUMENTARY: SOME BASIC QUESTIONS... 187
as Homay King writes, ‘questions of ownership and self-possession are to

some extent inseparable from the problem of cross-cultural representa-
tion’ (2010: 103).3 As a point of departure, let me turn to a controversial
documentary, made in and of contemporary China at a time when it was
still largely closed to the rest of the world.
A CONTESTED DOCUMENTARY ABOUT CHINA

In 1972, at the invitation of the then-Chinese premier Zhou Enlai, the
Italian director Michelangelo Antonioni made a documentary film known
as Chung Kuo (Cina).4 Approaching his subject matter in a mixture of
mystification and humility, believing that he is observing and record-
ing the daily lives of Chinese people (see his remarks as quoted in Sun,
2009: 49), Antonioni takes us on a 3-hour 40-minute tour of scenes rang-
ing from the extraordinary to the banal. For example, we are shown a
woman giving birth by caesarian section with the help of acupuncture
and no anaesthesia, acrobatic performances, ordinary people shopping for
daily food and cooking at home, men practising taiji in the park, factory
workers weaving cotton cloth at mechanised looms and men and women
chatting over tea at a teahouse. From today’s standpoint, when China is
caught up in a relentless series of urban demolitions and reconstructions,
Antonioni’s film, made during the Cultural Revolution, comes across as a
fascinating piece of ethnography, offering glimpses of a way of life and a
pace of life that seem to have all but vanished.5
Antonioni’s camerawork aroused strong feelings of disapproval on the
side of Chinese viewers from the start (the film was banned in China until
2004). As Umberto Eco sums it up in good humour, after the film’s sen-
sationally awkward showing at the 1974 Venice Biennal Exposition: ‘the
anti-Fascist artist who went to China inspired by affection and respect …
found himself accused of being a Fascist, a reactionary in the pay of Soviet
revisionism and American imperialism, hated by 800 million persons’ (Eco
1977: 9).6 Zhu Qiansheng, a young party cadre assigned to accompany
Antonioni on his shoot, recalls vividly decades later (in a teahouse scene
in Jia Zhangke’s film Hai shang chuanqi/I Wish I Knew, 2010, a scene set
in such a manner as to reconnect us to a similar scene in Chung Kuo): ‘It
struck me that he was filming a lot of bad things, things that reflected our
backwardness. It seemed totally unfair.’ Afterwards, according to Zhu,
whenever he and his coworkers revisited locations that had been shown in
the film, they were criticised as traitors and counter-revolutionaries.
188 R. CHOW
The intense reactions of the Chinese (including those of the Gang of

Four) stage the encounter between native informant and foreign observer
in familiar terms, capturing the idealism, risks, ironies and conflicts embed-
ded in the phrase ‘looking after Europe’. While for Antonioni the film
suggests ‘the search for China as a potential utopia by the frenetic, neu-
rotic West’, against which even something like poverty may take on lyrical
qualities of simplicity and austerity (see Eco 1977: 10), for the Chinese, a
Western filmmaker who knows little about China’s history and its people
has presented China in a distorting, callously cold manner. How is this
so? His images denigrate the Chinese by portraying them as shabby and
backward; they highlight things in China that are old and dilapidated; they
catch people in indecorous and embarrassing moments, such as when they
are picking their noses, going to the bathroom and so forth.
In her classic study, On Photography (1978), Susan Sontag offers a
thoughtful discussion of this confrontation, defending the subtle, pluralist
and modernist aesthetic meanings produced by Antonioni’s camerawork,
while also attempting to rationalise the Chinese difference (see Sontag
1978).7 In the main, Sontag accounts for the Chinese reactions by draw-
ing attention to the culturally different attitudes towards photography,
corroborating Eco’s view that ‘when political debate and artistic represen-
tation involve different cultures on a worldwide scale, art and politics are
also mediated by anthropology and thus by semiology’ (Eco 1977: 9). As
Sontag states:
While for us photography is intimately connected with discontinuous ways

of seeing (the point is precisely to see the whole by means of a part – an
arresting detail, a striking way of cropping), in China it is only connected
with continuity … In China, where no space is left over from politics and
moralism for expressions of aesthetic sensibility, only some things are to be
photographed and only in certain ways. (Sontag 1999[1978]: 88, 92)
Written in the 1970s, Sontag’s comments are consistent with an enlight-

ened Euro-American understanding of the repressive constraints placed
by communist orthodoxy on all forms of expression, including images.
Although this understanding leads to a pretty fair assessment (supported,
notably, by an equally sceptical view about capitalism’s insatiable demand
for more images), something else in Sontag’s discussion intimates a mode
of cultural response that goes beyond the immediate context of Chinese
communism. For example, the following remarks position the Chinese as

primitive:
There is no interest in catching a subject in movement. This is, one sup-

poses, partly because of certain old conventions of decorum in conduct and
imagery. And it is the characteristic of visual taste of those at the first stage of
camera culture, when the image is defined as something that can be stolen
from its owner; thus, Antonioni was reproached for ‘forcibly taking shots
against people’s wishes’, like ‘a thief’. (Sontag 1999[1978]: 89; emphasis
added)
By thus assimilating the Chinese to those at the first stage of camera cul-
ture—an association which, by the mid-twentieth century, is hardly his-
torically accurate—Sontag has shifted the discussion about Antonioni’s
undertaking to the plane of an ethnographic encounter, in which behav-
iour around the photographic image becomes a way of charting a culture’s
degree of advancement and sophistication. The Chinese, Sontag points
out, do not share the ‘good manners of a camera culture’ in which
one is supposed to pretend not to notice when one is being photographed

by a stranger in a public place as long as the photographer stays at a discreet
distance – that is, one is supposed neither to forbid the picture-taking nor to
start posing. (Sontag 1999[1978]: 89)
In China, ‘taking pictures is always a ritual; it always involves posing and,

necessarily, consent’ (Sontag 1999[1978]: 89). If one follows people
around and shoots their picture while they are unaware of one’s intention
to do so, one is depriving people of ‘their right to pose, in order to look
their best’ (Sontag 1999[1978]: 89).
In her ethnographic reading, Sontag reveals a characteristically mod-
ernist logic based on an implicit narrative of development in which Europe
and North America stand in the forefront, while the rest of the world is
arrested or frozen at an earlier stage.8 Thus, what is intended as a study of
photography is entangled unwittingly with a different type of discourse,
in which photography itself—as a practice not only among photographers,
but also among those who are photographed—is turned into a yardstick for
cross-cultural comparison and evaluation. The Chinese difference in this
case is remarkable because it exposes the neoliberal limits of this modern-
ist logic: however generous and tolerant such a logic is of other cultures,
190 R. CHOW
in the end it cannot but reduce them to an earlier and out-moded—that

is to say, less developed—condition of being. Photographic realism in this
instance has become another name for civilisation.
THE FRONTAL POSE

What if we were not to follow this modernist, developmental logic? How
otherwise to account for the Chinese difference, the Chinese claim to the
real? To begin to do so, it would be necessary to argue a different premise
about photography, including questioning the (desire for) pictorial realism
that lies at its historical foundation. Without attributing primitivism to the
Chinese, let me suggest that we begin with a different set of assumptions
and practices, whereby we do not treat the photographic image—perhaps
the image in general—as natural or free.
The cross-cultural photographic and documentary situation, then,
brings to the fore a seasoned politics of combat and antagonism. Indeed,
as historians tell us, the arrival of photography was greeted with resistance
in China, as it was in many parts of the world, including Europe, because
people felt aggression from a foreign apparatus, which many believed was
robbing them of something precious, like their soul or spirit. Although sci-
ence and technology have distanced us from such superstitions, something
of those reactions to being photographed still seemed to be evident in the
way that Chinese people behaved around the camera in 1972, but is this
necessarily a sign of underdevelopment, as Sontag seems to suggest with
her phrase ‘first stage of camera culture’? Rather than a friendly process
of enquiry based on a poetic vision, as was intended by Antonioni and
his crew, the situation involving the camera lens was approached by the
Chinese as containing the risk of a possible violation, an assault. The way
they responded was meant to return, and to counter, the aggression that
they sensed in the camera’s gaze. The act and the art of striking a pose in a
dignified manner may be residual practices from earlier times, but they are
also eminently modern, indeed contemporary, as many of us can tell from
our own personal experiences with being photographed. Accordingly,
instead of letting down one’s guard before the camera, one needs to con-
sciously perform, showing (off) one’s most positive features. Rather than
being primitive in the derogatory sense, the Chinese reactions indicate that
photography is about social space—in particular, how social space is, or
should be, negotiated in relation to one’s surroundings and fellow human
beings, including in particular the photographer, who is armed.
At the root of this incident of cultural confrontation is the problematic

of photographic realism, of what is considered real in a picture. Antonioni
wished with modesty and goodwill to capture China in as natural a man-
ner as possible, because such naturalism—which in this instance is another
name for realism—has been assumed since photography’s arrival to be
what the photographic image is all about. Consider the snapshot: the pic-
ture taken when people are in motion, engaged in their activities, unaware
of or simply not heeding the camera’s existence, is often condoned as a
real or natural rendering. One could go so far as to say that in spite of its
technicality as a skilled act of capture, the snapshot has been thoroughly
fetishised in existential terms. What the snapshot releases, it is believed, is
the other—the object of capture—in an unrehearsed and unself-conscious
state, the state when it is most (freely and spontaneously) ‘itself’. If such
belief in photographic realism may be compared to a kind of religion (one
that replaces God’s gaze with a machinic gaze), the Chinese distrust of the
camera may be understood as healthy scepticism. Instead of treating and
trusting the camera’s gaze as benignly transparent, the Chinese continue
to treat it as an alien object before which they tend to feel guarded, self-
conscious and in need of protection.
Pierre Bourdieu’s (1999[1965]) insights into the social norms and
conventions that structure photographic practices are uniquely helpful
at this juncture.9 Perhaps unsurprisingly, being mindful of class distinc-
tions, Bourdieu is sensitive to the meanings of the pose in front of the
camera. Unlike Sontag, for whom the pose suggests a rehearsed or pre-
programmed relation to photographic realism as it has been developing
in the capitalist West, Bourdieu sees the pose as a type of embodiment
that is inextricable from social politics: ‘the meaning of the pose adopted
for the photograph can only be understood with relation to the sym-
bolic system in which it has its place and which … defines the behaviour
and manners suitable for [one’s] relations with other people’ (Bourdieu
1999[1965]: 166). Bourdieu’s grasp of the power dynamics involved in
photography leads him to remark: ‘To strike a pose is to offer oneself to be
captured in a posture which is not and which does not seek to be “natu-
ral” … Striking a pose means respecting oneself and demanding respect’
(Bourdieu 1999[1965]: 166). In the realm of photography, whatever may
come across as natural is already ‘a cultural ideal which must be created
before it can be captured’ (Bourdieu 1999[1965]: 166).
In particular, Bourdieu discusses frontality—the posture of carry-
ing oneself face on, forehead held high, head straight, clothes in good
192 R. CHOW
order and so forth—as a deep-rooted cultural value that, in photogra-

phy, is expressed most readily in the conventional genre of the portrait.
According to Bourdieu, people of the lower classes such as peasants are
often ill at ease when they are photographed, because they have been con-
demned to internalise the pejorative image that society has of them and,
as a result, have a poor relationship to their own bodies. Thus, to pose
consciously amounts to an attempt to transform oneself into an agreeable
image (like wearing a beautiful mask). ‘Like respect for etiquette, frontal-
ity is a means of effecting one’s own objectification: offering a regulated
image of oneself is a way of imposing the rules of one’s own perception’
(Bourdieu 1999[1965]: 168). The upright body and alert facial expres-
sion are ceremonial signs of composure, orderliness and propriety. Above
all, they signal a readiness to stand up to a challenge.
Returning to Antonioni’s film, we can now say that it is precisely fron-
tality understood in these terms—not simply as a face and a body, but also
as the chance to strike a pose—that Antonioni’s camerawork does not per-
mit his objects. Instead, the camera insists on its own existential freedom,
its own enjoyment at roaming around in unnoticed spaces, looking at
dirty alleys, decrepit living compounds, peeling walls, poor livelihoods and
so forth, which rather deflate the potency of the frontal pose, of the act
and the art of (con)fronting. Deprived of the possibility of reciprocation
by none other than the camera’s caprice (and, shall we say, digression),10
China is incapacitated from meeting the camera’s gaze head-on in a ritual
of combat. To Antonioni’s Chinese critics, this denial of the frontal pose
comes across as an affront and a humiliation.
PHOTOGRAPHIC OR DOCUMENTARY REALISM AS LOSS OR

THE PLACE OF THE SUBJECT
When the conflict between Antonioni and his Chinese audience is recast
in these terms, the question that surfaces is a basic and double one: what
constitutes ‘the real’ in a documentary, and who is to say so? It seems fair
to say that whereas for European directors such as Antonioni, documen-
tary realism is the outcome of not letting the camera appear intrusive—
of letting the object unfold ‘spontaneously’—for Antonioni’s Chinese
audience, realism has to do with an aesthetic-cum-political arrangement,
whereby signifiers such as a body must be carefully anchored to a desired
signified. As in the case of socialist realism in the realm of literary writing,
documentary realism has to do with securely fastened social and collective

meanings.
In his semiological studies of mass culture, in particular of commodi-
fied images such as those featured in advertisements, the French literary
critic Roland Barthes has deconstructed such ‘fastening’ of meanings by
way of the opposition (derived from structural linguistics) between deno-
tation and connotation. In his analyses, Barthes consistently demonstrates
the untenability of a pure, first level of signification—in short, the unten-
ability of a denotative or indexical meaning that is assumed to be fixed in
such a manner as to pre-exist connotation (see Barthes 1977[1964]).11
Notably, in the book that he devoted to photography, Camera Lucida,
Barthes becomes invested in a different kind of critical focus. Instead of
being concerned, as he had been in the previous stages of his career, with
the controversy over denotation and connotation, he proposes an alterna-
tive typology for reading photographic images. Barthes gives the name
‘studium’ to the kind of photographic signification that fulfils general cul-
tural purposes in such ways as to produce affect of a civil, educated kind.
He gives the name ‘punctum’ to the kind of photographic signification
that disturbs the studium with a more subjective appeal, which he analo-
gises to a wound, a prick and a mark made by a pointed instrument, as
well as a ‘sting, speck, cut, little hole – and also a cast of the dice. A pho-
tograph’s punctum is that accident which pricks me (but also bruises me,
is poignant to me)’ (Barthes 1981: 26–27). With the notion of the punc-
tum, Barthes is calling attention to a supplementary, off-centred some-
thing that imbues the photograph with another kind of reality, one that is
often at odds with the realism that is enthusiastically constructed for, and
readily consumed by, a community.
Writing in the late 1970s, Barthes was not taken with mechanical
automatism as the occasion for utopian political possibilities (as were
media theorists of an earlier generation, such as Walter Benjamin). To
many readers, the punctum signals his turn to a decidedly personal way of
engaging with representational issues.12 Be that as it may, we should not
forget something important: namely, that despite concerted semiological
efforts (such as those made by Barthes himself) to demonstrate otherwise,
the investment in denotative meaning has hardly waned; rather, it tends
to intensify (if only as a source of dispute) each time a set of photographic
images is advanced in a documentary fashion for public consumption.
Insofar as realism—which may be broadly defined here as the wish for, or
insistence on, a denotative or indexical coincidence between the picture
194 R. CHOW
and its signified—has remained a bone of contention in the theorisation of

photographic practices, Barthes’s move of introducing an explicitly sub-
jective and literally eccentric dimension amounts to nothing short of a cat-
egorical intervention. With this move, photographic realism is articulated
not so much to the real as such, as to its disappearance, its voiding.13
Camera Lucida is, among other things, a book of mourning in which a
picture of Barthes’s mother as a little girl in the ‘Winter Garden’ is recalled
with melancholic tenderness.14 While numerous photographs are displayed
in the book, the Winter Garden picture is withheld, as though its appear-
ance would detract from or diminish its emotional significance. Thus for
Barthes, photography’s interest lies much less in the promises enabled by
technical ‘reproducibility’ (as envisioned by Benjamin in the 1930s) than
in its instantiation of an absolute break between temporalities, between
the time when the object is right there before the camera and the time that
the picture is viewed (when that object has become, in Barthes’s words,
what has been). As he writes elsewhere:
The type of consciousness the photograph involves is indeed truly unprec-

edented, since it establishes not a consciousness of the being-there of the
thing (which any copy could provoke) but an awareness of its having-been-
there. What we have is a new space–time category: spatial immediacy and
temporal anteriority, the photograph being an illogical conjunction between
the here-now and the there-then. (1977[1964]: 40)
In the act of looking at a photograph, Barthes traces a simple but

unbridgeable rupture between the moment of machinic capture and the
moment of personal review or contemplation: a rupture that makes pal-
pable the imprint of another time, an otherness to the time of the present
in which his own looking takes place. As Barthes suggests, this temporal
non-correspondence, or literally out-of-sync-ness, is what makes up the
materiality of photography as a practice about life and death. In the age
of the camera, as photographs mediate or cut into even the most mun-
dane episodes of cognition, what constitutes knowledge, what allows us
‘to know’, is none other than this temporal non-correspondence, this lac-
eration between different times that, for the spectator, translates into an
awareness of the loss of the other (or the other’s being). It is in such loss,
rather than in the object’s ontological fullness, that the ‘intractable’ reality
of the photograph—what Barthes (1981) calls ‘madness’—is to be found.
By articulating cognition to an awareness of loss, Barthes’s reading pro-

duces a partial, because vulnerable, viewing subject. Barthes’s intervention
makes it possible to ask of Antonioni’s Chung Kuo: where is the viewing
subject in this set of frames, and what kind of affect accompanies it? The
documentary orientation of Antonioni’s genre at first makes these ques-
tions sound like a non-sequitur. After all, the camera’s lens is serving as an
instrument of exploration, with the aim of observing and recording how
a social experiment (communism) has fared in a non-Western country. Is
there really a need to locate the subject and his/her affect?
Importantly, in Chung Kuo, the camera’s wandering pursuit of places,
faces, daily activities and communal interactions is accompanied by a male
voiceover, which supplies a narrative account—part-myth, part-history
and part-exotica—of the images on the screen. In one brief moment,
as the crew arrives in a village, this voice becomes reflexive. As though
the camera lens were suddenly turned back on the film crew, we hear
these comments on how strange the Europeans must have looked to the
Chinese:
These Chinese have never seen a westerner. They come to the doorways:
amazed, a bit scared and curious, they can’t resist the temptation to stare
at us.
We go on filming, but soon, we realise that it’s us who are peculiar and
foreign. For the people, to the other side of the camera, we’re completely
unknown and perhaps a bit ridiculous. A hard blow against our European
arrogance. For one-fourth of the earth’s population, we’re so unfamiliar
that it fills us with awe. Our big eyes, curly hair, big long noses, pale skin,
extravagant gestures, outlandish costumes …
They are taken aback, but very courteous, afraid to offend us by flee-
ing. They come out and stand still in front of the camera, often motionless,
as if petrified. Driving our brief digression into the highland, we’ve wit-
nessed a gallery of astounded faces, but we’ve never noticed any expression
of hostility.
Despite its potential for humanising and subjectivising the filmmakers,

however, this gentle and humble gesture of reflexivity is left at the level
of the voiceover. At the level of images, the Chinese remain the only
people being objectified. Indeed, no matter how much various Chinese
people are shown to stare at the camera—that is to say, no matter how
long the camera lens lingers on their faces, eyes and bodies—we cannot
see what or whom they are looking at: the blocking of the filmmak-
196 R. CHOW
ers’ visual presence in this acousmatic manner turns the subject who
investigates into a transcendental subject, an origin that organises every-
thing that is seen, but that stays hidden.15 By contrast, those who are
visible on the screen, by virtue of the effect of isolation produced by
the diegetic framing, are rendered abject, in the shapes of creatures
moving about in their own time, a time that is segregated from ours.
In relation to the commanding stance of the camera lens and narra-
tive voice, which hover over them but remain unlocatable, the Chinese
faces, despite being filmed during daylight, seem at some moments to
be groping in the dark, looking mystified as to what is really happening
right before them.
If both Barthes and Antonioni rely on the invisibility of certain images
for the effectiveness of their presentations, the results produced are
remarkably different. Whereas by withholding the Winter Garden pic-
ture, Barthes amplifies the subjective dimensions accompanying the (re)
viewing stance, by removing the investigating subject from the scene,
Antonioni rather endows on the latter an objectifying power that cor-
roborates photography’s traditional claim to realism. The documentary
style, supposedly based on a spirit of curiosity and a respect for the object,
seems in his hands to have created the aesthetic effect of a differentiation
between ‘here’ and ‘there’, between ‘now’ and ‘then’ and between ‘us’
and ‘them’. Just as the faces and bodies of the film crew are carefully
hidden from view even when they are alluded to, so too are the voices
of the Chinese going about their activities presented only as muted or
incomprehensible sounds. If the point of these technical decisions is to
emphasise a non-correspondence between the picture’s contents and the
viewing stance, the contrast with Barthes also is revealing. In Barthes’s
case, that non-correspondence (between the Winter Garden picture and
himself) heightens the sense of what is being framed as ephemeral, as
what has been but no longer is. In Chung Kuo, what is heightened is
the sense of a boundary, apparently uncrossable, between us and another
culture, which is caught in a different time rather than being coeval with
us.16 The loss of the other from the subject’s own time, which in Barthes’s
case helps to dismantle photography’s traditional claim to a realism based
on imagistic positivity, becomes in Antonioni’s case an exercise in render-
ing such imagistic positivity exotic. Antonioni’s enigmatic expressions so
baffled Chinese audiences that the entire exercise had to be written off as
an offence.
DOCUMENTARY REALISM IN THE AGE

OF DIGITAL SYNCHRONISATION: PRELIMINARY
REMARKS ON JIA ZHANGKE
Closely parallel to the question of the subject, then, is the question of
time, which demands a probing of what documentary realism signifies in
the age of digital hypermediality. More than one theorist has discussed the
transformation wrought by digitisation on the ongoing, modern practices
of image capture. In particular, digitisation’s capacity for instantaneous
transmission and storage, a capacity that increasingly reduces or eliminates
the factor of time lag that used to define photographic practices—the
duration from taking images to their chemical processing, finish and pres-
ervation in the form of print—has been the focus of thought-provoking
scholarly debates (e.g., Grace 2007; Hadjioannou 2008, 2012). One can
see why this steady reduction and elimination of the time lag factor is so
pertinent in the case of the documentary. Unlike the feature film, which
is generally organised around a fictional storyline, the documentary is
driven conceptually by the principle and agenda of producing a record.
This realist code under which it operates is the reason that the documen-
tary is associated regularly with specific types of knowledge production,
such as capturing events of a politically sensitive or educational nature.17
The implicit assumption is that the documentary as a record offers the
second-best access to such events: if one cannot be present on the scene, a
documentary is a good substitute because it offers one a chance to experi-
ence those events after the fact. In this sense of a stand-in version of what
is ‘live’, documentary work may be compared to translation. In both cases,
there is usually the assumption of an original occurrence which authorises
the documentary or translation, and in relation to which the documentary
or translation must occupy a hierarchically secondary, derived status as
evidence. As we have seen with Antonioni’s film, the most poignant issues
of interpretative conflict and incommensurability also begin with this set
of premises and presuppositions.
In the age of digitisation, when there are often just split seconds
between the event happening and the act of documentary making, when
documentary making can become virtually simultaneous with the event
taking place, and when the time of watching, the time of shooting and
the time of whatever is happening easily collapse into one synchronised
time rather than being neatly differentiated, what is happening to the
fundamental rationale of the documentary? In other words, when flaw-
198 R. CHOW
lessly synchronised (or just-in-time) ‘recording’ has become the norm,

the epistemic, aesthetic and moral grounds on which the documentary
exists—that it is a vicarious experience legitimated by a pre-existent reality
to which it is linked by the factor of the time lag—seem to be dissolving.
As time shifts are expunged with advanced high-tech capture, there seems
to be an interesting reversal of the conventional order of things: the time
of the documentary effort now may coincide with, or even precede, the
reality that it is supposedly recording. To capture, in these days of digital
synchronisation, is to actively produce: taking a picture is really making
one. Yet precisely at a time when its hitherto guiding principle of being a
record or a recording has become moot due to the disappearance of the
time lag guarantee, the documentary seems to have become a ubiquitous
impulse in contemporary global filmmaking. Why is this the case?
This is the point at which the work of a director such as Jia Zhangke
becomes provocative. Among well-known contemporary Chinese direc-
tors, Jia stands out insofar as his filmmaking seems to be inextricably bound
up with the documentary genre. One thinks of memorable films such as
Platform (2000), The World (2004), Still Life (2006), Useless (2007), I
Wish I Knew (2010) and 24 City (2008). Even though the occasional
deployment of the documentary style can be traced in the works of other
Chinese directors such as Tsai Ming-liang, Hou Hsiao-hsien, Edward
Yang, Zhang Yimou, Wong Kar-wai, Ann Hui, Li Yang and their inter-
nationally renowned cohort, Jia’s approach is, I believe, quite unique.18
Of course, in contrast with Antonioni it would be temptingly conve-
nient simply to see Jia as a native informant whose films are belatedly
vindicating the reality of China that Western directors have repeatedly
missed. Were we to follow this line of logic, we would need to criticise Jia
in a similar manner as the Gang of Four had criticised Antonioni: namely,
that he is filming ungainly rather than positive views of China, calling
attention at a time of national prosperity to embarrassing issues such as
poverty, unemployment, dislocation, alienation, broken family ties and
transient urban relations. If, unlike Antonioni, Jia allows Chinese people
the opportunity to face the camera and talk about their lives, as in the cases
of I Wish I Knew and 24 City, does the presence of these Chinese-speaking
voices make for a different kind of documentary undertaking, and how?
To address these questions, it is necessary to come to terms with the spe-
cific nature of the provocation of Jia’s work.
That provocation lies, I contend, in Jia’s fondness for making films that
look like documentaries. This ambiguous look—pertaining both to a way
of gazing, which is invisible and to the visible appearance of people and

things on the screen—is found not only in the explicitly documentary pro-
ductions such as Useless, I Wish I Knew and 24 City, all of which are inter-
mixed with fictive elements; it is also a constant in Jia’s feature films such
as The World and Still Life. Consider, for example, I Wish I Knew: the bulk
of the film comprises interviews with descendents of well-known Chinese
personalities, historic photographs, eyewitness accounts, footage from
previous films and media, chronological reminders of key national events
and named geopolitical locations such as Shanghai, Taiwan and Hong
Kong. However, these factographic elements, the conventional makings of
a documentary record, are interspersed with a fictional design such as the
choice of Shanghai as a point of focalisation, the use of a female character
(played by Zhao Tao) who reappears at regular intervals (as if to remind us
that the film is an artificial construct), and narrative spotlights on monu-
mental happenings such as the Cultural Revolution, the exodus from the
mainland to Taiwan around 1949, the Shanghai diaspora in Hong Kong
and the rise of the new rich in China in the early twenty-first century.
Similarly, in 24 City, amid historical details about a former military fac-
tory in Chengdu, having been demolished to make way for a commercial
real estate development, we watch professional actors such as Lü Liping,
Joan Chen and Zhao Tao performing the roles of interviewees, recalling
personal life experiences in intimate relations to the factory. Except for the
name recognition of the female stars, the distinction between acting in real
life and acting in fiction does not seem to matter.
In this enmeshment of documentary and fictional frames, what exactly
is the status of documentary realism in Jia’s films? While the historicity of
much of his content is verifiable, these films also come across as stylised
renditions, but of what? Should these films be understood as a belated
on-screen engagement with postmodern practices of collage, pastiche,
appropriation and citation? On the one hand, these postmodern practices
rely for their aesthetic effects on none other than the factor of the time
lag: for collage, pastiche, appropriation and citation to work, the viewer
needs to be able to grasp a temporal difference embedded in the playful
gesture of parody so that, for example, a visual cliché can be recognised as
such because it has been seen before. On the other hand, in Jia’s films, the
documentary style serves a different function: it presents the past not as a
bygone reality, already well understood, but rather as a (re) collection, one
that curates materials in fragmented form from different media. The past is
literally a matter of found footage, but this found footage (what supposedly
200 R. CHOW
offers documentary evidence) is now delinked from the authoritative or

authorising claim of a prior reality.
To borrow from Marcel Duchamp, the past in Jia’s films is presented
in the form of a ready-made, an ordinary object that has been discovered
(by accident, perhaps) and is now repurposed for a new kind of event:
the globalised exhibition. What renders the past ‘ready-made’ is precisely
this phenomenological condition or quality of being recognisable, of
having-been-looked-at: the past is given to view as what has been cut into
countless times already, by processes and apparatuses of (audial, visual and
narrative) recording. The acute senses of ephemerality, loss and ultimately
melancholy which characterise many a moment in Jia’s films are the results
of this deeply-felt sensation of hypermediality—indeed, of the composite
material, tracks and symptoms left on human perceptions and interactions
by media forms such as print, photography, film, newsreel images, histori-
cal reportage, popular songs, interviews and, most importantly, storytell-
ing. Rather than a unified reality, it is these motley tracks and symptoms,
with their ever-unfolding, ever-multiplying questions about origins and
destinations as well as identities, that Jia’s documentary approach invites
us to contemplate.
In an age when digital synchronicity has replaced temporal distinc-
tions in the production of meanings, Jia’s films astutely make hypermedial
display their central attraction. To this extent, the Chinese voices, narra-
tives and life stories are part of this display, arranged in accordance with
self-conscious codes of global medial literacy. In addition, just as these
‘native informant’ contributions may be formally archived as national his-
tory, so too are they eminently transmissible to popular global sites such
as YouTube, the access to which turns every viewer, Chinese or non-Chi-
nese, into a de facto ‘foreign observer’. What Jia’s films make explicit,
then, is much less documentary realism in the old-fashioned sense, with
its acute problem of contending claims to ‘the real’, as exemplified by the
Chinese reception of Antonioni’s film. Rather, Jia’s works showcase the
cultural politics of a new kind of conceptual project: a project of imagin-
ing modern China not simply as a land, a nation or a people, but first and
foremost as medial information, a project that takes the very notion of
‘China as documentary’ as its will to knowledge, its discursive force field.
FUNDING
This research is supported by the Trinity College of Arts and Sciences,
Duke University.
NOTES
1. For a classic set of explorations of this inexhaustible topic, see Kracauer
(1997[1960]), in particular chapters 1, ‘Photography’; 2, ‘Basic Concepts’
and 11, ‘The Film of Fact’.
2. Sophisticated discussions of photography often provide excellent pointers
to this effect of the eminently manipulable nature of photographs, pre-
cisely on account of their claim to the real. For a few examples of inspiring
discussions, see Crimp (1993), Krauss (1989), Sekula (1989) and Tagg
(1987).
3. See King (2010: 102–111) for her sensitive reading of Antonioni’s Chung
Kuo.
4. For an informative account in English of the circumstances of the film’s
making and receptions in China, Italy and the USA, as well as of the con-
tents, shots and commentaries within the film, see Sun (2009).
5. For a set of notes on the typical, obligatory itineraries and points of interest
that Western visitors were allowed to follow in China during the 1970s, see
Barthes (2012).
6. Eco’s analysis is keenly perceptive of the semiotic, aesthetic, symbolic and
political nuances that constitute the cultural differences between
Antonioni’s perspective and that of his Chinese critics.
7. This chapter is excerpted in Visual Culture: A Reader (1999: 80–94); hereafter
page references to this chapter will be taken from this volume.
8. A quick reminder of a key theoretical connection seems salutary at this
juncture. In his work on the epochal changes introduced by the camera,
Walter Benjamin draws attention to ‘technical reproducibility’: not only
numerically repeatable copies of a technically reproducible image, but also
the expansion and multiplication of previously unknown spaces within
familiar images. Benjamin’s argument is about the historic transformation
of the way we see in the age of mechanical automatism. As the machine
now provides a degree of accuracy that was hitherto unimaginable, the
human way of looking has been irreversibly supplanted and made obsolete.
Produced photographically, even the most natural-seeming of sights bear
the imprint of an inhumanism that is the result of machinic intervention.
Paradoxically, this also means that even the most objective-looking images,
such as those captured under the category of the documentary, are open to
controversy: their objectivity and verifiability, as well as the positions of the
image producer and viewer, are all subject to debate and dispute (see
Benjamin, 1969). In their reactions to Antonioni’s film, did not the
Chinese enact precisely this implosion of the assumption about photo-
graphic objectivity from within the bounds of the documented image? If
the native informants seem primitivist (as Sontag suggests) on account of
202 R. CHOW
their residual superstition and irrationality—they behave as though the

Western director or ethnographer had usurped something from them and
thus injured them—such primitivism ‘performs’ the unpredictable alterity
and infinite potentiality that, according to critics such as Benjamin, techni-
cal reproducibility is all about.
9. This chapter is excerpted in Visual Culture: A Reader (1999: 162–80);
hereafter, page references to it will be taken from this volume.
10. For a reading that sees Antonioni’s camera style as wandering and guided
by contingency, see King (2010: 105, 108).
11. For related interest, see also Barthes’ works such as Mythologies, Writing
Degree Zero and Elements of Semiology.
12. As Barthes himself asks in humour: ‘why mightn’t there be, somehow, a
new science for each object? A mathesis singularis (and no longer universa-
lis)?’ (1981: 8).
13. A reference to the Buddhist notion of the void comes up early in Barthes’
text, with suggestive connections to notions of contingency, pure deictic
language and the referent (see Barthes 1981: 5).
14. For the notes he took on his suffering over a period of two years, see
Barthes (2010).
15. I am using the word ‘acousmatic’ to indicate the separation of the sound
of the voiceover from the source of the sound. For explications of the
acousmatic in film, see Chion (1999).
16. For a well-known discussion of the denial of coevalness in the ethnographic
situation, see Fabian (1983).
17. For example, for John Grierson, who initiated and promoted the British
documentary movement, the ‘documentary idea’ was about public educa-
tion, while the film element was incidental. As Kracauer comments: ‘To
[Grierson] and his collaborators, then, film, in particular documentary, is a
medium of mass communication like the newspaper or the radio, a propi-
tious means of spreading civic education in a period and world in which the
strength of democracy more than ever depends upon the spread of infor-
mation and universal goodwill’ (1997[1960]: 210).
18. For discussions of Jia and contemporary documentary film practices in the
People’s Republic of China, see Berry (2007) and McGrath (2007). See
also Berry et al. (2010).
BIBLIOGRAPHY
An Jingfu (1994) The Pain of a Half Taoist: Taoist Principles, Chinese Landscape
Painting, and King of the Children. In Linda C. Ehrlich and David Desser
(eds.). Cinematic Landscapes: Observations on the Visual Arts and Cinema of
China and Japan. Austin: University of Texas Press, 117–25.
Anderson, Marston (1990) The Limits of Realism: Chinese Fiction in the
Revolutionary Period. Berkeley: University of California Press.
Anon (1937) “Yueyu pian zhengming yundong” [“Jyutpin zingming wandung”
or Cantonese film rectification movement]. Lingxing [Ling Sing] 7, no. 15
(June 27, 1937): no page.
Appelo, Tim (2014) ‘Wong Kar Wai Says His 108-Minute “The Grandmaster” Is
Not “A Watered-Down Version”’, The Hollywood Reporter (6 January), http://
www.hollywoodreporter.com/news/wong-kar-wai-says-his-668633.
Aristotle (1996) Poetics, trans. Malcolm Heath (London: Penguin Books).
Arroyo, José (2000) Introduction by José Arroyo (ed.) Action/Spectacle: A Sight
and Sound Reader (London: BFI Publishing), vii-xv.
Astruc, Alexandre (2009) ‘The Birth of a New Avant-Garde: La Caméra-Stylo’ in
Peter Graham with Ginette Vincendeau (eds.) The French New Wave: Critical
Landmarks (London: BFI and Palgrave Macmillan), 31–7.
Bao, Weihong (2015) Fiery Cinema: The Emergence of an Affective Medium in
China, 1915–1945 (Minneapolis: University of Minnesota Press).
Barthes, Roland (1968a) Elements of Semiology (trans. Annette Lavers and Colin
Smith). New York: Hill and Wang.
Barthes, Roland (1968b) Writing Degree Zero (trans. Annette Lavers and Colin
Smith). New York: Hill and Wang.
Barthes, Roland (1972) Mythologies (trans. Annette Lavers), New York: Hill and
Wang.

DOI 10.1057/978-1-137-55309-6
204 BIBLIOGRAPHY
Barthes, Roland (1977 [1964]) Rhetoric of the Image. In S. Heath (ed.) Image,
Music, Text. New York: Hill and Wang, pp.32–51.
Barthes, Roland (1981) Camera Lucida: Reflections on Photography (trans. R.
Howard). New York: Hill and Wang.
Barthes, Roland (2010) Mourning Diary: October 26, 1977 - September 15,
1979 (trans. R. Howard). New York: Hill and Wang.
Barthes, Roland (2012) Travels in China (ed. AH Pierrot , trans. A. Brown).
Cambridge: Polity Press.
Benjamin, Walter (1969) The Work of Art in the Age of Mechanical Reproduction.
In H. Arendt (ed.) Illuminations (trans. H. Zohn). New York: Schocken,
pp.217–251.
Berliner, Todd (2010) Hollywood Incoherent: Narration in Seventies Cinema.
Austin: University of Texas Press.
Berliner, Todd (2013) “Hollywood Storytelling and Aesthetic Pleasure,” in Arthur
P. Shimamura (ed.) Psychocinematics: Exploring Cognition at the Movies, Oxford
& New York: Oxford University Press, 195–213.
Berry, Chris (1994) “Neither One Thing Nor Another: Toward a Study of the
Viewing Subject and Chinese Cinema in the 1980s.” In Nick Browne, Paul
G. Pickowicz, Vivian Sobchack, and Esther Yau (eds.). New Chinese
Cinemas: Forms, Identities, Politics. Cambridge: Cambridge University
Press, 88–113.
Berry, Chris (2007) “Getting Real: Chinese Documentary, Chinese Postsocialism.”
In Z. Zhen (ed.) The Urban Generation: Chinese Cinema and Society at the
Turn of the Twenty-First Century. Durham, NC: Duke University Press,
pp.115–134.
Berry, Chris (2012) “Every Colour Red? Colour in the Films of the Cultural
Revolution Model Stage Works.” Journal of Chinese Cinemas 6:3, 233–246.
Berry, Chris and Mary Ann Farquhar (1994) “Post-Socialist Strategies: an Analysis
of Yellow Earth and Black Cannon Incident.” In Linda C. Ehrlich and David
Desser (eds.). Cinematic Landscapes: Observations on the Visual Arts and
Cinema of China and Japan. Austin: University of Texas Press, 81–116.
Berry, Chris and Mary Ann Farquhar (2006) China on Screen: Cinema and Nation.
New York: Columbia University Press.
Berry, Chris, Lu X. and L. Rofel (eds) (2010) The New Chinese Documentary
Film Movement: For the Public Record. Hong Kong: Hong Kong University
Press.
Berry, Michael (2005) “Chen Kaige: Historical Revolution and Cinematic
Rebellion.” In Speaking in Images: Interviews with Contemporary Chinese
Filmmakers. New York: Columbia University Press, 82–106.
Bettinson, Gary (2013) “Sounds of Hong Kong Cinema: Johnnie To, Milkyway
Image, and the Sound Track.” Jump Cut: A Review of Contemporary Media 55
(Fall). http://www.ejumpcut.org/archive/jc55.2013/BettinsonToAudio/
index.html.
BIBLIOGRAPHY 205
Bettinson, Gary (2014) The Sensuous Cinema of Wong Kar-wai: Film Poetics and
the Aesthetic of Disturbance. Hong Kong: Hong Kong University Press.
Bettinson, Gary (2015) “Hong Kong Stars and Stardom,” in A Companion to
Hong Kong Cinema, edited by Esther M.K. Cheung, Gina Marchetti, and
Esther C.M. Yau, West Sussex: Wiley-Blackwell, 379–387.
Bettinson, Gary (2016) “Once Upon a Time in China and America: Transnational
Storytelling and the Recent Films of Peter Chan,” in Felicia Chan and Andy
Willis (eds), Chinese Cinemas: International Perspectives, London & New York:
Routledge, 37–52.
Bordwell, David (1985) Narration in the Fiction Film (Madison: The University
of Wisconsin Press).
Bordwell, David (1988) Ozu and the Poetics of Cinema (Princeton, N.J: Princeton
University Press).
Bordwell, David (1989) “Historical Poetics of Cinema,” in The Cinematic Text:
Methods and Approaches, edited by R. Barton Palmer, New York: AMS Press,
369–398.
Bordwell, David (1993) The Cinema of Eisenstein. Cambridge, Mass.: Harvard
University Press.
Bordwell, David (1995) “Contemporary Film Studies and the Vicissitudes of
Grand Theory,” in David Bordwell and Noël Carroll (eds) Post-Theory:
Reconstructing Film Studies, Madison: University of Wisconsin Press, 3–36.
Bordwell, David (2000/2011 [2nd ed.]) Planet Hong Kong: Popular Cinema and
the Art of Entertainment. Cambridge, Mass; London: Harvard University Press.
Bordwell, David (2001) “Transcultural Spaces: Toward a Poetics of Chinese
Film.” Post Script 20:2/3, 9–24.
Bordwell, David (2001a) “Aesthetics in Action: Kungfu, Gunplay, and Cinematic
Expressivity,” in Esther C.M. Yau (ed.) At Full Speed: Hong Kong Cinema in a
Borderless World, Minneapolis; London: University of Minnesota Press, 73–93.
Bordwell, David (2002) “Richness Through Imperfection: King Hu and the
Glimpse,” in The Cinema of Hong Kong: History, Arts, Identity, edited by
Poshek Fu and David Desser, Cambridge: Cambridge University Press,
113–136.
Bordwell, David (2003) “Louder Than Words: The Films of Johnnie To.”
Artforum International 41:9 (May), 154–157.
Bordwell, David (2005) Figures Traced in Light: On Cinematic Staging. Berkeley,
Los Angeles, London: University of California Press.
Bordwell, David (2006) “The Departed: No Departure.” Observations on Film
Art: David Bordwell’s Website on Cinema. October 10. http://www.davidbor-
dwell.net/blog/2006/10/10/the-departed-no-departure/.
Bordwell, David (2006) The Way Hollywood Tells It: Story and Style in Modern
Movies. Berkeley: University of California Press.
Bordwell, David (2008) Poetics of Cinema. New York & London: Routledge.
206 BIBLIOGRAPHY
Bordwell, David (2009) “Another Shaw Production: Anamorphic Adventures in

Hong Kong,” David Bordwell’s Website on Cinema: Observations on Film Art
(October). http://www.davidbordwell.net/essays/shaw.php
Bordwell, David (2013) “VIFF 2013 Finale: The Bold and the Beautiful,
Sometimes Together.” Observations on Film Art: David Bordwell’s Website on
Cinema (October 21), http://www.davidbordwell.net/blog/2013/10/21/
viff-2013-finale-the-bold-and-the-beautiful-sometimes-together
Bordwell, David; Janet Staiger, and Kristin Thompson (1985) The Classical
Hollywood Cinema: Film Style and Mode of Production to 1960 (New York:
Columbia University Press).
Bourdieu, Pierre (1999[1965]) “The Social Definition of Photography.” In J.
Evans and S. Hall (eds) Visual Culture: A Reader. London: Sage,
pp.162–180.
Bowman, Paul (2010) Theorizing Bruce Lee: Film-Fantasy-Fighting-Philosophy
(Amsterdam and New York: Rodopi).
Bowman, Paul (2013) Beyond Bruce Lee: Chasing the Dragon Through Film,
Philosophy, and Popular Culture (London and New York: Wallflower Press).
Browne, Nick (1980) “Film Form/Voice Over: Bresson’s Diary of a Country
Priest.” Yale French Studies 60, 233–240.
Buckland, Warren (2006) Directed by Steven Spielberg: Poetics of the Contemporary
Hollywood Blockbuster. New York & London: Continuum.
Buhot, Jean (1967) Chinese and Japanese Landscape Art; with Sections on Korea
and Vietnam. Translated from the French by Rémy Inglis Hall. Edited by
Charles McCurdy. New York: Anchor Books [Gallimard, 1961].
Burch, Noël (1979) “Mizoguchi Kenji.” In To the Distant Observer: Form and
Meaning in the Japanese Cinema. London: Scholar Press, 228–34.
Cahill, James (1997) “Approaches to Chinese Painting, Part II.” In Yang Xin, Nie
Chongzhen, Lang Shaojun, Richard M. Barnhart, Jamesa Cahill, and Wu
Hung. Three Thousand Years of Chinese Painting. New Haven: Yale University
Press, 131–96.
Ch’en Yun-ru (2003) “The Five Dynasties and Sung Period.” In Wang Yao-t’ing
(ed.). The Art and Aesthetics of Form: Selections from the History of Chinese
Paintings. Translated by Donald E. Bix. Taipei: National Palace Museum,
24–35.
Chan, Feinung (Chen Feinong, 2007). Yueju liushi nian [Jyutkek luksap nin or
Sixty years in the Cantonese theater]. Hong Kong: The Cantonese Opera
Research Programme of the Music Department, the Chinese University of
Hong Kong.
Chan, Joseph M. (2009) “Policies for a Sustainable Development of Hong Kong
Film Industry.” Public Policy Digest (July). http://www.ugc.edu.hk/rgc/
ppd1/eng/04.htm.
BIBLIOGRAPHY 207
Chang, Hsiao-hung (2011) ‘Fade in-Fade Out of the Body-City: Hou Hsiao-
hsien and Café Lumière’, Chung-Wai Literary Monthly, vol.40, no.3, 7–37.
(Chinese)
Chen Kaige (1993) Interview with Klaus Eder (24 May, Cannes). In Klaus Eder
and Deac Russell (eds.). New Chinese Cinema, a dossier. London: The National
Film Theatre, 89–92.
Chen Kaige (2001) “Les vertes collines.” In Une jeunesse chinoise. Translated from
Chinese by Christine Corniot. Arles, France: Editions Philippe Picquier [1989],
161–203.
Chen Kaige and Wan Zhi (1989) “King of the Children,” film script, based on the
short story, Haizi Wang, by Ah Cheng. Edited and Translated by Bonnie
S. McDougall. In Chen Kaige and Tony Rayns, King of the Children and The
New Chinese Cinema. London: Faber and Faber, 59–121.
Cherchi Usai, Paulo (2000) Silent Cinema: An Introduction, revised and expanded
edition. London: British Film Institute [1994].
Cheung Mong-wan (Zhang Wangyun, c. 1960 [1930]). Renhai leihen [Janhoi
leoihan or Traces of tears in the sea of humans]. Hong Kong: Nantian chuban-
she or Naamtin cheotbaanse.
Cheung, Esther M. K. (2015) “The Urban Maze: Crisis and Topography in Hong
Kong Cinema,” in A Companion to Hong Kong Cinema, edited by Esther
M.K. Cheung, Gina Marchetti, and Esther C.M. Yau, West Sussex: Wiley
Blackwell, 51–70.
Chi, Robert (2008) “The Red Detachment of Women: Resenting, Regendering,
Remembering.” In Chris Berry (ed.) Chinese Films in Focus: 25 New Takes,
London: British Film Institute, 152–159.
Chiksan Jichen (2009). “Yueyu pian gexin yundong” [“Jyutjyu pin gaaksan wand-
ung” or Cantonese film reform movement]. Yilin [Art Lane], no. 11 (1937):
no page.
Chion, Michel (1999) The Voice in Cinema (ed. and trans. C. Gorbman). New
York: Columbia University Press.
Chiu, Melissa and Zheng Shengtian (eds.) (2008) Art and China’s Revolution.
New York: Asia Society.
Choi, Kimburley Wing-yee, and Steve Fore (2015) “Animating the Translocal:
The McDull Films as a Cultural and Visual Expression of Hong Kong,” in A
Companion to Hong Kong Cinema, edited by Esther M.K. Cheung, Gina
Marchetti, and Esther C.M. Yau, West Sussex: Wiley Blackwell, 140–167.
Chow, Rey (1991) “Male Narcissism and National Culture: Subjectivity in Chen
Kaige’s King of the Children.” Camera Obscura, 25/26: 9–40.
Chow, Rey (1996) ‘We Endure, therefore We Are: Survival, Governance, and
Zhang Yimou’s To Live’, South Atlantic Quarterly, 95 (4): 1039–64.
Chow, Rey (2006) ‘Sacrifice, Mimesis, and the Theorizing of Victimhood (A
Speculative Essay)’, Representations, 94 (1): 131–49.
208 BIBLIOGRAPHY
Chung Po-yin (Zhong Baoxian, 2004 [2007]). Xianggang yingshi ye bainian

[Hoenggong jingsijip baaknin or One hundred years of Hong Kong film and
television industry]. Hong Kong: Joint Publishing.
Clark, Paul (1987) Chinese Cinema: Culture and Politics Since 1949. Cambridge:
Cambridge University Press.
Clark, Paul (2008) The Chinese Cultural Revolution: A History. New York:
Cambridge University Press.
Clark, Paul (2010) “Model Theatrical Works and the Remodelling of the Cultural
Revolution.” In King, R. ed. Art in Turmoil: The Chinese Cultural Revolution,
1966–1976. Hong Kong: Hong Kong University Press.
Cook, Pam (2007) ‘For a New French Cinema: The Politique des Auteurs’ in Pam
Cook (ed.) The Cinema Book, 3rd edn (London: BFI), 390–409.
Cooper, Rhonda, and Jeffrey Cooper (1997) Masterpieces of Chinese Art.
New York: Todtri Productions Ltd.
Crimp, Douglas (1993) “The Museum’s Old, The Library’s New Subject.” In On
the Museum’s Ruins. Cambridge, MA: MIT Press, pp.66–81.
Darley, Andrew (2000) Visual Digital Culture: Surface Play and Spectacle in New
Media Genres (Abingdon, Oxon: Routledge).
Daruvala, Susan (2007) “The Aesthetics and Moral Politics of Fei Mu’s Spring in
a Small Town.” Journal of Chinese Cinemas, 1:3, 171–187.
De Carle Sowerby, Arthur (1940) Nature in Chinese Art. New York: The John
Day Company.
Denton, Kirk A. (1987) “Modern Drama as Myth: A semiotic Analysis of Taking
Tiger Mountain by Strategy.” In Tung, C. and MacKerras, C. (eds) Drama in
the People’s Republic of China. Albany: State University of New York Press.
Durgnat, Raymond (1990) “The Long Take in Voyage to Cytheria: Brecht and
Marx vs. Bazin and Go.” Film Comment 26.6 (November/December), 43–46.
Dutton, Michael (2015) (unpublished) Cultural Revolution as Method.
Eco, Umberto (1977) “De Interpreatione, or The Difficulty of Being Marco Polo
(on the occasion of Antonioni’s China Film).” Film Quarterly 30:4, pp.8–12.
Eder, Klaus (1993) “New Chinese Cinema: A Preliminary Introduction.” In Klaus
Eder and Deac Russell (eds.). New Chinese Cinema, a dossier. London: The
National Film Theatre, 8–17.
Ehrlich, David, and Tianyi Jin (2001) “Animation in China.” In John Lent (ed.)
Animation in Asia and the Pacific. Eastleigh, UK: John Libbey Publishing,
7–29.
Elley, Derek (1995) “Introduction: Chinese Catalogue of Films, 1905–1937,”
Cheng Jihua, Li Shaobai, Xing Zuwen (eds.). Griffithiana 54 (October): 4–91.
Elsaesser, Thomas (2009) Hollywood heute: Geschichte, Gender und Nation im post-
klassischen Kino (Berlin: Bertz + Fischer).
Elsaesser, Thomas (2012) “Film as System: Or How to Step Through an Open
Door.” In The Persistence of Hollywood. New York: Routledge, 105–26.
BIBLIOGRAPHY 209
Eng, Clare Sher Ling (2009) “Red Detachment of Women and the Enterprise of
Making ‘Model’ Music During the Chinese Cultural Revolution.” voiceX-
change 3:1. Available at: https://letterpress.uchicago.edu/index.php/voicex-
change/article/view/29 (Accessed 29 October 2015)
Evans, Harriet and Donald, Stephanie (eds) (1999) Picturing Power in the People’s
Republic of China: Posters of the Cultural Revolution. Lanham: Rowman and
Littlefield.
Fabian, Johannes (1983) Time and the Other: How Anthropology Makes Its
Object. New York: Columbia University Press.
Fan, Victor (2015) Cinema Approaching Reality: Locating Chinese Film Theory
(Minneapolis: University of Minnesota Press).
Farquhar, Mary (2010) ‘Visual Effects Magic: Hero’s Sydney Connection’ in Gary
D. Rawnsley and Ming-Yeh T. Rawnsley (eds.) Global Chinese Cinema: The
Culture and Politics of Hero (Abingdon, Oxon: Routledge), 184–97.
Fei Mingyi 費明儀 (1996) “Mr. Fei Mu – My Father,” in The Film Aesthetics of
Spring in a Small Town 小城之春的電影美學 (Taipei: National Film Archives):
8–10.
Fitzgerald, Carolyn (2013) Fragmenting Modernisms: Chinese Wartime Literature,
Art, and Film, 1937–1949 (Leiden, Boston: Brill).
Grace, Helen (2007) “Monuments and the Face of Time: Distortions of Scale and
Asynchrony in Postcolonial Hong Kong.” Postcolonial Studies 10:4,
pp.467–483.
Gunning, Tom (1986) ‘The Cinema of Attraction: Early Film, Its Spectator and
the Avant-Garde’, Wide Angle, 8 (3 & 4): 63–70.
Hadjioannou, Markos (2008) “Into Great Sillness, Again and Again: Gilles
Deleuze’ ‘Time and the Construction of Digital Cinema’,” Rhizomes 16.
http://www.rhizomes.net.issue16/hadji/index.html.
Hadjioannou, Markos (2012) From Light to Byte: Towards an Ethics of Digital
Cinema. Minneapolis, MN: University of Minnesota Press.
Hall, Sheldon, and Neale, Steve (eds.) (2010) Epics, Spectacles, and Blockbusters: A
Hollywood History (Detroit, Michigan: Wayne State University Press).
Hansen, Miriam (1999) “The Mass Production of the Senses: Classical Cinema as
Vernacular Modernism,” Modernism/Modernity 6:2, 59–77.
Hasumi, Shigehiko (2008) ‘The Eloquence of the Taciturn: An Essay on Hou
Hsiao-hsien,’ Inter-Asia Cultural Studies 9:2, 184–194.
He, M. (何明) (1964) “Supporting modern theatre (提倡现代戏剧),” Red Flag
(红旗) nos.2–3, 58–60.
Hegel, R. E. (1984) “Making the Past Serve the Present in Fiction and Drama:
From the Yan’an Forum to the Cultural Revolution.” In MacDougall, Bonnie
S. (ed.) Popular Chinese Literature and Performing Arts in the People’s Republic
of China, 1949–1979, University of California Press.
210 BIBLIOGRAPHY
Hesemann, Sabine (1999) “China: The Song Period and the Aesthetics of
Simplicity.” In Gabriele Fahr-Becker (ed.). The Art of East Asia. Chris Murray
English edition ed. Cologne: Könemann, 142–65.
Hjort, Mette (2010) “Ruan Lingyu: Reflections on an Individual Performance
Style,” in Chinese Film Stars, edited by Mary Farquhar and Yingjin Zhang,
London: Routledge, 32–49.
Hsü Kuo-huang (1999) The Landscape Painting Tradition of Li Ch’eng and Kuo
Hsi. Taipei: National Palace Museum.
Hu, Jubin (2003) Projecting a Nation: Chinese National Cinema Before 1949.
Hong Kong: Hong Kong University Press.
Hunt, Leon (2003) Kung Fu Cult Masters: From Bruce Lee to Crouching Tiger
(London: Wallflower Press).
Huters, T. (1993) “Ideologies of Realism in Modern China: The Hard Imperatives
of Imported Theory.” In Liu, K. and Tang, X. eds. Politics, Ideology, and
Literary Discourse in Modern China. Durham: Duke University Press.
Ingham, Michael (2009) Johnnie To Kei-Fung’s PTU. Hong Kong: Hong Kong
University Press.
Kermode, Mark (2014) ‘The Grandmaster Review: Thrilling But Often Incoherent
Martial Arts Epic’, The Guardian (7 December), http://www.theguardian.
com/film/2014/dec/07/the-grandmaster-review-thrilling-incoherent-
martial-arts-epic.
King, Homay (2010) Lost in Translation: Orientalism, Cinema and the Enigmatic
Signifier. Durham, NC: Duke University Press.
Koepnick, Lutz (2014) On Slowness: Toward an Aesthetic of the Contemporary
(New York: Columbia University Press).
Kong, Tong. “Yueyupian de bujie yu gualong” [Jyutjyupin dik pukgaai jyu gwaat-
lung]. Lingxing ribao [Lingsing jatbou or Stage and movie stars daily].
December 7, no page.
Kracauer, Siegfried (1997[1960]) Theory of Film: The Redemption of Reality.
Princeton, NJ: Princeton University Press.
Kraicer, Shelly (2008) “Rediscovering the Fourth Generation.” Cinema Scope, 33
(Winter): 29–33.
Krauss, Rosalind (1989) “Photography’s Discursive Spaces.” In R. Boston (ed.)
The Contest of Meaning: Critical Histories of Photography. Cambridge, MA:
MIT Press, pp.286–301.
Lai, Kin (Li Jian, 2010). Xianggang Yueju xulun [Hoenggong Jyutkek zeoileon or
A discourse on the Hong Kong Cantonese theater]. Hong Kong: Joint
Publishing.
Lam Nin-tung (Lin Niantong, 1978). “Wushi niandai Yueyu dianying yanjiu
zhong de jige wenti” [“Ngsap nindoi Jyutjyu dinjing jingau zung dik geigo
mantai” or Several questions on the studies of the Cantonese cinema of the
1950s]. Daai dakse, no. 59 (June 9): 2–5 and 47–51.
BIBLIOGRAPHY 211
Lan, Z. (2004) Café Lumière, http://mypaper.pchome.com.tw/russiablue/

post/1242758215.
Landsberger, Stefan (2001) Chinese Propaganda Posters: From Revolution to
Modernization. Amsterdam: The Pepin Press [1995].
Law, Kar (2011). “Union Film: From Noon Till Dusk.” Trans. Agnes Lam. In
One for All: The Union Film Spirit. Ed. Grace Ng. Hong Kong: Hong Kong
Film Archive, 36–56.
Lee, Vivian (2007) ‘Virtual Bodies, Flying Objects: The Digital Imaginary in
Contemporary Martial Arts Films’, Journal of Chinese Cinemas, 1 (1): 9–26.
Lee, Vivian P. Y. (2009) Hong Kong Cinema Since 1997: The Post-Nostalgic
Imagination. Hampshire & New York: Palgrave Macmillan.
Lee, Bono (2012) “The Possibility of China for Hong Kong Directors: The
Transformation of Peter Chan’s Identity,” in Peter Ho-Sun Chan: My Way,
edited by Li Cheuk-to, Hong Kong: Joint Publishing (HK), 187–195.
Leese, Daniel (2011) Mao Cult: Rhetoric and Ritual in China’s Cultural
Revolution. New York: Cambridge University Press.
Leung Yuk Ming, Lisa (2013) “Re-nationalizing the Transnational? The Cases of
Exiled and Warlords in Hong Kong-China Film Co-Production,” in Popular
Culture Co-Productions and Collaborations in East and Southeast Asia, edited
by Nissim Otmazgin and Eyal Ben-Ari, Singapore and Japan: NUS Press and
Kyoto University Press, 115–135.
Li Shaohong (1993) Interview with Klaus Eder (2 March, Beijing). In Klaus Eder
and Deac Russell (eds.). New Chinese Cinema, a dossier. London: The National
Film Theatre, 99–103.
Li Cheuk-to (ed.) (2012) Peter Ho-Sun Chan: My Way. Hong Kong: Joint
Publishing (HK).
Li, Cheuk-to and Shu M. (eds.) (2003) Ozu Yasujiro: the 100th Anniversary
(Hong Kong: Hong Kong Arts Development Council).
Liao, Ping-hui. (1989) ‘Women and typhoon: an analysis of Hou Hsiao-hsien’s
Dust in the Wind (Nuxing yu taifeng--guangkuei Hou Xiaoxian de Liangliang
fengchen),’ Chung-Wai Literary Monthly 17:10, 61–70. (Chinese)
Lim, Song Hwee (2014) Tsai Ming-liang and a Cinema of Slowness (Honolulu:
University of Hawaii Press).
Lin Yutang (1967) The Chinese Theory of Art: Translations from the Masters of
Chinese Art. New York: G.P. Putnam’s Sons.
Lin Niantong 林年同 (1991) Chinese Film Aesthetics 中國電影美學 (Taipei:
Yunzhen).
Lin, Y. (2005) ‘A man of culture between gaps: a study of Jiang Wen-yeh and his
time (Jiafeng zhong de wenhuaren--riju shiqi Jiang Wenye ji qi shidai yanjou)’,
Doctoral Thesis (Taiwan: National Tainan University), 102–106. (Chinese)
Lin, Wenchi (2009) ‘Poetics of the mise en abyme in The Sweet Hereafter and Café
Lumière’, Journal of Art Studies (Yishuxue yanjou) 5, 81–112. (Chinese)
212 BIBLIOGRAPHY
Liu, Ping (2012). Zhong Ri xiandai yanju jiaoliu tushi [An illustrated history of the
exchange between the Chinese and Japanese theaters]. Beijing: Joint Publishing.
Lu, X. (2004) Rhetoric of the Chinese Cultural Revolution: The Impact on Chinese
Thought, Culture and Communication. Columbia: University of South Carolina
Press.
Ludden, Y. (2013) China’s Musical Revolution: From Beijing Opera to Yangbanxi.
Ph.D thesis. University of Kentucky.
Mak, Siu-ha (Mai Xiaoxia, 1940). Guangdong xiju shilue [Gwongdung heikek siloek
or Brief history of the Cantonese theater]. Hong Kong: Zhongguo wenhua
xiejinhui or Zunggwok manfaa hipzeonwui.
Marchetti, Gina (2007) Andrew Lau and Alan Mak’s Infernal Affairs - The Trilogy.
Hong Kong: Hong Kong University Press.
Massumi, Brian (2002) Parables for the Virtual: Movement, Affect, Sensation.
Durham: Duke University Press.
McDonough, Tom (2004) ‘Introduction: Ideology and the Situationist Utopia’ in
Tom McDonough (ed.) Guy Debord and the Situationist International: Texts
and Documents (Cambridge, Massachusetts, and London, England: The MIT
Press), ix-xx.
McDougall, Bonnie S. (1991) “The Process (November 1983 to August 1984).”
In The Yellow Earth: a Film by Chen Kaige - With a Complete Translation of the
Filmscript. Hong Kong: The Chinese University of Hong Kong, 25–54.
McGrath, Jason (2007) “The Independent Cinema of Jia Zhangke: From
Postsocialist Realism to a Transnational Aesthetic.” In Z. Zhen (ed.) The
Urban Generation: Chinese Cinema and Society at the Turn of the Twenty-
First Century. Durham, NC: Duke University Press, pp.81–114.
Mi, Z. (1991) ‘Can women enter history? On female characters in City of Sadness
(Nuren wufa jingru lishi? Tan Beiqing chengshi zhong de nuxing jiaose)’ in Mi
Z. and Liang X. (eds.) Death of the new cinema (Xindianying zhi si) (Taipei:
Tanshan). (Chinese)
Misawa, M. (2012) Sandwiched between the ‘empire’ and the ‘motherland’: Taiwan
film workers’ negotiation and border-crossing during Japanese occupation (Zai
“diguo” yu “zuguo” de jiafeng jian: Rizhi shiqi Taiwan dianyingren de jiaoshe
yu kuajing). Trans. Li W. and Hsu S. (2012) (Taipei: National Taiwan University
Press). (Chinese)
Mittler, Barbara (2012) A Continuous Revolution: Making Sense of Cultural
Revolution Culture. Cambridge: Harvard University Press.
Möller, Olaf (2011) “Among the Ruins: The Two Poles of Fei Mu.” Cinemascope
45 (Winter), 19–21.
Morris, Meaghan (2004) ‘Transnational Imagination in Action Cinema: Hong
Kong and the Making of a Global Popular Culture’, Inter-Asia Cultural
Studies, 5 (2): 181–99.
N.A. (2003) “Girlgeek of the Week: Ellen Poon,” GirlGeeks (February). http://
www.girlgeeks.org/innergeek/gkwk/gkwk_poon.shtml
BIBLIOGRAPHY 213
Ni Zhen (2002) Memoirs From The Beijing Film Academy: The Genesis of China's
Fifth Generation. Translated by Chris Berry. Durham, NC: Duke University
Press [1995].
North, Dan (2005) ‘Virtual Actors, Spectacle and Special Effects: Kung Fu Meets
“All That CGI Bullshit”’ in Stacy Gillis (ed.) The Matrix Trilogy: Cyberpunk
Reloaded (London: Wallflower Press), 48–61.
Nye, Joseph S., Jr. (2004) Soft Power: The Means to Success in World Politics (New
York: Public Affairs).
Ouyang, Yuqian (1954). “Shitan Yueju” [An attempted discussion on the
Cantonese theater]. In Zhongguo xiqu yanjiu ziliao chuji [A preliminary anthol-
ogy of research materials on the Chinese theater and music]. Ed. Ouyang.
Hong Kong: Xiju yishu chubanshe or Heikuk ngaiseot ceotbaanse, 109–57.
Pang, Laikwan (2011) “Shanghai Films of the 1930s.” In Song Hwee Lim and
Julian Ward (eds.). The Chinese Cinema Book. London: British Film Institute,
56–64.
Pang, Laikwan (2012) “The Dialectics of Mao’s Images: Monumentalism,
Circulation, and Power Effects.” In Henriot, C. and Yeh, W. (eds) Visualizing
China, 1845–1965: Moving and Still Images in Historical Narratives. Leiden:
Brill.
Pasolini, Pier Paolo (1980) “Observations on the Long Take” October 13
(Summer), 3–6.
Rancière, Jacques (2006) The Politics of Aesthetics: The Distribution of the Sensible,
translated with an introduction by Gabriel Rockhill (London and New York:
Continuum).
Rayns, Tony (1989) “Chinese Vocabulary: an Introduction to King of the Children
and the New Chinese Cinema.” In Chen Kaige and Tony Rayns, King of the
Children and The New Chinese Cinema. London: Faber and Faber, 1–58.
Rayns, Tony (2014) “The Second Generation.” In James Bell (ed.). Electric
Shadows: A Century of Chinese Cinema. London: British Film Institute, 16–28.
Richie, Donald (1974) Ozu. Berkeley: University of California Press.
Rist, Peter (2006a) “The Presence (and Absence) of Landscape in Silent East
Asian Films.” In Martin Lefebvre (ed.). Landscape and Film. New York:
Routledge, 189–212.
Rist, Peter (2006b) “Visual Style in the Silent Films made by the Lianhua Film
Company [United Photoplay Service] in Shanghai: 1931–35.” In Haili Kong
and John Lent (eds.). 100 Years of Chinese Cinema: A Generational Dialogue.
Norwalk CT: EastBridge, 3–28.
Rist, Peter (2007) “The Presence (and Absence) of Landscape in Silent Chinese
Films.” The Centennial Celebration of Chinese Cinema and the 2005 Annual
Conference of ACSS - National, Transnational, International: Chinese Cinema
and Asian Cinema in the Context of Globalization - Proceedings (II). Beijing
and Shanghai, 372–82.
214 BIBLIOGRAPHY
Roberts, Rosemary (2008) “Performing Gender in Maoist Ballet: Mutual Subversions

of Gender and Ideology in The Red Detachment of Women.” [online] Intersections:
Gender and Sexuality in Asia and the Pacific 16. Available at: http://intersec-
tions.anu.edu.au/issue16/roberts.htm [Accessed 29 October 2015]
Robertson, Patrick (ed.) (1991) The Guinness Book of Movie Facts and Feats. 4th
ed. Enfield, Middlesex: Guinness publishing Ltd. [1980].
Sato, Tadao (1994) “Japanese Cinema and the Traditional Arts: Imagery,
Technique, and Cultural Context.” In Linda C. Ehrlich and David Desser
(eds.). Cinematic Landscapes: Observations on the Visual Arts and Cinema of
China and Japan. Austin: University of Texas Press, 165–86.
Schager, Nick (2005) ‘Café Lumière,’ Slant Magazine, 30 Dec.
Schoonover, Karl (2012) ‘Wastrels of Time: Slow Cinema’s Laboring Body, the
Political Spectator, and the Queer’, Framework, 53 (1): 65–78.
Sekula, Allan (1989) Photography Against the Grain. Halifax: Nova Scotia Press.
Shek, Kei (1966). “Gangchan zuopai dianying ji qi xiao zichan jieji xing”
[Gongcaan zopaai dinjing kap kei siu zicaan gaaikap sing]. Zhongguo xuesheng
zhoubao [Zunggwok hoksaang zaubo or Chinese Students Weekly], no. 740
(September 23): 11.
Shen, S. (2013) ‘Back Seat of My Bike: Youthful Dreams in Three Times’,
International Conference, Chinese Language Cinema: Text, Context and History
(Hong Kong Baptist University, June 6–7, 2013). (Chinese)
Sickman, Lawrence and Alexander Soper (1961) “The Masters of Landscape.” In
The Pelican History of Art: The Art and Architecture of China. Harmondsworth,
Middlesex: Penguin Books, 1st int. edition, 203–22.
Silbergeld, Jerome (1999a) “Drowning on Dry Land: Yellow Earth and the
Traditionalism of the 'Avant-garde.” In China into Film: Frames of Reference in
Contemporary Chinese Cinema. London: Reaktion Books, 15–52.
Silbergeld, Jerome (1999b) “The Children of Melodrama: No-drama, Pseudo-
drama, Melodramatic Masquerade and Deconstruction Drama,” In China into
Film: Frames of Reference in Contemporary Chinese Cinema. London: Reaktion
Books, 255–60.
Sing, Song-yong (2010) ‘The Phantom-Effect of ‘Cinema Revised’: On ‘Trans-
imageity’ of Hou Hsiao-hsien’s Café Lumiére and Le Voyage du Ballon Rouge’,
Chung-Wai Literary Monthly 39:4, 135–169. (Chinese)
Sobchack, Vivian (1991) The Address of the Eye: A Phenomenology of Film Experience.
Princeton: Princeton University Press.
Sontag, Susan (1978) “The Image-World.” In On Photography, Harmondsworth:
Penguin, pp.153–180.
Sontag, Susan (1999[1978]) “The Image-World.” In J. Evans and S. Hall (eds)
Visual Culture: A Reader. London: Sage, pp.80–94.
Sternberg, Meir (1978) Expositional Modes and Temporal Ordering in Fiction.
Bloomington & Indianapolis: Indiana University Press.
BIBLIOGRAPHY 215
Stuckey, G. Andrew (2014) “The world out there: spectacle and exposure in
Perhaps Love.” Journal of Chinese Cinemas 8:1, 17–36.
Sullivan, Michael (1973) Chinese Art: Recent Discoveries. London: Thames and
Hudson.
Sullivan, Michael (1980) Chinese Landscape Paintings in the Sui and T’ang
Dynasties. Berkeley and Los Angeles: University of California Press.
Sun, Hongyun (2009) “Two Chinas? Joris Ivens Yukong and Antonioni’s China.”
Studies in Documentary Film 3:1, pp.45–59.
Sung Manlei (Song Wanli, 1938). “Huanan dianying yu Guangdong jingshen”
[“Waanaam dingjing jyu Gwongdung zingsan” or South China cinema and its
Kwantung spirit]. Art Lane, no. 35 (August 1, 1938): no page.
Szeto, Mirana M., and Yun-chung Chen (2013) “To work or not to work: the
dilemma of Hong Kong film labor in the age of mainlandization.” Jump Cut:
A Review of Contemporary Media 55 (Fall). http://www.ejumpcut.org/
archive/jc55.2013/SzetoChenHongKong/index.html.
Tagg, John (1987) The Burden of Representation. London: Macmillan.
Tamura, S. (2007) Li Xianlan’s Lover - Film and War (Li Xianlan de lianren -
dianying yu zhanzheng) Trans. Shi G. and Wang J. (2010; Taipei: Wu-Nan
Book). (Chinese)
Taylor, Richard (2002) October, “BFI Film Classics.” London: British Film
Institute.
Teo, Stephen (2008) “Promise and perhaps love: Pan-Asian production and the
Hong Kong-China interrelationship.” Inter-Asia Cultural Studies 9:3,
341–358.
The Pride of China: Masterpieces of Chinese Painting and Calligraphy of the Jin,
Tang, Song and Yuan Dynasties from the Palace Museum. Exhibition catalogue.
Hong Kong: Hong Kong Museum of Art. 2007.
Thompson, Kristin (1988) Breaking the Glass Armor: Neoformalist Film Analysis.
Princeton: Princeton University Press.
Treager, Mary (1997 [1980]) “Court and Chan Buddhist Arts.” In Chinese Art,
rev. ed. London: Thames and Hudson.
Truffaut, François (1966) ‘A Certain Tendency of the French Cinema’ in Cahiers
du Cinéma in English, vol. 1 (New York: Cahiers Publishing Co.), 30–40.
Tsuji, H. (1998) A Chronicle of China Film Company, 1939-1945 (Chuka denei
shiwa: ichi heisotsu no Nitchua eiga kaisoki, 1939-1945) (Tokyo: Gaifusha
Publishing). (Japanese)
Udden, James (2009) No Man an Island: The Cinema of Hou Hsiao-hsien. Hong
Kong: Hong Kong University Press.
Udden, James (2012) “In Search of Chinese Film Style(s) and Technique(s),” in
Yingjin Zhang (ed.) A Companion to Chinese Cinema, Malden, Ma:
Wiley-Blackwell.
Viénet, René (2004) ‘The Situationists and the New Forms of Action against
Politics and Art’, trans. Tom McDonough, in Tom McDonough (ed.) Guy
216 BIBLIOGRAPHY
Debord and the Situationist International: Texts and Documents (Cambridge,

Massachusetts, and London, England: The MIT Press), 181–5.
Wada-Marciano, M. (2012) ‘Café Lumière: Hou Hsiao-hsien’s memory and dia-
logue (Kohi jiko: Ho shaoshien no memori to taiwa)’ in Maeno, M. (eds.) Hou
Hsiao-hsien’s poetics and the prism of time (Ho shaoshien no shigaku to jikan no
purizumu) (Nagoya: Arumu). (Japanese)
Wang, B. (2003) “Revolutionary Realism and Revolutionary Romanticism: The
Song of Youth.” In Mostow, J. ed. The Columbia Companion to Modern East
Asian Literature. New York: Columbia University Press.
Wang, Q. (2014) Memory, Subjectivity and Independent Chinese Cinema.
Edinburgh: Edinburgh University Press.
Wark, McKenzie (2013) The Spectacle of Disintegration: Situationist Passages Out
of the 20th Century (London: Verso).
Watson, William (1974) Style in the Arts of China. Harmondsworth, Middlesex:
Penguin Books.
Watson, William (2000) The Arts of China 900–1620. New Haven: Yale University
Press.
Whissel, Kristen (2010) ‘The Digital Multitude’, Cinema Journal, 49 (4): 90–110.
Williams, Linda (1991) ‘Film Bodies: Gender, Genre, and Excess’, Film Quarterly
44 (4): 2–13.
Wong Ain-Ling (ed.) (2009) Restored Classic Confucius. Hong Kong: Hong
Kong Film Archive.
Woo, Catherine Yi-Yu Cho (1991) “The Chinese Montage: From Poetry and
Painting to the Silver Screen.” In Chris Berry (ed.). Perspectives on Chinese
Cinema. London: British Film Institute, 21–9. [first published as no. 39 in the
Cornell East Asia Papers series.]
Wood, Aylish (2002) ‘Timespaces in Spectacular Cinema: Crossing the Great
Divide of Spectacle Versus Narrative’, Screen, 43 (4): 370–86.
Wu Hung (1997) “The Origins of Chinese Painting (Paleolithic Period to Tang
Dynasty).” In Yang Xin, Nie Chongzhen, Lang Shaojun, Richard M. Barnhart,
James Cahill, and Wu Hung. Three Thousand Years of Chinese Painting. New
Haven: Yale University Press, 15–85.
Xu, Gary (2013) “Edification Through Affection: The Cultural Revolution Films,
1974–1976.” In Rojas, C and Chow, E. eds. The Oxford Handbook of Chinese
Cinemas. New York: Oxford University Press.
Yang Xin (1997) “Approaches to Chinese Painting, Part I.” In Yang Xin, Nie
Chongzhen, Lang Shaojun, Richard M. Barnhart, James Cahill, and Wu Hung.
Three Thousand Years of Chinese Painting. New Haven: Yale University Press,
1–4.
Yau, Esther C. M. (1991) “Yellow Earth: Western Analysis and a Non-Western
Text.” In Chris Berry (ed.). Perspectives on Chinese Cinema. London: British
Film Institute, 62–79.
BIBLIOGRAPHY 217
Yau, Esther (2015) “Watchful Partners, Hidden Currents: Hong Kong Cinema
Moving into the Mainland of China,” in A Companion to Hong Kong Cinema,
edited by Esther M.K. Cheung, Gina Marchetti, and Esther C.M. Yau, West
Sussex: Wiley Blackwell, 17–50.
Yeh, Emilie Yueh-Yu (1999) “A Life of Its Own: Musical Discourses in Wong Kar-
wai’s Films.” Post Script 20 (2/3), 61–76.
Yeh, Y. (2001) Politics and Poetics of Hou Hsiao-hsien’s Films. Post Script
20:2/3, 61–76.
Yeh, Emilie Yueh-Yu (2005) “Poetics and Politics of Hou Hsiao-hsien’s Films,” in
Chinese-Language Film: Historiography, Poetics, Politics, edited by Sheldon
H. Lu and Emilie Yueh-yu Yeh, Honolulu: University of Hawai‘i Press,
163–185.
Ying Xiong 应雄 (1993) “Spring in a Small Town and Eastern Cinema, (pt. 1)”
“《小城之春》与”
Yu, H. (于会泳) (1968) “May Culture Always Be a Base for the Dissemination of
Mao Zedong Thought (让文艺界永远成为宣传毛泽东思想的阵地).” Wenhui
News (文汇报) 23 May.
Yue, Genevieve (2008) ‘The Light outside the Window’, Reverse Shot, http://
reverseshot.org/symposiums/entry/873/cafe_lumiere.
Zhang Ziliang (1991) “The Filmscript: The Yellow Earth.” Revised by Chen Kaige.
Translated by Bonnie S. McDougall. In The Yellow Earth: a Film by Chen
Kaige - With a Complete Translation of the Filmscript. Hong Kong: The Chinese
University of Hong Kong, 173–263.
Zhang, Xudong (1997) Chinese Modernism in the Era of Reforms: Cultural Fever,
Avant-garde Fiction, and the New Chinese Cinema. Durham, NC: Duke
University Press.
Zhang Zhen (2001) “An Amorous History of the Silver Screen: The Actress As
Vernacular Embodiment in Early Chinese Film Culture.” Camera Obscura, 48:
229–63.
Zhang, Yingjin (2004) “Cinema and the Nation State in the PRC, 1949–78.” In
Chinese National Cinema. New York: Routledge, 189–224.
Zhou, Chengren (2011). “Glory be with Cantonese Films: Ng Cho-fan and
Union Film.” Trans. Tam King-fai. In One for All: The Union Film Spirit. Ed.
Grace Ng. Hong Kong: Hong Kong Film Archive, 9–18.
Zhu, T. (2009) Flight of the Red Balloon: A Sequel to the Documentation of the
Cinema of Hou Hsiao-hsien (Hong qiqiu de luxing: Hou Xiaoxian dianying jilu
xupian) (Jinan: Shangdong haubao chuban she). (Chinese)
INDEX
A Au Kin-yee, 131, 133, 135, 136,

Aberdeen (2014), 142 144n21
Ah Cheng, 65 Avenging Eagle, The (1978), 129
Allen, Woody, 133 average shot length (ASL), 5, 58, 83
All’s Well End’s Well (1992), 141 Azalea Mountain (1974), 9
All’s Well Ends Well 2011 (2011), 126
All’s Well Ends Well 2012 (2012),
126 B
Althusser, Louis, Bada Shanren, 60
American Dreams in China (2013), Badham, John, 63
122 Bai Yan, 173
anamorphic widescreen, 23 Ballet mécanique (1924), 76n31
An Autumn Afternoon (1962), 107 Banquet, The (2006), 149
An Empress and the Warriors (2008), Barthes, Roland, 13, 193, 194–6,
120 201n5, 202n11–14
Angelopoulos, Theo, 26, 83 Battle of Wits, A (2006), 120
An Inspector Calls (2015), 126 Bazin, André, 83
An Jingfu, 65 Beetlejuice (1988), 63
Antonioni, Michelangelo, 11, 13, 68, Beijing (1938), 115
76n34, 80, 86, 91, 94, 185–202 Beijing Film Academy (BFA), 51
Aoyama Shinji, 25 Benjamin, Walter, 193, 194, 201n8
Aristotle, 151, 152 Berliner, Todd, 13n2, 140
Ashes of Time (1994), 19, 141 Berlin: Symphony of a Great City
Asian minimalism, 25, 28n8 (1927), 76
Assassin, The (2015), 149 Berry, Chris, 9, 10, 29–49, 61, 62,
Astruc, Alexandre, 152, 153 202n18

DOI 10.1057/978-1-137-55309-6
220 INDEX
Bettinson, Gary, 1–14, 119–45 Chang Hsiao-hung, 99

Big Blue Lake (2011), 142 Chan, Jackie, 148
Big Parade, The (1986), 63 Chan, Peter Ho-Sun, 142
Black Coal, Thin Ice (2014), 3 Cheerful Wind (1981), 113
Black Swan (2010), 121 Cheng, Sammi, 134, 138
Blind Detective (2013), 121, 134, Chen, Joan, 199
136–40, 142, 145n27 Cheung Mong-wan, 175
Blissfully Yours (2002), 25 Che, Zhang, 122, 129, 144n15
Bloodbath on Wolf Mountain (1937), Chinese Communist Party (CCP), 48,
57 173
Blooded Treasury Fight (1979), 129 Chinese Fifth Generation, 10
Bordwell, David, 1, 2, 4, 5, 7–9, 13, Chinese Fourth Generation, 59, 60
13n1, 13n3, 14n4, 15–28, 30, Chinese painting, 10, 52, 57, 61
31, 48, 62, 74n17, 77n39, 111, Chi, Robert, 41
121, 122, 130, 132, 136, 137, Chow, Rey, 13, 64, 162, 163, 164n8,
142, 143n4, 152, 168 185–202
Bourdieu, Pierre, 13, 191, 192 Chow, Stephen, 27, 149
Brave Archer, The (1977), 129 Chronicle of a Love Affair (1950), 86
Bresson, Robert, 11, 80, 86, 91, 93, Chuang Tzu, 65
94 Chungking Express (1994), 7, 19
Bride Wars (2015), 6 Chung Kuo/Cina (1972), 13
Broadway Danny Rose (1984), 62, 63 cinematography, 8, 31, 40, 41, 45, 59,
Browne, Nick, 91 61
Buckland, Warren, 13n2 City of Sadness (1989), 22, 25
Buhot, Félix, 52 Clark, Paul, 30–3, 46, 48, 59
Burton, Tim, 63 Closer Economic Partnership
Arrangement (CEPA), 119
Cognitivism, 1
C Communism, 51, 66, 189, 195
Café Lumiére (2003), 11, 97–118 Confucius, 53, 58, 114
Cai Chusheng, 56 Confucius (1940), 58
Can Dialectics Break Bricks (1973), 12 Connected (2008), 5
Cannes International Film Festival, coproduction, 3–5, 11, 100, 120–4,
60 129, 130, 131, 133, 135, 139,
Cantonese Opera, 12, 169 141, 142, 144n15
Cassavetes, John, 62 Cowherd’s Flute, The (1963), 60
Cathay film company, 173 Crossing Hennessy (2010), 142
censorship, 61, 62, 113, 122, 123, Crouching Tiger, Hidden Dragon
136, 142n1, 175 (2000), 4
Center Stage (1991), 2, 141 Crouching Tiger, Hidden Dragon:
Chan, Fruit, 142 Sword of Destiny (2016), 5
Chang Chen, 165n12 Cuaron, Alfonso, 83
INDEX 221
Cultural Revolution, 2, 9, 10, 29–49, Eco, Umberto, 187, 188

51, 59, 60, 65, 113, 162, 199 Eder, Klaus, 59, 60
Cultural Studies, 1, 39 editing, 3, 8, 9, 12, 20–3, 41, 45, 66,
cyber-fu, 148 71, 88–90, 93, 95, 156, 171, 178
Eisenstein, Sergei, 17, 21
Elley, Derek, 56
D Elsaesser, Thomas, 169, 175, 177
Darley, Andrew, 164n5 Enlai, Zhou, 76n36, 187
Daruvala, Susan, 86, 90 Equinox Flower (1958), 107
Daughter of the Earth (1940), 115 Eureka (2000), 25, 140
Daughters of China (1949), 59 Everlasting Regret (1948), 173
Dawn Must Come (1950), 173 Eye, The (2002), 5
Day a Pig Fell in the Well, The (1996), Eye, The (2008), 5
25
Daybreak (1933), 56
Days of Being Wild (1990), 19 F
de Carle Sowerby, Arthur, 56 Fantasia (2004), 132
Deleuze, Gilles, 99 Fan, Victor, 12, 167–83
Demme, Jonathan, 63 Farewell My Concubine (1993), 65, 66
Deng Lijun, Teresa, 115 Farquhar, Mary Ann, 61
Deng Xiaoping, 30 Fei Mingyi, 79
Departed, The (2006), 5 Fei Mu, 7, 9, 10, 57, 74n13, 79–95
Diary of a Country Priest (1951), 91 Feng Xiaogang, 149
digital technology, 148–50, 153–4, Feuillade, Louis, 22
163 Fight Club (1999), 129
Ding, Shan, 132, 134, 135, 142, Firestorm (2013), 142
144n18, 144n20 Fist of Fury (1972), 148
documentary, 9, 13, 60, 75n25, 107, Fitzgerald, Carolyn, 80, 82, 90, 91
115, 185–202 Five Golden Flowers (1959), 60, 74n21
Dot 2 Dot (2014), 142 Flaming Swords/Strife for Mastery
Doyle, Christopher, 153 (1977), 129
Dracula (1979), 63 Flight of the Red Balloon (2007), 113
Driver No. 7 (1958), 173 Flowers of Shanghai (1998), 24, 84
Duchamp, Marcel, 200 formalism, 1, 15
Dust in the Wind (1986), 22 French New Wave, 22
Dutton, Michael, 40 Furious 7 (2015), 5
E G
Eastern Zhou dynasty, 52 Gallants (2010), 142
Eat Drink Man Woman (1994), 5 Gang of Four, 32, 188, 198
Echoes of the Rainbow (2010), 142 German Expressionism, 24, 170
222 INDEX
Godard, Jean-Luc, 67, 68, 76n34 Hu, Jubin, 74n13

Goddess, The (1934), 8 Hunt, Leon, 148
Golden Chickensss (2014), 142 Hu Peng, 172
Golden Era, The (2014), 130 Huston, John, 63
Good Men, Good Women (1995), 99 hypermediality, 185, 197, 200
Grandmaster, The (2013), 3, 149,
150, 158–62, 164n4
Grand theory, 13, 17 I
Grandview studio, 172 Il Mare (2000), 129
Gravity (2013), 83 Infernal Affairs (2002), 4, 5, 19, 120,
Great Wall, The (2016), 5 121
Guangxi Film Studio, 63 Ingeborg Holm (1913), 22
Gu Changwei, 64 In the Face of Demolition (1953), 12,
Gunning, Tom, 40, 153 168, 175–83
In the Mood for Love (2000), 19, 121
In the Wild Mountains (1985), 64
H Ip Man, 150, 159, 161, 162, 164n4,
Hansen, Miriam, 168, 173 165n13, 165n15
Hegel, Georg Wilhelm Friedrich, 24, Ip Man (2008), 3
37 Iron Man 3 (2013), 5
Hero (2002), 4, 27 Italian Neorealism, 24, 94
Heroes Shed No Tears (1980), 129 I Wish I Knew (2010), 13, 187, 198,
Hesemann, Sabine, 73n5, 73n6 199
historical poetics, 4, 6, 7, 9, 13n1, 16,
17
Hitchcock, Alfred, 83 J
Hjort, Mette, 8, 14n5 Jiang Qing, 48
Hollywood cinema, 13, 16, 18, 28n4, Jiang Wenye, 99–106, 113–17
167, 168, 172, 182 Jin, Zhang, 160
Hong Kong Film Archive, 58, 168
Hong Kong International Film
Festival, 60 K
Hong Sangsoo, 25, 26, 28n8 Kaige, Chen, 9, 10, 51–77, 149, 163n2
Horse Thief, The (1986), 64 Kamei Fumio, 115
Hot Wind (1943), 115 Karate Kid, The (2010), 5
Hou Hsiao-hsien, 7–9, 22, 28n7, 84, Kid, The (1950), 175
86, 95, 97–118, 149, 198 Kill Bill (2003), 153
House of Flying Daggers (2004), 4, Kill Bill 2 (2004), 153
149, 150, 154, 155, 157 King, Homay, 187
Huang Manli (Mary Wong), 173 King Hu, 7, 21, 28n6
Hui, Ann, 7, 130, 142, 198 King of the Children (1987), 10, 52,
Hui, Kara, 130 64–7, 71, 72, 76n35
INDEX 223
Kitano Takeshi, 25 Li Han-hsiang, 7

Koepnick, Lutz, 159 Li, Jet, 148, 156
Ko Lo-chuen, 179 Li Keran, 60
Kong Tong, 176 Lim, Song Hwee, 147–65
Kore-eda Hirokazu, 24 Lin Kaihou, 115
Kraicer, Shelly, 75n24 Lin Niantgong, 83
Kuleshov, Lev, 20 Lin, Wenchi, 99
kung-fu genre, 19–21 Li Ping-bing, Mark, 109
Kung Fu Hustle (2004), 3, 149 Li Pingqian, 115
Kung Fu Killer (2014), 142, 145n28 Li Shaohong, 59
Kung Fu Panda 3 (2016), 5 Li Ssu-hsün, 52, 73n1
Kuomintang (KMT/Nationalist Li Tianji, 80
Party), 173 Li Yang, 198
Kurosawa, Akira, 94, 124 Loach, Ken, 133
Kwan, Stanley, 141 Lo Duen, 173, 177
long take, 11, 22, 24–6, 76n37, 80,
83–5, 88, 89
L Looper (2012), 5
Lacan, Jacques, 186 Love in the Buff (2012), 3, 130
Lam, Aubrey, 122 Love Streams (1984), 62
Lao Tzu, 65 Loving Blood of the Volcano (1932),
Late Autumn (1960), 107, 108 56–7
Late Spring (1949), 107, 108 Ludden, Yawen, 30
Lau, Andy, 132, 135–7, 155 Lü Liping, 199
Lau Ching-wan, 132, 134, 135 Lumière brothers, 74n18, 98, 99
Lee, Bono, 120 Luo Mingyou, 57
Lee, Bruce, 7, 12, 142, 148–50, Lu, Sheldon, 14n4, 28n7
164n4 Lu Xing, 30
Leese, Daniel, 30
Lee Tit, 12, 168, 175, 182
Lee, Vivian P.Y., 144n13, 148, 154, M
164n6 Maborosi no hikari (1995), 24
Léger’s, Fernand, 76 MacLean, Douglas, 57
Leung Chiu-wai, Tony, 159–60 Mad Detective (2007), 121, 134–6,
Leung, Lisa, 122 141, 144n20, 144n24
Lianhua Film Company, 56, 73n11 Magnificent Ambersons, The (1942),
Liao Yiyuan, 173 23
Li Chao-tao, 52 Malick, Terrence, 91
Life Gamble (1979), 129 Man with a Movie Camera (1929),
Life on a String (1991), 65 76n31
Life Without Principle (2011), 132 Mao Zedong, 215
Light of Women (1937), 170 Marchetti, Gina, 120
224 INDEX
Marxism, 17 norms, 3–7, 9–12, 16, 18, 20, 23, 56,

Massumi, Brian, 40 121, 130, 131, 135, 136, 191
Matrix, The (1999), 148, 153, 155, Northern Song dynasty, 55
159
McDougall, Bonnie S., 61, 72
Melville, Jean-Pierre, 11, 80 O
Midnight After, The (2014), 142 October (1928), 71
Milkyway Image studio, 11, 130 Oh, Sadaharu, 115
Ming dynasty, 60 Oldboy (2003), 129
Mingxing (Star) film company, 56 Orphan Island period, 58
mise-en-scène, 40, 45, 152, 171 Overseas Chinese Films company,
Mission, The (1999), 19 173
Mittler, Barbara, 30–2, 34, 37 Ozu Yasujiro, 97
Mizoguchi, Kenji, 24, 25, 83, 86
model works, 9, 10, 30, 31–3, 35,
39–49 P
Möller, Olaf, 82 Pang Ho-cheung, 130
Monk Comes Down the Mountain, The Pang, Laikwan, 37, 74n19
(2015), 149 Paris Texas (1984), 62
Morris, Meaghan, 147 Pasolini, Pier Paolo, 83
M/Other (1999), 25 Path to the Peace of East Asia (1938),
Murder Plot (1979), 129 115
music, 15, 31, 35, 37, 38, 41, 45, 55, Peking Opera, 30, 33, 35, 170
101–4, 114–17, 152, 170 Pen-ek Ratanaruang, 25
My Left Eye Sees Ghosts (2002), 134 People’s Liberation Army, 29
Mystery (2012), 129 Perhaps Love (2005), 129, 144n13
Myth, The (2005), 120 photographic realism, 191, 192, 194
Pirates of the Caribbean (2003), 16
planimetric composition, 26
N Platform (2000), 25
Nai-hoi, Yau, 131 Po Fung, 171
Nanjing (1938), 52, 73n5, 115 Political Science, 40
narration, 3, 10, 13, 58, 91, Power of Kangwon Province, The
123–30, 136–40, 142, 143n10, (1998), 25
168–72 Prague Structuralism, 1
Needing You (2000), 134 Preminger, Otto, 83
neoformalism, 1, 27n3 primacy effect, 126
New One-Armed Swordsman, The Promise, The (2005), 149
(1971), 122, 129 psychoanalytic theory, 17
New Taiwanese Cinema, 24 Puppetmaster, The (1993), 117
Ng Cho-fan, 171, 173, 176 Purple Rose of Cairo, The (1985), 63
Ni Zhen, 59, 76n34 puzzle films, 119–45
INDEX 225
Q Sino-Japanese War, 56, 174

Qiansheng, Zhu, 187 6ixtynin9 (1999), 25
Sixth Sense, The (1999), 129
Sklovsky, Viktor, 182
R slow cinema, 12, 164n3
Rancière, Jacques, 158, 186 Small Toys (1933), 57
Rashomon (1950), 94, 124 Sobchack, Vivian, 40
Rayns, Tony, 62, 64, 74n20, 75n24, socialist realism, 34, 59, 61, 192
76n32 Sokurov, Alexandr, 83
Red Cliff (2008), 3 Song dynasty, 51–77
Red Detachment of Women, The Song of China (1935), 57
(1971), 9, 30, 32, 35, 36, 40–2, Song Wanli (Sung Man-lei), 172
44, 45 Sontag, Susan, 13, 188–91, 201n8
Red Sorghum (1987), 64 Soper, Alexander, 54, 73n4, 73n8
Renoir, Jean, 66, 76n34, 81, 83 Sophocles, 152
Rist, Peter, 10, 51–77 Southern Song dynasty, 53, 55
River Without Buoys, The (1983), 64 Spring in a Small Town (1948), 10,
Rooftop, The (2013), 3 57, 58, 74n13, 79–83, 85, 86,
Rose, The (1979), 62, 63 88, 93, 94
Ruan Lingyu, 8 Springtime in a Small Town (2002),
Rules of the Game (1939), 81 94
Running on Karma (2003), 134, 135 State Administration of Radio, Film
Russian Ark (2002), 83 and Television (SARFT), 119–20
Russian Formalism, 1 Sternberg, Meir, 126
Ruttmann, Walter, 76n31 Still Life (2006), 198
Rydell, Mark, 63 Sui dynasty, 73
Sullivan, Michael, 73n1
Sun Chun, 129
S Sun Yu, 56, 57
Sacrifice (2010), 65 Suwa Nobubiro, 25
Salt, Barry, 63, 75n29 Suzhou River (2000), 129
SARS virus, Suzuki Shigeyoshi, 115
Saving General Yang (2013), 120
screenwriting, 59, 131–6
Shanghai Animation Film Studio, 60 T
Shaw Brothers film company, 7, 129 Takeshi, Akimoto, 115
Shek Kei, 174 Takeshi Kaneshiro, 124, 125
Sher Ling Eng, Clare, 36 Taking Tiger Mountain by Strategy
Shochiku film studio, 97 (1970), 9, 29, 30, 32, 34, 37, 39,
Shutter Island (2010), 121 40, 44
Silence of the Lambs, The (1991), 63 T’ang dynasty, 52
Sing Song-yong, 100 Tao, Zhao, 199
226 INDEX
Tarantino, Quentin, 153 Umbrella Movement, 142

Tarr, Bela, 95 Union Film Enterprise, 12, 168,
Tcherepnin, Alexander, 114 172–5
Television Broadcast Limited (TVB), Useless (2007), 198
131 Usual Suspects, The (1995), 121
Ten Minutes Older: The Trumpet
(2002), 65
Te Wei, 60 V
The Weinstein Company, 4, 143n9, Vertov, Dziga, 76n31
143n11 Viénet, René, 12, 147, 150
Third Sister Liu (1960), 75 Visitor on Ice Mountain (1963), 60
Thompson, Kristin, 18, 27n3, 28n4, 172 voiceover narration, 58
Tiananmen Square massacre, 5 Vulgaria (2012), 142
Tian Zhuangzhuang, 10, 64, 79–95
Time to Live and a Time to Die, A
(1985), 22 W
Tin-shing, Yip, 131 Wachowski Brothers, 153
Tiny Times 3.0 (2014), 3 Wai Ka-fai, 19, 121, 130, 131
To Kei-fung, Johnnie, 19 Wang Qi, 40
Tokyo Family (2013), 97 Wang Xueqi, 62
Tokyo Story (1953), 11 Wang Yu, Jimmy, 130
Tokyo Twilight (1957), 11, 107–13 Warlords, The (2007), 120, 122,
To Live (1994), 162 144n15
Tortilla Soup (2001), 5 Watson, William, 53, 55, 56, 73n2,
Transformers: Age of Extinction 73n3
(2014), 5 Way We Are, The (2008), 142
Tropical Malady (2004), 25 Weerasethakul, Apichatpong, 25
Truffaut, François, 153 Wenders, Wim, 62
Tsai Ming-liang, 7, 25, 26, 158, 198 Western Han dynasty, 73n1
Tsi Lo-lin, 176 What Women Want (2011), 6
Tsui Hark, 7, 29 Where is Mama (1960), 60
Twin Sisters of the South (1939), 170 Whissel, Kristen, 154, 161, 164n7
2/Duo (1997), 25 wire-fu, 148, 150
24 City (2008), 13, 198 Wise Blood (1979), 63
Two Punks (1996), 25 Woman, a Gun, and a Noodle Shop, A
2046 (2004), 121 (2009), 5
Tykwer, Tom, 159 Wong Ain-Ling, 74n15
Wong Cho-san, 176
Wong Fei-hong, 20
U Wong Kar-wai, 4, 7–9, 19, 27, 141,
Udden, James, 8, 10, 11, 30, 58, 149, 150, 158, 159, 164n2,
74n16, 79–95, 99, 107 164n4
INDEX 227
Woo, Catherine Yi-Yu Cho, 57 Yang Xin, 52

Wood, Aylish, 152 Yan Yan, Chen, 57
Woo, John, 7, 27 Yau, Esther, 121
Woo-ping, Yuen, 153 Yeh, Emilie, 8, 14n4
World, The (2004), 198, 199 Yellow Earth (1984), 10, 61–3
Wu Hung, 73n1 Yen, Donnie, 124, 126–8,
Wu Pang, 172, 173 132, 158
Wu Tianming, 64 Yimou, Zhang, 12, 60, 61, 153–7,
Wu Tip-ying, 171 163, 164n9, 198
Wu Xia (2011), 121–38, 141, 143n3, Ying, Cheung, 173, 176
143n9, 143n12, 144n15 Young Bruce Lee (2010), 142
wuxia pian (swordplay film), 19, 122 Yuan, Dong, 52, 73n5
Yuan, Xie, 64
Yuen, Chor, 129, 144n14
X Yuen Kuei (Corey), 21
Xi’an studio, 64
Xi Jinping, 149
Xue Bai, 62 Z
Xu, Gary, 124–8, 138, 143n10 Zhangke, Jia, 7, 9, 13, 25, 95,
185–202
Zhang, Xudong, 64
Y Zhang, Yingjin, 59, 75n22
Yamada, Yoji, 97 Zhen, Zhang, 58
Yamamoto Satsuo, 115 Zhou dynasty, 52, 65
Yang, Edward, 24, 198 Ziyi, Zhang, 155, 160, 164n9

The Poetics of Chinese Cinema

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Poetics of Chinese Cinema

Uploaded by

Copyright:

Available Formats

EAST ASIAN POPULAR CULTURE

More information about this series at

East Asian Popular Culture

Library of Congress Control Number: 2016947977

© The Editor(s) (if applicable) and The Author(s) 2016

Cover illustration: © Rawpixel Ltd / Getty

Printed on acid-free paper

This Palgrave Macmillan imprint is published by Springer Nature

1 Introduction: The Poetics of Chinese Cinema 1

2 Five Lessons from Stealth Poetics 15

3 Red Poetics: The Films of the Chinese Cultural Revolution

4 Renewal of Song Dynasty Landscape Painting Aesthetics

5 Poetics of Two Springs: Fei Mu versus Tian Zhuangzhuang 79

6 Remaking Ozu: Hou Hsiao-hsien’s Café Lumière 97

7 Hong Kong Puzzle Films: The Persistence of Tradition 119

8 Can Poetics Break Bricks? 147

9 Poetics of Parapraxis and Reeducation: The Hong Kong

10 China as Documentary: Some Basic Questions (Inspired by

Chris Berry is Professor of Film Studies at King’s College London. His publica-

Fabulations, Contemporary Chinese Films (2007), Entanglements: Transmedial

Fig. 4.1 Ma Yüan’s A Mountain Path in Spring 54

Table 3.1 Hong Changqing discovers Wu Qinghua in Red

Introduction: The Poetics of Chinese

From Russian Formalism and Prague Structuralism to neoformalism and

© The Editor(s) (if applicable) and The Author(s) 2016 1

Why is a poetics of Chinese cinema important? For critics of a cul-

quintessentially Chinese film language falters on several fronts. For one

to mass audiences in Western territories. (Hence, the efforts by Western

blossoms, American studios court PRC producers for collaborations and

What Women Want (2011/2000), Bride Wars (2015/2009)—demand

and theoretically informed analyses of particular films, genres, and oeu-

inseparable; as Berry demonstrates, the films shared certain aesthetic

an imagined traditional Chinese discourse. Among Tian’s putatively tradi-

4. Bordwell sets forth this research program in “Transcultural Spaces” (2001),

Five Lessons from Stealth Poetics

I can only be happy that poetics, as a perspective on studying films, has

© The Editor(s) (if applicable) and The Author(s) 2016 15

Once I had a better sense of this local variant of episodic construction,

LONG TAKES, LONG LENSES

kitchen scene of The Magnificent Ambersons [1942]) or arranging them in

congenial way to work with nonactors and eventually realized he could

The mid-1990s brought international audiences a great many Asian film-

but also to meet, sit on panels, participate in marketing, and serve on

ALL TOGETHER NOW

restless, teeming film culture. In my 2000 book, I tried to convey the

Red Poetics: The Films of the Chinese

Tsui Hark’s three-dimensional (3D) action film, The Taking of Tiger

© The Editor(s) (if applicable) and The Author(s) 2016 29

is meant by the “model works.” Second, it considers the genealogy of the

THE “MODEL WORKS”

Efforts to reform the traditional Chinese opera repertoire had been

remaining “friendly” countries such as Albania, North Korea, Yugoslavia,

Chinese to foreign viewers looked equally exotic to Chinese viewers at

ballets, a whole new array of movements was developed for narratives in

THE THREE PROMINENCES

CINEMATIC POETICS OF THE MODEL WORKS

investigation of meaning through semiotic analysis and to also attend to

ending on CU, Changqing’s sidekick, they repeat the dance of checking

then fade to they freeze in a pose and the camera moves

Shot no. Duration Size, angle Movement Music Narrative

Renewal of Song Dynasty Landscape

Surprisingly, an interest in the glories of Chinese art history was virtu-

© The Editor(s) (if applicable) and The Author(s) 2016 51

vast natural environment.4 Predominantly, these works are hanging scrolls,